Testing an old amp.

Forum for discussing the construction, renovation and testing of equipment
Post Reply
JoeB
Posts: 27
Joined: Thu Aug 28, 2025 9:38 pm

Testing an old amp.

Post by JoeB »

15 or 18 years ago, I built myself a little stereo amp for my own listening. It outputs around 8 watts RMS per channel, and is actually very loud into my loudspeakers. In fact its too loud at full tilt. Its been working perfectly up untill a couple of months ago when it started crackling when adjusting volume.
This is a fairly common problem and can happen to any amp, regardless of what topology it uses, especially when it ages somewhat and gets used fairly often. This little amp is vacuum, so it has somewhat higher voltages than those found in most sand amps. Higher voltages stress components to a somewhat larger degree, even if good quality and correctly specced parts are used. So I suspected maybe a coupling capacitor or two might have gone leaky, which usually results in crackling when adjusting volume controls, or for that matter, any potentiometer within an amplifier. This is then a story of what I do when testing and repairing an amplifier with noisy controls. This applies equally to Hi-Fi or guitar or even PA amplifiers. By PA I mean the types that you normally cant understand at the railway station or airport.
Stereo1.JPG.jpg
Stereo2.JPG.jpg
Stereo4.JPG.jpg
Note the authentic vintage dust that can be purchased on Ebay :lol: . Yes, there are some shonky ebay operators.

So when examining the underside of the amp, I noticed that the four cathode bypass capacitors on the cathodes of the output valves had popped their corks. By that I mean, for some reason they had overheated, and thats generally caused by over voltage, or being too close to a heat source. The last is not the reason, as I spend a lot of construction time ensuring that nothing can cause this. i.e. no component that generates heat is anywhere near electrolytic capacitors.
Said capacitors were RIFA 100 uF @ 25 volt units made in Sweden, cost an arm and a leg and fail very rarely. The bias voltage "normally " found on a 6BW6 (or 6V6 in a nine pin jar ) is around 18 volts, give or take. These have been removed in the pic, as they were leaking electrolyte all over. It can be corrosive, and is defiantely not healthy to handle, although not as bad as some old electrolytics which contain PCB.

( PCB = Poly Chlorinated BiPhenols which is now banned the world over because it DOES cause cancer in humans ).

Yes, read that again and remember it.

New electrolytics will be fitted. These are different to what was origionally in place, but are close enough and no sound change will happen. New caps are Siemans low leakage long life 68 uF 40 volt units. In fact its getting quite hard to source modern RT series caps. RT means they have a tail out both ends of the cap suitable for point to point wiring, rather than RB series which are designed for printed circuit board mounting, and have both legs out one end of the cap.
Also they are uselessly short and dont fit back where the origional cap was removed.

I also did some clean up of my shoddy work when I built the amp origionally.

Most notably the DC heater supply for heating the valves in the integrated phono stage. It was cobbled together for testing, and as it worked marvelously well I ledt it thus, but was a bit shamed when I saw it again.
In the photo Stereo4 right on the back panel facing the viewer is the end on view of two 10,000 uF 16 volt capacitors, and two five watt wirewound resistors. One of those resistors was "hanging " on full length lead length, as it was just dagged in there to see what heater voltage I obtained by using it. So after testing for 15 or 18 years its finaly been fitted in its final position.


More tomorrow.

Joe
JoeB
Posts: 27
Joined: Thu Aug 28, 2025 9:38 pm

Re: Testing an old amp.

Post by JoeB »

I just discovered the forum doesnt like inline photos.
Here they are again.

JB
Attachments
Stereo4.JPG.jpg
Stereo4.JPG.jpg (252.45 KiB) Viewed 19071 times
Stereo2.JPG.jpg
Stereo2.JPG.jpg (234.09 KiB) Viewed 19071 times
Stereo1.JPG.jpg
Stereo1.JPG.jpg (238.9 KiB) Viewed 19071 times
duncanamps
Site Admin
Posts: 43
Joined: Tue Aug 26, 2025 9:26 pm

Re: Testing an old amp.

Post by duncanamps »

Hi Joe,

Nice amp, it looks well laid out and well made. And nice to see 6BW6's in use, don't often see them. Spookily, I've just bought 3 new 6BW6 for my little Codar AT5 - one to replace the RF power output, one to replace the modulator, and one as a spare 'cos they do run a little hot... :D

Regards,
Duncan
Attachments
PXL_20251105_110854921 (Small).jpg
PXL_20251105_110854921 (Small).jpg (128.73 KiB) Viewed 19067 times
PXL_20251105_110544135 (Small).jpg
PXL_20251105_110544135 (Small).jpg (198.47 KiB) Viewed 19067 times
JoeB
Posts: 27
Joined: Thu Aug 28, 2025 9:38 pm

Re: Testing an old amp.

Post by JoeB »

I dont know IF this will work. I just discovered that the forum wont let me use PDF files. so again, a test.

JOe
Attachments
Screenshot from 2025-11-06 07-51-04.png
Screenshot from 2025-11-06 07-51-04.png (235.79 KiB) Viewed 19059 times
JoeB
Posts: 27
Joined: Thu Aug 28, 2025 9:38 pm

Re: Testing an old amp.

Post by JoeB »

seems pretty useless.
HELP Duncan,
why cant I submit PDF files ?.

Joe
JoeB
Posts: 27
Joined: Thu Aug 28, 2025 9:38 pm

Re: Testing an old amp.

Post by JoeB »

So LOTS of mucking about produces this:.


Another test.

Joe
Attachments
Circuit.jpg
Circuit.jpg (250.92 KiB) Viewed 19057 times
JoeB
Posts: 27
Joined: Thu Aug 28, 2025 9:38 pm

Re: Testing an old amp.

Post by JoeB »

OK, it works and now everybody can see the circuit. Its OK for discussion, but will fall apart if one tries to enlarge it.
( Anybody that wants the full article can email me and I will send a complete PDF file ).
This circuit has not changed greatly since the mid 1930's as far as design goes. It does have negative feedback, which is perhaps newer than that.
Negative feedback consists of a single resistor running from the output transformer, to the cathode of the gain stage, in this case a 6AU6.
This case its 3000 ohms.

( just as a note, the value of 3000 was very common in the 1930's, and also up to about 1960, when the E12 series of resistors was produced. Prior to that we had " dogbone " resistors that only had 6 values per decade. The colours were read from left to right,
Body. colour
End. of the body which was different
and Dot, a blob of paint applied to show the multiplier.
In this particular case, Orange, orange, orange. Simple huh ?.
Further reading for newbies:
https://www.radioremembered.org/rescode.htm )


The circuit is as explained, as "old as the hills ". That doesnt mean its no good. 1954 was the rise of Hi-Fi, the name actually coming from DTN Williamsons circuit of 1948, when he dicovered, or invented negative feedback.
His origional design, using his own output transformers had a high frequency response extending to 100,000 cycles,
or Hertz as they have named it today. His actual amplifier is still very hard to beat today using convetional circuitry, although I must admit modern MOSFETS can beat it easily, BUT they were not around 75 or 80 years ago.

The basic circuit is very standard, by having as the first stage of the left of circuit, a voltage amplifier. This is a very simple circuit using a Pentode that has been strapped to make a Tetrode ( by tying grid number 3, to the cathode ) and making a large amount of gain. But making gain can lead to instability where the amplifier wants to oscillate and that means it has lost control, to counter this, the 100 pF capacitor connected to the plate of the first 6AU6 and connected to ground, cuts off the topmost frequencies, to prevent that oscillation.

Following this stage is the phase splitter/ phase inverter which splits the signal into two parts, the positive part of the cycle, and the negative part of the cycle. This is achieved by using another 6AU6, in this case strapped as a triode.
i.e grid three is strapped to the cathode, and grid two is strapped to the plate. In this configuration the valve acts very like a 12 AU7 with much less gain, but more power output. This power output is used to drive the output pair consisting of two 6BW6's, which provides about 10 watts RMS of useable power. This ten watts is quite low distortion and of wide bandwidth. The plates of the 6BW6's are connected to the output transformer and thence to the speaker.
The top 6BW6 only amplifies the positive half of the cycle, and the bottom 6BW6 amplfies the negative half of the cycle.
These two halves are joined back together in the output transformer.

That makes four only active elements, that I dont think can be done with solid state devices, of ANY variety, and still supply the power, distortion levels and frequency response that this very old circuit can achieve.
duncanamps
Site Admin
Posts: 43
Joined: Tue Aug 26, 2025 9:26 pm

Re: Testing an old amp.

Post by duncanamps »

JoeB wrote: Wed Nov 05, 2025 9:59 pm seems pretty useless.
HELP Duncan,
why cant I submit PDF files ?.
Hi Joe,

Forums are generally not that great for including or handling non-text items. I would suggest the best way to deal with it is to store the file on your website if you have one, or on a storage medium like Dropbox, Onedrive etc. and provide a link to it.

There's Onedrive Basic which gives you 5GB for free, there's others out there too. There used to be a free Dropbox, not sure if they still do it.

Here's an example link to one of my PDFs -> https://www.dropbox.com/scl/fi/cql4wz8r ... od9qm&dl=0

I'll put something in the FAQ when I get a moment, thanks for bringing this up.

Regards,
Duncan
duncanamps
Site Admin
Posts: 43
Joined: Tue Aug 26, 2025 9:26 pm

Re: Testing an old amp.

Post by duncanamps »

Forgot to mention the bit about PDFs, I believe they can contain code which is maybe why the server is rejecting them. It would be the same for .EXE files etc.
Post Reply