Need help

[Rob11966]

Hi folks, need some words of wisdom -

I have just entered the testing phase for a stereo version of the dynaco MK III that I have made.

OK, heres the thing. Connected the first channel up using EL34 (as I have a couple of old ones hanging around and didn't want to smoke my new KT88's). Cautious power up measuring the bias voltage across a 10 R resistor (adjusting for 0.7v = 70 ma). All the tubes light up, no smoke, no funny smells etc, all looks good. Put in an input signal from CD player through a 100 K pot and fantastic clear sound - a great victory!! All looks good, listing to the music and very pleased with myself. I decide to crank up the volume, goes OK but as I am cranking up the volume, the bias voltage shoot up to about 3v and at 'midrange' volume, I get a sudden abrupt, loud and disturbing buzzing through the speaker and about 2 seconds later, bang, the 3A fuse blows.

Hmm, so I let everything cool down, replace the fuse and give it another quick go. Hey presto, all seems well at low volume - amp working fine. I then shut it down, not game to crank it up for fear of damaging something!

I have attached a circuit diagram. All is the same except I was using EL34's instead of the KT88's. Should also mention I was using a 6R test speaker on 8R tap.

Any ideas?? Maybe just too much current for the EL34's? As I mentioned, sounds great at lower (not quiet but not loud) volumes but the response to cranking the volume was quite spectacular.

Cheers,

Rob

 

[pmillett]

Sounds to me like it's breaking into HF oscillation, maybe at the onset of clipping. Nasty things can happen w/ the feedback loop at clipping...

You really need to look at the output with a scope and I bet whatever it is will be obvious.

Pete

 

[Rob11966]

Thanks Pete,

Some further info - I did try the KT88's. They worked OK and did tolerate the higher volume but as I turned up the volume the bias voltage jumped from 1v resting to over 3v. This was a dramatic jump not a slow climb and represents 300mA cathode current - too much for these tubes and that was just at medium volume. I did not push it any further but have shut the whole thing down.

I do have a scope but I am fairly new and an amateur at trouble shooting, so I will need to be talked through things I'm afraid. I will treat his as a learning experience if you guys have the patience!

Cheers,

Rob

 

[Rob11966]

A bit more info

A little bit more info -

Tested the other channel using the same tubes, and the same power and output transformer - no problems at all. Works fine with acceptable cathode current (as determined by bias voltage) at all volumes.

I guess this eliminates the tubes, transformers and my basic interpretation of the circuit (ie my wiring in one channel is fine). So this leaves -

1. Faulty components - but all brand new (including caps) and tested during construction

2. Incorrect resistor value somewhere - but unlikely as all were double checked and were matched with the other channel

3. A wiring error - most likely but it must be a small problem as the channels were made concurrently (and very carefully) and they are a mirror image of each other.

Now, I need is some advice as to where I should begin looking apart from the obvious (double check everything) which I will begin to do this evening when I get a chance to sit down and look at it properly again.

In the meantime, keep the advice simple, as I am still struggling to keep up with the level of knowledge that most of you guys have!

Thanks again,

Rob

 

[Burnedfingers]

This is probably a stupid question but are the coupling caps new?
Bias pot?

 

[SHiFTY]

Try swapping around the anode leads to the output tubes, you might be getting positive feedback.

Also you could try disconnecting the negative feedback resistor from the output transformer secondary and see if the problem is related to that or not- also check your feedback cap is 750pF (MMF = 1x10-12 or picofarads)

Also double and triple check all your component values and solder joints.

 

[Rob11966]

OK,

Coupling and Bias caps new.

Feedback cap is actually an 820mmf (couldn't get hold of a 750) but definitely mmf. Same as the channel that works OK.

I will swap the anode leads in the output tubes and give it another whirl.

Report to follow

Thanks for the replies so far.

Rob

 

[Rob11966]

Sorry, correction - meant to say coupling caps and bias pots new.

Rob

 

[Rob11966]

Problem solved

Thanks Shifty,

Great advice, Problem solved - In my original test configuration, I had the anode leaps connected the wrong way. Works perfectly well now - both channels up and running. Shifty, I would love a brief explanation of the positive feedback issue if you have time but will be satisfied with having just solved the problem.

I will post some photo's soon. It is an absolute monster - it must weigh 25kg and I have just received the top-plate back from the electroplaters which I had done in 18ct Gold - It's way over the top but I love it!!

Cheers all and Happy Christmas.

Rob

 

[chrish]

Hello Rob,

Glad you have success with your new amp. I have just finished a flea powered single ended 6P15P point to point that is to be a Christmas gift for my girlfriend. I know the feeling of a successful scratch built project! Looking forward to the photos! Unfortunately, based on your previous project, the neatness and attention to detail will put my effort to shame

Merry Chrisrmas,

Chris

 

[ray_moth]

I would love a brief explanation of the positive feedback issue

I'm not Shifty (at least, I try not to be ) but I can explain the problem. Put simply, global negative FB is used on your amp to lower output impedance, reduce distortion, reduce hum and noise, stabilize the gain and broaden the audio frequency response. Another result of NFB is reduction in gain.

A fraction of the signal from the secondary of the output transformer is fed back into the input at the cathode of the first stage. This fed back signal must be out of phase with the input signal for the feedback to be negative.

If you have the connections to the primary (or secondary) of the OPT reversed, the feedback will be positive, with exactly the opposite effects listed above for negative FB. If the amount of positive FB is sufficiently high, the amp will oscillate as soon as it warms up. In your case, the fraction of signal fed back was evidently too low to cause that, but the amp was on the verge of instability and it only took a large enough signal (when you turned up the volume) to send it over the edge into oscillation.

 

[Wavebourn]

Try to increase capacitance from anode of output toob to cathode of the 1'st one. For the beginning take twice higher and if it helps decrease gradually and check.

Also, check if you connected this cap to the proper toob's anode...

 

[Rob11966]

Thanks for the explanation Raymoth, makes sense to me now. Silly mistake, those blue and blue/white striped leads can fool the eye and the brain.

Hi Chris, I am sure it won't put yours to shame. I have yet to earth it properly (temporary earthing at the moment for the testing phase) but when I do it will have a great big thick copper bus-bar thanks to your very kind donation of the thickest earthing wire I have ever seen. I will post a photo for sure.

Have a great Christmas,

Rob

PS Merry Christmas to all of diyaudio including Sy and all the moderators for keeping it going. There is no way I would have been able to successfully complete my amps without the help that I have received here.

 

[SHiFTY]

Good to hear you got it sorted. I have had this problem before, especially when the primary leads are the same colour and you have a 50% chance of getting it wrong and getting 50 watts of howling square waves

Have a good listening session over Xmas!