[Builder Brad]

Help!

I have recently noticed that the LHS channel in my system seems a little "dead" compared to the other side. To investigate I have switched around the RCA leads until I narrowed the source problem down to the GK1 R, this was a good thing for me as I knew that the fault finding from there would be really easy, and if I got really stuck Hugh is allways there when I really make a mess.

Having looked inside the case I immediately noticed that the Tubes looked different, I dont remember seeing a difference when I built the unit, and I am sure that I would of noticed. One of the tubes has white letering on it and the other one is completely clean, physically they look identical!.

First thing I tried was to switch the tubes - and the problem seemed to move from left to right - for about 30 seconds, and then my system exploded!!!!. well it sounded like it to me standing right next to one of the speakers when the LHS of the volume port shorted.

with everything disconnected I tested the volume pot resistances and found that one 1/2 of its output jumps from 300 ohms to 36k ohms, talk about pump up the volume!

What has happened?

ALSO..

should I worry about the valve fairy?

Is this a good opportunity to cahnge the pot for a stepped attenuator? I am not concerned with the Rs Remote volume control.

[AKSA]

Hi Brad,

I shall answer later tonight;  sorry about delays, frantic at present!!

Hugh

[AKSA]

Brad,

This looks to me like the failure of the pot.  I am not sure why, but you might check if there is any dc voltage across it.  Disconnect the output of the preamp, switch on, then measure voltage across the outer leads of the two pots.  There should be no more than 10mV TOTAL DC across these terminations.

It's possible that DC has found it's way into the pot either from the SS section (input), which could be faulty, or from the tube grid, which sits around 45V.

The cap, C19, is polystyrene, and if heated too much on installation can sometimes fail in precisely this manner.  To check, pull of one lead, and measure with an ohmeter, but first check there's no voltage across the pot.

Hope this helps,

Cheers,

Hugh

[Builder Brad]

Hi Hugh,

cheers

you are spot on!

There is 1.5v across the faulty 1/2 of the volume pots outer terminals.
does this indicate the failure of c19, or will I need to investigate further?

I have found a 10k stepped atenuator on e-bay here

http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=110029197442&sspagename=ADME%3AB%3AAAQ%3AUK%3A1&rd=1

I know that 22k is ideal, but do you think this will be ok

[AKSA]

Brad,

This could indicate that the SS section of the GK1 (left channel) has a fault, and is applying voltage to the pot.  Over time, this voltage has worn down the pot and damaged it.  We need to isolate and repair the fault in the SS section.

First up, disconnect the output from the SS section which leads to the top of the pot.  Is the 1.5V still across the pot terminals?  If yes, the problem is actually coming from the tube, and C19 will need replacement.  If the voltage has now vanished, then it's coming from the SS section.

If the voltage has vanished, indicating a SS section fault, measure the voltage across R12, which feeds the emitter of T4.  It should be 1.25V +/-0.1V.  This is an analog of the current through the output stage, in this case 12.5mA +/-1mA.  The green LED should be glowing, of course, indicating current source (T4) operation.

If this is OK, we've reached first base and the output stage is probably OK, pointing to the the diff pair.

A simple replacement of the diff pair should fix this.  It's likely it's been damaged by a savage input spike;  they are very low stressed devices in normal operation.

The stepped attenuator you selected should be fine but first we need to find what is happening which caused the fault.

Hope this helps, tell us your results as you find them!

Cheers,

Hugh

[Builder Brad]

Hugh,

all done, when I removed the ss output connection to the volume pot there was a very small change to the voltage across the remaining 2 connections. I measured 1.5v before and 1.7v after I removed this connector.

does this indicate the failure of C19, as described? I think heat damage is the likeliest reason if this is the fault

you will be spot on with this Hugh, there are minor scortch marks of the surface of C19 from my attempts to solder Cat5 wire onto the adjacent wiper pins.

[AKSA]

Brad,

Yes, you've found it.......  C19 is sprung.

These polystyrene caps are preternaturally sensitive to overheating.  They often fail in insidious ways, like this one.

These are Suflex types, made in your bejewelled isle, so if you can find any of 10nF to 22nF to replace it, the problem should be solved very easily!

I'm glad we found it.  BTW, what is the offset at the SS section output, at the collector of T3?  It should be less than 10mV, but you'd have to measure it disconnected from the pot.

Cheers,

Hugh

[Builder Brad]

those polystyrene caps are really hard to obtain, even if they originated here in the UK!, and as I will only be purchasing 1 or 2 I doubt if any supplier will be that interested what with the minimum ordering policys ect.

I know this is crazy, but would you be happy to sell me half a dozen from your stock?

[Felipe]

Brad,

I hope i am not completly off here, but Maplin stores a lot of caps . Hugh will correct me if i am wrong, but try these out :

http://www.maplin.co.uk/Module.aspx?ModuleNo=26583&criteria=suflex%20capacitor&doy=11m9

For 0.24£ a piece, you couldnt go cheaper than that ! I dont know about shipping charges, but within UK ou should be fine.

hope this helps,

Filipe

[Rom]

Brad,

Try RS component  part # 113-409 to start with. I think this is the one you looking for , I'm not sure about the voltage rating, no diagram in front of me to check, so I'm speaking from just memory.

Felipe,

I think that's another kind caps on maplin.

cheers
rom

[Felipe]

oops..sorry. I said i was not sure...

Yes RS is a viable one too....

[rabbitz]

What actually does C19 do?

I run a non tube GK-1 and it certainly doesn't like C19 which I tried after the pot.... the 100R with a 100K added going to ground works fine.

[AKSA]

Hi Pierre,

Welcome to the thread!

The tube is a self biased cathode follower (CF), with grid at 46V DC.  The signal has to come off the wiper, which is at 0V DC, and be level shifted up to the grid.  This can be done with SS devices, but they introduce needless additional processing, so I simply use a cap.

The Zin of the CF is a calculated 4.31 megohms.   This is so high that a very small cap may be used without loss of bass;  I specify 10nF, which is at the upper practical limit of the cylindrical polystyrene chosen for its very good sonics.  To drive anything substantial, like a preamp, you'd need a much, much larger cap - more like 1uF, indeed the same as the Auricap used at the output of the tube.

But the downside of the polystyrene caps is their extreme sensitivity to soldering heat.  Overdo it, and they eventually fail - usually about six months later.  Nuttin' for nuttin'......

Cheers,

Hugh

[rabbitz]

Thanks for the explanation Hugh as most of this is beyond my old brain.

[AKSA]

Peter,

Your brain and my brain were minted in the same vintage!!  I find as I get older that the memory fails, but the logic is as good as ever.  I dream up some wunderkind new circuit, look at it with smug satisfaction, talk airily and self-importantly to those who will listen, then lose my glasses, and after a visit to the loo, forget to do up my trousers properly.  (Good thing I'm no threat to womenkind!!)

How many times have I driven/rode down to the Post Office with packages to deliver, and forgotten my chequebook for the shipping charges?

Last night, after a heady fume session in the workshop building modules, I forgot to switch off the iron and the desolding station.  That hurts;  it does the tips no good at all.

Peter, good advice - don't give up!  Tell your family - as we age, we become children again, and THE CHILDREN ARE NEVER WRONG!!

Cheers,

Hugh

[Builder Brad]

Fantastic!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!


I have just replaced C19 capacitor with a 10nf polystyrene one sourced from RS and threw out the broken voltage pot, installing one new 20k stepped attenuator as per these specs:

http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=110029197442&sspagename=ADME%3AB%3AAAQ%3AUK%3A1&rd=1

the e.bay user kyc111 is selling these units with Holco and Vishay Dale resistors and he was happy to supply one with a 20k range, even though he does not list this. I asked for this after the auction finished and recieved the item 1 week later.

While I was waiting for the parts from RS components I wired all of the resistors in the stepped attenuator in series and played white noise through them for 4 days using my PCs sound card, in order to burn in every volume step before fitting to the pre amp. Here is a very helpfull page about burning in passive components using your PC sound card:

http://homepages.tcp.co.uk/~nroberts/article80.pdf#search=%22vishay%20stepped%20attenuator%22

my system sounds better than I ever remember, I do wonder if C19 had failed when I assembled the amp. the volume control allways crackled a little so this may be the case. Its hard to tell how much of an improvement is comming from the stepped attenuator, but I am very happy now.

Hugh,

The Left channel SS ofset is 8mv and the right chanel is 0.3mv, both measured with the pot removed.

I would love to get the LHS ofset down to similar values as the RHS