I need some help troubleshooting a low frequency oscillation in a tubeamp SOLVED

[Planobilly]

Hi guys,

I just finished the build of a AB763 Fender style Deluxe Reverb amp. Everything works well and the amp sounds good to my taste.

The issue is that I have a low frequency oscillation which can be heard at idle. I think this sound would be considered "motorboating" It appears with all gain controls set at zero. I seems to be associated with channel one. When the channel one gain is increased to about 60% it goes away and when the gain is increased to about 65% it comes back.

THe amp is built very close to this schematic. http://el34world.com/Hoffman/images/Hoffman_AB763.pdf

All the voltages are normal and I have changed all the tubes. I have moved all the wires around to no effect. I tried installing a low wattage tube (5751) in V1 to no effect. I have re-flowed many solder connections. I have changed 95% of the capacitors in the amp to no effect on the oscillation. The cap changes did make the amp sound better....used orange drop caps.

Everything is built per the schematic except the cathode caps are 25uf and not 22uf. 25uf is what Fender put in their amps. There is a error on the schematic with the value of the resistors in the power supply which I changed to 10K to provide correct plate voltages.

I have pretty much done everything I know how to do to try to resolve the issue. At this point I really need some guidance on how to proceed.

Thanks,

Billy 

[uncle_bob]

Hi

Do you have an oscilloscope or other test gear? Best bet is that it actually is a high frequency oscillation that impacts the bias. That gets it down to the point you can hear it. Probe through the circuit with a scope that has at least a 10 MHz bandwidth (100 would be better) and you likely will find the problem stage. Then it's a matter of fixing the bypassing and grounding to make it go away.

Bob

[Planobilly]

Hi Bob,

Yes I have a 100 MHz scope, a newer Owan digital. I also used a probe connected to another amp to hear the sound it was making. The sound does not appear on the plates of V1 and V2.

I am not very skilled at using my scope. When looking at the signal at the output transformer there is a voltage both positive and negative that appears with the sound.

Perhaps you could give me some guidance on how to set up the scope and just what to look for.

Thanks,

Billy

[Planobilly]

Here is a photo if that helps.



Billy

[uncle_bob]

Hi

Start at the front end of the circuit (V1 and V2) and work your way to the output (V7). At some point, you will find a tube with junk on the plate. Once you know which tube that is, you can start trying to figure out what the cause is. Equally, you can start from V7 and work back towards the front end. There are a couple of branches in the circuit, so it's not entirely a linear process. An alternative is to make a list of all the plate sections (V1A, V1B, V2A, V2B ....) and check each and every one of them. Check off which ones have junk and which ones do not have junk.

As always - be careful to keep things grounded and avoid touching the signal leads (High Voltage ....).

Bob

[Andy Watson]

Please forgive the obvious question (but you never know) - have you tried removing or otherwise disabling V5?

[uncle_bob]

Please forgive the obvious question (but you never know) - have you tried removing or otherwise disabling V5?

Hi

To the point: Since this is basically an AC coupled circuit, you can "randomly" pull tubes. There is no worry about pulling one and frying another. That is an OK way to isolate which tube (or section) is creating the problem.

Bob

[Planobilly]

Thanks Bob,

LOL...I will use due care when probing the circuit!! I sure don't want any high voltage drama...lol

On the V5 plate there is a 3.5 Hz signal at 1.7 VRMS

V4 plate 312KHz 23mv

V3 plate 107 KHz 50mv

V2 plate 10 Hz at 5 mv...I assume that is what I am hearing

V1 plate 10 Hz at 11mv

V6 grid 10 Hz at 14mv

V7 output tube grid 107 Hz and 120 Hz at 3.1 V

Not sure just what all that indicates.

Billy

[Planobilly]

I just pulled V5 with little change except for how loud the signal is.

I will go pull each tube one at a time.

Billy

[uncle_bob]

Thanks Bob,

LOL...I will use due care when probing the circuit!! I sure don't want any high voltage drama...lol

On the V5 plate there is a 3.5 Hz signal at 1.7 VRMS

V4 plate 312KHz 23mv

V3 plate 107 KHz 50mv

V2 plate 10 Hz at 5 mv...I assume that is what I am hearing

V1 plate 10 Hz at 11mv

V6 grid 10 Hz at 14mv

V7 output tube grid 107 Hz and 120 Hz at 3.1 V

Not sure just what all that indicates.

Billy

Hi

Well if the main component on the output is 120 Hz, that's an easy one .... your power supply is getting into the output.  Unless you see the 10 Hz on V7, that's not what you are hearing.

The 312 KHz on V4 is the sort of thing that I suspected at first. The only way for that to happen is if the stage is oscillating. You either have a bum ground or a bad bypass. In this case the location of the ground and the size / type of the bypass could be what is wrong.

Bob

[Planobilly]

Pulling the tube at V3 stopped the oscillation. V3 is the reverb driver.

Billy

Be back in five min

[uncle_bob]

Pulling the tube at V3 stopped the oscillation. V3 is the reverb driver.

Billy

Hi

That would suggest that you poke around V3 and see what's going on. In addition to ground and bypass, lead dress (which wire goes where) can be an issue. Did the low frequency crap go away with V3 as well?

Bob

[woodchips]

Wow, what a lot of valves!

My very rusty experience is that motorboating is very often caused by a naff or missing cathode bypass capacitor, might be worth a look at the 22uF one on V3.

[Andy Watson]

The point of pulling V5 is that it is a low-frequency oscillator. I was thinking that may be it wasn't being fully disabled by the foot switch. However, looks like events have overtaken me!

I agree with Ian. Check that there are no unintended DC paths to the gird of V3 and from the cathode - it looks very close to the metalwork of the chasis in the photo.

[Planobilly]

I am looking at the circuit now.

I am getting a plate voltage of 424V which seems perhaps a little high to me.The current must not be to much due to the 2.2k cathode resistor ground. Why would there be a need for such a high plate voltage??

The voltage at the power supply choke is 435V, and what I would consider normal. The data sheet primary impedance  on the reverb driver transformer is 22800 ohms and the inductance is 95 H and 8 ohm secondary

I moved the heater wire a bit and re-soldered. No change. I also removed V4 with  all the other tubes in with no effect, so this seems confined to V3. I also moved the reverb transformer around to no effect. I also replaced the V3 tube with a new one with no change. I also removed V2 and the sound is still there.

As I remember the 435V on the V3 plate was indicated on the original Fender schematic.....perhaps that is correct...
Billy

EDIT: No unintended path to ground on the cathode

[Andy Watson]

I am getting a plate voltage of 424V which seems perhaps a little high to me.

There is no plate resistance as such, just the DC resistance of the transformer - so I would expect to see almost the power supply voltage on the plate.

The current must not be to much due to the 2.2k cathode resistor ground. Why would there be a need for such a high plate voltage??

And what is the voltage on the cathode?
And (whilst you're there) what is the voltage on the grid?

[Planobilly]

High Andy,
8.5V cathode
8mv on the grid

Billy

[Planobilly]

One other perhaps unrelated issue.

The small amount of hum that exist is reduced by touching the scope ground to the case. Checking the mains ground wire attachment point now.

[Andy Watson]

Voltages looks very reasonable.
The case is connected to the ground of the circuit?

[uncle_bob]

Hi

For proper measurements:

1) The chassis of the amp needs to be grounded.

2) The chassis of the scope needs to be grounded.

3) The ground lead on the scope probe needs to be clipped to a ground lug on the chassis near the point you are looking at.

If things are floating relative to each other ... who knows what you might see.

Best gess right now:

The plate lead on V3 is to close to the grid lead.

Bob

[Planobilly]

Hi guys,

Under the mains plug there is a earth ground to the chassis. Three inches away there is a chassis ground where everything else is grounded to.

The scope is grounded to the mains through the power cord. I could not duplicate the reduction in hum again.

I will remove and resolder the plate lead and change the position of the plate lead to put it on top and have the grid lead closest to the cathode lead. Having three wires that cross in close proximity jumping across the tube I hope will solve the issue....I hope!!

Thanks,

Billy

[Planobilly]

Moved the wires well far enough apart. No change.... I also disconnect the grid lead from V3 with no change. I don't think I have any more 25uf caps but I could change out the cathode cap for some higher value that I have to verify the cathode cap.

I think I will take a break and come back to this in a hour or so.


Thanks for all the help so far.

Billy

[uncle_bob]

Hi

Something like a 0.1uf ceramic from the cathode to ground may help.

Bob

[Planobilly]

Hi Bob,

Took a break and thought about this. Could I disconnect the plate voltage from V3 to see the effect?  Also I will place a 1 uf cap to ground. I assume you mean place a 1uf in series with the 25uf cathode cap.

Billy

EDIT: I disconnected the plate at the V3 tube and the sound is still there......So now there is no conduction through the tube...I assume the issue must be somewhere else...perplexing..lol

[uncle_bob]

Hi

The ceramic bypass goes in parallel with the electrolytic. Ceramic does better at high frequency than electrolytic.

Bob

[Planobilly]

Back...ok Bob I understand now...I will go dig up a Ceramic cap

[Planobilly]

That did make some slight change. The oscillations are more intermittent and  I have to turn up the gain to hear them well.

Any value in placing ceramics on V4a and b??? 

[uncle_bob]

Hi

If the electrolytics you got have high frequency issues adding another ceramic to V4 might be a good idea. I would go slow and check after each change.

Bob

[Planobilly]

I found the problem.

A few days ago AT&T installed a new router in my shop. I was not here at the time and it got installed under my workbench.

Interference from the router caused the amp to display a sound that sounded just like motorboating.

What a pain in the ass problem to have to deal with!!

Cheers,

Billy