Fluke 45 repair

[tbodor]

Hi,

Fluke 45, was working perfectly until suddenly stopped switching to the resistance function while in use. The resistance measurement doesn't work, mV and 3V DC ranges also, as well as continuity (always on/beeping). Higher voltage ranges are fine, current measurements as well. Pulled out the service manual...

I've traced it to the K1 relay, which seems to be stuck in the set position. Unfortunately, it doesn't appear to be a problem with the relay. I'm currently guessing the U1 chip isn't sending the reset signal when switching ranges or functions. I can capture the set/reset signals for the other relays and the set pulse for K1, but no reset. It's not the driver (U2), there's just nothing on U1's FC0/28 pin at all.

I've found no mention of this failure mode anywhere. I presume U1 will need replacing? Any idea where I can find one?

Thanks,
T.

[lowimpedance]

OK, without downloading the manual to look at what U1 is , post a picture of it showing the part number, or at least indicate the details on it may help with locating one!.

[Samogon]

No source of Fluke’s custom IC

[lowimpedance]

If U1 is the main custom IC then your only source will be a parts unit.., hopefully with a good U1 !.

[tbodor]

The service manual is available here: https://www.seattleu.edu/media/college-of-science-and-engineering/files/departments/electricalandcomputerengineering/Fluke_45_ServiceManual3ec7.pdf

There's a schematic on page 167 (9-3), and I'll try to attach a photo as well. U1 is the analog processor, marked 776195.

[tbodor]

I suspected it would be a long shot to expect to be able to find one without buying the entire board. It just seems strange that the chip is functional apart from that one pin. I was hoping someone might have seen this before and that perhaps there's something I'm missing.

[plurn]

...
I've found no mention of this failure mode anywhere.
...

Sections "5-9. Analog Troubleshooting" and "5-10. Uart Test" of the service manual sound like they might be related to your issue. There is a bunch of tests in those sections that might narrow down the cause. Hopefully it is not due to some difficult to replace part like U1.

[wolfp]

Maybe the FC0-Port is held down by a defective input of the ULN driver or a short circuit to GND.
Wolfgang

[TheDefpom]

The traces from pins 13 and 15 look a bit odd, you might want to check those.

[tbodor]

Maybe the FC0-Port is held down by a defective input of the ULN driver or a short circuit to GND.
Wolfgang

Interesting idea, definitely something I wanted to try as it would have been easy to fix; the driver is a generic part. Unfortunately, doesn't seem to be the case: I lifted the input pin on the driver, but U1 is still not outputting the pulse on FC0. It's also not shorted.

I can flip the relay by providing a short pulse on the driver input. I'm thinking it should be possible to hack something around this. I've connected a logic analyser to the optocoupler that is used to send commands from the main processor. It's some sort of a UART, running at 4800 baud, but I haven't found any documentation on the protocol. However, I think I could spy the line using a small microcontroller, recognise the couple of commands that would normally cause the relay to reset, and send a pulse to the driver.

[tbodor]

Sections "5-9. Analog Troubleshooting" and "5-10. Uart Test" of the service manual sound like they might be related to your issue. There is a bunch of tests in those sections that might narrow down the cause. Hopefully it is not due to some difficult to replace part like U1.

There is an automated UART test function (accessed by holding the RATE button). I get the PASS result for this.

There are no error codes that are the usual entry point into the analogue troubleshooting. But I had followed the input to the relay; that's how I found the relay is not switching.

[tbodor]

The traces from pins 13 and 15 look a bit odd, you might want to check those.

On U1? I don't see anything, unless you mean the guard traces as marked in the attached photo - they are not very level (except those are not 13 and 15; pin 1 is in the middle of the top side for this package).

[PA0PBZ]

I can flip the relay by providing a short pulse on the driver input. I'm thinking it should be possible to hack something around this.

RESET K1 = (SET K2 OR SET K3) ?

[tbodor]

In an interesting twist, I checked the 5V power supply on U1. The allowed range is 4.95-5.45. Mine was 5.62V. Since everything worked apart from the relay drive pulse, I initially didn't think much of it. However, I decided to tweak it - just in case. The two resistors R35 and R36 can be used to move the voltage up and down a bit. Cutting/removing R36 raises the voltage by 7%. Cutting/removing R35 lowers it by 5%. I found R36 had already been cut (perhaps in factory). So I cut R35 as well.

The supply is now 5.28V, K1 reset pulse is generated as expected, relay flips, and everything works. I'm almost disappointed there was no need for hacking, but there. Maybe this helps someone else.

Thanks everyone who replied with suggestions and ideas, much appreciated.

[Samogon]

Yeah that is what we check first, power rails lol.
Good you found it before swapping/soldering spree.

[lowimpedance]

 Interesting failure mechanism, I wonder how long its been sitting close to failure and what changed !.  Just had a look at the power supply cct., not your average simple linear is it !!. Have any of the resistors around the controller drifted out of tolerance?, or C32,34,35,41 dried out ?. All the rails in spec.?.

[Ilot]

Hi,
I recently bought a defective Fluke 45 for 50€...
In my case, the only working ranges were 3 VDC and milliamp DC.
After checking the power supply lines, i starded tracing the input signal through the meter.
I didn't have to go far : the R5 (3.5 kOhms / 5W) resistor is "open circuit" (soldering is fine ; the component is dead).