Page 2 of 2

Re: HP 5970 / 5890

Posted: Tue Aug 11, 2015 10:15 pm
by James_Ball
I recently had an HP 5971 refurbished. During the refurb, it was 'upgraded' to a 5972 and supposedly the power supply was replaced. Honestly I was not really happy with the finished product - the EM was putting out <2500mv during BFB tuning (meaning he didn't bother to replace it while it was apart). Still, it powered up, pumped down, and performed an acceptable tune.

Now several months after putting it into operation I am seeing some unusual activity. It was first observed as flat areas within a run where even noise was washed out - occasionally in areas i was expecting peak activity. It happened for several runs while trying to calibrate. The next day while checking a CCV run I got the 'Fault 8 error' = no emission current. When running diagnostics, I can get reasonable spectra with both filaments = they are not burnt out. When I try and run a tune program, it starts out fine but 2-3 minutes into it, I get the Fault 8 No emission current error.

Can I assume that the power supply is going bad?, and that when the spectrum drops out it is because of low volatge to the filaments? The problem appears to be intermitttent but getting slowly worse. I have seen the power supply available for sale on the internet. The manual indicates that it may be a chore do it in house. I'd appreciate any advice regarding checking the PS and if it is worth attempting the replacement myself.
I have replaced one before, and it was the upgrade since they made the power supply more reliable with the 5972 and all the connections are the same. You will have to pretty much completely disassemble the unit, but it isn't too terrible of a job, just takes time and patience.

Another thing to look at is the ceramic pass through where the filament pins pass through from the connector to the board inside the vacuum chamber. I had one once that had burns on it from where it was welded in and intermittently it would allow the filament current to pass to ground before going through the filament. The service rep was able to make a temporary fix by sanding between the pins, but we got a new top plate to make sure it was fixed for good. It would show up as good when checking with a ohm meter, but the higher current of the filament would sometimes jump the gap to ground.

Re: HP 5970 / 5890

Posted: Tue Sep 01, 2015 3:19 am
by cleh
Are you using the 5972 filaments?

Re: HP 5970 / 5890

Posted: Tue Sep 01, 2015 8:19 pm
by James_Ball
Are you using the 5972 filaments?
I think we still used the 5971 filaments on that one since only the power supply was changed. We don't have it anymore or I would check it to be sure. I do know we still had a few 5971 filaments around last I looked, but with new people they may have been discarded without knowing what they were, just like my old 5970 manuals :(

Re: HP 5970 / 5890

Posted: Wed Sep 09, 2015 8:24 pm
by stevek3
Wow. Thanks to all whom have replied. It is more than I expected.

I pulled the covers off of the MS and have been checking the Filament and power supply test points during operation.
The power supply is generating the proper voltages 5,15,24, and -15Vdc. When the MS is on, it is showing -75Vdc (and 1.10Vac) at the filaments. At some intermittent point during operation, the -75Vdc at the filaments stops and the Fault 8/No emission current error shows up. When this happens, the PS voltages still show the proper voltages. If I am reading the Tech manual properly, the -15Vdc supply from the PS is the actual source of the -70Vdc seen at the filaments. This suggests to me that I am losing the Main Board Filament Drive circuit.

The block diagram in the manual does not directly identify the -70V power supply but it looks like it is the small transformer marked T1 (05971-80020) on the board itself. I am not familiar enough with transformers to check if the transformer itself is the point of failure or if it is something else in the circuit. Is it possible to check the soldered leads at the edges of the transformer to see if the -70Vdc is present? Does testing the Emission Test point for 0/-.6V actually confirm if this is the point of failure?

Answer to a ? - It is supposed to be using the 5972 filaments. It is what I paid for during the refurb back this past spring 2015. I have never opened it to look, just hooked it up and started running. This unit was not operating when I arrived at this position 4 yrs ago. It was old and gave 'no emission current' msg when plugged in, so I had it sent in for repairs. Looking at it now, it appears the PS board was not upgraded during the refurb, nor was the EM replaced (>2500EMV during BFB tune). I am not sure exactly what was performed on it and communication with the refurb guy was difficult.