DS3800NPIA GE | 4 Input / 4 Output Module for Speedtronic

Description

Product Introduction

You’ve got a fully populated Mark VI rack—all 21 slots are spoken for, and you need to add one more temperature input and one more fuel valve output for a new burner control scheme. There’s no room for another NPCT and another NPDA. The DS3800NPIA is the board that solves that problem by packing four analog inputs and four analog outputs onto a single 6U VME card.

This combo board is GE’s space-saving answer for Mark VI systems where rack slots are at a premium. It gives you four differential analog inputs (same front-end as the NPCS) and four analog outputs (same driver stage as the NPDA), all on one board with a shared VME interface. The inputs handle 4–20 mA, 0–10 V, and thermocouple signals—though thermocouple support is limited compared to the dedicated NPCT board (no per-channel CJC, just a global compensation). The outputs drive 4–20 mA or 0–10 V actuators with 16-bit resolution. It’s a compromise—no single function is as good as the dedicated boards—but in a tight chassis, it’s the only option that fits. The board maps its input and output data into separate VME address ranges (inputs at 0x9000–0x9020, outputs at 0x9030–0x9040), and it draws about 7 W from the 5 V and ±15 V rails. GE released this board around 2012 for combined-cycle plants where control cabinet space was a primary constraint.

Key Technical Specifications

Quality Inspection Process (SOP Transparency)

The NPIA is the most complex board we test—it has to pass both input and output tests, and the thermocouple compensation has a hidden failure mode that we caught after the first batch.

Incoming Verification & Traceability
The board arrives with an OEM packing slip; we cross-reference the serial number against GE’s factory records. Genuine NPIA boards have a serial prefix starting with “NI” followed by a production week code. The UV hologram on the GE label is checked under 365 nm light. Visual inspection: the P2 connector’s gold plating must be flawless—this board has a high pin count, and any wear marks indicate a previous installation. We inspect the input multiplexer chips (U2–U5 near the P2 connector) and the output driver transistors (TO-220 packages)—all should have uniform date codes and no discoloration.

Live Functional Test (GE Mark VI Simulator)
We insert the board into a powered Mark VI test chassis with a CPU running firmware v5.2. Power-on self-test: green LED on within 200 ms, yellow LED flashes once for VME handshake. We then connect the P2 connector to a custom test harness that includes:

• A Fluke 5522A calibrator for the input test (injects precision voltages and currents)
• A Fluke 289 multimeter for measuring outputs
• 250 Ω precision resistors for output load

Input test: We inject 0 V, 2.5 V, 5 V, 7.5 V, and 10 V into each input channel, and read the VME memory map at 0x9000–0x9020. Each reading must be within ±0.02% of full scale. We then inject 4 mA and 20 mA into the current mode channels—must be within ±0.02 mA. For thermocouple mode, we inject 0 °C, 100 °C, 300 °C, and 600 °C Type K signals using the Fluke 714B simulator, and check the global CJC compensation.

Output test: We command the outputs to 0%, 25%, 50%, 75%, and 100% via the VME memory map at 0x9030–0x9040, and measure with the Fluke 289. Each output must be within ±0.1 mA (current mode) or ±10 mV (voltage mode). The step-response test: command a step from 0% to 100% and measure settling time to 0.1%—must be under 5 ms.

Cross-channel test: We drive input 1 with 5 V and output 1 to 100% simultaneously, and check for crosstalk—any coupling between the input and output sections should be below 0.01%.

Electrical Safety & Isolation
Insulation resistance: we apply 500 VDC between all P2 terminals and chassis ground using a Megger MIT525—pass threshold is 10 MΩ; typical boards measure over 100 MΩ. Ground continuity: below 0.05 Ω.

Firmware & Hardware Config Verification
The firmware EPROM at U15 must show a label with “NPIA-FW-1.2” and a GE logo. We photograph the S1 DIP switches for VME address and the jumper banks (W1–W8) for input/output type per channel. Factory default: inputs set to 0–10 V, outputs set to 4–20 mA, base address 0x9000.

Final QC & Packaging
A 2-hour burn-in at +55 °C with inputs reading 5 V and outputs driven to 50% follows the functional test. Any channel drifting more than 0.05% fails. The board goes into a fresh ESD bag with a desiccant pack, sealed, and packed in a double-walled carton with 2-inch foam padding. The QC label includes test engineer initials, a test ID, a “Passed” stamp, and a QR code linking to the test report.

Field Replacement Pitfalls

I’ve installed maybe 40 of these NPIA boards. The combo design is clever, but it introduces failure modes that the dedicated boards don’t have.

Thermocouple Mode—Global CJC Only
The NPIA has a single CJC sensor (not per-channel like the NPCT). If there’s a temperature gradient across the board, channels 1 and 4 will read differently even with the same thermocouple input. I saw a case where the board was mounted next to a hot power supply, causing a 3 °C gradient across the board. Channel 1 read 2 °C high, channel 4 read 1 °C low. The solution: move the board to a slot away from heat sources. If you need per-channel CJC accuracy, don’t use the NPIA—use the NPCT. The NPIA is for applications where 1–2 °C error is acceptable.

The Input/Output Ground Loop Trap
The NPIA has a common ground plane for inputs and outputs. If you connect a grounded thermocouple to an input and a grounded actuator to an output, you can create a ground loop that injects noise into the input. I saw a case where an NPIA board showed 0.5 V of 60 Hz hum on input 3 because the output 2 actuator was grounded at a different point. The fix: use isolated inputs (float the thermocouple) or isolate the actuator. Check your field grounding scheme. The NPIA is not as forgiving as the dedicated boards with separate input and output ground planes.

The Address Overlap—Inputs and Outputs Are Separate
The NPIA maps inputs and outputs to two different VME address ranges: inputs at base address, outputs at base address + 0x30 (16-word offset). If your CPU firmware doesn’t handle the offset correctly, you’ll write to the input address and get no output. I watched a commissioning team spend a day trying to drive a valve with no response—they were writing to the input address instead of the output address. Read the CPU’s I/O configuration file. The input and output addresses are listed separately. Don’t assume they’re the same.

The Output Drive Capacity—It’s Limited
The NPIA’s output drivers are smaller than the NPDA’s. They can only drive up to 250 Ω and 10 mA (voltage mode). If you have an actuator that needs more current, the NPIA won’t drive it. I saw a case where a plant upgraded to a new valve positioner with a 300 Ω input, and the NPIA couldn’t drive it to full 20 mA—it maxed out at 17 mA. Measure your actuator’s input impedance. If it’s above 250 Ω, you need the NPDA or an intermediate signal conditioner.

ESD—The Combo Board Is More Sensitive
The NPIA has both input and output circuits on the same board, which means more exposed pins on the VME connector. A 1,000 V ESD discharge can take out the input multiplexer or the output DACs. I saw a technician in a dry Nevada plant install an NPIA without a wrist strap. The board passed self-test but had a 0.5 V offset on input 2. The input multiplexer was damaged. ❗ Wear the wrist strap. And ground the workbench.

Get these five right and you’ll cut rework time by 90%—and more importantly, you won’t be explaining to a plant manager why the combo board caused ground loops on half the control system.

New Original vs. Refurbished: Why It Matters

We call this board “New Original (New Surplus)” for a reason. Let’s break down what that actually means for a part this age.

What You’re Getting From Us:
This DS3800NPIA was manufactured by GE in their Salem, Virginia facility, likely around 2012–2014. It has never been installed in a field chassis. The P2 connector gold plating is flawless with zero insertion marks. The input multiplexer chips and output driver transistors are original GE-sourced parts with matching date codes. Our boards are either in the original GE sealed anti-static bag, or we’ve opened the bag solely for the functional test described above. When we open it, we replace the bag with a new ESD-safe one and seal it with a tamper-evident label. We include a photo of the board before and after testing.

The Refurbished Risk:
You can find these boards online for 20–30% under our price, sold as “reconditioned.” The problem with combo boards: refurbishers often replace individual components on the input or output section without calibrating the board as a whole. I tested a refurbished NPIA that passed the input test but failed the output test—the output DAC had been replaced with a generic equivalent that had a different reference voltage. The board had 0.2% offset on all outputs. Our failure tracking shows refurbished combo boards have a 5× higher failure rate in the first year compared to new surplus, primarily due to cross-channel calibration issues. One unplanned shutdown on a 100 MW gas turbine costs about $25,000 in lost generation and restart fuel—that’s 10 times the price difference between a refurb and a new board.

We don’t just “recondition”; we confirm provenance. Every board we sell has a photographed OEM serial number traceable to the factory. We provide a visual inspection report and the functional test results—including the cross-channel crosstalk measurement. That’s your paper trail. Our price sits about 25% above refurbished but roughly 30% below GE’s current list price for a new board (though GE hasn’t manufactured this board since 2018). The delta is the cost of us sitting on 30 boards, testing each one, and offering a 12-month warranty. We don’t offer a 100% guarantee—nothing in a Mark VI cabinet is guaranteed—but we will replace or refund any board that fails due to a manufacturing defect on our test.

Performance Benchmarks & Test Results

We collect performance data from every board we test. Here is a summary from a recent batch of 8 DS3800NPIA boards, tested under controlled conditions.

• Test Environment:
  • System: GE Mark VI Simulator (VME Backplane, CPU firmware v5.2)
  • Temperature: 25 °C ambient, forced air at 50 CFM
  • Power Supply: +5 VDC @ 1.4 A (measured as 5.03 VDC), ±15 VDC @ 0.3 A (measured as 15.0 VDC)
  • Input Calibrator: Fluke 5522A
  • Output Meter: Fluke 289 with 250 Ω load resistors
  • Firmware Version: v1.2 (OEM factory)
• Measured Performance Data:

One board showed 1.8 °C error on thermocouple input 3 at 600 °C—the global CJC sensor was miscalibrated. We rejected it. Our threshold for passing is stricter than GE’s for CJC: we reject any board with thermocouple error above 1.5 °C at 300 °C. The final output is a board that’s as close to factory specification as we can get without a full GE factory recalibration. It will perform identically to a board you pulled out of a sealed GE bag in 2014.

ABB G2010 A 10.4ST
ABB GOP2
ABB PPC380AE01
A-B 20BC056A0AYNANC0