Ultra high spec. opamp MC/MM phono, warp "elliptic" filter, line, headphone amps.

The protection circuit does not work as expected.
I understand why, just a stupid oversight on my part.
When the supply plug is pulled the -15v rail, which triggers the operation, starts falling about 100ms or so after the +15v rail, so by the time the protection circuit operates some of the "damage" is done and the output deviates significantly.
I'm pondering how best to deal with it.
When I simulate the circuit with supplies that behave in the actual way the problem manifests itself.
I'll be building any changes on the prototype board to ensure proper operation.
 
The protection circuitry is now much better.
20241123_125809.jpg
This shows what happens when the AC plug is pulled out.
The unit is in MC mode, nominal gain.
Yellow is a line output, blue is the -15v supply.
The protection circuit takes 30ms to operate once the -ve rail starts to fall, which occurs about 200ms after the plug is pulled.
Multiple switched on/offs are fine and are essentially inaudible.
Bill has been sent the modified kicad schematic.
I think that I'm done.
I plan to box up the board and send it out for listening.
 
GKTAUDIO said:
Cool. May I ask what you did to correct this?
Click to expand...
Compared to what baseline?
Screenshot 2024-11-23 150520.png

This is the up-to-date protection/line output relay circuit. The maximum voltage on the gate of the relay FET is now clamped, with the zener diode being replaced with a reverse biassed small signal diode (needed to protect the electrolytic cap) and the turn off circuit is slightly different.
In addition, a 750 ohm R was placed between the -15v rail and ground to make the load on the two supply rails asymmetric and force the -15 rail to droop first, triggering the protection before an excessive reaction to the overall supply reduction occurred.
 
It would work better (earlier) if the detection was performed earlier in the supply- say at the output of the -18.5 v regulator or after the negative supply active filter, but that would require the supply to be routed to the output stage and I'm reluctant to do that for such a rare problem.
 
I have a functional prototype.
20241130_145507.jpg
20241130_145529.jpg

Everything works as it's supposed to. The box is the only one that I had that was of an acceptable size. It doesn't have continuous metal sides which is a problem, but I tried to ground everything although it's still a bit of a work in progress. The MM/MC switch is internal but can be activated with the blunt end of a pencil.
Next is to install it in the home system and see how it does.
 
I have measured the noise of the boxed prototype stage in MC mode 54dB gain, with the input shorted.
MCshorted 54dB.jpg

The A weighted value, relative to the nominal 6dBv output is c. 94dB. This is very close to what was expected.
Note that the mains pickup is a very low c. 120dB below rated output.
It should also be noted that the short is at the end of c. 6" of phono cable.
The converter noise is at -100dBVA and the hum is at -140dBVA with its input shorted.
Now, 69dB gain (+6dB at the input and +9dB at the output) with the input shorted.
69dB gain.jpg

All seems as it should be.
This should be inaudible with even the lowest output cartridges.
 
wyn palmer said:
I have measured the noise of the boxed prototype stage in MC mode 54dB gain, with the input shorted.
View attachment 3359373

The A weighted value, relative to the nominal 6dBv output is c. 94dB. This is very close to what was expected.
Note that the mains pickup is a very low c. 120dB below rated output.
It should also be noted that the short is at the end of c. 6" of phono cable.
The converter noise is at -100dBVA and the hum is at -140dBVA with its input shorted.
Now, 69dB gain (+6dB at the input and +9dB at the output) with the input shorted.
View attachment 3359379

All seems as it should be.
This should be inaudible with even the lowest output cartridges.
Click to expand...

What he said. I can’t wait.
 
I see that mods are editing out any reference to the actual reason why using Paypal has become onerous.
It seems that anything even obliquely political is now unacceptable, even if factually correct.
I expect that this post will be excised. Who knows, perhaps I will be banished from AK for my impudence in pointing this out.
In any case, I'm in the process of giving the "production" board the once over to ensure that all the changes to improve the "manufacturability" have been included.
The current prototype board is operating in my audio system, even as we speak, and I have asked the local Audio club to do a listening session in a little while.
There are some caveats in the use as, for example, switching from MC to MM on the fly with the power on is a really bad idea due to the need for input calibration.
 
wyn palmer said:
I see that mods are editing out any reference to the actual reason why using Paypal has become onerous.
It seems that anything even obliquely political is now unacceptable, even if factually correct.
I expect that this post will be excised. Who knows, perhaps I will be banished from AK for my impudence in pointing this out.
In any case, I'm in the process of giving the "production" board the once over to ensure that all the changes to improve the "manufacturability" have been included.
The current prototype board is operating in my audio system, even as we speak, and I have asked the local Audio club to do a listening session in a little while.
There are some caveats in the use as, for example, switching from MC to MM on the fly with the power on is a really bad idea due to the need for input calibration.
Click to expand...
This means powering down stereo amps/bass bin plate amps first. So be it.
 
mkane said:
This means powering down stereo amps/bass bin plate amps first. So be it.
Click to expand...
You need to power down the phono stage, the power amps are optional. If you do that the 20sec delay is operational.
If the unit was relay controlled the delay would be automatic, but that wasn't going to fly.
I still wouldn't operate the stereo/mono/warp switches for about a minute after switching over and powering up to let everything settle.
 
wyn palmer said:
I'd offer to sell boards at cost but only to those who are willing to commit to building/evaluating them as the Paypal restrictions are punitive.
@GKTAUDIO also mentioned that selling abroad has become particularly problematic.
In any case the Gerbers will be made available.
Click to expand...
Unless somebody else steps in: if exporting boards seems too hairy, I suppose I could sell them at cost at the European side. The price will be what it gets after VAT and all customs fees.

Since I want a couple myself, I could just as well order a batch from the Gerbers once we get there. I've already got the routines in place for handling sales both within and outside EU, so that part shouldn't get too tricky.

Of course, any buyers will have to sort out their local customs and tax as required.
 
I think that I'll add a mute function that disables the output then enables it after c. 20 seconds- just like the power on/off switch does. This should be used whenever any of the functions (MC/MM, warp/mono, stereo/ (warp or mono) are switched as a protection.
Is this a reasonable thing to do?
This can actually be done by just using the two connections already used for the relay control portion of the on/off switch, and I can modify the prototype accordingly.
 
wyn palmer said:
I think that I'll add a mute function that disables the output then enables it after c. 20 seconds- just like the power on/off switch does. This should be used whenever any of the functions (MC/MM, warp/mono, stereo/ (warp or mono) are switched as a protection.
Is this a reasonable thing to do?
Click to expand...
Seems like a good idea.
 
You must log in or register to reply here.
Back
Top Bottom