It seems to me to be a capacitor with a damaged dielectric. It could be a failing cathode junction inside a valve, but you said you replaced the set. It could also be a dirty PCB where the high DDP between close tracks attracted some dust which, from time to time, becomes a conductor.
Does the pop generate on both channels?
Once solved, if you want a good sound from Aikido, I suggest reading this. https://www.diyaudio.com/community/threads/warmth-of-the-sound-wots.380373/page-8#post-7206622 post #159, at the end of the post.
Does the pop generate on both channels?
Once solved, if you want a good sound from Aikido, I suggest reading this. https://www.diyaudio.com/community/threads/warmth-of-the-sound-wots.380373/page-8#post-7206622 post #159, at the end of the post.
Those are two diodes - they shield the rectifier valve from the inrush when first turned on. The rectifier (6X5) sees a 3.3uF first, then choke, then 180uF-1k-47uF after that per channel.
I ran the preamp with peaking turned on. I got peaks from -300mV to +300mV, and if I sit and watch on the scope, both channels track perfectly together. However, those peaks were directly related to other consumer items in the workshop turning on/off. In reality, when all is quiet, it floats -100mV to +100mV.
I then went a hooked up my bench HV supply to try rule out the onboard supply - but I see the same thing there.
I also added more decoupling on the aikido PCB itself - 100uF - again no change.
I changed the output caps in case they were leaky - I now have 680//680nf on each channel - no difference
Last thing to do now is remove C5 as suggested and see does that make a difference.
The linked video shows what I am seeing.
I ran the preamp with peaking turned on. I got peaks from -300mV to +300mV, and if I sit and watch on the scope, both channels track perfectly together. However, those peaks were directly related to other consumer items in the workshop turning on/off. In reality, when all is quiet, it floats -100mV to +100mV.
I then went a hooked up my bench HV supply to try rule out the onboard supply - but I see the same thing there.
I also added more decoupling on the aikido PCB itself - 100uF - again no change.
I changed the output caps in case they were leaky - I now have 680//680nf on each channel - no difference
Last thing to do now is remove C5 as suggested and see does that make a difference.
The linked video shows what I am seeing.
Despite trying a good few different things I still haven't managed to sort this.... I have a bit more to do before I report back again.
I ran it from a bench supply already, but still the same.....
The thing that's driving me on is that sonically, its very good (and dead quiet) - and that is without even trying the arrangement mentioned above in the WOTS link.
C5 replaced
C6 replaced
Added more decoupling local to the board
Upped the value of grid stoppers
Tried heaters on AC (were on DC)
Tried separate bench supply
Changed output caps
Changed heater elevation and verified decoupling with 47uF (~80V)
Checked, and rechecked all connections to make sure there wasn't a dry joint etc
One other thing I thought of - in this amp, I have two transformers. One transformer supplies the heater for the 6X5 rectifier and HV, and I have a second 4A single secondary 12.6V transformer for the 6SN7 heaters which is elevated ~80V (200k-68k//47uF). I'm now running on AC, with each set of 2 valves sharing 12.6V, voltage at each valve within 0.1-0.2V of 6.3V target.
So with the elevation, that means that the heater (supplied via one transformer) is elevated/referenced to the HV which comes from a different transformer. Could there be something in that which is causing the offset to wander up and down? If I track the HV voltage, I see it varying up an down by a volt or two, and slowly whereas the offset is moving quicker than that.
The thing that's driving me on is that sonically, its very good (and dead quiet) - and that is without even trying the arrangement mentioned above in the WOTS link.
C5 replaced
C6 replaced
Added more decoupling local to the board
Upped the value of grid stoppers
Tried heaters on AC (were on DC)
Tried separate bench supply
Changed output caps
Changed heater elevation and verified decoupling with 47uF (~80V)
Checked, and rechecked all connections to make sure there wasn't a dry joint etc
One other thing I thought of - in this amp, I have two transformers. One transformer supplies the heater for the 6X5 rectifier and HV, and I have a second 4A single secondary 12.6V transformer for the 6SN7 heaters which is elevated ~80V (200k-68k//47uF). I'm now running on AC, with each set of 2 valves sharing 12.6V, voltage at each valve within 0.1-0.2V of 6.3V target.
So with the elevation, that means that the heater (supplied via one transformer) is elevated/referenced to the HV which comes from a different transformer. Could there be something in that which is causing the offset to wander up and down? If I track the HV voltage, I see it varying up an down by a volt or two, and slowly whereas the offset is moving quicker than that.
You can remove the tubes then do a measurement at each pt, then compare to both channels. I have different pcb, I couldn't remember hum or pop sound when I got R10 damage but still working. My R10 was 200r but it still works with measurement 110r. After replacing R10, it has been working without problem.
How did you connect the chassis to the mass and to the earth? Try to remove all those connections, do not connect the house earth, connect the chassis to the mass with a 1KOhm resistor and test it with an ohm meter before powering up. Connect to the chassis all metal parts (do you have any semiconductors on any heatsink?)
Did you try to lower the voltage to 250V, for example?
Are you sure your tubes are good quality? Using the 6H30 at 330V in my first Aikido (which has one power line for upper heathers and another for the lower heathers) I had some trouble with one tube that made the woofer move while turning the volume pot. No problem at 250V, even if the tube is a (semi) defective one.
During the building of my first Aikido, I tried using a PS with a GZ34 and a choke input followed by a capacitor. It didn't work because of too much hum. Then I tried with capacitor-choke-capacitor. No hum this time, but I observed the woofer moving slowly, with some pops (not able to trigger the DC protection) from time to time, due to home appliances. On the scope, I observed exactly the same behavior you observed. Then, after the CLC filter, I applied a Virtual Battery stabilizer and all problems disappeared. My second Aikido dual mono uses two PS21s and never had those problems.
We all know the Aikido scheme, even if the index of the component could be different in a different type of scheme/PCB (I have the noval mono). But we don't know the complete scheme of your power supply. If I understood and remember well how you are using the 6X5, it seems to me a stressful condition for it. Check it out. If your bench PS uses the same tube, you are in the same condition.
Can you try to replace the 6X5 with a couple of diodes, if the voltage does not rise too much?
It would be better if you post the scheme of your PS.
Did you try to lower the voltage to 250V, for example?
Are you sure your tubes are good quality? Using the 6H30 at 330V in my first Aikido (which has one power line for upper heathers and another for the lower heathers) I had some trouble with one tube that made the woofer move while turning the volume pot. No problem at 250V, even if the tube is a (semi) defective one.
During the building of my first Aikido, I tried using a PS with a GZ34 and a choke input followed by a capacitor. It didn't work because of too much hum. Then I tried with capacitor-choke-capacitor. No hum this time, but I observed the woofer moving slowly, with some pops (not able to trigger the DC protection) from time to time, due to home appliances. On the scope, I observed exactly the same behavior you observed. Then, after the CLC filter, I applied a Virtual Battery stabilizer and all problems disappeared. My second Aikido dual mono uses two PS21s and never had those problems.
We all know the Aikido scheme, even if the index of the component could be different in a different type of scheme/PCB (I have the noval mono). But we don't know the complete scheme of your power supply. If I understood and remember well how you are using the 6X5, it seems to me a stressful condition for it. Check it out. If your bench PS uses the same tube, you are in the same condition.
Can you try to replace the 6X5 with a couple of diodes, if the voltage does not rise too much?
It would be better if you post the scheme of your PS.
I'm also scratching my head about that low voltage low frequency output signal which shouldn't be expected from an idling amplifier. Does this happen with shorted inputs also?
Best regards!
Best regards!
Thanks guys.
All of this is with shorted inputs I'm afraid.
Circuit ground is connected to earth via a single point, with back to back diodes, 10R power resistor and a 10nF cap.
I've tried 3 sets of tubes - but I see this on both channels at the same time, so I don't think its tubes only.
I will try without the 6X5, using some diodes instead for sure. If the 6X5 is being stressed that could be it. I can also make up a new solid state HV supply, I have some boards here for that.
Power supply is:
DC 6.3v supply to the rectifier heater
HT to 6X5 diodes via a pair of diodes to lessen start up stress.
Then to 3.3uF film cap
Then 10H choke
Then the HV splits into two separate sections - each with a further 180uF-1k-47uF filter for left and right. However, I have both of these joined together after the second 47uF.
Heater elevation is also on this power supply board, 200K-68K decoupled with 47uF.
DC sits around 300V, and elevation is ~80-82V
All of this is with shorted inputs I'm afraid.
Circuit ground is connected to earth via a single point, with back to back diodes, 10R power resistor and a 10nF cap.
I've tried 3 sets of tubes - but I see this on both channels at the same time, so I don't think its tubes only.
I will try without the 6X5, using some diodes instead for sure. If the 6X5 is being stressed that could be it. I can also make up a new solid state HV supply, I have some boards here for that.
Power supply is:
DC 6.3v supply to the rectifier heater
HT to 6X5 diodes via a pair of diodes to lessen start up stress.
Then to 3.3uF film cap
Then 10H choke
Then the HV splits into two separate sections - each with a further 180uF-1k-47uF filter for left and right. However, I have both of these joined together after the second 47uF.
Heater elevation is also on this power supply board, 200K-68K decoupled with 47uF.
DC sits around 300V, and elevation is ~80-82V
Why a rectified heater voltage for your HV rectifier? Of course this doesn't need to correlate with your issue...
Best regards!
I used a rectified supply because the transformer I have has a 12V secondary for heater - so I used a DC PSU to drop that to 6.3V.
Just to add - the PSU I'm using is this one:
https://www.diyaudio.com/community/threads/aikido-linestage-help-needed.98142/post-1158365
Exception - C1 is 3.3uF, and I'm using 10H choke instead of 1K/25W. C2/C3 are 180uF and C4/C5 are 47uF
https://www.diyaudio.com/community/threads/aikido-linestage-help-needed.98142/post-1158365
Exception - C1 is 3.3uF, and I'm using 10H choke instead of 1K/25W. C2/C3 are 180uF and C4/C5 are 47uF
Attachments
This is what my PS for 6SN7: Either Choke/Resistor doesn't cause hum/pop. B+ is stable, doesn't cause any problem when the AC move from 120Vac-115Vac. I configure B+ at 120Vac where it works fine at 115Vac if it drops. My home AC is stable at 120-118.
Either Choke or Resistor at J7 has no problem with hum.
I have tested both with/without ground house, doesn't cause hum either. You can wire the GND to the chassis.
The hum/pop I was encountered were a steel chassis and a damage resistor. After found out the steel chassis causing hum with toroidal transformer, I changed to the aluminum. Then a damage resistor causing a pop or a cracking sound I think. Finally, it has been working without problem.
Aikido article: https://tubecad.com/2017/01/blog0368.htm
Either Choke or Resistor at J7 has no problem with hum.
I have tested both with/without ground house, doesn't cause hum either. You can wire the GND to the chassis.
The hum/pop I was encountered were a steel chassis and a damage resistor. After found out the steel chassis causing hum with toroidal transformer, I changed to the aluminum. Then a damage resistor causing a pop or a cracking sound I think. Finally, it has been working without problem.
Aikido article: https://tubecad.com/2017/01/blog0368.htm
This makes the amplifier receiving spikes of too high voltage being amplified and showing at the outputs. It is not good to continue with the device as the amplifier will be at risk of going to semiconductor fields too soon. Please be aware of that as you may end up with 2 defective devices. If your amplifier would not have the µpc1237 modules your woofers probably already would have burnt.
The same risk may apply during power on/off so a muting circuit with a small shorting to GND relay (a continuing novelty since at least 40 years) will protect that same amplifier. Please measure and then define what time is adequate. You will be safe with 30 to 60 seconds power on delay I guess. At power off the relay should short outputs to GND immediately. In both outputs a series resistor of 47 to 100 Ohms will limit possible currents and prolong the relays contacts usable life.
*There should be a sticky here warning that marrying (often completely superfluous) tube preamplifiers to solid state amplifiers is OK but that it requires low gain/high drive capability in most cases and that the situation requires DC protection of the solid state amplifier by means of muting outputs of the tube preamplifier at power on/off. Semiconductors are normally not very fond of high DC voltages.
Last edited:
Completely agree. I don't see the advantage of mixing technologies completely different. Or tubes or sand.
This just happens to be the amplifier currently in the system, and the preamp in question is not my daily driver either. So I suppose, this is not so much a discussion on the merits or otherwise of valves or solid state, but more about an obvious issue with a preamp I'm trying to fix (and failing so far I might add!) so that the things outlined by @jean-paul don't happen!
Anyway, in todays experiment, I swapped in a pair of 1n4007 for the 6X5 and also added a dropping resistor, but with no change. The experiments will continue until I run out of options... and that is probably pretty soon!!
Well, I do. Even in a tube amplifier there are some tasks that can be done well by SS without compromising the overall performance. Think of rectifiers, CCS's, active plate loads, maybe emitter or source followers - you name it 😉 .
Best regards!
All the functions you mentioned can be done with tubes. If very complex radar, tv and computer systems had been done with tubes, I don't see the necessity for mix both in a humble audio amplifier, that also had been done using only vacuum tubes.
Ok folks, maybe we might stay away from the (interesting) discussion on the suitability or otherwise of tubes and solid state..... If we go down that rabbit hole we might never recover!
Last night I hooked up a new transformer, a new crcrc supply, lowered B+ to 260V and the issue is still there. I also checked just in case there is weird DC at low levels running across the chassis, but I see a perfectly steady 0.5mV from circuit ground to earth. Given that I have a pair of back to back diodes, power resistor//10nf cap I think that is ok.
I also checked current through the valves, and as specified by Broskie, one set of tubes is running at 5mA and the other is hotter at 10mA.
The last thing is to disassemble the aikido board completely and start from scratch.
Last night I hooked up a new transformer, a new crcrc supply, lowered B+ to 260V and the issue is still there. I also checked just in case there is weird DC at low levels running across the chassis, but I see a perfectly steady 0.5mV from circuit ground to earth. Given that I have a pair of back to back diodes, power resistor//10nf cap I think that is ok.
I also checked current through the valves, and as specified by Broskie, one set of tubes is running at 5mA and the other is hotter at 10mA.
The last thing is to disassemble the aikido board completely and start from scratch.