JMI-Vox AC15 build PI buzz problem

[Interstellareye]

I’m new to tube amp repair. I have a JMI-style AC15 built from the layout in Stephen Grosvenor's "Service Engineer's Guide to Vox AC30", which follows the 1960 JMI-Vox AC15 schem. I can attach the schem but it’s the standard Vox schem and on this forum elsewhere.

having 2 issues with this amp that I'm hoping someone can help me figure out. The worst of the two is a high pitched hum/buzz/oscillation that won't resolve no matter what I've tried. I used an iPhone audio analyzer app to determine that the frequency of the buzz is about 1580Hz—- If I'm reading/understanding app correctly.

It's not an overpowering buzz, but it's more than just hiss, and I don't think it should be buzzing like this.

The buzz is inaudible with volumes down, and gets louder as I turn up the volume on either channel. It's stronger on Channel 1 (vibrato channel), weaker and muffled on channel 2 (EF86 channel).

I've pulled all of the tubes one at a time to see which if any effects the buzz, and only by pulling the Phase Inverter tube is the sound eliminated, removing other tubes does not effect the buzz.

I've tried 4 different ECC83's in this PI position, including a vintage Mullard pulled from another working Vox AC10 PI, but they don’t affect the buzzing much. Maybe slightly louder or quieter with different tubes.

I used a rubber dental pick/brush and DeOxit to clean the Phase Inverter socket, but didn't notice any change. I've wiggled the socket pins and wires a bit, adjusted the tube in socket a bit while powered, didn't notice any difference in buzz.

I've checked all of the continuities and layout many times over. Everything looks right. The grounds don't seem to be looped anywhere, and are arranged just like a JMI AC15 as far as I can tell.

I have a cheap little digital oscilloscope that I used to try to track down the noise sig using a 1KHZ sine wave (from a keyboard synthesizer), but I'm very inexperienced with this method. I noticed that the buzzing doesn't appear to come into the signal chain until after the volume pots. I don't think I see the interfering signal until after the post-potentiometer caps that lead to the Phase Inverter grids. [edit: might be confused about this— maybe true for ch2 (EF86 ch) only]

I'm not sure how else to go about isolating where the interference signal is coming from.

--------

The second problem is that I don't hear any difference between Vibrato and Tremolo when I turn the switch. There’s a thud when the switch is turned, and the lead goes to ground, but doesn’t change the effect.
This leads me to think the tremolo side of the dual effect is already grounded somehow before it’s even switched out. Just a guess.

[Jalmeida]

Hello,

In my experiences Stephen’s books ha had errors. I would check it against the schematic.

I used this layout, but this one swapped the wires at C7 and R20. You can maybe compare to this one. And then against the schematic.

[Jalmeida]

Also, make sure C42 is actually a .033-.050uf capacitor. Most schematics including Stephen’s shows the wrong value which was transplanted from an error JMI made on some versions of their schematics that actually was from the AC30.

[Interstellareye]

I used this layout, but this one swapped the wires at C7 and R20. You can maybe compare to this one. And then against the schematic.

I noticed this switcheroo on Stephen’s layout, as well as a few other discrepancies with the schematic. I made the changes with C7 and R20 hoping it was part of my problem, but unfortunately was not.

[Interstellareye]

Also, make sure C42 is actually a .033-.050uf capacitor. Most schematics including Stephen’s shows the wrong value which was transplanted from an error JMI made on some versions of their schematics that actually was from the AC30.

Yep, found this one also and replaced with a .033 which is what was in a 1960 Vox Schem I saw. Was really hopeful here also, but again, no dice.

There’s also a missing link to the depth trim on his layout that disconnects the vib oscillator. That caused me a lot of frustration until I noticed it.

I wrote him about these things to make sure he knew and found out that he had a corrections page on his website already.

[Jalmeida]

I’m new to tube amp repair. I have a JMI-style AC15 built from the layout in Stephen Grosvenor's "Service Engineer's Guide to Vox AC30", which follows the 1960 JMI-Vox AC15 schem. I can attach the schem but it’s the standard Vox schem and on this forum elsewhere.

having 2 issues with this amp that I'm hoping someone can help me figure out. The worst of the two is a high pitched hum/buzz/oscillation that won't resolve no matter what I've tried. I used an iPhone audio analyzer app to determine that the frequency of the buzz is about 1580Hz—- If I'm reading/understanding app correctly.

It's not an overpowering buzz, but it's more than just hiss, and I don't think it should be buzzing like this.

The buzz is inaudible with volumes down, and gets louder as I turn up the volume on either channel. It's stronger on Channel 1 (vibrato channel), weaker and muffled on channel 2 (EF86 channel).

I've pulled all of the tubes one at a time to see which if any effects the buzz, and only by pulling the Phase Inverter tube is the sound eliminated, removing other tubes does not effect the buzz.

I've tried 4 different ECC83's in this PI position, including a vintage Mullard pulled from another working Vox AC10 PI, but they don’t affect the buzzing much. Maybe slightly louder or quieter with different tubes.

I used a rubber dental pick/brush and DeOxit to clean the Phase Inverter socket, but didn't notice any change. I've wiggled the socket pins and wires a bit, adjusted the tube in socket a bit while powered, didn't notice any difference in buzz.

I've checked all of the continuities and layout many times over. Everything looks right. The grounds don't seem to be looped anywhere, and are arranged just like a JMI AC15 as far as I can tell.

I have a cheap little digital oscilloscope that I used to try to track down the noise sig using a 1KHZ sine wave (from a keyboard synthesizer), but I'm very inexperienced with this method. I noticed that the buzzing doesn't appear to come into the signal chain until after the volume pots. I don't think I see the interfering signal until after the post-potentiometer caps that lead to the Phase Inverter grids.

I'm not sure how else to go about isolating where the interference signal is coming from.

--------

The second problem is that I don't hear any difference between Vibrato and Tremolo when I turn the switch. There’s a thud when the switch is turned, and the lead goes to ground, but doesn’t change the effect.
This leads me to think the tremolo side of the dual effect is already grounded somehow before it’s even switched out. Just a guess.

I am not a sharp or experienced enough Tube amp technician to know what would cause that frequency of buzz/hum. So take that for what it’s worth.

Not thinking this would cause that frequency, but check the polarity of the 8uf filter cap on the Anode side of the PI tube. Then check all the values of the resistors on the cathode end. If any of those are not correct then I would imagine the noises you get could be a bit variable. Maybe ring those values. If they are NOS resistors they may have drifted. Maybe reflow those solder joints in the off chance you have a resistive or capacitive solder connection.

[Jalmeida]

The second problem is that I don't hear any difference between Vibrato and Tremolo when I turn the switch. There’s a thud when the switch is turned, and the lead goes to ground, but doesn’t change the effect.
This leads me to think the tremolo side of the dual effect is already grounded somehow before it’s even switched out. Just a guess.

I know you said you checked layout. But looking at this layout diagram and matching against the schematic I see these underboard traces that if you got one tag off would cause that Tremolo switch to be always grounded.

[Interstellareye]

check all the values of the resistors on the cathode end. If any of those are not correct then I would imagine the noises you get could be a bit variable. Maybe ring those values.

Thanks for all these suggestions, much appreciated.

What does it mean to say the noise could be “variable” ? Also, what do you mean by “ring” the values?

I did indeed use NOS resistors and poly caps throughout the build (along with new F&T electrolytics), but I was very careful measuring and selecting exact values only.

I probed with my scope again, this time w/o audio input, and it seems like there’s a lot of signal(or voltage?) coming into the vibe channel pot as I raise the volume.

[Interstellareye]

underboard traces that if you got one tag off would cause that Tremolo switch to be always grounded.

I will re-examine that section against your layout, thanks.

[Jalmeida]

check all the values of the resistors on the cathode end. If any of those are not correct then I would imagine the noises you get could be a bit variable. Maybe ring those values.

Thanks for all these suggestions, much appreciated.

What does it mean to say the noise could be “variable” ? Also, what do you mean by “ring” the values?

I did indeed use NOS resistors and poly caps throughout the build (along with new F&T electrolytics), but I was very careful measuring and selecting exact values only.

I probed with my scope again, this time w/o audio input, and it seems like there’s a lot of signal(or voltage?) coming into the vibe channel pot as I raise the volume.

By ring, I mean check resistance with a multimeter.

By variable I mean if one value of the resistors are off and only one phase side of the PI’s signal is affected, then the frequency of the buzz or hum may vary. It may be shifting phase slightly on one side only. As opposed to say chasing a 60hz or 120hz mains noise.

[HotBluePlates]

... I have a JMI-style AC15 built from the layout in Stephen Grosvenor's "Service Engineer's Guide to Vox AC30", which follows the 1960 JMI-Vox AC15 schem. ...

... The worst of the two is a high pitched hum/buzz/oscillation that won't resolve no matter what I've tried. ...

The buzz is inaudible with volumes down, and gets louder as I turn up the volume on either channel. ...

I have that book, but it's in a box somewhere.

Do you have the Vox-style layout where the output transformer wires come through the chassis near the preamp input jacks?  IF yes, you probably need to add a piece of metal bolted to the chassis to screen the jacks from the output transformer wires.

[Interstellareye]

It’s a Vox style two part chassis and I used a screen below the jacks.

The Service Manual book is great, but I’ve learned there are some discrepancies/errors to watch out for.

[Interstellareye]

By ring, I mean check resistance with a multimeter.

By variable I mean if one value of the resistors are off and only one phase side of the PI’s signal is affected, then the frequency of the buzz or hum may vary. It may be shifting phase slightly on one side only. As opposed to say chasing a 60hz or 120hz mains noise.

it seems like it’s been a stable 1580hz oscillation ever since I first powered up the amp.

I’ve now tried disconnecting both of the channels before their respective volume pots and it still did not alter the buzz.

[Jalmeida]

... I have a JMI-style AC15 built from the layout in Stephen Grosvenor's "Service Engineer's Guide to Vox AC30", which follows the 1960 JMI-Vox AC15 schem. ...

... The worst of the two is a high pitched hum/buzz/oscillation that won't resolve no matter what I've tried. ...

The buzz is inaudible with volumes down, and gets louder as I turn up the volume on either channel. ...

I have that book, but it's in a box somewhere.

Do you have the Vox-style layout where the output transformer wires come through the chassis near the preamp input jacks?  IF yes, you probably need to add a piece of metal bolted to the chassis to screen the jacks from the output transformer wires.

👆🏻👆🏻👆🏻100%.


My AC15 build with Vox style chassis would feedback(not buzz, but oscillation of sort) really bad until I added that.

[Interstellareye]

I’m starting to wonder if this buzz is from a lack of heater ground. Someone on another forum brought this up, and it got me thinking about many builders who mention adding a center tap to their heaters to eliminate noise.

I’m reading that Vox grounded one heater directly to chassis on the AC30 (off one ECC83).

I also have an AC10 that I just examined and sure enough, there is a wire to ground on one heater wired here also.

[Interstellareye]

to update, in case someone with similar noise issues comes across this topic, it seems the floating heater voltages are what was causing this oscillation/noise.

Instead of grounding directly to chassis as I believe Vox did, I used an artificial center tap consisting of two 100ohm resistors wired from both heater lines to ground.

According to the sources I read, the artificial center tap is more efficient than simply wiring one filament to ground.

I don’t know why it is that the heater ground is left off of all of the JMI-style AC15 layouts I’ve seen. I think it’s very unlikely that AC15s were built with floating heaters.

[pdf64]

The 1960 OA/031 schematic shows the heater circuit connected to circuit common at V1 pin 4 https://el34world.com/charts/Schematics/files/Vox/Vox_ac151960.pdf

As mentioned elsewhere, some / any DC elevation of the heater circuit will take a bit of stress off the EZ81 heater - cathode insulation.
And will enable audio valves that would otherwise be too buzzy (due to their h-k insulation being leaky) to be usable.
The output valve cathodes being a suitable and readily available elevation node.