Marine Geology 196 (2003) 73^89 www.elsevier.com/locate/margeo
Holocene sea-level history on the Rio Grande do Norte State coast, Brazil
Francisco H.R. Bezerra a; , Alcina M.F. Barreto b , Kenitiro Suguio c
Departamento de Geologia, CCET, Universidade Federal do Rio Grande do Norte, Campus Universitario, Natal, RN 59072-970, Brazil Departamento de Geologia, Centro de Tecnologia e Geociencias, Universidade Federal de Pernambuco, Cidade Universitaria, Recife, PE 50740-530, Brazil Departamento de Geologia Sedimentar e Ambiental, Instituto de Geociencias, Universidade de Sao Paulo, Rua do Lago 562, Sao Paulo, SP 05508-900, Brazil Received 6 December 2001; accepted 27 January 2003
Abstract The Rio Grande do Norte State coast, northeastern Brazil, lies in an intraplate region which displays elevated Holocene shorelines, abandoned tidal flats and other intertidal deposits. A field and chronological investigation, which used 48 radiocarbon dates on unaltered mollusk shells, peats, coral reefs, and vermetids in living position and death assemblages encrusted on beachrock, tidal flat and peat deposits, was carried out along two littoral zones, one trending east^west and the other north^south. The sea-level envelope curve for the region indicates at least one sealevel oscillation after the maximum Holocene highstand, which occurred at V5000 cal. yr BP. This curve shows notable deviations from the mean sea-level curve for the Central Brazilian coast in the early and mid Holocene. However, in general, this envelope curve fits a glacio-isostatic prediction for the area. A significant deviation from this prediction, related to a small sea-level oscillation, occurred from V2100 to V1100 cal. yr BP. The comparison between the glacio-isostatic prediction and the sea-level observation enables us to conclude that local events may have disturbed the sea-level record. Minor Holocene sea-level oscillations, which were mainly associated with climate changes or tectonics and superimposed on a major smooth pattern, were observed in South Africa and Australia. But in northeastern Brazil, neotectonic controlled crustal movements or variation in marine and wind currents could provide the answer. The data indicate that local events may have an important role in the history of coastal progradation and retreat. 2003 Elsevier Science B.V. All rights reserved.
Keywords: sea-level change; radiocarbon dating; Holocene; Brazil
1. Introduction The debate on oscillating vs. smooth sea-level curves is becoming increasingly important because of possible sea-level rise related to human-induced global warming (Baker and Haworth, 2000). Sev-
* Corresponding author. E-mail addresses: bezerrafh@geologia.ufrn.br (F.H.R. Bezerra), alcina@npd.ufpe.br (A.M.F. Barreto).
0025-3227 / 03 / $ ^ see front matter 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0025-3227(03)00044-6

MARGO 3295 13-3-03

F.H.R. Bezerra et al. / Marine Geology 196 (2003) 73^89
eral factors are generally taken into account when interpreting postglacial sea-level changes, such as eustasy, glacio-isostatic rebound, wave and wind pattern, climate, and tectonics. Coastal studies carried out in recent decades along the Brazilian coast have revealed a vast amount of diverse information that contributes to the understanding of processes in the coastal region. Several coastal studies, mainly concentrated along the Central Brazilian coast, from latitude 11S to 26S, have described sea-level changes in the Holocene (e.g., Bittencourt et al., 1979; Suguio et al., 1985; Martin et al., 1997). In some localities there is probably evidence for more than one sea-level highstand after 6400 calendar years before present (cal. yr BP) and in others major highstands were probably disturbed by high-frequency oscillations as small as 2^3 m on time scales of no more than 300 years (Suguio et al., 1985, 1988). The existence of these sea-level oscillations was questioned by Angulo and Lessa (1997). Dierences between sea-level curves along the Brazilian coast are ascribed to geoid variation (Martin et al., 1985),

neotectonic movements (Martin et al., 1986a; Bezerra et al., 1998), and low-frequency climatic variations (Suguio et al., 1985; Suguio, 1993). Progress in global glacio-isostatic modeling and in eld techniques now makes it possible to separate the contribution of several components in the sea-level record. For example, discrepancies between eld data and glacio-isostatic predictions may reveal the local eect of other factors. The main objective of this study is to compare age^height data with published glacio-isostatic predictions and mean sea-level behavior for the Central Brazilian coast in an attempt to decouple eustasy from other causes of sea-level changes. We also compare our results with sea-level curves from tectonically stable mid- to low-latitude regions. The following questions are specically addressed: (1) What can cause contrasts in relative sea-level curves from neighboring areas ? (2) How far is regional sea level disturbed? The study focuses on emerged shorelines and superbly preserved coastal deposits along the Rio Grande do Norte State coast, northeastern Brazil.
Fig. 1. General geological map and location of study area. Inset: South American plate and location of the sea-level curve for the Central Brazilian Coast (curve 1) (Bittencourt et al., 1979; Suguio et al., 1985) and the glacio-isostatic model by Peltier (1998) (curve 2).
2. Coastal and geological settings 2.1. Coastal setting The study area is located in latitudes 500PS^ 630PS and longitudes 3710PW^3510PW. It ex tends from Ba|a Formosa to Areia Branca, comprising approximately 300 km of shoreline trending partly north^south and partly east^west (Fig. 1). The two coasts have similar morphologies. The littoral zone is characterized by relatively low relief, and by clastic, dominantly siliciclastic shorelines. Beaches are usually backed by vegetated and non-vegetated Pleistocene to Holocene sand dunes (Yee et al., 2000), behind which lie lagoons, tidal ats, and fossil dune ridges. These coastal dunes occur conspicuously as far as 30 km inland. The littoral zone has experienced erosion or very limited progradation since the Pleistocene (Dominguez and Bittencourt, 1996). Coastal erosion apparently bears no relationship to a rise in sea level, but instead is intrinsically related to sediment budget from the littoral zone (Dominguez and Bittencourt, 1996). The area is drained by small streams, with a few larger rivers. Landwards, tidal ats and lagoons are connected to the sea by tidal inlets. There is no signicant tidal range and climate dierence between the two littoral zones. The area has a mesotidal, semi-diurnal regime, where normal tides attain a maximum of 2.0 m and spring tides have a range of 3.2 m (Hayes, 1979). The present climate is characterized by an alternation of dry and pluvial seasons. The mean annual temperature is ca. 30C and the rainfall is 600^1000 mm/yr (Nimer, 1989). But there are important wind pattern dierences. Owing to the orientation of both the N^S-trending and E^W-trending coasts, the incoming waves and daily winds induce drift to the north and west respectively. Wind velocity ranges from 3.8 to 5.4 m/s on the former and from 2.0 to 5.0 on the latter (Nimer, 1989). 2.2. Geological setting The stratigraphic scheme for the region starts

with a deformed crystalline basement of Precambrian age, composed mainly of gneisses, schists, marbles, quartzites, migmatites, and granites (Jar dim-de-Sa, 1994). The basement is overlain by marine and continental sedimentary rocks of Cretaceous age, and unconformably by the Miocene^ Pliocene continental deposits of the Barreiras Formation (Suguio and Nogueira, 1999). Quaternary alluvial, eolian, and marine deposits overlie older stratigraphic units along river valleys and the littoral zone. Late Pleistocene marine terraces dating from yr BP to yr BP form clis 7^20 m high (Barreto et al., 2002). Late Pleistocene and Holocene coastal sediments (beachrock, peat, tidal at deposits) usually form narrow strips along the current littoral zone and along the river valleys. In addition, late Quaternary faults that oset coastal deposits have been mapped and dated along the littoral zone (Bezerra and Vita-Finzi, 2000; Bezerra et al., 2001). A simplied geological map is shown in Fig. 1.
3. Previous sea-level studies Higher than present Holocene sea levels are well documented in the Southern Hemisphere. They are found in Africa (e.g., Miller et al., 1993; Ramsay, 1995), Australia (e.g., Belperio, 1995), and South America (Suguio et al., 1985; Isla, 1989). Suguio et al. (1985, 1988) and Suguio (1993) have provided several sea-level curves for the Brazilian coast on the basis of more than 700 radiocarbon dates. According to Suguio (1993), sea level along the Central Brazilian coast rose above its present level V7490 cal. yr BP. It reached a maximum at 5460 cal. yr BP and fell to an altitude slightly below present mean sea level (msl) between 3990 and 3860 cal. yr BP, rose again to 3 m above present msl V3470 cal. yr BP, and fell to slightly below the present msl between 2530 and 2350 cal. yr BP, rose to V2.5 m about 2140 cal. yr BP, and nally fell steadily to its present position. The best constrained sea-level curve described by sea-level studies along the Brazilian coast is the Salvador curve, which indicates that the last transgression
reached a maximum at 5460 cal. yr BP when sea level rose 5 m above msl, and then fell to its present position. Two regressive periods occurred around 3990 and 2350 cal. yr BP (Suguio et al., 1985, 1988) (Fig. 1). A valuable discussion on the smooth vs. oscillating pattern of Brazilian sea-level curves occurred in the last few years. Angulo and Lessa (1997) presented evidence contrary to the existence of 2^3-m sea-level oscillations. They presented a smooth curve based on vermetids from V24S to V27S latitude. Martin et al. (1998), however, questioned this smooth curve and presented a review on the Brazilian sea-level data, where they found evidence to conrm the highfrequency oscillations. Finally, Lessa and Angulo (1998), also revising Brazilian sea-level studies, presented vermetid and sedimentary deposit age^ height data that indicate a gradual sea-level fall since 5000 yr BP. They concluded that their position arguing against these oscillations remained unabated in view of the existing evidence. Another approach to explain sea-level changes on the Brazilian coast used several theoretical models based on glacio-isostatic rebound. The rst glacio-isostatic models proposed for northeastern Brazil were presented by Clark et al. (1978) and Peltier (1988). They were based on the sea-level curve for the Brazilian coast of Fairbridge (1976). These sea-level predictions have been supplanted by recent glacio-isostatic studies. The glacio-isostatic prediction generated by the Peltier (1998) global model for the Touros area did not identify a radically dierent eustatic regime that could distinguish the E^W- from the N^S-trending littoral zones. In the absence of neotectonic eects, this prediction indicates that during late Pleistocene times, Vcal. yr BP, the glacial maximum produced a sea level as low as 107 m below present sea level. Sea level rose above the current level slightly after 7000 cal. yr BP in the Touros area, reached a maximum of +2.0 m V5000 cal. yr BP, and then fell gradually to its present position (Fig. 1). Two studies have investigated the Holocene sea-level evolution of the study area. Oliveira et al. (1990) made an attempt at a sea-level curve for the N^S-trending littoral zone but armed with

only four 14 C ages they were forced to adopt the Central Brazilian curve. Bezerra et al. (1998) have used the mismatch between sea-level predictions and the radiocarbon coastal chronology to dene the extent of tectonic uplift and submergence. Neither study proposed a new relative sea-level curve for the area.
4. Methods used to collect and analyze samples 4.1. Sample collection and estimation of vertical uncertainty The determination of heights of sampled sites from former sea-level positions took into account several sources of error. The vertical uncertainty range included (1) the elevation error due to transportation of local port tidal predictions to sampled sites, (2) possible dierences between the observed sea levels and tide predictions, and (3) the accuracy of the sea-level indicator used. The elevation of sampled sites from the current sea level was determined by theodolite leveling. Possible errors from this leveling were negligible. Sample height and location were determined using the same procedures. They mainly followed procedures recommended by the Admiralty (1996). The zero level to which all measurements and corrections were made was the Brazilian Corrego Alegre National datum. Geographic coordinates were determined using a pocket global positioning system device, with an estimated accuracy of 25 m. Altitudes discussed in text are with reference to msl. Sample heights were surveyed to tide levels and normalized to msl using tide-table predictions by the Brazilian Navy (2002). These tables give daily predictions of the times and heights of high and low waters at a selected number of Brazilian ports. The application of the tidal prediction from local ports (Macau and Natal) to other areas was made by linear interpolation. Heights of tidal levels for local ports were obtained from the Brazilian Hydrographic and Navigation Division web site (Brazilian Navy, 2002). We assumed that moving this datum could be a source of error of about 20 cm.
The Admiralty table (Admiralty, 1996) was used to nd the height of tides at times between high and low water. The method is based on the assumption that the duration of tide rise and fall lies within the scope of a graph (height vs. time diagram), which approximates to a cosine curve. A possible dierence between the observed sea level at the sampled sites and the sea level interpolated from the local ports may occur. Mean sea level may remain as much as 0.3 m above or below the average for as long as a month (Admiralty, 1996). This possible source of vertical error was also added to the estimated vertical uncertainty. The accuracy of the sea-level indicators is discussed further below. 4.2. Radiocarbon dating Most problems aecting radiocarbon age related to post-mortem transportation and burial history were minimized by dating material in liv-

ing position, or material which was unambiguously younger or older than the feature in question. Shells selected for dating were mainly articulated and presented primary colors, in addition to the absence of signs of abrasion, which indicated that they lived and died near their burial site. Broken and highly abraded or worn shells from death assemblages as well as specimens with clear signs of diagenetic alteration were not used for dating. Ages of aragonitic shell samples with secondary calcite were considered minimum values and were therefore not included in the relative sea-level curves. In a few locations, however, age discordances observed in paired beachrock samples may indicate that the shells accumulated over periods of tens or hundred of years (samples GR1, GR2, GR3, C14-6, Table 2) or trace contamination detected by X-ray diraction analysis (BR1, BR2, C14-1, Table 2) (Bezerra et al., 2000). Two groups of radiocarbon ages were used.
Table 1 List of radiocarbon ages used in this study from the E^W-trending littoral zone Sample/source C14-18/(1) C14-19/(1) C14-20/(1) C14-21/(1) C14-22/(1) C14-23/(1) C14-24/(1) C14-25/(1) C14-26/(1) C14-27/(1) C14-28/(1) C14-29/(1) C14-30/(1) C14-31A/(1) C14-31B/(1) PG/(2) MC1/(2) MC2/(2) GA/(2) FSA1/(2) REC1/(2) GU/(2) SAL/(2) Lab number Beta 121264 Beta 121265 Beta 121266 Beta 121267 Beta 121268 Beta 121269 Beta 121270 Beta 121271 Beta 121272 Beta 121273 Beta 121274 Beta 121275 Beta 121276 Beta 121277 Beta 121278 UCL 423 UCL 345 UCL 418 UCL 416 UCL 410 UCL 397 UCL 431 UCL 417 Height (m above msl) 0.2 1.0 2.0 1.0 2.1 1.0 2.1 1.0 1.2 0.5 0.0 1.0 0.0 1.0 1.0 1.0 1.8 0.5 3.9 1.0 0.6 1.0 4.0 1.0 4.0 1.0 4.0 1.0 4.0 1.0 0.6 1.0 1.8 1.0 1.0 1.0 1.1 1.0 30.5 1.0 3.9 1.0 0.6 1.0 1.2 1.0 Sea-level indicator/rock dab/br dab/br dab/br dab/br dab/br wd/pd blp/pd dab/br vm/br oy/br cr/br dab/atf dab/atf dab/atf oy/tf dab/br dab/br mtf oy/mtf dab/br dab/br dab/br dab/br

C/12 C

14 C age (yr BP)
Calibrated age (yr BP at 2c) 3640^3330 2960^2720 4140^3760 2850^60 870^610 2670^2300 5450^5130 3260^2860 5600^5300 4980^4630 5610^5310 1690^1390 2680^2190 1250^1060 3680^3210 6900^6080 5730^5060 3040^2690 4240^3640

0.0 31.00 31.00 31.10 30.50 325.40 327.20 30.70 2.00 0.50 30.60 32.10 32.00 31.10 37.30 0.54 0.48 32.37 0.30 31.23 ^ ^ 30.19

live 3950 110

Sample source: (1) this study; (2) Bezerra et al. (1998). Sea-level indicator code: das, death assemblage of bivalve shells; wd, wood; blp, bivalve shell in living position; vm, vermetid; oy, oyster; cr, coral reef. Host rock code: br, beachrock; pd, peat deposit; atf, abandoned tidal at; mtf, modern tidal at.
The rst group represents ages published by Bezerra et al. (1998) and Bezerra and Vita-Finzi (2000) (Tables 1 and 2). It is made up of 21 samples dated by rst-order radiocarbon assay at University College London (UK) according to the method developed by Vita-Finzi (1983, 1991). This group also includes one age analyzed by accelerator mass spectrometry at Beta Analytic (USA). The second group consists of 26 new ages, which were analyzed by the conventional radiocarbon method at Beta Analytic (Tables 1 and 2). Conventional results were calibrated to cal. yr BP using the program CALIB (version 4.3), written and distributed by the Quaternary Isotope Lab, University of Washington (Stuiver and Reimer, 1993). Calendar calibration results were rounded to the nearest 10 years. Age ranges are reported as time intervals of the extremes of 2c. Ages younger than 450 yr BP were considered
invalid for calibration with the data of Stuiver and Reimer (1993). The calibrated ages adopted here do not include hard water or reservoir corrections, although CALIB incorporates a time-dependent global ocean reservoir correction of about 400 years. But as the study area is not subjected to restricted marine circulation, the reservoir eect is not considered important (Bezerra et al., 2000). Furthermore, the activity determined for a specimen collected live (MC2, Table 1), 1 cpm above the pre-bomb level of 7.7 cpm above background, shows that the residual eect of the atmospheric bomb tests swamps any apparent residual reservoir eect within the resolution of the method (Bezerra et al., 1998). In addition, similarly, there is no indication that dead carbon from the con tinent, mainly from limestones of the Janda|ra Formation, which have been continuously washed
Table 2 List of radiocarbon ages used in this study from the N^S-trending littoral zone Sample/source C14-1/(1) C14-2/(1) C14-3/(1) C14-4/(1) C14-6/(1) C14-7/(1) C14-8/(1) C14-10/(1) C14-11/(1) C14-14/(1) C14-17/(1) PB/(2) RF/(2) s PJC/(2) JC/(2) VC/(2) s PR2/(2) s PR3/(2) BR1/(2) BR2/(2) GR1/(2) GR2/(2) GR3/(2) CH2/(2) CH1/(2) Lab number Beta 121249 Beta 121250 Beta 121251 Beta 121252 Beta 121253 Beta 121254 Beta 121255 Beta 121257 Beta 121258 Beta 121260 Beta 121263 UCL 420 UCL 409 UCL 424 UCL 413 UCL 430 UCL 425 UCL 361 UCL 403 UCL 404 UCL 419 UCL 421 UCL 405 UCL 432 UCL 414 Height (m above msl) 2.1 1.0 0.6 0.5 0.5 1.0 0.5 1.0 0.8 1.0 0.4 0.5 0.5 0.5 2.3 0.5 1.0 0.5 30.2 0.5 0.2 1.0 0.1 1.0 30.2 1.0 30.4 0.5 1.0 30.7 1.0 0.1 0.5 2.2 1.0 1.8 1.0 0.2 1.0 0.0 1.0 0.7 1.0 1.7 1.0 1.5 1.0 Sea-level indicator/rock dab/br vm/br wd/pd oy/pd dab/br vm/BFs vm/BFs vm/BFs dab/br vm/br md/pd dab/br blp/pd cr dab/br dab/br cr cr dab/br dab/br dab/br dab/br dab/br dab/br dab/br

based upon changes in texture, grain size, and sedimentary structures have been presented by Shipp (1984), Dupre (1984), and Bezerra et al. (1998). Two beachrock facies were identied by Bezerra et al. (1998) : facies A, which represents the lower foreshore and upper shoreface with a precision of 0.5 m (Fig. 6) and facies B, which corresponds to the middle to lower foreshore with a precision of 1.0 m (Fig. 7). Death assemblages were also collected in tidal ats (Figs. 5 and 8), where they were sometimes found still articulated. Tidal ats are
Fig. 5. Modern beach cross-section and Holocene coastal deposits. (A) Guajiru beach showing site of sample REC-1 and GU (modied from Bezerra et al., 1998). (B) Rio do Fogo beach showing site of sample RF.
Fig. 6. Trough cross-stratication at Guajiru beachrock dated 3040^2690 cal. yr BP (sample GU, Table 1), which represents facies A of Bezerra et al. (1998).
poorly to moderately well sorted and commonly possess alternating layers of mud and sand. They occur on the E^W-trending coast, mainly in the Acu River Delta and the Mossoro River Estuary (Fig. 9). The tidal ats represent msl with a precision of 1.0 m (Bezerra et al., 1998). Fragments of wood in peat deposits were collected at least at one site on the N^S-trending
coast (Fig. 10). The term peat is used in this study for lithofacies which appears to consist of at least 50% organic material and which is dominated by roots and tree branches at the top and black clay at the bottom. The upper part of the peat deposit that was sampled represents middle to upper foreshore with a precision of 1.0 m (Geyh et al., 1979; Morzadec-Kerfourn, 1979; Martin et al., 1986b).
Fig. 7. Swash cross-stratication at Guajiru beachrock, which corresponds to beachrock facies B of Bezerra et al. (1998).
Fig. 8. General view of the deposit with death assemblage of mollusk shells dated 5610^5310 cal. yr BP (sample C14-31A, Table 1) and 1690^1390 cal. yr BP (sample C14-31B) at a tidal at to the south of Macau.
5.2. Field data Two sea-level patterns, which show small dierences, appear when the E^W- and N^S-trending littoral zones are compared. Tables 1 and 2 list age^height data from sea-level indicators sampled in the study area. We drew envelope curves that enclosed potential errors: the horizontal errors (radiocarbon ages) and the estimated vertical errors (sample and sea-level indicator heights). Both curves were hand-tted and have large error envelopes that represent the adequate assessment of age and height uncertainty. They are plotted in Fig. 11 using calibrated radiocarbon time scale. We set the N^S-trending against the E^Wtrending envelope curve. The area common to
both curves was highlighted and sorted out as a third and nal envelope curve (Fig. 11). We suggest this nal envelope curve represents the sealevel trend in the study area. There is evidence for multiple phases of sea-level uctuation in the Holocene. The envelope curve for the study area indicates that relative sea level rose between V7150 and 5800 to V5000 cal. yr BP to about 2.5^4.0 m above present sea level. After this sea-level highstand (V5000 cal. yr BP), one important positive oscillation occurred from V2100 to V1100 cal. yr BP. The curve also indicates that a lower than actual sea level could have occurred in the region between these major oscillations. The last sea-level oscillation is based mainly on samples C14-1 and C14-31A, MC1, which are located on the N^S-trending and E^W-trending coasts respectively. Samples C14-1 and MC1 are death assemblages of bivalve shells from extensive beachrock bodies, whereas sample C14-31B is composed of oysters from a large and well-exposed tidal at deposit. These samples represent reliable sea-level indicators from carefully mapped sedimentary deposits. The envelope curve for the study area is plotted in Fig. 12 together with the eustatic prediction by Peltier (1998) and the relative mean sea-level curve for the Central Brazilian coast (Suguio, 1993). There are remarkable discrepancies between the curve for the Central Brazilian coast and the envelope curve for the study area. On the other hand, the glacio-isostatic curve of Peltier (1998) falls almost entirely within the envelope curve for the study area. But dates from V2100 cal. yr BP to V1100 cal. yr BP depart by 2^3 m from the glacio-isostatic prediction of Peltier (1988).

6. Discussion Over the last years, the dominant factors controlling the position of sea level have generally been eustatic and glacio-isostatic, but as regional dierences in the isostatic signal will have declined towards the present, variations in the Holocene sea-level history should highlight local controls.

Fig. 9. Location map of

C samples collected along the E^W-trending littoral.
A review of the smooth vs. oscillating sea-level curve discussion (Angulo and Lessa, 1997; Lessa and Angulo, 1998; Martin et al., 1998) is necessary to address this point. Vertical uncertainty related to paleosea-level indicators and sample positions is a central point for this debate. The drawing of lines joining age^height data gives a false sense of accuracy to several curves that are claimed to represent sea-level changes along the Brazilian coast. We suggest these opposing views can be accommodated if age^height uncertainty, regional glacio-isostatic rebound and local factors that inuence sea level are taken into account. A more cautious approach is presented in this study. Our data indicate that there are few dierences between the envelope curves of the N^Sand the E^W-trending littoral zones. The error envelope, however, is too large to claim that the dierences between these curves are signicant. The envelope curve for the study area, derived from the previous curves, departs from Peltiers (1998) glacio-isostatic prediction at V2100^1100 cal. yr BP. It suggests that local factors that disturb the sea-level record are present in the region. Sea-level data from the low to middle latitudes of the Southern Hemisphere indicate that minor
peaks or variability in the Holocene highstand are mainly caused by climate changes or tectonics. Oscillations in sea level were identied in the southwestern Cape area and southern coast, South Africa (e.g., Illenberger and Verhagen, 1990), which were associated with neoglaciation periods (Jerardino, 1995). In addition, the investigation of Baxter and Meadows (1999) at Western Cape, South Africa, indicate three major sealevel transgressive and regressive phases from V8000 yr BP to the present. In Western Australia, spatial and temporal variability in the Holocene sea-level highstand was associated with local tectonic eects (Belperio et al., 2002). But an ambiguous pattern was suggested by the study of Baker and Haworth (2000). They analyzed data from southeast and northeast Australia, the Pacic Islands, and southern Brazil by carrying out cross-regional statistical regressions of xed biological indicators. They concluded that both smooth and oscillating sea-level curves are possible in these regions. Belperio et al. (2002) reached a similar conclusion on the ambiguous pattern of sea-level curves from the South Australian coastline. In northeastern Brazil, local factors could also

Fig. 10. Location map of

C samples collected along the N^S-trending littoral.
be the main cause for sea-level oscillations. First, the assumption that the Brazilian coast is a highly stable region has been scarcely questioned. But several lines of evidence indicate that the answer for the oscillation that departs from the glacioisostatic model could lie in tectonics. Coseismic uplift may have occurred to the east of Carnaubais fault, along the E^W-trending littoral zone (Fig. 9) (Bezerra et al., 1998; Bezerra and VitaFinzi, 2000), where rapid emergence of at least 4 m occurred ca. 4080^2790 cal. yr BP. In the Acu
Delta (Fig. 9), Silva (1991) used auger-hole and vibracore data to identify NW- and NE-trending normal fault movements which seem to represent the reactivation of Cretaceous faults. These NWand NE-trending faults cut across marine and uvial sediments dated to 370 yr BP with normal osets of 30 m. He suggested that normal faults controlled both Pleistocene and Holocene marine and uvial sedimentation by providing the topographic gradients and barriers for distributing sediments. In addition, a yr BP ma-
Fig. 11. Sea-level curves showing error envelopes based on 2c age ranges and estimated vertical uncertainty. (A) Samples of the E^W-trending littoral zone. (B) Samples of the N^S-trending littoral zone. (C) Final sea-level curve for the study area (gray area) derived from curves A and B.
rine terrace that forms sea cli as high as 20 m in the area, when compared with marine oxygen isotope stages and deposits of similar age described 1000 km to the south (Martin et al., 1982), reects at least 10^12 m uplift of the southeastern block of the Carnaubais fault since deposition (Barreto et al., 2002). The NE-oriented Jundia| fault which occurs on the N^S-trending coast (Fig. 10) also cuts across the Holocene coastal record. Bezerra and VitaFinzi (2000) and Bezerra et al. (2001) concluded that the fault osets alluvial sediments which yielded a 14 C age of 4860^4570 cal. yr BP. Structural data indicate that normal fault movement
was followed by dextral strike-slip movement in the Holocene. These fault movements caused coastal uplift and subsidence that may have disturbed the sealevel record. However, the high-frequency oscillations between 4000 and 2500 cal. yr BP may correspond to sea-level changes resulting from lowmagnitude climatic variations as pointed out by Suguio et al. (1985) and Suguio (1993) for the Central Brazilian coast. Second, the wind and wave pattern dierence between the N^S- and the E^W-trending coasts could be another factor that disturbs the local sea-level record and produces minor sea-level os-
Fig. 12. Final sea-level curve for the study area (gray area) against (A) the relative sea-level curve for the Central Brazilian coast (solid line) of Bittencourt et al. (1979) and Suguio et al. (1985) and (B) the glacio-isostatic prediction of Peltier (1998) (solid line).

cillations. Usually, a strong wind blowing straight onshore would pile up the water and cause high tides to be higher than predicted (Admiralty, 1996). This pattern could occur along the N^Strending coast, where wind currents blow from southeast to northwest. In addition, the variation in wind direction could also result in storm surges which can raise water levels.
to about 2.5^4.0 m and induced an overall coastal retreat in the region. Sea level fell immediately and eventually rose again about 2100^1100 cal. yr BP, resulting in a second coastal retreat in the Holocene. This envelope curve does not match Peltiers model entirely nor the curve for the Central Brazilian coast. Holocene sea-level oscillations have been observed in South Africa and probably in Australia. These oscillations were mainly associated with minor climate changes or tectonics. But in northeastern Brazil, eld data suggest the local sea-level record may have been disturbed by tectonics or wind^wave patterns. This relative inconsistency between prediction and eld data indicates that local factors that contribute to sea-level changes cannot be ignored. It follows that sea-level curves along many stretches of northeastern Brazil must be constructed for small areas rather than vast littoral zones along which variability in factors such as tectonics or wind and wave patterns makes generalization on sea-level changes dicult. We also suggest that sea-level curves along the Brazilian coast have only regional, or even local, validity. But whether or not the discrepancies between prediction and eld data found in the study area are the result of tectonics or wind and wave patterns is worthy of further investigation.
Acknowledgements 7. Conclusion In this paper we attempted to establish the sealevel history of the Rio Grande do Norte State coast, northeastern Brazil, based on 48 radiocarbon dates used to model relative sea-level changes. When analyzed, ages and heights were accompanied by error margins. A glacio-isostatic prediction generated by the global sea-level model of Peltier (1998) and the relative mean sea-level curve for the Central Brazilian coast were compared with eld data. Based on the envelope curve for the study area, a relatively rapid sea-level rise occurred between V7100^5800 cal. yr BP and about 5000 cal. yr BP Support for this research was partly provided by Grant Fapesp Geociencias-97/09974-3. We thank P. Pirazzoli, J.T. Wells, and an anonymous referee for valuable suggestions and manuscript review. We also thank C. Vita-Finzi for an early review of the manuscript and help with radiocarbon dating carried out at University College London. References
Admiralty, 1996. Admiralty Tide Tables: The Atlantic. Admiralty Hydrographic Department, The Hydrographer of the Navy, Southampton. Angulo, R.J., Lessa, G.C., 1997. The Brazilian sea level

F.H.R. Bezerra et al. / Marine Geology 196 (2003) 73^89 Geyh, M.A., Kudrass, H.-R., Streif, H., 1979. Sea-level changes during the late Pleistocene and Holocene in the Strait of Malacca. Nature 278, 441^443. Hayes, M.O., 1979. Barrier island morphology as a function of tidal and wave regime. In: Leatherman, S.P. (Ed.), Barrier Islands. Academic Press, New York, pp. 1^27. Illenberger, W.I., Verhagen, B.T., 1990. Environmental history and dating of coastal duneelds. South Africa. South Afr. J. Sci. 86, 311^314. Isla, F.I., 1989. Holocene sea-level uctuations in the Southern Hemisphere. Quat. Sci. Rev. 8, 359^368. Jardim-de-Sa, E.F., 1994. A Faixa Serido (Prov|ncia Borbor ema) e seu signicado geodinamico na Cadeia Brasiliana/ Panafricana. Ph.D. Thesis, Universidade de Bras|lia, Bras|lia, 804 pp. Jerardino, A., 1995. Late Holocene neoglacial episodes in southern South America and South Africa: a comparison. Holocene 5, 361^368. Kempf, M., Laborel, J., 1967. Formations de vermets et dalgues calcaires sur les cotes du Bresil. Rec. Trav. Stat. Mar. Endoume 59, 9^23. Kikuchi, R.K.P., Leao, Z.M.A.N., 1998. Rocas (southwestern equatorial Atlantic, Brazil): An atoll built primarily by coralline algae. 8th International Coral Reef Symposium, Panama. Laborel, J., 1979. Les gasteropodes vermetides: Leur utilisation comme marqueurs biologiques de rivages fossiles. Oceanis 5, 221^239. Lessa, G.C., Angulo, R.J., 1998. Oscillations or not oscillations, that is the question ^ Reply. Mar. Geol. 150, 189^ 196. Martin, L., Bittencourt, A.C.S.P., Dominguez, J.M.L., Flexor, J.-M., Suguio, K., 1998. Oscillations or not oscillations, that is the question: Comment on Angulo, R.J. and Lessa, G.C. The Brazilian sea level curves a critical review with empha sis on the curves from Paranagua and Cananeia regions [Mar. Geol. 140, 141^166]. Mar. Geol. 150, 178^187. Martin, L., Bittencourt, A.C.S.P., Vilas-Boas, G.S., 1982. Pri meira ocorrencia de corais pleistocenicos da costa brasiliera. ltima transgressao. Ci. Terra 1, Datacao do maximo da penu 16^17. Martin, L., Flexor, J.-M., Blitzkow, D., Suguio, K., 1985. Geoid change indication along the Brazilian coast during the last 7,000 years. In: Coral Reef Congress 5, Tahiti, Proceedings, IGCP, Project 200, 3, pp. 85^90. Martin, L., Flexor, J.M., Bittencourt, A.C.S.P., Dominguez, J.M.L., 1986a. Neotectonic movement on a passive continental margin, Salvador region, Brazil. Neotectonics 1, 87^ 103. Martin, L., Morner, N.A., Flexor, J.M., Suguio, K., 1986b. Fundamentos e reconstruco de antigos n|veis marinhos do es Quaternario. Bull. Geosci. Inst. Univ. Sao Paulo, Sao Paulo, Brazil 4, 1^161. Martin, L., Suguio, K., Dominguez, J.M.L., Flexor, J.M., 1997. Geologia do Quaternario Costeiro do litoral norte do Rio de Janeiro e do Esp|rito Santo. Publicacao FA gicos escala 1:200.000. PESP-CPRM, 104 pp. Mapas Geolo

curves: a critical review with emphasis on the curves from Paranagua and Cananeia regions. Mar. Geol. 140, 141^166. Baker, R.G.V., Haworth, R.J., 2000. Smooth or oscillating late Holocene sea-level curve? Evidence from cross-regional statistical regressions of xed biological indicators. Mar. Geol. 163, 353^365. Barreto, A.M.F., Bezerra, F.H.R., Suguio, K., Tatumi, S.H., Yee, M., Paiva, R., Munita, R., 2002. Late Pleistocene marine terrace deposits in northeastern Brazil: sea-level changes and tectonic implications. Palaeogeogr. Palaeoclimatol. Palaeoecol. 179, 57^69. Baxter, A.J., Meadows, M.E., 1999. Evidence for Holocene sea level change at Verlorenlei, Western Cape, South Africa. Quat. Int. 56, 65^79. Belperio, A.P., 1995. The Quaternary. In: Drexel, J.F., Preiss, W.V. (Eds.), The Geology of South Australia, vol. 2. Geol. Soc. South Aust. Bull. 54, 218^281. Belperio, A.P., Harvey, N., Bourman, R.P., 2002. Spatial and temporal variability in the Holocene sea-level record of the South Australian coastline. Sediment. Geol. 150, 153^169. Bezerra, F.H.R., Amaro, V.E., Vita-Finzi, C., Saadi, A., 2001. Pliocene^Quaternary fault control of coastal plain morphology and sedimentation in northeastern Brazil. J. South Am. Earth Sci. 14, 61^75. Bezerra, F.H.R., Lima-Filho, F.P., Amaral, R.F., Caldas, L.H.O., Costa-Neto, L.X., 1998. Holocene coastal tectonics in NE Brazil In: Stewart, I., Vita-Finzi, C. (Eds.), Coastal Tectonics. Geol. Soc. London Spec. Publ. 146, 279^293. Bezerra, F.H.R., Vita-Finzi, C., 2000. How active is a passive margin? Paleoseismicity in northeastern Brazil. Geology 28, 591^594. Bezerra, F.H.R., Vita-Finzi, C., Lima-Filho, F.P., 2000. The use of marine shells for radiocarbon dating of coastal deposits. Rev. Bras. Geoci. 30, 211^213. Bittencourt, A.C.S., Martin, L., Vilas-Boas, G.S., Flexor, J.M., 1979. Quaternary marine formations of the state of Bahia (Brazil). 1978 International Symposium on Coastal Evolution in the Quaternary Proceedings, Sao Paulo, pp. 232^253. Brazilian Navy, 2002. Tide tables for Brazilian ports. http:// www.mar.mil.br/Vdhn/tabuas. Campos-e-Silva, A., Silva, D.D., Vasconcelos, M.D.T., 1964. Informaco es sobre a malacofauna dos beach rocks de Touros e Sao Bento do Norte, Rio Grande do Norte. Arq. Inst. Antropol. Univ. Fed. Rio Grande do Norte 1, 79^89. Clark, J.A., Farrel, W.E., Peltier, W.R., 1978. Global changes in post-glacial sea level: a numerical calculation. Quat. Res. 9, 265^287. Dominguez, J.M.L., Bittencourt, A.C.S.P., 1996. Regional assessment of long-term trends of coastal erosion in northeastern Brazil. An. Acad. Bras. Ci. 68, 655^671. Dupre, W.R., 1984. Reconstruction of paleo-wave conditions during the Late Pleistocene from marine terrace deposits, Monterey Bay, California. Mar. Geol. 60, 435^454. Fairbridge, R., 1976. Shellsh-eating preceramic indians in coastal Brazil. Science 191, 353^359.

Setting Drop Addresses

RIO Network Cable System
Overview The RIO processor at the controller head-end is connected to an adapter at each of the remote drops via a network cable system. Starting at the RIO processor and running the entire length of the network are one (linear) or two (dual or redundant) trunk cable(s). Taps are installed along the length of the trunk cable(s), and a drop cable is run from a tap to a drop adapter. The trunk cable may be an approved flexible or semirigid coaxial type. See RIO Network Hardware Components, p. 67 for more details. The taps connect the drop adapter at each drop to the trunk cable via a drop cable, providing each adapter with a portion of the signal that is on the trunk. The taps also isolate each drop adapter from all other drop adapters on the network so that they wont interfere with each other. Extending from a tap to an adapter is a drop cable. The drop cable connects to the tap with an F connector, and it connects to the adapter with either an F connector or a BNC connector, depending on the type of RIO adapter at the drop (see Planning RIO Drops, p. 62). The drop cable may be an approved coaxial type, as specified in RIO Network Hardware Components, p. 67. Splitters are used to create a branch in the network cable trunk. They provide isolation between the branches and allow the cable to be laid out in two directions. One trunk splitter is allowed in a network. Hot Standby systems are allowed a second splitter to connect the two RIO heads.

Trunk Cable

Drop Cable

Splitters

Terminating the Cable System
A proper impedance match is maintained across the network with 75 terminators. You must install a 75 terminator:
l in the unused trunk port of the last tap on the network to terminate the trunk cable l in any open drop cable ports on taps that have been installed for future system l in-line on cables running from the primary and standby controllers to the splitter
in a Hot Standby system; this allows you to disconnect one of the two Hot Standby controllers while the other one maintains primary control Terminators are present inside most drop adapters to automatically terminate each drop connectionthe exceptions are some older J890/J892 Adapters and the 410 and 3240 Motion Control products:
RIO Adapters that Do Not Have Internal Termination RIO Drop Adapters AS-J890-001 AS-J890-Motion Controllers 110-230 110-Motion Controllers 100-265-815 100-265-816 100-265-825 110-231 110-233 AS-J892-001 AS-J890-002

expansion

The devices listed above require an in-line terminator (part number 60-0513-000) installed in the drop cable. Note: The J890/J892-10x Adapters contain internal termination. When a drop cable without in-line termination gets disconnected from an adapter while the network is running, the possibility of network errors and data transfer delays is introduced. When internally terminated adapters are installed, you may want to consider designing mechanical self-termination into your drop cables, particularly if a time-critical application is being run on the network. For more details on this and other aspects of cable system termination, see Tap Connections and Locations, p. 49.

RIO Network Node Part Number Summary

RIO Devices

RIO Device Type Head Processor in a 16K 984A chassis (standard) in a 32K 984A chassis (standard) in a 32K 984B chassis (standard) in a 64K 984B chassis (standard) in a 128K 984B chassis (standard) in a 984X chassis (standard) on an AT-984 (standard) on an MC-984 (standard) on a Q984 for MicroVAX II (standard) on a 984-485E (standard) on a 984-48K (standard) option module for 984-685E and 984-785E/K/D
The following table shows RIO device types.
One RIO Port Px-984A-816* Px-984A-832* Px-984B-832* Px-984B-864* Px-984B-828* S929-001 AM-0984-AT0 AM-0984-MC0 AM-0984-Q20 PC-E984-485 PC-K984-485 AS-S908-110 140CRP93200 AS-J890-102 ASP890300 AS-J892-102 ASP890300 ASP890300 ASP890300 ASP890300 Px-984A-932* Px-984B-932* Px-984B-964* Px-984B-928* Two RIO Ports
option module for Quantum all CPUs 140CRP93100 Drop Adapter for 800 Series I/O for 800 Series I/O with two ASCII ports for 800 Series I/O with built-in P/S for 800 Series I/O with two ASCII ports, built in P/S AS-J890-101 ASP890300 AS-J892-101 ASP890300 AS-P890-000 ASP890300 ASP890300
for 800 Series I/O with ASCII, built in AS-P892-000 P/S ASP890300 J291 conversion for 200 Series I/O J290 conversion for 200 Series I/O with ASCII without ASCII for Quantum I/O AS-P451-581/-681 AS-P453-582/-682 AS-P453-581/-681 140CRA93100
AS-P453-592/-692 AS-P453-591/-691 140CRA93200
*These part numbers are for the entire chassis mount PLC system, including the chassis itself; x = 1 for a four-card chassis; and x = 5 for a seven-card chassis.
Planning and Designing an RIO Cable System
Overview Whats in this Chapter? This chapter provides information on planning and designing an RIO cable system. This chapter contains the following topics:
Topic Linear Cable Topologies Hot Standby Cable Topologies Trunk Splitter Use Illegal Coaxial Cable Topologies Using Fiber Optics in an RIO System RIO System Design Choosing Coaxial Cables for an RIO Network Coaxial Cable Characteristics Electrical Characteristics of Coaxial Media Components EMI/RFI Considerations in a Coaxial Cable Routing Plan Tap Connections and Locations Grounding and Surge Suppression Terminating a Coaxial Cable System Designing a Coaxial Cable System to an Attenuation Limit Attenuation Considerations in an Optical Path Maximum Number of Repeaters and Jitter Considerations Planning RIO Drops Page 62
Planning and Designing RIO Cable System

Linear Cable Topologies

Overview There are many possible topologies that may be used for RIO networks. The most common RIO networks use one or two coaxial trunk cables with taps that connect via coaxial drop cables to a series of remote I/O drops. At the head-end of a trunk cable is the PLC with an RIO processor, and at each remote drop is an RIO adapter. These topologies are linearthey do not use any branches or loops in the cable layouts. A single-cable linear topology, as shown in the following illustration, is the simplest and most commonly used RIO cable system:

RIO System Design

Overview When designing an RIO cable system, consider:
l whether you will route one or two cables to the remote drops l the node limitationse.g., single-port or dual port, ASCII device support l the expansion capabilities of the PLCsi.e., the maximum number of drops l the number of nodeshead processors and drop adapters l the locations and the environmental conditions in which these nodes must
operate Key Elements in a Cable System Plan The following are the key elements in a cable system plan: supported
l The cable system must be dedicated to RIOno other signals or power can be l The attenuation between the head processor (or the last fiber optic repeater, if an l l l l l
optical link is used) and any drop adapter must not exceed 35 dB at 1.544 MHz (32 dB for the host-based 984 PLCs) minimum bend radiuses specified for the trunk and drop cables must not be exceeded expansion and contraction loops should be put into the cable system to allow for temperature changes band marked trunk cable is useful for determining tap placement the cable system should be single-point grounded within 20 ft. of the RIO processorthe central ground point may be a tap, a splitter, or a ground block the physical cable installation must be well supported, and cable pull strength must be considered; some manufacturers suggest that RG-6 and RG-11 cable be supported at least every 50 ft; contact the manufacturer to ensure that you do not exceed the strain limit of the cable. where rodents may be a problem, protect the cable installation by using conduit or a similar material precautions should be taken when the media components are installed in hostile environments where high temperatures or corrosives existconsult cable manufacturers and/or CATV suppliers for other special products for harsh environments applied or transmitted on this network
Note: Document your decisions for the installer and for future reference by maintenance personnel. Use the forms provided in Planning RIO Drops, p. 62 to document the system.
Planning for System Expansion
The potential for system expansion should be considered in the initial design. It is less costly to provide for expansion in the original RIO network plan than to redesign the network later. If your PLC is able to support more RIO drops than your current plan requires, consider installing additional taps along the network trunk cable. If, for instance, you intend to use a Quantum CPU, which could support up to 31 remote drops, and your current plan calls for only 10 remote drops, you can install as many as 21 extra taps for future expansion. Remember that the unused expansion taps need to be terminated (see Network Terminators, p. 83).

EMI/RFI Considerations in a Coaxial Cable Routing Plan
Overview Electromagnetic interference (EMI) and radio frequency interference (RFI) sources can be avoided by using effectively shielded cable and by using the cable away from troublesome locations.
Guidelines for Interference Avoidance
l Avoid installation of RIO cables in trays or conduits that contain AC or DC power l Separate RIO cable from power cable or power sources; trunk cable runs should
avoid panels, trays, and other enclosures that contain power wires. Note: We recommend that a spacing of 12. 14 in./kV of power be maintained between the RIO cable installation and power cables. cable or power services
l Make sure that any RIO cable power cable crossings are at right angles only l Do not route trunk cable into equipment cabinets or panelstrunk cable and taps
should be mounted away from cabinets or panels in a separate enclosure (One satisfactory method is to install the trunk cable in the ceiling of the facility and mount the taps within an enclosure up in the ceiling. The drop cable can then be installed down to the node.) l Do not exceed the cables minimum bend radius and pull strength l Install cable in steel conduit in high noise environments
Tap Connections and Locations
Overview Each tap has three portsa trunk-in port, a drop cable port, and a trunk-out port; the RIO cables connect to the tap ports via F connectors. The taps come mounted to a plastic block that is used to isolate them from ground. They must be surface mounted to a wall or an enclosure. Make sure that no tap in the RIO system is grounded or touched by a grounded metallic surface unless it is being used intentionally as the single grounding point for the entire system. Improper placement of taps can cause signal reflections and distortion of the signal waveform. Proper placement will keep these reflections to a minimum and avoid problems with waveform distortion. The preferred method of tap placement is on cable band markers. Note: If taps are placed too close to each other (or too close to a splitter in a Hot Standby system), a cumulative reflection will result. To avoid this problem, install taps at least 8 ft 2 in. (2.5 m) away from one another. Trunk cable with band markers applied at regular intervals should be purchased from the manufacturer. Intervals will vary based on the propagation of the cable. Modicon RG-11 trunk cable is band marked at 8.86 ft (2.7 m) intervals; RG-6 cable is not band marked. If you are not using Modicon RG-11 for trunk cable, you can instruct your cable manufacturer to apply marker at the required intervals. The cost to perform band marking is very small. Tap Port Connections An RG-11 cable can connect directly to a tap port F connector via a Modicon 490RIO00211 F Connector installed on the end of the cable (see F Connectors for Coaxial Cables, p. 78). Quad shield RG-6 cable can be connected to a tap port F connector via a Modicon MA-0329-001 F Connector (see F Connectors for Coaxial Cables, p. 78). Semirigid cable is more difficult to connect to the two (trunk-in and trunk-out) F connector ports on the tap. Because there is only a 1 in. space between the two ports, you may not be able to fit semirigid connectors directly on both ports. To avoid this problem, we recommend that you use high quality 90 right angle F adapters such as the Modicon 52-0480-000 Right Angle F Adapter (see F Adapters for Semirigid Cable, p. 80).

RIO Coaxial Cable System Hardware Components
The following table shows the RIO coaxial cable system hardware components.
Description Tap Splitter Hot Standby system use used for trunk splitter F Connectors quad shield RG-11 (6/bag) Part Number MA-0185-100 MA-0186-100 MA-0331-000 490RIO00211
quad shield RG-6 (10/cassette) MA-0329-001 Right angle F connector BNC connectors non-quad shield RG-6 quad shield RG-6 F-to-BNC Adapter BNC Jack to male F connector Tap port terminator Trunk terminator BNC In-line terminator Self-terminating BNC Adapter Hot Standby Processor Warning Label Self-terminating F Adapter Hot Standby system and drop use Hot Standby system use Hot Standby system and drop use Hot Standby system and drop use Ground block Surge suppressor Semirigid Connectors QR540JCA Cable QR869JCA Cable 52-0480-000 52-0487-52-0614-000 52-0724-000 52-0402-000 52-0422-000 60-0513-000 52-0370-000 MD-9423-000 52-0399-000 (non-quad shield RG-6) 52-0411-000 (quad shield RG-6) 60-0545-000 CBT-22300G (Relcom) AI540FMQR (CommScope) AI860FMWQR (CommScope)

Tap Specifications

Overview Modicon MA-0185-100 Taps connect the drop cables to the main trunk cable and isolate the RIO drop adapter from the rest of the network. This tap is nondirectionalit allows signals to be propagated in both directions along the trunk cable. An MA-0185-100 tap has one drop port and two trunk ports.

AN AEG COMPANY

1.00 in.

2.75 in. 3.1 in.

2.00 in. OUT

.650 in.97 in.

Note: Although the trunk ports are labeled IN and OUT, these labels can be ignoredi.e., the tap is not directional. An MA-0185-100 tap is supplied with a plastic isolator on its back. The tap isolates the drop adapter from the trunk cable by 14 dB. Unused ports on the taps must be terminated with a Modicon 52-0402-000 Port Terminator, and the last (trunk-out) port of the last tap on the network must be terminated with a Modicon 52-0422-000 Trunk Terminator (see Network Terminators, p. 83).
The following table shows the specifications for the MA-0185-100 Tap.
MA-0185-100 Tap Specifications Impedance Frequency Range Tap Loss Trunk Insertion Loss Trunk Return Loss Tap Return Loss Temperature Range Humidity Sealing Interconnections kHz. 30 MHz 14 dB (+0.5 dB) 0.8 dB maximum 26 dB maximum -18 dB minimum -40. +60 C 95% at 85 C RFI/EMI sealed F Connectors torque up to 90 in./lb
Note: Taps not supplied by Modicon are not supported by Modicon.
Note: The Modicon MA-0185-000 Tap can be used on an RIO network if it is at Revision C. Do not use a lower revision of the MA-0185-000 tap.

Note: Do not ground a tap unless you are using it specifically as the single-point ground for the entire RIO cable system.

Splitter Specifications

Overview The Modicon MA-0186-100 Splitter is used as a signal combiner in a Hot Standby cable system; each programmable controller has the ability to transmit onto the network using the splitter. The Modicon MA-0331-000 splitter is used as a branching device in certain trunk cable topologies, as defined in Planning and Designing an RIO Cable System, p. 19. The following illustration shows the splitter dimensions.
IN.9 in. AN AEG COMPANY TRUNK SPLITTER OUT OUT
.875 in. minimum 1.75 in. 2.8 in.

.7 in. 1.0 in.

Note: When not in use, splitter ports must be terminated with a Modicon 52-0402000 Port Terminator.
The following table shows the specifications for the MA-0186-100 and MA-0331-000 specifications.
MA-0186-100 Impedance Frequency Range Trunk Insertion Loss Trunk Return Loss Temperature Range Humidity Sealing Interconnections kHz. 5 MHz 6.0 dB 18 dB -40 C. +60 C 95% @ 60C RFI/EMI sealed MA-0331-100 kHz. 5 MHz 3.5 dB 30 dB -40C. +85 C 95% @ 85 C RFI/EMI sealed
F connector, torque 90 in./lb F connector, torque 90 in./lb max max
Note: Splitters not supplied by Modicon will not be supported by Modicon.
Note: The Modicon MA-0186-000 splitter can be used in an RIO network if the splitter is at least Revision B. Do not use a lower revision of the MA-0186-000 splitter.
Note: Existing systems that utilize the MA-0186-X00 as a trunk splitter are not required to upgrade to MA-0331-000 if performance is acceptable. The MA-0331-000 splitter provides higher port isolation.
F Connectors for Coaxial Cables
Overview Flexible cables (RG-6 and RG-11) use F connectors to make the tap port connections; F connectors are also used to make the drop cable connection to certain drop adapters (see Planning RIO Drops, p. 62). F connectors use a 3/8-32 thread. Always use industrial grade F connectors in RIO cable systems commercial grade F connectors should not be used. The Modicon MA-0329-001 F Connector is recommended for quad shield RG-6 cable; it is packaged in a plastic cassette that contains ten connectors. These connectors can be purchased only by the cassette.

F-to-BNC Adapters for RG-11 Cable
There is no approved BNC connector for RG-11 cable. Where a BNC connection is required, use an approved F connector for the RG-11 cable followed by an adapter connection such as the Modicon 52-0614-000 F-to-BNC Adapter. Note: The S901, S908, or S929 head processors used in the 984A, 984B, and 984X Programmable Controllers require the use of a 52-0614-000 F-to-BNC Adapter. The following illustration shows the F-to-BNC Adapter.
.566 in. Diameter over Knurl.437 in. Diameter 3/8 - 32 THD Accepts Standard Female F Connector Female F Terminal Accepts 18-22 AWG Center Conductor

.403 in.

.546 in. Dia.

1.153 in.

The 52-0614-000 Adapter permits the F connector on an RG-11 trunk cable to be attached to the BNC connector on an RIO processor at the network head-end or the F connector on an RG-11 drop cable to be connected to a J810/J812 or J890/J892 drop adapter at the drop. BNC Jack to Male F Connector The 52-0724-000 Jack is supplied with the J890/J892-10x RIO drop adapters to terminate cables with BNC connectors. Consult Sales; this product is available by special order only.

7/16 Hex

.032 Diameter 3/8 - 32 THD

Network Terminators

Overview All terminators used on the RIO network must have a power handling capability of at least 1/4 W. Terminators designed for power-handling, CATV applications, or broadband cable applications cannot be used on an RIO networkthey do not work in the RIO frequency range and will cause signal distortion. All unused drop connectors on taps must be terminated with a standard 75 tap port terminator. The Modicon 52-0402-000 Tap Port Terminator provides suitable termination for this purpose, with a return loss of 22 dB and a frequency range from 100 kHz. 30 MHz.

3/8 - 32 THD 7/16 Hex

Tap Port Terminators

.025 DIA

0.12 0.275 0.57

Trunk Terminators

The trunk cable must be terminated at its tail-end point (in the trunk-out port of the last tap in the trunk cable) with a trunk terminator. The Modicon 52-0422-000 Trunk Terminator is a precision 75 , 1% tolerance, 14 dB terminating resistor specifically designed for trunk termination. Do not use the 52-0402-000 Tap Port Terminator to terminate the trunk cable. The return loss of the 52-0422-000 Trunk Terminator is 40 dB or better at 10 MHz, and its frequency range is from 100 kHz. 30 MHz.

Install the F connector onto the cable port of the RIO drop adapter, tap, or other cable hardware device using a 9/16 in. open-end wrench. Note: Finger tightening is not sufficient.
Semirigid Cable Connections
Overview The following products are recommended for making F connections on a semirigid cable:
l the LRC two-piece cable adapter, catalog number AI540FMQR made by l the LRC three-piece cable adapter, catalog number AI860FMWQR made by
Thomas & Betts, for QR 860 JCA cable Actual cable installation is not trivial because of the cable size and shield material. We recommend that you contact CommScope, the QR series cable manufacturer, for installation tool information, instruction, and assistance. For interconnection, Thomas & Betts/LRC and Gilbert Engineering, among others, carry a full line of QR type cable hardware including F adaptors, terminators, and entry hardware. See RIO Cable Material Suppliers, p. 141 for contact information for these manufacturers. Required Tools Tools are required to strip the cable aluminum sheath and jacket, core the dielectric, and trim the conductor to the appropriate length to accept the F connector. Two standard 1" or 1 1/2" open end wrenches are used to assemble the connector. Thomas & Betts, for the QR 540 JCA cable
Semirigid Cable Installation Tools
Overview The Ripley Company Cablematic JCST-QR Jacket Coring Stripping Tool performs all required operations to prepare the recommended cables for connector attachment. Instructions for use and component replacement part numbers are included with this product. Tool part numbers are:
Cable QR 540 JCA QR 860 JCA Handle (Standard) JCST 540QR JCST 860QR Handle (Ratchet) JCST 540QR-R JCST 860QR-R Coring Bit Kit CB143K CB127K
Replacement components are:
Name Jacket Blade Sheath Cutting Blade Part Number CB6667 CB60
Other Cablematic tools are available for performing the same functions independently. Insure that the tools purchased apply to the various types of QR cables that are offered. Additional Installation Tools The following are additional installation tools that can be obtained from Cablematic:
CC-100 Center Conductor Cleaner CC-200 Center Conductor Scraper CXC Cable Cutter (0.75 in. maximum) CXC-1 Cable Cutter (1 in. maximum)
Preparing a Semirigid Cable for a Connector
Overview Instructions for use are included with the JCST-QR Jacket Coring Stripping tool. Preparation may be done using a power drill if the ratchet handle has been purchased. A drill adapter is included with this part. Take the following steps to prepare the semirigid cable for a conductor.
Step 3 Action Cut the cable, keeping the end as round as possible. Insert the cable into the tool and rotate the tool clockwise with a slight forward pressure. This will remove first the dielectric, then the jacket and sheath. Discard the stripped material and use the cleaner or scraper to remove any remaining dielectric material from the center conductor.

Related Documentation Mounting a Repeater
See the Fiber Optic Repeaters Users Guide, part number GM-FIBR-OPT, more more detailed information on installing fiber optic repeaters. The 490NRP954 Repeaters bottom surface is fitted with pads. Brackets for bolting the unit to a vertical panel are also provided. Your choice of horizontal or vertical mounting should provide access to the device for observing the LED indicators on the front panel and to the rear panel connectors for ease of installation and future servicing. Horizontal Mounting To mount the unit on a horizontal surface, place it at or below eye level to allow viewing the network indicators. Secure it to the surface to prevent it from shifting its position. Do not allow the unit to pull or strain on the network cables and power cable. The mounting brackets supplied with the unit for vertical panel mounting can also be used to secure the unit on a horizontal surface.
Vertical Mounting For vertical mounting, use the brackets supplied with the unit for bolting to a panel. The brackets have tabs that insert into slots on the repeaters bottom panel. No additional hardware is required for securing the brackets. You will have to furnish hardware for bolting the repeater brackets to your panelfour bolts are required. Typically, standard 1/4-20 (10 mm) bolts are satisfactory. The repeaters indicators will usually be readable at or slightly above eye level when the unit is installed in the vertical position. Connecting the Network Cables The fiber optic cables should already be run to the site, with connectors installed. If they are not in place, install them using the manufacturers installation guidelines. Each cable should be labeled to identify the transmit/receive link to which it connects. Connect the RIO coaxial cable and the fiber optic cables to the repeaters rear panel connectors. Secure the coaxial cable to the F-connector.
Fiber Port 1 Tx Rx Fiber Port 2 Rx Tx
RIO Coaxial Cable Connection
If the network links are active, the remote I/O and fiber port LEDs on the front panel of the unit will be in a steady ON state, indicating that receive activity is under way (see Fiber Optic Repeater, p. 90 for details). WARNING Danger to Personnel Do not view the ends of fiber optic cable under magnification while a transmit signal is present on the cablesevere eye damage may result. Use white light only! Failure to follow this precaution can result in death, serious injury, or equipment damage.
RIO Shield-toChassis Jumper Connecting Power
Set the shield-to-ground jumper switch appropriately to specify the repeaters relationship to chassis ground (see Fiber Optic Repeater, p. 90 for details). The repeater operates either from 110/220 VAC line power or from 24 VDC. The AC and DC power connections are located on the back of the panel. Connecting AC Power The repeater is supplied with an AC power cable 6 ft (2 m) long for use with either 110/120 VAC or 220/240 VAC single-phase power. The power cable connects to a socket on the rear panel. Grounding is supplied through the power cable. The AC power cable is keyed for North American 110/120 VAC power outlets. If necessary, install a different plug on the cable for the power source at your site. Turn the power switch OFF and remove the AC power cable from the repeater. Set the power selector plug to the 110/120 VAC or 220/240 VAC position for the power source at your site. To do this, remove the power selector plug by prying under its tab with a small screwdriver. Set the plug to the proper voltage position as shown on the plug body, then reinsert it. Insert the power cable into the rear panel connector. Secure the power cable under the strain relief. Plug the cable into the AC power source. Connecting DC Power Your DC power source must supply 1 A at 24 V. Switch the DC source OFF. Connect the source to DC power terminals, observing the proper polarity. Secure the power wiring under the strain relief.

24 VDC Connection

Power Selector Plug and Fuse
Power Switch Power Cable Connector
CAUTION Possible equipment damage Fiber optic repeaters cannot be operated with both 115 VAC and 24 VDC power applied at the same time. Failure to follow this precaution can result in injury or equipment damage. Grounding The repeater obtains its ground in the AC power cord via the green gnd wire or through the DC wire. Using a continuity tester, verify the repeater chassis is grounded to the site ground. To ensure proper grounding, connect the chassis ground to the site ground by direct chassis to ground connection. Applying AC Power If you are using AC line power, reapply AC to the fiber optic drop site. The main power switch controls the power to the unit. Set the power switch to the (ON) position. The units power OK LED will illuminate. Applying DC Power If you are using DC power, switch on your DC to the repeater. The units power ok LED will illuminate.
Overview The trunk cable must be terminated by inserting a Modicon 52-0422-000 Trunk Terminator in the trunk-out port of the last tap on the RIO network:
Trunk-In Port IN Drop Cable Trunk-Out Port OUT
52-0422-000 Trunk Terminator
Last Tap on the RIO Network
Installing the Ground Point
Overview The cable system should be grounded at a point within 20 ft of the RIO processor at the head-end of the network. A Modicon 60-0545-000 Ground Block, a single Modicon MA-0185-100 Tap, a Modicon MA-0186-100 Splitter, or a Modicon MA0331-000 Splitter may be used, assuring that the cable system will be permanently grounded even when disconnected from the RIO processor. Note: Do not disconnect the cable system from the central ground point disconnecting the system from ground will create an unfavorable floating ground condition. A screw is provided on taps, splitters, and ground blocks as the grounding point. If you use a ground block, mount it in a small enclosure. To install a 60-0545-000 Ground Block:
Cut the cable Install two F connectors on the cable Attach two F connectors to the ground block Wire the ground block to an appropriate ground (typically building steel)
Testing and Maintaining an RIO Network
Overview Whats in this Chapter? This chapter provides information on testing and maintaining an RIO network. This chapter contains the following topics:
Topic Maintenance and Testing Requirements Coax RIO System Network Integrity Problem Sources on an RIO Network On-line and Off-line Error Isolation Troubleshooting Fiber Optic Repeaters Page 137
Maintenance and Testing Requirements

Rx FR2 (Head)

Coaxial Cable RIO Drop #2 P/S RIO I/O I/O I/O
There are well documented procedures for analyzing the wire side characteristics of this type application, and it is recommended that they be used as a first line of attack and afterward whenever trouble is suspected. If the coaxial system is working properly, it will cause the remote I/O LED on FR1 to illuminate. If that LED illuminates as expected, then the fiber port 1 LED on FR2 should illuminate and the fiber port 2 LEDs on FR1 and FR2 should be OFF. If the fiber port 1 LED on FR2 does not illuminate, check the Tx and Rx connections on the fiber link. If the problem persists, substitute a known good repeater for FR2 and repeat the procedure. If the problem still persists, check the drop adapter and coaxial link at drop #2. If all this still checks out properly, then you have isolated the problem to faulty fiber cable, and manufacturers test procedures must be used.
Broken Cable Detection and Remedies
Unlike coaxial cable, fiber cable contains physically separate transmit and receive lines. It is possible to lose communications through the Rx line while the Tx line remains intact. A break in the Rx line will deprive the PLC of input data. Under ordinary circumstances, the PLC continues to drive outputs via the intact transmit line. This could lead to outputs turning ON or OFF due to invalid (INPUT STATE: 0) input data. A method to prevent this from happening uses STAT and SENS instructions in ladder logic to detect the loss of input communication and inhibit improper output state changes:
40101 STAT #0187 #00097 SENS #0001
STAT and SENS monitor the I/O status of Drop #2 and inhibit output 00001 if communications are lost. STAT provides access to the systems status, including the status of S908 communications. The status information is stored in a table starting at register 40101 and has a length of 187 words (as shown in the top and bottom nodes of the STAT instruction). SENS senses the first (communications health) bit (SENS top node value = 1)of the 185th word in the status table (SENS middle node value = 40285). This bit is the communications health for Drop #2 of the S908. Coil 00001 has been configured as an output in the I/O Map. If the PLCs Rx line is broken, the sensed bit becomes 0 (OFF). The middle node output to coil 00097 is set to 0 (OFF). Coil 00097 controls a normally open relay which, when power is removed, opens the circuits to coil 00001, thus inhibiting this output.The coils can now be used in ladder logic to inhibit specific output writes. As an alternative, the coil can be used to control a SKP instruction to prevent execution of that portion of the network which would ordinarily output data.

Overview These appendices provide information on RIO cable material suppliers as well as a glossary of terms related to remote I/O cable systems. The appendix contains the following chapters:
Chapter A Chapter Name RIO Cable Material Suppliers Page 141

Whats in this Appendix?

RIO Cable Material Suppliers
RIO Cable Material Suppliers Belden Wire and Cable Company 2200 U.S. Hwy. 27 South P.O. Box 1980 Richmond, IN 47374 Telephone: (765) 983-5200 or (800) 235-3361 Fax: (765) 983-5294 Website: www.belden.com CommScope, Inc. Digital Broadband Division P.O. Box Lenoir-Rhyne Blvd. Hickory, NC 28603 Telephone: (800) 982-1708 (828) 324-2200 Fax: (828) 328-3400 Website: www.commscope.com Gilbert Engineering (now known as Corning Gilbert, Inc.) 5310 West Camelback Road Glendale, AZ 85301 Telephone: (623) 245-1050 or (800) 528-5567 Website: www.corning.com/CorningGilbert Relcom, Inc. 2221 Yew Street Forest Grove, OR 97116 Telephone: (800) 382-3765 Website: www.relcominc.com
Ripley Company Cablematic Tool Division 46 Nooks Hill Road Cromwell, CT 06416 Telephone: (860) 635-2200 Website: www.ripley-tools.com Rostra Tool Company 30 East Industrial Road Branford, CT 06405 Telephone: (203) 488-8665 Fax: (203) 488-6497 Website: www.rostratool.com Square D Services - Automation 1960 Research Drive Troy, MI 48083 Telephone: (888)-SQUARED Website: www.squared.com Thomas & Betts World Headquarters 8155 T & B Boulevard Memphis, TN 38125 Telephone: (901) 252-5000 Website: www.thomasandbetts.com 3M Telecom Systems Division 6801 River Place Blvd. Austin, TX 78726-9000 Telephone: (800) 426-8688 Website: www.3m.com/market/telecom

Glossary

amplitude application armor attenuation A measure of the strength of a signal. A user program A metal wrapping around a coaxial cable used for mechanical protection. Signal loss through an electrical circuit or conductor ( see also signal loss).
bandwidth baseband A range of frequencies. A type of network having a single communications channel. RIO is a baseband communications network. The radius of the arc along which a cable may be bent. The number of bits received in an error divided by the total number of bits received. A wire mesh used to construct the shield of a coaxial cable A single cable connecting multiple ports.

tap insertion loss TDR (time domain reflectometer) terminator

through loss

topology
transfer impedance trunk cable

trunk terminator

velocity of propagation VSWR (voltage standing wave ratio) The speed of the signal in the cable, expressed as a percentage of the speed of light in free space. The measure of the signal reflected back from a transmitted signal. Lower ratios indicate better impedance match and cause less signal to be reflected back at the transmitting source.
wavelength The distance between the same point on adjacent waves.
zero crossing The condition when the wave form crosses 0 V, either on a voltage rising or on a voltage falling. See also phase continuous signaling.
52-0399-000 Self-terminating F Adapter illustration, 85 52-0402-000 Tap Port Terminator disconnecting from port, 120 terminating splitter ports, 76 terminating unused drop connectors, 83 terminating unused ports, 74 52-0411-000 Self-terminating F Adapter illustration, 85 52-0422-000 Trunk Terminator terminating a trunk cable, 52 terminating trunk cable, 83, 126 52-0480-000 Right Angle F Adapter for semirigid cable, 49, 80 52-0487-000 BNC Connector for non-quad shield, 81 52-0614-000 F-to-BNC Adapter illustration, 82 52-0724-000 Jack to Male F Connector illustration, 82 60-0513-000 BNC In-line Terminator running drop cables to, 118 terminating end of drop cable, 84 60-0545-000 Ground Block description, 87 illustration, 87 installing, 127 60-0558-000 Cable Cutters for RG-6 connectors, 97 600-558-000 Cable Cutters illustration, 99

Numerics

043509432 Crimp Tool for RG-6 connectors, 97 illustration, BNC Connector for quad shield, 81 490NRP954 Fiber Optic Repeater alternative communication link, 72 communication between two or more RIO nodes, 90 horizontal mounting, 122 LEDs illustration, 91 use in RIO cable topology, 35 vertical mounting, 123 490RIO00211 F Connector connecting RG-11 cable, 107 for RG-11 cable, 49 490RIO00400 Installation Tool for RG-6 cable, 98 for RG-6 connectors, 97 490RIO00406 Installation tool replacement blade packs, 98 490RIO0211 F connector for RG-11 cable, 78 490RIO0C411 Installation tool installing connectors on RG-11 cable, 112 490RIO0S411 stripping RG-11 cable, 108 52-0370-000 Self-terminating BNC Adapter optional use, 119 use in Hot Standby systems, 84
AS-MBII-003 Pre-assembled Drop Cable 50 ft length, 68 AS-MBII-004 Pre-assembled Drop Cable 140 ft length, 68 attenuation bandwidth, 71 cable type, 53 calculation equation, 54 calculation example, 55 description, 53 maximum in RIO networks, 46 minimum distance between repeaters, 57 on fiber optic link, 57 on point-to-point optical link, 58 parameters, 71 tap, 53 typical coaxial cable losses, 46