Bicron MicroREM Geiger Counter - Meter pegs at power on

[elementip]

I picked up a Bicron MicroREM Geiger counter that was advertised to be working. Cosmetically it's in great shape.
Unfortunately the meter movement pegs to the full scale mark as soon as the unit is powered on and stays there. The behavior is the same no matter what setting the knob is set it. Does anyone have experience troubleshooting these types of devices?

I haven't been able to locate a schematic for it, just sales literature.

[ManateeMafia]

Is this the same device?

https://www.nukeworker.com/study/instruments/manuals/Micro_Rem_Manual.pdf

Schematic is at the bottom.

[elementip]

Thanks ManateeMafia, I'm not sure how I missed this.
Time to start measuring component values and voltages!

[SparkyBruce]

It could of course have a contaminated front window......

Bruce

[stj]

or you could be in japan! 

[txescientist]

It is not a Geiger counter, it is a scintillation counter. You should check voltages and see if there is a small hole on the probe tip that is allowing light to hit scintillator crystal. Also, maybe PMT tube is damaged, or you have very high ripple on -HV rail of the PSU. It would help if you could post some images of the unit. If the above schematic is the right one, you should check -5V also and 7660 IC.

[elementip]

If the above schematic is the right one, you should check -5V also and 7660 IC.

The schematic and board diagrams are slightly different (a newer revision), but the change list includes the changes from rev to rev. I was able to determine that the 7660 is not outputting -5V on pin 5. I pulled it out and tested it on a breadboard, it functions just fine. Moreover, the meter behaved exactly the same regardless of whether or not the 7660 was inserted or not.

I believe the charge pump capacitor for the 7660 is at fault. I was able to simulate the same voltage reading if when mocked up in the breadboard if I substituted the charge pump capacitor for a 10K resistor. The board has a 33uf/16V tantalum capacitor working as the external cap for the 7660. What's the typical failure mode for tantalum caps? I don't encounter them often in the equipment I repair.

I'm going to be replacing the cap tomorrow night. There are a total of 4 of this capacitor in the device, would it be likely all should be replaced?

[txescientist]

See http://www.kemet.com/Lists/TechnicalArticles/Attachments/199/2014%20EDFA%20Tantalum%20Cap%20Failure%20Analysis%20Review%20by%20Javaid%20Qazi.pdf

Think that tantalum caps just short out or explode :-) Pure evil parts...

Replace them all, replacement is cheap, and they tend to loose specs over time. They are the main reason why many old test equipment stop working. There are a bunch on them in old Tek scopes. Integrator opamp when not supplied with -5V is going crazy and pegs the scale. Please let us know did you fixed it by replacing tantalums around 7660  .

[elementip]

I replaced all 4 33uf/16v tant caps but the unit still behaves the same. I'm measuring a positive 0.2V on the output of the 7660, where I should be seeing -5V.

I pulled the 7660 from the circuit and measured +4.6V on the socket pin where the 7660 would be outputting -5V.  This leads me to think that something may have failed shorted between the +5 and -5 rails and is letting this voltage across. Bad opamp perhaps?

[txescientist]

Here you have res better schematics. Hope that there are no major changes in the revisions.
http://www.qsl.net/k0ff/Bicron%20MicroRem%20Audio/Bicron%20Micro%20Rew%20W%20Audio/Image.JPG
http://www.qsl.net/k0ff/Bicron%20MicroRem%20Audio/Bicron%20Micro%20Rew%20W%20Audio/Image0001.JPG
http://www.qsl.net/k0ff/Bicron%20MicroRem%20Audio/Bicron%20Micro%20Rew%20W%20Audio/Image0002.JPG

You are right to think that. Two opamps are connected to -5V and an analog switch. If opamps are socketed, you should pull them out and then check the voltage again. Opamps are the only parts connected directly to +/-5V rails.
Maybe you should also disconnect the meter indicator microampermeter, just in case, because you can fry it easily. It has a buffered output with current limiting resistor to 5-7mA, so it should be fine, but in any case.

I hope we can help you fix it 

[elementip]

I pulled both opamps and then measured the voltage on the negative rail. Unfortunately positive voltage is still present on the negative rail. The negative rail measures 1VDC (with the 7660 pulled (leaving the 7660 in circuit causes the voltage to jump to 2VDC, clearly the 7660 cannot cope when biased that way).

Could there be a shorted diode? I think I'm down to testing individual components at this point.

The high res schematics you linked to are actually mostly correct - That unit seems to have some additional audio and response controls that mine does not. However substantial parts are identical, it was helpful in determining some of the harder to read parts of the schematic in the manual linked to in the earlier post.

[txescientist]

I still think that only ICs can take -5V rail to positive voltage, because there are only ICs connected to the -5V rail and resistors. Resistors are less likely to fail. You should check/desolder U3, U4, U5, U9 and U15 - opamps and an analog switch IC. Then if you don't get -5V, you should check individual resistors connected to -5V rail with ommeter.