IS200ISBBG1AB I/O Pack | Dual Bus Segment Controller

Description

Product Introduction

The bus bridge is a niche module—you only need it when you’ve got two separate electrical zones and you need to keep them isolated while sharing data. The IS200ISBBG1AB is the standard-temperature version of that bridge, but with a twist: GE gave it the same power supply and isolation improvements as the extended-temp “AAB” without the cold-rated components. If your control room is climate-controlled but you’ve got VFDs and contactors making noise on the backplane, this is the bridge that doesn’t glitch when the 5 V rail gets dirty.

Two independent bus segments, each with a redundant A/B port pair running at 100 Mbps. The propagation delay is 2 ms—transparent to the controller’s scan. The isolation between segments is 1,500 V, and the “AB” revision bumps the common-mode transient immunity to 25 kV/µs, up from 15 kV/µs on the earlier “A” version. It doesn’t have the conformal coating or the cold-rated oscillator of the “AAB,” but if you’re in a clean, temperature-controlled environment, you don’t need them. This is the bridge for the indoor racks that still have to deal with electrical noise from heavy equipment.

Key Technical Specifications

Quality Inspection Process (SOP Transparency)

The “AB” gets the same EMI stress test as the “AAB,” but without the thermal chamber extremes—we run it at 25 °C and 60 °C ambient instead. Our 30-point inspection verifies the improved filtering and isolation.

Incoming Verification. OEM packing slip matched to GE’s serial database. We log the serial and photograph the anti-static bag before cutting. The holographic GE label gets a UV check. The PCB edge must read “–ISBBG1AB” clearly.

Visual Inspection. Magnifying lamp, full board scan. The pi-filter components near the power input are visible—we verify they’re present and correctly soldered. The four RJ45 connectors show zero wear. The isolation barrier (optocouplers/transformers) is intact.

Live Functional Test. Mark VIe test rack with two remote I/O simulators and an EMI injector on the 5 V rail.

• Standard test: Segment A and B throughput (>95 Mbps each), bridging functional, redundancy switchover (<10 ms), propagation delay (<2.5 ms).
• Isolation test: Fault on one segment—other stays active.
• EMI stress test: Inject 100 kHz, 100 mV peak-to-peak noise on the 5 V backplane rail—the “AB” must maintain full throughput with zero errors over 8 hours.
• 24-hour soak at 50 °C: Both segments active, continuous data exchange—log any errors.

Electrical Parameters. Insulation resistance: 500 VDC via Megger MIT420, >10 MΩ between segments and backplane. Isolation between segments: >10 MΩ at 500 V. Ground continuity: <0.1 Ω.

Firmware Verification. Read the FPGA firmware via ToolboxST—verify checksum.

Final QC & Packaging. The QC report includes throughput, propagation delay, EMI test results, isolation resistance, and a photo. Into an anti-static bag with desiccant, 2″ foam, double-wall carton. “QC Passed” label with date.

Field Replacement Pitfalls

The “AB” handles EMI better than the “A” version, but it’s still a bridge—installation mistakes happen. I’ve seen these across the fleet.

Isolation—Don’t Defeat It with Shared Switches. The “AB” has better transient immunity, but if you share a switch between segments, you bypass the isolation entirely. One site in Texas connected both segments to the same managed switch—noise from a VFD on Segment A coupled into Segment B, corrupting data on both buses. Keep the cabling physically separate. Segment A gets its own switches, Segment B gets its own.

EMI—The “AB” Handles It, But Not Without Limits. The pi-filter and faster optocouplers can handle 100 mV of 100 kHz ripple. If your backplane has 500 mV of noise (from a failing power supply), the bridge will eventually reset. I saw this at a plant in Ohio—the 5 V supply was original from 2008. The “AB” worked for three months before it started showing CRC errors. The fix: replace the rack’s power supply. Don’t rely on the bridge’s filtering to fix a bad backplane.

Propagation Delay—It Adds Up. Two ms per bridge is fine. But if you chain two or three bridges, the delay adds up. One site in Pennsylvania had three “AB” bridges in a daisy chain—6 ms total delay. Their fast flow control loop started oscillating. The fix: use a star topology instead of daisy-chaining, or adjust the loop timing. The “AB” doesn’t change the propagation delay math.

Power Budget. The “AB” draws 12 W. If you’re using it in a rack with multiple comms modules, the power adds up. I’ve seen a rack with two “AB” bridges (24 W), two ISBAs (20 W), and a CPU (25 W)—total 69 W, fine. But they added three analog modules and a discrete pack, pushing it to 145 W—close to the 150 W limit. At startup, the 5 V rail sagged and the bridges reset. Leave 20% headroom.

Firmware Mismatch. The “AB” uses the same firmware as the standard ISBBG1A. Requires CPU v5.0 or later. If you install it into an older system running v4.2, the bridging logic won’t work—the CPU won’t see the remote racks on the second segment. Check your CPU version before installation.

ESD. The PHY chips and isolation barrier are CMOS. I watched a tech handle a bare “AB” on a dry day in Arizona—he discharged through an RJ45 connector, and the A port on Segment B stopped working. Strap up.

New Original vs. Refurbished: Why It Matters

The “AB” has the pi-filter and faster optocouplers—refurbishers often can’t replicate these improvements.

What “New Original (New Surplus)” means. This IS200ISBBG1AB came from GE’s factory with the pi-filter, the faster optocouplers, and factory-fresh PHY chips. We break the seal only for testing.

Refurbished risk in plain terms. The pi-filter is a board-level addition—it requires extra components. A refurbisher may buy a standard “A” version, clean it, and sell it as an “AB.” But they won’t add the pi-filter or upgrade the optocouplers. So you get a module that fails the EMI test. I’ve tested refurbished “AB” units that had no pi-filter—they showed CRC errors when we injected 100 mV of ripple. Failure rate on refurbished bridges runs 4× higher than new, based on our service data.

Real cost of a refurbished failure. Let’s say a refurbished “AB” (actually a standard “A” bridge) is installed in a switchgear room with noisy VFDs. The 5 V rail has 150 mV of ripple—the bridge starts resetting. The CPU loses the remote racks on both segments—the turbine trips. Lost generation: 30,000. The refurbished module saved you 1,500. The outage cost you 20× that.

What we provide as proof. For every IS200ISBBG1AB we ship: a photo of the OEM packing slip, serial traceability to GE’s records, a full test report that includes EMI test results, throughput, propagation delay, isolation resistance, and a sealed anti-static bag.

Pricing context. Our price sits 30–50% above refurbished, 20–30% below GE’s current list price. The delta covers our sourcing, our EMI testing, and a 12-month warranty.

Performance Benchmarks & Test Results

Data from our Mark VIe test rack (ambient 45 °C, supply +5.0 VDC), EMI injector on the 5 V rail. Two remote I/O simulators. Firmware v5.3.

• Port throughput—Segment A: 97.2 Mbps, zero CRC errors over 24 hours.
• Port throughput—Segment B: 97.1 Mbps, zero CRC errors.
• Propagation delay: 2.1 ms (A to B), 2.0 ms (B to A).
• Redundancy switchover: 8.2 ms—under 10 ms.
• EMI stress test (100 mV, 100 kHz ripple): Throughput held at 97.0 Mbps, zero CRC errors over 8 hours. The “A” revision we tested alongside showed 15 CRC errors under the same stress.
• Common mode transient immunity: Tested at 25 kV/µs—the bridge held isolation with no data errors.
• Isolation resistance: >100 MΩ at 500 V.
• Thermal performance: At 60 °C ambient with both segments active, the FPGA ran at 66 °C—under the 85 °C rating.
• Reliability estimate: MIL-HDBK-217F gives a demonstrated MTBF of 51,000 hours at 40 °C—that’s 5.8 years. Refurbished units with missing pi-filter show a demonstrated MTBF around 9,000 hours in noisy environments—the filter stress kills them prematurely.

140DD036400 SCHNEISER
140DD035310 SCHNEISER
140D035301 SCHNEISER
14DD015310 SCHNEISER DCSPLC