Restoring Tektronix Attenuator Switch Deck Assemblies

[med6753]

The object of this thread is to demonstrate an alternate cleaning method for the Tektronix attenuator switch decks used on multiple plug-ins and scopes in the 1970's and 1980's. These decks use gold plated “finger contacts” against a special board to achieve specific attenuation ranges. The cleaning method recommended by Tektronix has been IPA only. But unfortunately after years of contamination buildup and wear and tear in many cases IPA will not restore normal operation. So I decided to perform an experiment on a 7A24 vertical plug-in that was hopelessly intermittent on multiple ranges on both channel 1 and 2 even after multiple cleanings with IPA and see if I could restore normal operation.     

Here is a typical 50 ohm attenuator switch deck. This is the 7A24 which will be used in this experiment. The boards will be removed and stripped down.



Here is a typical 1MEG attenuator switch deck. This is from a 7A15A. This cleaning method will involve removing the assembly and stripping down the board so be sure to keep the attenuator blocks with the same board and location to minimize the amount of compensation required after reassembly.   



Here's the attenuator boards removed. There are finger contacts on both sides of the board so have the boards elevated as shown. Those contacts can be easily damaged.



Remove the attenuator blocks. It's best to keep track of which board they came from but not absolutely necessary for 50 ohm input. For 1MEG input keeping track will minimize compensation after reassembly.



Prepare a small piece of  fine sandpaper or emory cloth as shown. What I did was super glue it to a toothpick. Use this to polish the board surface under the finger contacts until shiny. On the backside of the board the finger contacts are open so this is easy. On the front of the board the contacts are closed. VERY CAREFULLY pry up one contact at a time to polish the board. I do NOT recommend attempting to polish the finger contacts themselves. They are easily damaged and you'll never restore proper operation. We will deal with any potential contamination on the fingers later.



After the polishing is complete soak the boards in 100% IPA for about 10 minutes. Agitate slightly then remove and allow to air dry or use a blow dryer.   



Now we'll deal with contamination on the finger contacts. Deoxit Gold G5. For gold plated surfaces. But don't go crazy spraying it on the board. I used a cotton swab soaked in the deoxit applied to both board surface and finger contact by prying up very carefully. The cotton swab can leave fibers so check carefully. If you have a syringe it will work even better.




The cleaning is complete. Reassemble the switch decks and check the results. In the case of this 7A24 vertical plug-in it was perfect and rock steady across all switch ranges. Here's an example of a 10MHz square wave from a Heath IG-4244 Oscilloscope Calibrator.

[mnementh]

Interesting. How did you retension the contact springs after bending them up so high? Do you have a demonstrable technique?

Seems to me getting the tension close to right and even across the entire board would be a total summitch, and would definitely rely on experience that can probably only be gained by ruining a board or three... 

mnem
 

[worsthorse]

Nice explanation. I am going to give this a try on the decks I pull from the trash 466.

[med6753]

Interesting. How did you retension the contact springs after bending them up so high? Do you have a demonstrable technique?

Seems to me getting the tension close to right and even across the entire board would be a total summitch, and would definitely rely on experience that can probably only be gained by ruining a board or three... 

mnem
 

I did not touch the contact fingers. When you pull the attenuator assembly off the fingers on the backside spring out as shown all by themselves. When assembled they are automatically tensioned against the board by the switch deck. The fingers on the front of the board automatically close against the board when disassembled. When assembled the switch deck has cams that open and close them.     

[tautech]

Nice Med. 
Added to the list here:
https://www.eevblog.com/forum/repair/vintageclassic-renovation-techniques/

Edit
Can I ask how fine you went with the emery paper and was it glass or carborundum based ?

[med6753]

Nice Med. 
Added to the list here:
https://www.eevblog.com/forum/repair/vintageclassic-renovation-techniques/

Edit
Can I ask how fine you went with the emery paper and was it glass or carborundum based ?

Actually, I'm not sure.  I think it was carborundum based.

[tautech]

Nice Med. 
Added to the list here:
https://www.eevblog.com/forum/repair/vintageclassic-renovation-techniques/

Edit
Can I ask how fine you went with the emery paper and was it glass or carborundum based ?

Actually, I'm not sure.  I think it was carborundum based.

And the grit # for those others playing along ? 400, 600, 1000 ?

[med6753]

Nice Med. 
Added to the list here:
https://www.eevblog.com/forum/repair/vintageclassic-renovation-techniques/

Edit
Can I ask how fine you went with the emery paper and was it glass or carborundum based ?

Actually, I'm not sure.  I think it was carborundum based.

And the grit # for those others playing along ? 400, 600, 1000 ?

600 grit.

[tautech]

Nice Med. 
Added to the list here:
https://www.eevblog.com/forum/repair/vintageclassic-renovation-techniques/

Edit
Can I ask how fine you went with the emery paper and was it glass or carborundum based ?

Actually, I'm not sure.  I think it was carborundum based.

And the grit # for those others playing along ? 400, 600, 1000 ?

600 grit.

Just fine enough IMO. For those delicate gold switches maybe 1000 would be better or possibly a custom shaped pencil eraser. 

[mnementh]

Interesting. How did you retension the contact springs after bending them up so high? Do you have a demonstrable technique?

Seems to me getting the tension close to right and even across the entire board would be a total summitch, and would definitely rely on experience that can probably only be gained by ruining a board or three... 

mnem
 

I did not touch the contact fingers. When you pull the attenuator assembly off the fingers on the backside spring out as shown all by themselves. When assembled they are automatically tensioned against the board by the switch deck. The fingers on the front of the board automatically close against the board when disassembled. When assembled the switch deck has cams that open and close them.

Ahhhh... I grok now. Thank you!     I should have guessed the springs might be preloaded differently; I've seen a lot of that in hydraulic logic.

Nice Med. 
Added to the list here:
https://www.eevblog.com/forum/repair/vintageclassic-renovation-techniques/

Edit
Can I ask how fine you went with the emery paper and was it glass or carborundum based ?

Actually, I'm not sure.  I think it was carborundum based.

And the grit # for those others playing along ? 400, 600, 1000 ?

600 grit.

Just fine enough IMO. For those delicate gold switches maybe 1000 would be better or possibly a custom shaped pencil eraser. 

For gold electroplate... even that heavy... I'd be afraid to touch it with anything coarser than 2000 grit wet-sanding film. I did cringe a little when I saw the way it dug up the surface on the example piece; the same way I used to cringe when I saw fellow "technicians" clean a crusty VCR capstan with 220 grit while the machine was running. 

This kind of switch is generally self-cleaning; relying on the geometry of the leaf spring and how the fingers meet the contact to produce a wiping action every time the switch closes. Those scores in the gold surface caused by the sandpaper give me concern for the long-term reliability of the repair.

2000-3000 grit wet/dry film is readily available at any auto parts store and larger Wally World locations with an extensive automotive dept; it is used for final sanding between layers of base coat/clear-coat automotive paint. I'd be inclined to wet it slightly with alcohol in this case.

mnem
 

[VK5RC]

I have cleaned some relay contacts using de-oxit soaked cartridge paper - with success.

[med6753]

Refinements to this process are most welcome and I agree that perhaps a much finer polishing method should be used on the contact surfaces. As I said at the get go this was an experiment and I had no idea if it would be successful. I was willing to sacrifice this plug-in because it was useless in it's current state.

But I did forget to include one crucial step in the reassembly which I'll correct now. Once the reassembly is complete you should perform a static DC continuity check of all ranges before applying power. BNC input, coupling switch to DC, other probe to output side of the attenuator. On 50 ohm attenuators you'll have readings anywhere from 2.0 ohms up to about 100 ohms. The reading should be rock steady and not bounce around. If a range reads 4K ohms or above it's open and you'll have to investigate. On 1MEG attenuators you'll have a reading anywhere from approx 500K ohms up to approx 1.8MEG. Again, should be rock steady and not bounce around. And any reading above 2.0MEG that range is open and again you'll have to investigate. 

 

[mnementh]

This... this is the contextual equivalent of fingernails on a chalkboard to me. 

But thanks for the rundown on diag-ing these attenuators. I suspect the common modular ones will troubleshoot similarly. 

mnem
 

[tautech]

This... this is the contextual equivalent of fingernails on a chalkboard to me. 

mnem
 

Why ?
Med likes a tidy bench and I think all pics we've seen of it, it's spick and span. 

Coiled mains lead, why bother ? When it shits out splice in another. 

[med6753]

Sorry, but I don't like cords all over the bench when not in use.  You best not come over here because there are more like that. You'll be cringing for a week. 

[med6753]

This... this is the contextual equivalent of fingernails on a chalkboard to me. 

mnem
 

Why ?
Med likes a tidy bench and I think all pics we've seen of it, it's spick and span. 

Coiled mains lead, why bother ? When it shits out splice in another. 

Thank you!   

[mnementh]

Sorry, but I don't like cords all over the bench when not in use.  You best not come over here because there are more like that. You'll be cringing for a week. 

Neither do I; you've all seen just how anal I can be about tidy wires.  

However in my workspace, I need smooth workflow. I loathe wiggly ill-behaved cords that hang up on everything and drag all manner of shit off my bench onto the floor. I hang all my cords up so they hang as straight as possible; even the ones from my multiple soldering stations.

The damage it does to the cord does also irk me, I'll admit. There's a reason Pomona makes that hanger thing that looks like a garden rake; and it's one of the reasons their leads last forever. I have 30+ year old tools with the original cord in perfectly serviceable and flexible condition.

But it's the wiggly ill-behaved cables thing that makes my hackles rise. 

mnem
 

[David Hess]

Prepare a small piece of  fine sandpaper or emory cloth as shown. What I did was super glue it to a toothpick. Use this to polish the board surface under the finger contacts until shiny. On the backside of the board the finger contacts are open so this is easy. On the front of the board the contacts are closed. VERY CAREFULLY pry up one contact at a time to polish the board. I do NOT recommend attempting to polish the finger contacts themselves. They are easily damaged and you'll never restore proper operation. We will deal with any potential contamination on the fingers later.

Whoa, sandpaper or any abrasive is way too rough for this!  It will only temporarily restore operation if the contact plating is damaged.

The Tektronix recommendation which works very well is to use a non-clay bearing paper and draw it through the contacts.  Even the clay in common paper is too abrasive.  Vellum would be a good choice if you have it available (it was a common drafting material) but I generally use thin card stock which is a little easier to manipulate.

[med6753]

OK, here's what I'll do. I'll put the 7A24 away in the equipment closet for about a month or so and not use it. After that time I'll pull it out and see what the results are. If it's gone intermittent again shame on me, I ruined the boards and I'll either have to find replacements or junk the plug-in. If it's working fine then perhaps the use of mild abrasives may not be such as bad thing and we'll have further discussions on it's merits.

So look for an update around the end of June.

[mnementh]

David -

The point of this experiment was to restore a board that did not respond to the "approved" methods of service. I think that's a valid course of inquiry. Yes, it is obvious that this method is fraught with peril; any measure of success depends entirely on using the finest grit abrasive and having a gentle hand here, while polishing absolutely as little as needed to restore contact. And THEN washing everything thoroughly to ensure no grit contamination. The ham-fisted gorilla who grinds away the gold-plating will be rewarded with a unit that is now irrecoverably garbage.

But remember... we ALL take our turn being the ham-fisted gorilla at one time or another. This is for science.

mnem

[bd139]

Nice write up.

I'm not ashamed to say I blasted a "circling the drain" attenuator module in a 465 with WD40 contact cleaner (not the normal WD40) after traditional methods failed. That worked a treat.

[med6753]

I have also resorted to using Deoxit D100 in past with mixed results. In my experience with these attenuator decks...especially with the 1MEG inputs.....it's the 20mV/.2V ranges that usually are the first to start going intermittent. And unfortunately they are the ranges that respond LEAST to cleaning with IPA. Not sure why it seems to be due to the switch design. And even if you think IPA fixed it sometimes less than a week later those 2 ranges are intermittent again.      

[med6753]

As I promised back on May 15th I would post a follow up on the polishing of the attenuator switch deck on a 7A24. I had to substitute a military 7603N for the 7904 used previously because it is currently down for repairs. The 7603N has a “stated” B/W of only 65MHz as opposed to the civilian 7603 100MHz. And the 7904 is 400MHz. But to keep things consistent from the prior experiment I used a 10MHz square wave so there WILL be some roll off due to the limitations of the scope. But enough banter. What are the results?


Absolutely rock steady across all switch ranges on both channels. As far as I'm concerned the polishing of the contacts has no ill effects and in fact is a benefit. I typically see IPA cleaned only decks start to go intermittent again sometimes as short as few days and usually within a few weeks. This 7A24 hasn't been used in almost 1.5 months. This has convinced me that the polishing works and I intend to apply it to a 7A26 which is giving me fits on channel 1. Even Deoxit on that 7A26 is a short term fix.



The only thing I would do differently in the future is use a much finer polishing technique. I suppose we could debate ad naseum the pros/cons of this cleaning. But results speak for themselves. Obviously it's up to you if you want to apply it to your equipment.       

[David Hess]

In my experience with these attenuator decks...especially with the 1MEG inputs.....it's the 20mV/.2V ranges that usually are the first to start going intermittent. And unfortunately they are the ranges that respond LEAST to cleaning with IPA. Not sure why it seems to be due to the switch design. And even if you think IPA fixed it sometimes less than a week later those 2 ranges are intermittent again.

I have repaired some very old 7A18s and 7A26s that behaved like that.  The problem was that the gold plated fingers on the switches had "unsprung" so they were barely touching the gold plated pads on the printed circuit board.  I suspect they were damaged by someone cleaning them too aggressively bending them upwards.  I fixed them by using tweezers to bend the base of the fingers back so they again wiped on the board like they are suppose to.

[med6753]

I have also had to clean up similar messes caused by prior techs....or I should say ham-fisted Gorillas. The worst was a 465B where someone in an attempt to fix contact issues torqued the shield stand off screws in so tight that the rear retention bushings were stripped out. Even after complete disassembly and replacement of the bushings it was obvious that the assembly was damaged beyond repair. I had to replace the entire assembly from a parts mule.