Need Your Assistance - Audire Andante

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2004
Paid Member
Hi Greg,
Sorry to be so far below your knowledge levels, but I am very much enjoying the opportunity to learn that you are both providing.
No reason to be sorry about this. We all started from a point where we didn't understand and had to learn. You're asking questions and that is a lot better than not. Never be sorry for asking questions.
I will then be more prepared for replacing them as Chris has suggested if the adjustments don't work.
By all means, take measurements at the starting control positions. Mark them too and take a picture so you don't have to rely on memory. These steps don't take much time and I follow them too when a situation begs for experimentation.

The controls should be able to compensate for unbalanced pairs, if not they are so far from balanced that they would be considered defective. One mark of an unbalanced pair that has been compensated for would be an offset that changes as everything warms up. The other issue would be higher distortion. It is ultimately always better to have well matched pairs in a circuit designed using a balanced matched pair. Never forget that a circuit that is designed using a diff pair (or long tailed pair), the first and overriding assumption made by the designer is that the diff pair is perfectly matched. This is obviously a very optimistic hope, and the reason why you will see either DC offset adjustments or DC servos.

If possible, try to trace the signal path from input to the output jack. I'm hoping you will run into a capacitor intended to block any DC voltage from the tone control circuit from entering any other component. This is especially true gfor any adjustable control like a tone control or volume/ balance control as any DC potential will make the control noisy as if if needed cleaning. So DC offsets are very unwelcome in any audio circuit. The point is, if there are capacitors to block DC, they often can get leaky, allowing some current to create potentials where they shouldn't be. Even tantalum capacitors suffer from this failure mode.

So I agree. Go step by step taking pictures as you go along with your notes. It will be an excellent "lab" to do and learn from.

Best, Chris
 
Gentlemen: Some progress!

First, Chris - I had previously started to trace the input to output path on a photo of the bottom of the board, but didn't finish proofing that against the top components, so can't yet be sure about which cap is there to block DC, but given the comments by as_audio and yourself, I'll be sure to complete this task.

Not sure it is definitive, but maybe one or both of you will find it more diagnostic.

I pulled the new electrolytic 470uF caps I'd installed and they test good for capacity and leakage. Reinstalled.

Did a 3rd round of testing all of the caps that would appear to be DC blocking; those are new 470uF electrolytic in the MM section, 7 in the MC section, and one film cap in the line section, one in the MM section, and 4 in the MC section. There are Mica caps in all sections but I didn't test test the ones in the line or tone sections again (did so originally).

Next I hooked up my Fluke to the output and measured DC on the right channel. After marking and photographing all trimmers (I learned this long ago building Lotus Twin Cam engines and other motors, as well as all of the electronic gear I've been working on, Chris :)) I adjusted the right channel tone section trimmer. Offset went from 890mV to 60-75mV, but was very jumpy. I then adjusted the line trimmer and eventually got offset as low as 9mV, but constantly fluctuating between 7.7 to 9.5mV.

Total adjustment for the tone trimmer was 1-4/5 turn. Adjustment for the line trimmer was 1 turn.

Also, adjusting the treble and bass controls caused offset to fall, then climb.

I then used a tiny amount of cold spray on the tone section dual matched pair, and offset climbed to 50mV, but then stabilized for about 30 seconds. I then sprayed the line section pair, and offset climbed to 90mV, then dropped quickly to 45mV, with offset stabilizing for about 30-45 seconds. I repeated this 3 times with similar results.

I then used cold spray on the trimmers, other transistors, and the caps. Small changes to offset occurred with spray on some of the transistors, but nothing dramatic, and I didn't record those changes.

Moving to the left channel, I made a small adjustment on the tone trimmer, and offset went from 260mV to 0.015mV and fluctuated only slightly between 0.015 and 0.020mV.

I used cold spray on the left channel pairs and offset rose 20-50mV then settled after 30 seconds or so. DC readings on the left channel are much more stable.

Could this point to the left channel simply needing adjustment to account for aging of the components? And could it be that the transistors in the right channel are indeed misbehaving, given the profound effects from the cold spray and the constantly fluctuating offset that never went below 7.5mV?

Thank you very much for your continued help!! First time I've felt like progress is being made.

Cheers,
Greg
 
Last edited:
Administrator
Joined 2004
Paid Member
Hi Greg,
Yes, I think that is certainly a good set of assumptions. When a pair isn't matched, temperature changes will cause a lot of shifting relative to the same circuit with matched diff pairs. You might need to buy more pairs than you actually need just so you can avoid using a part that is substandard. If they all turn out to be great, then change the left channel (but I'm not in a hurry to change those).

-Chris
 
Administrator
Joined 2004
Paid Member
Hi as_audio,
Any good reason for claiming one or two dual fets need replacement ? Thanks.
I think you are asking me why I suggested to replace them. If you read the posts, I was clear on that. Greg's later measurements and observations confirmed what I was thinking.

Now, I did ask you why you didn't agree and you failed to answer that. You simply asked me to explain myself. Now it is your turn.

-Chris
 
" I hooked up my Fluke to the output and measured DC on the right channel ...
adjusted the right channel tone section trimmer. Offset went from 890mV to
60-75mV, but was very jumpy. I then adjusted the line trimmer and eventually
got offset as low as 9mV, but constantly fluctuating between 7.7 to 9.5mV."

I think I wrote "try a sensible adjustment here, for each of the four amp stages
and checking their respective outputs for dc" and "take care of the tc offset first".

This means at the tone control output first and line (main output) after.
 
Last edited:
Administrator
Joined 2004
Paid Member
Hi as_audio,
He measured between the base connections of the diff pair. That measurement should be within very low mV between the two bases. The fact that there was a large voltage between them tells me they are not in balance. Even creating an offset with the trimmer would still show the bases to be very close to each other if they were truly closely matched. I also knew it would be possible (or should be) to correct for the DC offset he was seeing. He did correct the offset before noticing the thermal behavior. Thermal drift is the trademark of unmatched diff pairs.
This means at the tone control output first and line (main output) after.
Well, assuming it is a good design, there ought to be DC isolation between the stages, and since there is a volume control before the amplifier normally, you shouldn't see much if any interaction between the two stages. Normally we would adjust the output stage first (preamp or power amp), then go after the various stages needing DC offset correction. In actuality, the tone stage shouldn't need an offset control. If it were designed properly, it really only needs a capacitor in the feedback leg to common (ground). Using matched pairs should give you less than 5 mV worth of DC offset. The same really holds true for the output section, no trimmer and less than 5 mV for DC offset. If you have trimmers, a balanced diff pair should give your DC null around 1/2 way, or mid-point on the trimmer. Not unless the DC resistance seen by each half of the diff pair differs. If that is the case (tone controls), then even a trimmed balance will vary.

Just for completeness, I did also say that the plan to adjust things was a good plan, but that unbalanced diff pairs would need to be replaced. I also decided to agree with you to work on the tone control stages first.

-Chris
 
" I hooked up my Fluke to the output and measured DC on the right channel ...
adjusted the right channel tone section trimmer. Offset went from 890mV to
60-75mV, but was very jumpy. I then adjusted the line trimmer and eventually
got offset as low as 9mV, but constantly fluctuating between 7.7 to 9.5mV."

I think I wrote "try a sensible adjustment here, for each of the four amp stages
and checking their respective outputs for dc" and "take care of the tc offset first".

This means at the tone control output first and line (main output) after.

Oh. I didn't understand this properly. I am not a professional, but a dedicated amateur trying to do good work, and with grateful appreciation for the help I'm receiving.

I measured the output jack, not the respective outputs. I understand now that wasn't according to your specific instructions. I also was not clear on what constitutes "sensible", but it sounds like you are indicating that my adjustments were not.

What I can reiterate is that the right channel adjustment to the tone trimmer was much less effective, measured at the output jack; that the offset fluctuated much more with the right channel; and that application of cold spray (my own initiative) caused the right channel transistors to stabilize for 30+ seconds.

Sorry if my inexperience is frustrating to you, and thank you for taking the time to comment.

Best,
Greg
 
He measured between the base connections of the diff pair.
There are no base connections.

That measurement should be within very low mV between the two bases. The fact that there was a large voltage between them tells me they are not in balance.
This may be a little bit different for gates of a jfet pair.

Even creating an offset with the trimmer ...
An offset is balanced, not created, if done properly.

Thermal drift is the trademark of unmatched diff pairs.
I think this is not true, it is a characteristic of the circuitry.
With a circuit like this you can expect to align the residual
offset to a few millivolts which should not change much in
thermal equilibrium (but it will if using freezer spray).

Well, assuming it is a good design, there ought to be DC isolation between the stages, and since there is a volume control before the amplifier normally, you shouldn't see much if any interaction between the two stages.
We leave the design as it is. Interaction between tone
and line stage was apparent when pushing "tone in".

In actuality, the tone stage shouldn't need an offset control. If it were designed properly, it really only needs a capacitor in the feedback leg to common (ground).
I think I explained the reason before why the tc does not have this cap
(how do you combine bass boost with a gain of one below some low
corner frequency ?).

Using matched pairs should give you less than 5 mV worth of DC offset.
A dual jfet is assumed to be made a matched pair.
I see no hint that some of these may be defective.

If you have trimmers, a balanced diff pair should give your DC null around 1/2 way, or mid-point on the trimmer.
We can not be sure about the layout of the dc balance trim
and it is not possible to see the midpoint using an opaque
ten turns trimmer.

Sorry gentlemen, I am not in a position to write lengthy essays in this thread.
I am pretty sure that the preamp is in working condition, shown by the screen
pictures, if no subsequent harm happened due to recapping and probing around
and it only needs some dc offset realignment.

OP may even have some distortion check function on his Rigol scope to confirm
proper operation.

However I will ship one dual jfets or two if it is still assumed these are required.
 
How could I know how much adjustment you need ? Keep cool please, said no wild rotation allowed.
These are ten turns trimmers. I can not see the full schematic on your pictures.

I wasn't asking for how much to adjust, but how much is "sensible", or "careful", or not "wild rotation". The right channel required 2.5 turns on the tone and 1 turn on the line trimmers. I don't know if this qualifies as any of those descriptions.

Cold spray was used to see if the noticeable fluctuations in offset changed (yes), not to see if offset changed, which of course I expected.

This preamp had high offset prior to recapping and no damage was done during recapping. Work has been done on a static grounding pad and with wrist strap. I held off probing until receiving advice from you both, and unit was malfunctioning before any probing.

I'll post a couple of photos that show the tone and line amp sections with traces visible through the board.

We have two knowledgeable, talented folks debating quite a few details above my level, but I'm trying.
 
This might show the traces and layout better.

Push button switches on left are input selectors.
Push button switch on right closest to bass pot is tone defeat.
Input jacks on upper left.
Output jacks far upper right.
Signal in to selectors across board in front of pots (trace not in photo) to tone defeat switch, etc.
Tone Out = to volume, Tone In to tone controls
Pots: Volume left, T middle, B right
Signal out to reed relay, then to output jacks
 

Attachments

  • ThroughBoard.jpg
    ThroughBoard.jpg
    985.6 KB · Views: 97
Administrator
Joined 2004
Paid Member
Hi as_audio,
I'm extremely busy now, but could you please do me the favor of sharing your schematic

Oh wait! You don't have one, do you? So your post (#53) is pretty much in your imagination.
There are no base connections.
Sorry, my mistake, but I meant gate connections. However you did know what I meant and are just taking a cheap shot. Bravo!
This may be a little bit different for gates of a jfet pair.
No, not for a linear circuit. If you had more experience, you would know that. Have a look at some real amplifiers using either technology. Amazingly they operate the same way, and both will have very low offset across the inputs, be they gates or bases.
An offset is balanced, not created, if done properly.
The only thing you can do with a trimmer or DC servo is unbalance the diff pair to correct for an imbalance. This will always drift with temperature.
I think this is not true, it is a characteristic of the circuitry.
With a circuit like this you can expect to align the residual
offset to a few millivolts which should not change much in
thermal equilibrium (but it will if using freezer spray).
Once again, you lack of experience is showing. A matched pair will have minimal offset changes with temperature. An unbalanced pair will always drift with temperature changes. If you are using a DC servo, monitor the output of the DC servo and you will see the drift there.
We leave the design as it is. Interaction between tone
and line stage was apparent when pushing "tone in".
Didn't say anything about changing the design. However, allowing the tone controls to affect the output amplifier is extremely poor design work. Every linear control must have zero DC across the control or it will be noisy, period. So the easiest way to make that happen is to use coupling capacitors. They are probably used there but are leaky.
I think I explained the reason before why the tc does not have this cap
(how do you combine bass boost with a gain of one below some low
corner frequency ?).
Of course you are not correct. Many tone control stages have this feedback capacitor. It is just large enough to not affect the audible range. It will not affect the ranges where you can hear bass, say down to 10 Hz or lower.
A dual jfet is assumed to be made a matched pair.
I see no hint that some of these may be defective.
They can be bought in different grades of mismatch, and they can certainly change over time. So when new (unless defective), they should be reasonably close to each other. The degree of mismatch is measured in mV between the gates. The fact that you don't see any evidence of mismatch is puzzling if you service this equipment. A voltage differential between the gates = mismatch. You simply need more experience to be able to understand this. It will come as long as you pay attention to these things.
We can not be sure about the layout of the dc balance trim
and it is not possible to see the midpoint using an opaque
ten turns trimmer.
I assume you can count the revolutions - right? In a good design, the trimmer if it exists will be centered when the DC offset is zero, or close to it. That's as long as the diff pair is matched and the circuit is working normally. Really poor designs might not, but we are assuming a good design as far as I'm concerned.
Sorry gentlemen, I am not in a position to write lengthy essays in this thread.
I am pretty sure that the preamp is in working condition, shown by the screen
pictures, if no subsequent harm happened due to recapping and probing around
and it only needs some dc offset realignment.
Of course not! I understand, you have to catch up on your courses first. You do have a lot of reading to do by the sounds of it.
OP may even have some distortion check function on his Rigol scope to confirm
proper operation.
The scope is only 8 bit, maybe 10. It will not give you a useful reading. Now, a signal analyzer, spectrum analyzer, network analyzer or a distortion analyzer will give you a useful answer. Sometimes you can use a sound card, but they are not calibrated so you will not have a useful answer using those for the most part. A sound card is also very easy to damage with excessive input voltage, and all preamps can manage those levels. I have all of these instruments and normally use a signal analyzer or distortion measuring set (HP 339A) to generate known, accurate readings. The signal analyzer will give you by far the most information. These all run about $30K each or more, but the RTX 6001 for $2,500 is a great instrument, a very good value. Which do you use as_audio? My other instruments are all HP / Agilent for signal analysis.

All I'm going to say about your post is that you should make sure you know what you are talking about, and who you are talking to before copping an attitude. Even then it's probably not a great idea.

-Chris
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.