EG&G PARC Model 113 Preamplifier repair/restoration

[vindoline]

I scored a new toy on the "bay" recently. An EG&G PARC Model 113 preamplifier. User manual an schematics here: http://web.mit.edu/8.13/8.13d/manuals/egg-par-ortec/par-113-preamplifier-manual.pdf.

I've been keeping my eyes out for a low noise pre-amplifier to look at low level signals and generally play around with. This one came up recently "for parts or repair" and the description said "internal battery corrosion" Hmmm, how bad could it be? I offered $30 and it was quickly accepted. Here is what it looked like upon receipt 

Yikes! The leaking NiCd battery pack really did a job on the board. Any one want to give me odds (or advice) on this being a happy ending? According to the manual, the NiCd battery pack hanging across the power supply provides the voltage regulation - something I've never heard of - and I will definitely not be powering this up until I find/build a replacement battery pack and clean the board.

[Vgkid]

Yikes , If it had just stayed at the passives, not a big deal, but it even attacked the adjustable cap.
Better start scrubbing, to see if you can even have a chance at fixing it. I wonder if you can do better with a a discrete low noise fet input, + modern low noise opamps. I have seen THD circuits built that way.

[vindoline]

Yikes , If it had just stayed at the passives, not a big deal, but it even attacked the adjustable cap.
Better start scrubbing, to see if you can even have a chance at fixing it. I wonder if you can do better with a a discrete low noise fet input, + modern low noise opamps. I have seen THD circuits built that way.

Yes, those beautiful variable air capacitors are bright blue-green! 
I'm sure one can do better now with modern low noise opamps, but I enjoy the challenge and I want something to compare future home brew preamps to.  Fingers crossed!

[TiN]

Mnom,nom,nom.. Sorry, it's just so colorful.
Best bet would be remove all components from PCB and clean everything thoroughly. Replace common parts with brand new ones, and rare parts try to clean-up and test if they still good. Lot of work. 

[edpalmer42]

It's hard to believe that the batteries did all that.  They don't even look that bad!  I've seen worse batteries that didn't cause nearly that much damage.  I think they must have had some help.  Is there any hint of water damage?

I have the same unit and there was no corrosion on mine even though the batteries looked about that bad.

Ed

[vindoline]

It's hard to believe that the batteries did all that.  They don't even look that bad!  I've seen worse batteries that didn't cause nearly that much damage.  I think they must have had some help.  Is there any hint of water damage?

I have the same unit and there was no corrosion on mine even though the batteries looked about that bad.

Ed

If th batteries don't look bad, that's due to my poor photography! I'm also surprised at the extent of the corrosion, but I think it's all due to the batteries. In person you can clearly see that the damage radiates out from the battery pack.

[vindoline]

Best bet would be remove all components from PCB and clean everything thoroughly. Replace common parts with brand new ones, and rare parts try to clean-up and test if they still good. Lot of work. 

I'm going to try a board level cleaning tomorrow and see how things look. It's going to be a PITA to get the pcb out because all the connections to it are soldered!  A quick count shows 27 wires to be desoldered 

[Kleinstein]

Quite some of the nice blue color is not from corrosion on the boards, but just salts coming from the leaking NiCd cells. The liquid contains quite some NaOH and thus is rather hygroscopic and tends to cause a film of water (brain) on all surfaces, even creping up and this way spread all over. For most parts NaOH is not promoting corrosion - it is even protecting iron from corrosion. But aluminum parts can be really damaged.

A lot of the salts can be just washed away with clean water. Possibly to the point of not having to replace many parts. The board needs good washing anyway to get leakage down.

This is an really old amplifier and the 2N5521 FETs noted in the plan are not an especially low noise type. There are lower noise modern parts available. So don't expect an instrument with top notch performance but more an historic one. This might mean keeping original parts might be preferred over slightly better performance of modern parts (e.g. metal film resistors instead of the old carbon ones).

[richnormand]

i have one of those. Works well and it is handy in many situations.

Before doing an in-depth cleaning take some good photos of the affected area. Depending on the pcb a good cleaning may wipe out some traces completely.
In your case that is not a real problem since you have the schematics and it is a relatively simple circuit however.

I can send you photos and voltage readings if you ever need them.

Have fun.

[vindoline]

i have one of those. Works well and it is handy in many situations.

Before doing an in-depth cleaning take some good photos of the affected area. Depending on the pcb a good cleaning may wipe out some traces completely.
In your case that is not a real problem since you have the schematics and it is a relatively simple circuit however.

I can send you photos and voltage readings if you ever need them.

Have fun.

Thanks for the offer. I'll definitely let you know if I need any readings.

[vindoline]

I managed to carefully unsolder all the wires from the PCB and get the board out. Some careful spot cleaning with IPA or water showed only marginal results on the corrosion. In a "what have I got to loose" moment I decided to give the entire board a thorough soaking in white vinegar. The results were almost instantaneous  All the visible blue-green salts dissolved. After a few minutes I changed to a fresh solution of vinegar, followed by multiple washings with clean water and then IPA. I think the board is looking pretty good now. Now it's going into the oven for a few hours to dry out completely. Hopefully everything survives this rather drastic treatment.

[SoundTech-LG]

Wow...

your work is cut out for you. I have an extremely rare Princeton Applied Research Model 110.
Oscillator/Notch Filter/Selective amplifier/Distortion Analyzer
Try finding that online

One thing you will want to do in your efforts is replace those electrolytics on the board. Mine had about 7, or 8 shorted ones!!!


All the best!

[vindoline]

One thing you will want to do in your efforts is replace those electrolytics on the board. Mine had about 7, or 8 shorted ones!!!

All the best!

Thanks for the heads-up. I certainly check into replacing those!

[vindoline]

OK, I ordered a set of 24 NiCd cells via eBay to rebuild the battery pack and they arrived today. It's not my finest work, but it was very "fiddley" to solder all the battery tabs while keeping the cells in the required footprint. However, it's done and a few wraps of kapton tape are holding things together. I don't see any way around having to re-solder everything back together to see if its working. Fingers crossed 

[edpalmer42]

Did you test the NiCd cells before using them?  I've got a few devices that need new batteries and I thought it would be a good idea to buy a few extra cells in case any were bad.  I don't know if that's necessary or not.

I've also thought about replacing the entire power supply with some kind of low noise supply that doesn't use batteries.  Any thoughts?

Ed

P.S.  There's a chapter in V3 of Art of Electronics that discusses low-noise amps like this one.  They do an analysis of a unit from Stanford Research and by comparing it to the 113, it looks like for best noise performance, the impedance of your source has to be around 1Mohm and the frequency range is only about 10 - 1000 Hz.  Definitely something to keep in mind!

[vindoline]

Did you test the NiCd cells before using them?  I've got a few devices that need new batteries and I thought it would be a good idea to buy a few extra cells in case any were bad.  I don't know if that's necessary or not.

I didn't do any specific testing of the cells. They did arrived charged and the assembled battery had the correct voltage.

I've also thought about replacing the entire power supply with some kind of low noise supply that doesn't use batteries.  Any thoughts?

I think it would be quite doable. It just needs +/- 12 V

[Vgkid]

There is a collection of low noise LDO's in this article.
http://www.analog.com/media/en/news-marketing-collateral/product-highlight/Low_Dropout_Regulators.pdf

[vindoline]

There is a collection of low noise LDO's in this article.
http://www.analog.com/media/en/news-marketing-collateral/product-highlight/Low_Dropout_Regulators.pdf

Vgkid, thanks for the reference. It should come in very handy.

[vindoline]

Alright, I've put the cleaned PCB back in and re-soldered all the connections. There is still some life in the amp, but it's one sick puppy  . The power rails seem fine now and the battery check function works. Using a sine wave from my function gen, there initially appeared to be no output    If you look at the block diagram, the unit consists of a preamp followed by the band pass filters then an output amp section. I can inject the signal at the input to the output amp and get an output  During initial troubleshooting of the preamp section I found that adjusting one of the variable caps caused an output signal to appear! However it's intermittent and distorted. I suspect I'll need to desolder and further clean or replace the variable caps and maybe trimmer pots