Share:
Notifications
Clear all

Tin whisker shaving trial perhaps in circuit in future

 
Refugee
(@refugee)
Famed V-Ratter Deactivated Account

I have successfully got three good ones out of four with this method.
After looking at many web sites all about this issue over time I have come to the conclusion that shaving these things off can be done kindly and without shorting all the leads together.
With a bench-top lash-up I tried it out.
I set the bench power supply to 3 volts with two limiting resistors one being 150 ohms and the other 47 ohms. An AVO was added between the resistor and the transistor. I wired the case to negative as the reverse voltage of the junctions is about 15 volts.
The first one was an AF115 that was short emitter to case. The 150 ohm resistor did not provide enough currant and I had to go again with the 47 ohm. I would guess that the currant was about 70ma.
This one recovered and showed a gain of 200 at 5ma when fixed.
The second was an NOS NKT603 short base to case. The 150 ohm resistor fixed this one first time and the gain after was 160 at 5ma.
The last was an NOS unmarked device from one of those 1970s by weight packs that was also short base to case.
This one also cleared with the 150 ohm resistor and had an after gain of 90 at 5ma.
The duff one was short base to emitter and proved to be a resistive short that just got a little bit lower after I blasted it with 15 volts either way with the 47 ohm resistor.
The next task is to have a look at doing them in circuit provided I can calculate the currant from the other components present in circuit and then we will have a very quick fix for this problem :thumb
The flash reflections were a bit of a pain but it does show most of the odd part numbers I have found om these TO7 transistors over the years however the 603 was hiding its part number in most attempts but not quite on this shot.

ReplyQuote
Topic starter Posted : 03/08/2014 6:02 am
Anonymous
(@anonymous)
Illustrious V-Ratter Deactivated Account

It's pointless as you are not removing the source of the tin. The whiskers will grow back.

ReplyQuote
Posted : 14/08/2014 12:47 pm
valvekits
(@valvekits)
Prominent V-Ratter Deactivated Account

Back in the 70's snipping the screen on an AF11X was often a quick neat solution to a problem - nobody knew why and frankly nobody cared as it got the job done.
Forty years on we know why but I bet that if service departments still repaired these today we'd still be snipping them because basically as engineers we are a lazy bunch. I like Refs' experiment, it might be pointless but not to us lazy lot who see it as a quick get out of jail card. :bba

Eddie

ReplyQuote
Posted : 14/08/2014 3:00 pm
Cathovisor
(@cathovisor)
Illustrious V-Ratter Registered

I like Refs' experiment, it might be pointless but not to us lazy lot who see it as a quick get out of jail card. :bba

Eddie

So do I: although I find Ref's method a bit subtle - I was going to try a charged electrolytic, C-B-E shorted together and connected to one side of said charged electrolytic, shield to the other... :bba

ReplyQuote
Posted : 14/08/2014 3:12 pm
Refugee
(@refugee)
Famed V-Ratter Deactivated Account

What I am trying to do is to slowly heat the whisker so that it melts all at once and with a bit of luck it will retract back into a little bead with a good gap so that it will take longer to reform.
I did it without shorting the leads together as I am using 2 to 5 volts with the base positive so that the junctions do not conduct.
I looked at the research others have done in science labs and they say that the rupture currant for whiskers is 50 to 100ma.
I just start at a low currant and creep it up until the whisker goes.

ReplyQuote
Topic starter Posted : 14/08/2014 4:36 pm
Cathovisor
(@cathovisor)
Illustrious V-Ratter Registered

What I am trying to do is to slowly heat the whisker so that it melts all at once and with a bit of luck it will retract back into a little bead with a good gap so that it will take longer to reform.
I did it without shorting the leads together as I am using 2 to 5 volts with the base positive so that the junctions do not conduct.
I looked at the research others have done in science labs and they say that the rupture currant for whiskers is 50 to 100ma.
I just start at a low currant and creep it up until the whisker goes.

Out of scientific interest Ref, have you ever opened a repaired transistor up to see what's happened, or is the act of examination likely to destroy the evidence?

ReplyQuote
Posted : 14/08/2014 5:04 pm
Refugee
(@refugee)
Famed V-Ratter Deactivated Account

I believe that NASA opened one up with a pipe cutter without destroying the whiskers.
I am estimating the amount of heat being applied to the whisker and its physical size.
Looking at how fuse wires blow at differing currants also gives a larger scale model of what is going on.
Too much currant will deposit a layer of metal over the inside making success less likely.
It needs to rupture and retract back to two little beads of tin with as big a gap as possible and no vaporized tin.
This is why I use a very low voltage to prevent arcing.

ReplyQuote
Topic starter Posted : 14/08/2014 5:22 pm
Terry
(@terrykc)
Famed V-Ratter Rest in Peace

It's pointless as you are not removing the source of the tin. The whiskers will grow back.

I think most would disagree with you Michael!

The whiskers grow very slowly so, if you can melt a suitable sized gap in the errant whisker, you may get years more service out of an otherwise useless device.

The problem is that you don't know how close to the vital parts any of the other whiskers in the can have grown. That is the gamble because that can considerably shorten the time to the next failure.

I believe that NASA opened one up with a pipe cutter without destroying the whiskers.

That would be Mr Stenning's contribution to the discussion on why satellites fail in orbit ...

http://nepp.nasa.gov/whisker/anecdote/a ... index.html

... Out of scientific interest Ref, have you ever opened a repaired transistor up to see what's happened, or is the act of examination likely to destroy the evidence?

NASA seemed to manage this quite neatly - perhaps it might work best with a pipe cutter in a suitable jig? If you can neatly slice the lid off, then scrape round the inside edge of the can to remove all the existing whiskers, then seal it again, I don't see why it shouldn't work as you have no need to disturb the active parts of the device.

Re-sealing it again afterwards - it must be light proof, rather than a hermetic seal - might be a problem.

Can anybody recommend a solvent which could be used to wash off the silicon grease? Then a dab of nail varnish or something similar would seal the active parts from both light and the air quite effectively ...

When all else fails, read the instructions

ReplyQuote
Posted : 14/08/2014 6:02 pm
Refugee
(@refugee)
Famed V-Ratter Deactivated Account

Perhaps after cleaning and refilling with silicon grease with a clean gap at the cut edge the transistor could be stood in a small tray of epoxy.
Once set the excess can be broken away and the bit in the top of the can painted black.

ReplyQuote
Topic starter Posted : 14/08/2014 6:43 pm
valvekits
(@valvekits)
Prominent V-Ratter Deactivated Account

So do I: although I find Ref's method a bit subtle - I was going to try a charged electrolytic, C-B-E shorted together and connected to one side of said charged electrolytic, shield to the other... :bba

I would love to see Chris's interpretation but fear Hunty would run a mile!

https://www.radios-tv.co.uk/?q=Hunty

Eddie

ReplyQuote
Posted : 14/08/2014 7:21 pm
crustytv
(@crustytv)
Vrat Founder Admin

Actually Hunty is quite an evil little blighter when it comes to tormenting iclkle transistors. Scroll to the bottom of the link above and you will find the new toon. :thumb

Crusty's Collection: Read the repair blogs
Crustys Youtube Channel: If you want to follow me on Youtube, please consider subscribing
Vrat FaceBook: Follow us

ReplyQuote
Posted : 14/08/2014 11:06 pm
Cathovisor
(@cathovisor)
Illustrious V-Ratter Registered

:bba :bba

ReplyQuote
Posted : 14/08/2014 11:21 pm
Anonymous
(@anonymous)
Illustrious V-Ratter Deactivated Account

I like it. (cartoon)

I have never had an AF11x transistor fail to recover and work reliably back in the circuit after zapping it.
There was one exception and that was my KB XR51, but There was something else wrong there as I could not get the FM to work properly even with new transistors.

The method I have used on what must be at least a dozen transistors now is simple, remove TX from circuit, solder e b c together and connect my bench power supply to the ebc and screen. I then turn the transistor round 90 degrees and flick the side hard with my finger to try and jolt any close whiskers to touch and fuse.

Michaels correct the whiskers could come back, but who knows how long, it could be another 1 to 30 years.

If the transistors have been very difficult to get to like in some of those Mullard modules, then I would normally replace with an AC125 or similar.
Mike

ReplyQuote
Posted : 15/08/2014 1:00 pm
Lloyd
(@lloyd)
Noble V-Ratter Registered

I opened one of these up a while back, just out of curiosity, it refused to work after zapping it. I discovered by accident you can pop the can off by heating the case around the bottom with a soldering iron. once opened, it was full of 'snot' which must have been the silicon grease, and this still had the tin whiskers visible (even after heating to over 400 degrees!), I washed the actual transistor in IPA to clean off as much of the old snot as possible, then just wired it back in circuit as it was, it surprisingly worked! I cleaned out the case and stuck it back on with fresh solder. Last time I checked it was still working in my Hacker Sovereign :bba

Lloyd.

ReplyQuote
Posted : 15/08/2014 8:09 pm
Boater Sam
(@boater-sam)
Reputable V-Ratter Registered

Heck, EHT, thats a real sledgehammer!
Glad I wasn't there, sparks frighten me. Still, if it works, OK. How do you avoid the header flashing over?
I have used a gas spark lighter ( piezo ) with good success, and never had a re-growth yet.
I have also de-canned some, usually for the oscillator transistor, as I have found silicon substitutes frequently don't work in this position.
Boater Sam.

Boater Sam
BVWS Member

ReplyQuote
Posted : 15/08/2014 10:14 pm
Anonymous
(@anonymous)
Illustrious V-Ratter Deactivated Account

I opened a faulty one up once to see the whiskers inside, my plan was to use a blow lamp on the end of the can, hoping that the heat would travel quickly around the can and melt the solder, before it travelled through the grease to the germanium. The can came flying off with a bang and flew across the room, it must have been the grease vaporising. The whiskers were quite visible, but the transistor never worked again.
Mike

ReplyQuote
Posted : 15/08/2014 10:38 pm
Refugee
(@refugee)
Famed V-Ratter Deactivated Account

That blow torch was even more of a sledge hammer as it boiled the silicon grease and made the poor thing go pop :cch
It looks like most of them are recovering though :aad
The one I tried in circuit had to come out and it worked again after my out of circuit treatment but it never went back into the set as one of the leads broke a little bit too short to accommodate a ferrite bead that was there in the original set I was fixing at the time. I had a Newmarket branded one that was NOS with long leads that got it sorted :thumb

ReplyQuote
Topic starter Posted : 16/08/2014 1:58 am
valvekits
(@valvekits)
Prominent V-Ratter Deactivated Account

Interestingly, at the end of a 1951 video article mentioned in a TV thread by Steve, there is another article by Bell Telephone Labs describing a new puzzling phenomenon namely Whiskering Zinc. (Page 50/88)
I hadn't realised the problem had been known for so long and begs the question how come they still ploughed on with zinc and tin plate coatings regardless?

Eddie

There is an article "Video Levels in TV Broadcasting" that covers early American practice in the magazine "Tele-Tech", 1951 August issue, page 45ff.

This is available on-line at:

http://www.americanradiohistory.com/Tel ... _Guide.htm.

Cheers,

Steve

ReplyQuote
Posted : 26/08/2014 2:30 pm