Seeking repair thoughts vintage VOX AC-100 guitar amp...

[ron_swanson]

abbey road d enfer​

>>> BTW, setting bias with a voltage is some kind of inadequate.

Yes, I understand. The bias voltage I was referring to was referencing the indicated schematic -35V.

>>> With the values you published, each EL34 dissipates 22W at idle, which is dangerously close to the 25W maximum.

I understand this too. My tech who has worked on this amp in the past was also not happy about how hot the OP tubes were running. Another friend who is also handy with repairs and who has worked on a number of AC-100's in the past says, it's how they were designed and built for better or worse. That was years ago and he's now retired so can't remember particulars, but did say that the 115V tap on the power I/P was improperly wound and these amps should be powered using the 245V tap in the US along with a step up transformer - which I'm doing. I tried once to run using the 115V tap which is even 'hotter'. I guess once I get this noise issue resolved, I may take a look at addressing this so the output tubes are not running near 100% at idle or possibly using a variac set at a slightly lower input voltage. TBD.

Do you have direct knowledge AC-100's running within acceptable parameters without having to modifying the circuit?

Cheers!

 

[rock soderstrom]

I'm still planning on giving all sockets a good, proper cleaning, re- solder of V3 at a minimum, and re-tensioning all as well as swapping out the preamp tubes with new tubes at least for testing purposes. Also checking the coupling cap voltages after being disconnected as suggested earlier in this thread.

This is basically a good approach for such old amps. As I said, I have had similar problems with bad contacts in the tube sockets.

I now believe that the voltages around the PI are okay, the problem may also be due to the biasing, since it is a modification. I would first implement all the tips mentioned and then see how the amp behaves.

Question...., when I was desoldering to measure and / or replacing the old resistors, many of them would bubble up with some sort of liquid. Would this suggest that this amp has been subjected to moisture in the past? If so, should I consider proactively replacing all of the carbon comp resistors? The remaining carbon comps at the moment measure in tolerance. Not a fan of shot gunning repairs, but maybe in this case justified?

Sounds strange, I haven't had that before.
I would first make all as described before I would exchange further components. We need a strategy

 

[ron_swanson]

abbey road d enfer​

>>> It's really the current that should be monitored. It is quite easy on this amp because of the 1r resistors in series with each plate.

1r? Do you mean, 100r?

 

[abbey road d enfer]

abbey road d enfer​

>>> It's really the current that should be monitored. It is quite easy on this amp because of the 1r resistors in series with each plate.

1r? Do you mean, 100r?

Yes, 100r. My keyboard has a sticky "0" key.

 

[abbey road d enfer]

Do you have direct knowledge AC-100's running within acceptable parameters without having to modifying the circuit?

My bet is Vox ordered their tubes from one vendor. The circuit was developped around this particular type, and Vox required tubes with similar characteristics from this vendor.
Today, some suppliers grade their tubes. My understanding is that the higher the number, the higher the idle current (i.e. the lowest cut-off voltage), so it seems, in your case, to promote using tubes with a low number, which is good news, because the higher the number, the higher the price generally.

 

[tkaz]

A couple of thoughts. I had a Fender Twin that would freak out when hit hard. Finally happened to walk behind it playing as it was getting dark & saw sparks under one of the 6L6s. Carbon track on the the socket that conducted when there was enough power flowing. The other one was a Silvertone Twin-Twelve. Worked ok quiet, freaked out bad when hit hard. Turned out the primary winding of the OT got hot enough to burn thru the cardboard bobbin. It shorted to the core when pushed. This second one could be your problem. I would think a 100 watt Marshall OT would work with that if you can't find the proper replacement. Don't remember how I figured that one out.

 

[CJ]

Bias voltage does not really affect sound when the amp is wide open, the tubes are going to be saturated and cutoff as the input voltage is hot and heavy, You might have a bit more crossover distortion between sq waves but who cares, the crazy drummer will drown that out

But you do not want your tubes to burn up while you are chugging whisky between sets,

Use 60 percent plate disp, so 15 watts per tube,

15 watts divided by plate voltage gives per tube current, so 30 ma,

Are those 100 ohm resistors still 100 ohms? They live a hard life ,
If they are 100 then you read 3 volts when tubes biased right.

60 percent seems cold but el34,s like to drift after a while. In a week you will be at 65 percent.

Usually when I see those plate resistors an alarm goes off saying "unstable amp"!
So you might have a layout problem, which is causing ultra sonic osc. You can not hear it directly, it shows up as terrible sound and hot tubes
Use a chop stick or pencil to move wires while running the amp and see what happens.

Fixed a Sunn Stagemaster today, what a weird amp,

 

[ron_swanson]

So, here's where I am atm...

As suggested above...

1. Replaced V3 with a new matched ECC82 (EDIT: Incorrectly indicated ECC83 in original post).
2. Cleaned, re-soldered V3 socket. All sockets actually, but only re-soldered V3.
3. Remeasured entire circuit. ( see attached schematic. NOTE -- My previous schematic with coupling cap mV measurements was incorrectly measured. Realized I had my meter set to ACmV, not DCmV. Sorry. )
4. Added BLUE boxed measurement before and after coupling caps. C= Connected measurement. D = Disconnected measurement.
5. Chop sticked the circuit looking for issues that may be causing noise. Helped a little, but find nothing that resolved.
6. If not mentioned previously, at this point ALL caps have been replaced with new, quality capacitors.



Notes...

Before cleaning V3 and re-soldering, I was unable to measure with any accuracy the DC mV at pin 7 due to extremely wild variations in voltage readings. Was fluctuating between low teens mV to 70mV-ish. After cleaning I was able to make a measurement, but it's still varying a few mV up and down. Maybe more cleaning required here?

I also just noticed that when I remove the guitar plug from the input jack, a loud ground hum is present. When plugged in, ground noise gone. Dirty jacks? Always something!

Anyway...

@ CJ. Yes, I undertand that the plate current on this amp at idle is running near 100%. I'm guessing this how all AC-100's were built? The few techs I know all say the AC-100's they have worked on all worked as the schematic indicates, although they could not say what exactly the current was from memory.

All things being equal, my goal here is to get this amp running clean through the entire volume range and then deal with 'modifications', but after all of the above, the amp is still producing nasty distortion / farting in the low frequency range (chugging power chords on low E/A strings for example) especially when the guitar strings are hit hard as it was before. Although, it does seem to be subjectively a bit less now after cleaning, etc. My thought is to tackle one issue at a time. Would the high plate current possibly be playing a part or even be the cause of the noise issue I'm chasing here?

Additionally, atm, It's not clear to me what changes I would want to do to reduce the plate current properly. I can find info on the Interwebs about fixed biased amps, but nothin that I find that exactly describes what to change and how to determine that change properly. Maybe I'm search or mis-understanding what I'm finding?

Thanks in advance!

 

[Tubetec]

I never saw an amp with 100 ohm resistors in the plate like that , can anyone show me an example or explain what use resistors in series with the anode are that would be great . Check the value of these resistors if you didnt already .
The smaller Vox amps tend to be class A cathode biased ,they run very hot , the AC-50 and AC-100 are fixed bias class AB much like a Marshall , its almost identical apart from a few differences .

Now that I look at it again ,its wrong , the cathode of the LTP sits at 11.5 v , your grids is not far off 0 volts , no wonder it sounds like crap .
I would change the phase invertor stage , make the LTP cathode resistor 10k , and use a 470 ohm resistor to get your grid bias drop ,
you can try grid 2 drive as is or you could ground it to the end of the tail resistor or even add the presence control if you like later , once you look after the LTP biasing problem this amp will roar into life is my guess .



see schem attached

 

[abbey road d enfer]

Before cleaning V3 and re-soldering, I was unable to measure with any accuracy the DC mV at pin 7 due to extremely wild variations in voltage readings. Was fluctuating between low teens mV to 70mV-ish. After cleaning I was able to make a measurement, but it's still varying a few mV up and down

There should be near-zero here. If it is confirmed, you could check R15 and C13. It may be grid leak but it should be negative.

I also just noticed that when I remove the guitar plug from the input jack, a loud ground hum is present. When plugged in, ground noise gone.

That's a serious concern. It shouldnt be. Are you sure the switched contacts are properly wired?

the amp is still producing nasty distortion / farting in the low frequency range

Are you sure it's the amp, and not the speakers?

Additionally, atm, It's not clear to me what changes I would want to do to reduce the plate current properly. I can find info on the Interwebs about fixed biased amps, but nothin that I find that exactly describes what to change and how to determine that change properly. Maybe I'm search or mis-understanding what I'm finding?

The thing is, Vox engineers chose to set bias by voltage, which is right only if the tubes are perfectly selected for this bias voltage.
In order to decrease the idle current, you need to increase the value of the bias voltage, which is regulated by the pair of zeners, so you would need to experiment with pairs of zeners that achieve a higher value.

 

[abbey road d enfer]

Now that I look at it again ,its wrong , the cathode of the LTP sits at 11.5 v , your grids will be around 0 volts , no wonder it sounds like crap .

Good catch, but still each triode section runs at 3 mA.

 

[musipol]

Re: the 100 ohm plate stoppers. I've seen those( but different values) on a few Ampegs back in the day when I used to do this stuff for a living. I couldn't give you a good explanation as to what it does other than waste power. Just guessing, but maybe it's to lower the Q to reduce parasitics in the output stage......that's a complete WAG on my part.

 

[mkiijam]

Just adding my 2 cents, which today is worth about 1.38 cents... Forgive me if this is stupid or as already be addressed. BUT- I have had the turret boards go conductive and cause all kinds of strange problems. Any chance of that?

 

[moamps]

1. Replaced V3 with a new matched ECC83.
........
Additionally, atm, It's not clear to me what changes I would want to do to reduce the plate current properly. I can find info on the Interwebs about fixed biased amps, but nothin that I find that exactly describes what to change and how to determine that change properly. Maybe I'm search or mis-understanding what I'm finding?

You can't use the ECC83 for V3 unless you change the anode resistors, I would advise you to put the ECC82 back in that position. After all, the -11.5V, 3mA, 230V operating point doesn't exist for the ECC83, IIRC.
You can change the anode currents of the output tubes by changing the negative bias voltage on the input grids, you mentioned that someone has already placed a trimmer in that position, maybe by adjusting it you can increase the negative voltage and thus reduce the anode currents.
Where did you measure the voltage in the video clip you showed, was the amplifier loaded with a speaker or a dummy load?

 

[ron_swanson]

You can't use the ECC83 for V3 unless you change the anode resistors, I would advise you to put the ECC82 back in that position. After all, the -11.5V, 3mA, 230V operating point doesn't exist for the ECC83, IIRC.
You can change the anode currents of the output tubes by changing the negative bias voltage on the input grids, you mentioned that someone has already placed a trimmer in that position, maybe by adjusting it you can increase the negative voltage and thus reduce the anode currents.
Where did you measure the voltage in the video clip you showed, was the amplifier loaded with a speaker or a dummy load?

Argh! I meant ECC82 / 12AU7, not ECC83. Sorry.

Measurements were made with a dummy load.

 

[user 130681]

Too many odd, random posts here so,
Have you still got instability ... if you have, did you measure the HT voltage with your scope, using a high voltage Hameg or similar probe to avoid damage?
Did you check the NTC feeding the choke?
Is this original in build? The original uses 470R current limiting on each side of the output stage.

 

[abbey road d enfer]

Is this original in build? The original uses 470R current limiting on each side of the output stage.

Where? Out of 4 different builds, the only 470r resistor is the one across the secondary.

 

[user 130681]

Copy of original AC80/100 output stage with arrows pointing to the screen feeds.
As it is a home brew, I cannot offer any further advice. Sorry.

 

[abbey road d enfer]

But this is part of the original cathode-bias version! Didn't you say yours is fixed bias?

 

[user 130681]

But this is part of the original cathode-bias version! Didn't you say yours is fixed bias?

If it is fixed bias, this dates it 1965 onwards.