Surplus Military Atlas Copco QAS38 Generator Issues

[hawk-aggie]

I can not but think that the AVR is the key here to getting the voltage right on all three phases. Is the new AVR the same as the old AVR?

Yes

 

[jmenende]

So I just acquired this same unit and it came with no filters. The manuals have the Atlas part numbers but im looking to cross reference. them. Oil filter: 2914 9200 00
Fuel filter: 2914 8092 00
Air filter: 2914 9302 00
Vbelt: 2914 9218 00

This is my second atlas branded unit. Got a qas 138 a few years back and really disliked the proprietary part number bs. Its quite challenging and i told myself id never buy another but here i am.
Thanks

 

[FarmingSmallKubota]

If you use the NAPA website for filter cross reference and put in the atlas part number without the spaces the NAPA equivalent will come up.

Filter Lookup

Web site created using create-react-app

napapartshub.com

 

[jmenende]

If you use the NAPA website for filter cross reference and put in the atlas part number without the spaces the NAPA equivalent will come up.

Filter Lookup

Web site created using create-react-app

napapartshub.com

Thanks. I went to my filter guy and he crossed all in donaldson. Will install to makensure they are correct and post results. Thanks

 

[hawk-aggie]

So I just acquired this same unit and it came with no filters. The manuals have the Atlas part numbers but im looking to cross reference. them. Oil filter: 2914 9200 00
Fuel filter: 2914 8092 00
Air filter: 2914 9302 00
Vbelt: 2914 9218 00

This is my second atlas branded unit. Got a qas 138 a few years back and really disliked the proprietary part number bs. Its quite challenging and i told myself id never buy another but here i am.
Thanks

I’m out of town until the 25th. I’ll send what I used when I get back.

 

[hawk-aggie]

Fuel Filter/Water Separator - Parker Racor R45P
Fuel Filter - Yanmar 123907 - 55810
Oil Filter - Yanmar 129150 - 35170
Atlas Copco Air Filter - 2914 9302 00 (Element) + 2914-9303-00 (Safety Cartridge). These were in good shape, so I didn't change them. All the other filters were not in place, so I had to do a LOT of searching/research.

 

[jmenende]

Thanks. I went to my filter guy and he crossed all in donaldson. Will install to makensure they are correct and post results. Thanks

Confirmed:
Oil filter: 2914 9200 00 Donaldson: P550162
Fuel filter: Wix 51334
Air filter: 2914 9302 00 DONALDSON P778984
Vbelt: 2914 9218 00
Avr: (UVR6-F)

 

[hawk-aggie]

I have another issue with our Atlas Copco QAS38. Originally, it came wired for 208 V. It has since been rewired (retapped) for 480V 3-phase. Each leg indicates 279-280V @ 60 Hz. It began service in May 2025 and ran reliably until it didn't in November. It would run a minute or less then would shut down. One indicator light would come on at shut down. That light indicated that the voltage being sensed was less than 75V. I replaced the AVR because of a discolored capacitor on the board. We tested it a handful of times over the past 7-10 days, and it ran great. Now, it's shutting down again. We're stumped and the Atlas Copco guys I've been working with don't have any experience on older units like this one. Thoughts?

 

[jmenende]

I have another issue with our Atlas Copco QAS38. Originally, it came wired for 208 V. It has since been rewired (retapped) for 480V 3-phase. Each leg indicates 279-280V @ 60 Hz. It began service in May 2025 and ran reliably until it didn't in November. It would run a minute or less then would shut down. One indicator light would come on at shut down. That light indicated that the voltage being sensed was less than 75V. I replaced the AVR because of a discolored capacitor on the board. We tested it a handful of times over the past 7-10 days, and it ran great. Now, it's shutting down again. We're stumped and the Atlas Copco guys I've been working with don't have any experience on older units like this one. Thoughts?

Is it showing the same undervoltage light or something else?

 

[hawk-aggie]

I can not but think that the AVR is the key here to getting the voltage right on all three phases. Is the new AVR the same as the old AVR?

I've sent you a private message, but wanted to respond here also to see what others might have to add.

Yes, the AVRs are exactly same.

The generator is a modified version built for the U.S. military. Our generator was originally wired/tapped for 208V, but we rewired/reconfigured it for 480V 3-phase. No issue there. I had some issues getting it running, but was finally successful. (Not really sure how I overcome the issues mentioned earlier in this thread other than changing out the AVR.) We've been using it solely to power a laser de-scaler that requires the 480V. It ran without issue from May 2025 to October 2025. In October, it started running for about 15-20 seconds, then shutting down. I installed another new AVR, but having the same results. When it shuts down, a fault indicator lights up on the control panel. The indicator lights up when no AC input is present, specifically when there is <75V present. Since the generator is putting out 480V during that 15-20 seconds it’s running, it makes no sense for it to shutdown unless it’s not VAC from the generator it wants, but rather VAC from the MPU. It’s looking for 75VAC or greater. This unit was missing the MPU when we got, and I’ve never found exactly how much VAC it’s supposed to be set at. When I first installed the MPU, I did so following your previous instruction to me about setting the MPU on an MEP generator. I got some pretty high voltage readings (maybe 40-50 volts or so) at the time, so I just adjusted the MPU to read in the 2-3V range. Now, I’m thinking those high readings weren’t so high afterall. Have I given you enough info to speculate? Let me know if you have questions. Thanks

 

[hawk-aggie]

Is it showing the same undervoltage light or something else?

When it starts, the red under voltage light comes on for just a second until full engine speed is reached, then the green (normal voltage)light comes on. The green light stays on until the unit shuts down. The red light comes on while the engine stops. Seems accurate to me. The indicator light the comes on when the unit shuts down is L4 on the control panel. That signifies that there <75V to the electronic speed control(?) or somewhere else in the circuitry.

 

[hawk-aggie]

I've sent you a private message, but wanted to respond here also to see what others might have to add.

Yes, the AVRs are exactly same.

The generator is a modified version built for the U.S. military. Our generator was originally wired/tapped for 208V, but we rewired/reconfigured it for 480V 3-phase. No issue there. I had some issues getting it running, but was finally successful. (Not really sure how I overcome the issues mentioned earlier in this thread other than changing out the AVR.) We've been using it solely to power a laser de-scaler that requires the 480V. It ran without issue from May 2025 to October 2025. In October, it started running for about 15-20 seconds, then shutting down. I installed another new AVR, but having the same results. When it shuts down, a fault indicator lights up on the control panel. The indicator lights up when no AC input is present, specifically when there is <75V present. Since the generator is putting out 480V during that 15-20 seconds it’s running, it makes no sense for it to shutdown unless it’s not VAC from the generator it wants, but rather VAC from the MPU. It’s looking for 75VAC or greater. This unit was missing the MPU when we got, and I’ve never found exactly how much VAC it’s supposed to be set at. When I first installed the MPU, I did so following your previous instruction to me about setting the MPU on an MEP generator. I got some pretty high voltage readings (maybe 40-50 volts or so) at the time, so I just adjusted the MPU to read in the 2-3V range. Now, I’m thinking those high readings weren’t so high afterall. Have I given you enough info to speculate? Let me know if you have questions. Thanks

UPDATE
I finally stumbled across a document for the electronic speed switch that indicates the MPU setting as at least 0.7 VAC. I removed the MPU, cleaned the bit of debris off of it, and reinstalled it with an output of 3 VAC. Voltage readings on the speed switch follow.
Battery voltage between terminals 1 and 2 = 27.2 VDC. No way to take this reading without the engine running, so voltage is engine alternator voltage.
Voltage between terminals 3 and 4 = 3.2 VAC. This is MPU voltage.
Voltage between negative side of battery (Terminal 2) and terminal 8 with power switch "ON" should be battery voltage. Reading was -1.5 VDC.
Voltage between terminals 2 and 6 should be battery voltage when cranking and zero when running. My readings were zero when cranking and -1.5 VDC respectively. (Those readings indicate the SW1 set point should be adjusted. From what I've read, that requires a signal generator and voltage supply with adjustments done on the bench.)
Readings seem suspect. Leaning towards replacing the speed switch.

I then checked the voltage readings on the AVR.
Terminal 2 to the + terminal should be 170-270 VAC. My reading was 215 VAC.
Sensing inputs between terms 1 and 2 = 139.9 VAC. No range given.
Sensing inputs between terms 3 and 4 = 139.7 VAC. No range given.
Sensing inputs between terms 5 and 6 = 140.5 VAC. No range given.
Readings seem satisfactory.

The only other component that could be suspect is the start switch controller. I'm wondering if the switch controller has some sort of timer in it, but I haven't found a circuit diagram yet. The reason I'm interested in the switch controller is that the engine is shutting down at about the same interval every time (15 seconds). The only other components include AC circuit breakers, AC fuses, and AC amp sensing blocks - none of those components have timers.

I'm still stumped, but making progress. Hesitating on buying 1 or 2 big ticket items until I'm more reasonably sure those items will permanently fix the issues.

 

[hawk-aggie]

UPDATE
I finally stumbled across a document for the electronic speed switch that indicates the MPU setting as at least 0.7 VAC. I removed the MPU, cleaned the bit of debris off of it, and reinstalled it with an output of 3 VAC. Voltage readings on the speed switch follow.
Battery voltage between terminals 1 and 2 = 27.2 VDC. No way to take this reading without the engine running, so voltage is engine alternator voltage.
Voltage between terminals 3 and 4 = 3.2 VAC. This is MPU voltage.
Voltage between negative side of battery (Terminal 2) and terminal 8 with power switch "ON" should be battery voltage. Reading was -1.5 VDC.
Voltage between terminals 2 and 6 should be battery voltage when cranking and zero when running. My readings were zero when cranking and -1.5 VDC respectively. (Those readings indicate the SW1 set point should be adjusted. From what I've read, that requires a signal generator and voltage supply with adjustments done on the bench.)
Readings seem suspect. Leaning towards replacing the speed switch.

I then checked the voltage readings on the AVR.
Terminal 2 to the + terminal should be 170-270 VAC. My reading was 215 VAC.
Sensing inputs between terms 1 and 2 = 139.9 VAC. No range given.
Sensing inputs between terms 3 and 4 = 139.7 VAC. No range given.
Sensing inputs between terms 5 and 6 = 140.5 VAC. No range given.
Readings seem satisfactory.

The only other component that could be suspect is the start switch controller. I'm wondering if the switch controller has some sort of timer in it, but I haven't found a circuit diagram yet. The reason I'm interested in the switch controller is that the engine is shutting down at about the same interval every time (15 seconds). The only other components include AC circuit breakers, AC fuses, and AC amp sensing blocks - none of those components have timers.

I'm still stumped, but making progress. Hesitating on buying 1 or 2 big ticket items until I'm more reasonably sure those items will permanently fix the issues.

UPDATE 2
I believe the start switch module is a Programmable Logic Controller (PLC). It “monitors” water temp, oil pressure and what appears to be voltage from the engine alternator. In turn, it controls power to the starter motor and fuel, as well as other stuff I can’t determine. It likely communicates with the Electronic Speed Module/Switch. I don’t know which may be bad, if either one, because I don’t have the equipment to bench test them. Based on the tests I could perform, the Electronic Speed Switch provided voltage readings that are inconsistent with what is in the published document. These two modules are the only devices that are "smart" parts, unlike voltmeters, ammeters, circuit breakers and fuses.

 

[LuckeyD]

I had to play on several of these. Big sand box. OK, after you got it going on 24V, you had NO AC. Guyfang has it right. Please check the fuse on the AVR. Then ensure the jumper is in the 60 HZ position. Other connections as they are supposed to be and give it a go. The regulators are weak in this application.

 

[hawk-aggie]

I had to play on several of these. Big sand box. OK, after you got it going on 24V, you had NO AC. Guyfang has it right. Please check the fuse on the AVR. Then ensure the jumper is in the 60 HZ position. Other connections as they are supposed to be and give it a go. The regulators are weak in this application.

After I got it running on 24V in April 2025, the unit put out 208V 3-phase power. We need 480V 3-phase power, so we retapped the alternator/generator for that voltage output. The unit was running well, and the laser de-scaler connected to it was operational - until October/November 2025.

Fuse on the original AVR (#1) was blown. Replaced it and it blew. Changed AVR (#2). After a few other adjustments, unit ran fine for 5 months. Then, it starting shutting down after about 15-30 seconds. Fuse on old AVR (#2) was good, but the green capacitor sticking up in one corner had some dark spots on it - like it had gotten overheated. Changed AVR (#3). Changed the jumpers and wires one at a time. Double checked the two jumpers and all other wiring against the photos I had taken before disassembly. Unit ran a handful of times for extended periods, then started acting up again. Fuse on current AVR (#3) is good.

After 3 AVRs, one would get the impression that the culprit is something other than the AVR. The only other "smart" parts are the PLC and the Electronic Speed Switch.