Author Topic: Advice please  (Read 5301 times)

Offline audiophool

  • Full Member
  • ***
  • Posts: 130
  • Liked: 3
Advice please
« on: March 03, 2011, 12:23:13 AM »
Casting around for any ideas on tuning out/eliminating or tracing the source of a particular problem I am hearing in my amp.
Given that I do not have access to an oscilloscope and cannot afford to either replace it or take it to an expert (if I knew one).
I understand the "can't polish a t**d" theory but I think the iron is pretty good, I have heard it "make magic" and I have had pretty good results so far in improving it by upgrading a number of parts from the fairly ordinary ones used in its original construction.
I know it is possible to get good sound from the right combination of fairly ordinary parts, but it is also possible to de-tune to mask issues such as I am now experiencing.  Whether that is the case now or whether the other improvements have just raised this issue to prominence I do not know.
What I am hearing seems to be an emphasis, resonance, peaking or ringing of a fairly limited range (I assume a range) of frequencies in the upper mid or lower treble, only occasionally obtrusive and most commonly on female vocals.  It is not noticeable to any particular degree on instrumental or orchestral pieces.
The effect is to give a hard, metallic, "edge".
I could possibly find close to the affected frequencies by recording the playback with a microphone and then inverting the source and summing the two in software.  I may try this.
Possibilities that occur to me as sources of this problem seem to be, microphonics - not highly likely as it is not proportionately greater with increased volume, increasing only at the same rate as the volume (all valves have dampers and chassis is on vibration absorbing supports), PSU electrolytic caps (these seem fairly ordinary), PSU switching/ripple or harmonics thereof (PS is SS rectified),  harmonic distortion being emphasised in the feedback loop - possibly due to value choices of components ? - do not know enough about how this works to form any theories.  O/P traffos - if these are the problem I cannot see a way around it.
Hints, tips, suggestions (other than vulgar) please.

Offline kajak12

  • Hero Member
  • *****
  • Posts: 2468
  • Liked: 78
    • http://killerdac.com/forum/index.php
Re: Advice please
« Reply #1 on: March 04, 2011, 12:38:17 AM »
Casting around for any ideas on tuning out/eliminating or tracing the source of a particular problem I am hearing in my amp.
Given that I do not have access to an oscilloscope and cannot afford to either replace it or take it to an expert (if I knew one).
I understand the "can't polish a t**d" theory but I think the iron is pretty good, I have heard it "make magic" and I have had pretty good results so far in improving it by upgrading a number of parts from the fairly ordinary ones used in its original construction.
I know it is possible to get good sound from the right combination of fairly ordinary parts, but it is also possible to de-tune to mask issues such as I am now experiencing.  Whether that is the case now or whether the other improvements have just raised this issue to prominence I do not know.
What I am hearing seems to be an emphasis, resonance, peaking or ringing of a fairly limited range (I assume a range) of frequencies in the upper mid or lower treble, only occasionally obtrusive and most commonly on female vocals.  It is not noticeable to any particular degree on instrumental or orchestral pieces.
The effect is to give a hard, metallic, "edge".
I could possibly find close to the affected frequencies by recording the playback with a microphone and then inverting the source and summing the two in software.  I may try this.
Possibilities that occur to me as sources of this problem seem to be, microphonics - not highly likely as it is not proportionately greater with increased volume, increasing only at the same rate as the volume (all valves have dampers and chassis is on vibration absorbing supports), PSU electrolytic caps (these seem fairly ordinary), PSU switching/ripple or harmonics thereof (PS is SS rectified),  harmonic distortion being emphasised in the feedback loop - possibly due to value choices of components ? - do not know enough about how this works to form any theories.  O/P traffos - if these are the problem I cannot see a way around it.
Hints, tips, suggestions (other than vulgar) please.
have you tried changing valves?
still discovering the link between electronics and audio reproduction.so much to learn and so little time

Offline audiophool

  • Full Member
  • ***
  • Posts: 130
  • Liked: 3
Re: Advice please
« Reply #2 on: March 04, 2011, 01:15:58 PM »
Thanks for the thought Mario, yes, I have, all that did was lose some bass and pull the anode cap off my preferred valve and probably render it useless  :( - I do have another pair of another make, but if memory serves me correctly they also had less bass than the pair I pulled, I know of no subs for the input pair.
I am suspecting that the improvements in clarity I have made to the input signal have shown up the problem.
Input caps have some time to burn in fully yet, that may ameliorate the issue somewhat.
I am coming more and more to think that resolving this will involve an oscilloscope and signal generator and a better understanding of the feedback circuit and the trimming loads on the transformers.

Offline vitavoxdude

  • Beauty is in the ear of the beholder
  • Hero Member
  • *****
  • Posts: 1175
  • Liked: 71
  • Caring and sharing
Re: Advice please
« Reply #3 on: March 05, 2011, 04:10:44 PM »
Hmmmmmmmm, tricky one that.  ::)
I would look for components which do not have copper lead outs first, you'd be suprised.
Valve's biased too high may be an issue.
Feedback loop - eeek,  :o get this wrong and everything is effected - spend lots on the resistor and cap here!!!!!!  :-X
As Mario said, some valves can be less good.
Distortion can give a hard ege to the sound too.
Time to break out the oscilloscope maybe.
Have you tried hooking up another pair of speakers?
How do you know its the amp?
V
We all like different things so lets all agree to disagree and if any common ground is found then worship it.  Mine is the KD hence being present on this forum.

Offline audiophool

  • Full Member
  • ***
  • Posts: 130
  • Liked: 3
Re: Advice please
« Reply #4 on: March 05, 2011, 08:29:05 PM »
Well a little progress even if no resolution in sight.
Found a pair of driver valves I have had a few years and never tried - gems - bass is back with interest and lots of other good things.  Their top end clarity reveals the issue clearly as a resonance or ringing at just one (or a small range) of frequencies.
Some years ago I purchased (as one does :D) a device called a Dakiom feedback stabiliser, at the time I did not find it particularly useful although it had a pronounced effect on a SS amp.  I dug this out and connected it and the problem seemed exacerbated.
Needless to say I unplugged it again.   This did indicate to me that the issue may well be in the feedback circuit and an investigation of its operating parameters might allow adjustment of component values to eliminate the problem.
While it could be the speakers, I do not have another set of equivalent sensitivity, resolution and accuracy against which to compare.  The only viable option might be to wire a pair of headphones to run off the speaker terminals for a reference.
Other suggestions filed for consideration as opportunity arises.  
I find it hard to be patient but the new input caps are big mofos and really should have another few hundred hours to settle in before I change anything else.
Maybe meant to be an opportunity to learn some more about exactly how the circuit is intended to work and the various operating points intended, if that is possible.
Meanwhile a lot of music that does not obviously excite this particular resonance is sounding v nice indeed.
I shall have to be content with that for now.

BTW and apropos of nothing in particular Bach's Art of the Fugue by the New Century Saxophone Quartet is a recent "find" and rather good - IMO
« Last Edit: March 05, 2011, 08:31:59 PM by audiophool »

Offline audiophool

  • Full Member
  • ***
  • Posts: 130
  • Liked: 3
Re: Advice please
« Reply #5 on: April 03, 2011, 11:52:58 AM »
Well, time for an update. (In case anyone is interested)
600 + hours of burn in on the input coupling and input to driver coupling caps has ameliorated most of the noted issue.
A slight increase in value of the feedback resistor seems also to have helped.
I have learned a lot more about how the circuit works although not yet to the stage of being able to calculate the actual or optimal operating points of the valves used.
The remainder of the observed issue is at least no longer seriously "in your face" but more a subtle overtone that can be observed when listening carefully to selected pieces.
More research is indicated but perhaps power supply caps might benefit from a few changes first.
thanks all,
TG

Offline vitavoxdude

  • Beauty is in the ear of the beholder
  • Hero Member
  • *****
  • Posts: 1175
  • Liked: 71
  • Caring and sharing
Re: Advice please
« Reply #6 on: June 23, 2011, 11:22:50 AM »
Well, time for an update. (In case anyone is interested)
600 + hours of burn in on the input coupling and input to driver coupling caps has ameliorated most of the noted issue.
A slight increase in value of the feedback resistor seems also to have helped.
I have learned a lot more about how the circuit works although not yet to the stage of being able to calculate the actual or optimal operating points of the valves used.
The remainder of the observed issue is at least no longer seriously "in your face" but more a subtle overtone that can be observed when listening carefully to selected pieces.
More research is indicated but perhaps power supply caps might benefit from a few changes first.
thanks all,
TG

OK so you tried your super dupper feedback fixer and things changed for the better?
So go all out and insert only the finest devices here in the way of caps and resistors as this is probably the single cheapest place to do so followed closely by the input decoupler cap.
Is this a pushpull amp or triode?  What valves are you using?  It sounds like an elaborate design allready.  Is it a one off or an off the shelf jobbie?  I'd seek first info from its designer as they know better than anyone where money was saved and how incremental improvements can be rung from all thier months of optomising etc.  We are guessing without the circuit!
V ;D
We all like different things so lets all agree to disagree and if any common ground is found then worship it.  Mine is the KD hence being present on this forum.

Offline audiophool

  • Full Member
  • ***
  • Posts: 130
  • Liked: 3
Re: Advice please
« Reply #7 on: June 23, 2011, 12:05:56 PM »
@V
thanks for thinking on this again, the state of play ATM is that the Dakiom made things worse and was quickly removed.
I have contacted the designer, who, since this was a prototype he built some years ago (the supplied hand drawn circuit diagram has a number of differences to the actual circuit as built) naturally has little recall of the actual circuit and no recommendations to make and without 'scope traces could offer no ideas.  I suspect he is, quite reasonably, dubious of my diagnosis.
I have changed a number of other components in the front end for better quality, all to good effect.
The issue is much reduced and the overall character much improved, changing all the cathode bypass caps made quite a change for the better.
I have yet to change the power supply 'lytics (finances rather stretched ATM and Jensens are not cheap).
The overall design is cascode input stage with cathode follower driver stage driving the screen grid of a pentode.
A couple of aspects of this arrangement have me puzzled, the cascode uses a resistor connecting the plate of T1 to the cathode of T2 and the grid of T2 appears to take its reference from the cathode of the output tube (through resistors - I also replaced these).
The main feedback circuit to the cathode of T1 has silver micas and I replaced the generic carbon films with Takmans - at least for the time being.
Not being able, ATM, to replace the plate supply 'lytic on the cascode stage, I bypassed them with small polystyrenes, not sure this was a good move, thinking of replacing them with oil fills, though with 10uf those are rather hefty.  Still, there is still plenty of room, just have to fit some hangers.
It may be that what I am now disliking is cascode phase distortion or even the pentode output signature.
It may be possible, with more knowledge, to detune/lower the output (it is a deal more than I need) to reduce this, or to vary the load in parallel to the O/P transformer on the plate of the output tube.
More research needed  :-\
Will change the remaining caps and replace the input wiring when funds allow and then take it from there.