Hoffman AB763-6G16: No tremolo (solved), hum, lack of power

[Wil]

Hi,
Thank you for accepting my registration.
EDIT : this message is only for the history of my recent "re-build". See the latest post for update since I always change the title according to the issues I'm encountering. Thank you.
EDIT : I changed the titled of this thread since someone answered my initial question. Unfortunately though, problems persist.
Some time ago (4 years), I managed to build my first amp. An Hoffman Ab763/6G16. Everything went fine for a while, then I started modifying things... then got some volume drop and couldn't find the solution.
I'm working a lot and haven't got much time to build and tweak amp. But recently, I had some time off. And I tried to heal this amp.
What I've done:
- Re-done the heater circuit (it was a mess...shielded with copper tubes).
- Re-done the ground scheme
- Change the filter caps (put them back in the dog house). It's a bit sketchy, but everything lands where it should. No shorts from + side to ground of any of the caps.
- Added some mods (Fat switch (treble cap value), tremolo controls are the back of the amp + switch on/off on the intensity pot) should've done that after everything is fine

Issue
- When I fired up the amp the first time I had a very loud buzz and a burning smell (didn't try to inject signal... turned it off immediately)
-> I identifed a problem : I accidentely connected the preamp plate resistors buss to ground... Connection is now clipped.
All ground connections are good but will double check tomorrow.
No when I power up the amp, the heater circuitry looks fine, PT voltages are fine, then, when I turn on the standby switch, I get a long increasing hum. See video:

VIDEO : https://tinyurl.com/yd7yo2ee
Sorry for the vertical orientation...

I'll try going upstream and adding the tubes one by one to locate the issue.

Thanks a lot in advance for anyone willing to help.
Have a nice day,
Wil from Belgium.

[sluckey]

Connectivity between chassis and heaters (both - center tap as well, but that's fine I reckon...) - but still shows 6.3/6.4 VAC - is that normal?

Perfectly normal.

[Wil]

Hi Sluckey,
Thank you for your answer.
Can you please explain to me why there is connectivity to ground from the heaters (when disconnected)?
What's the use of a center tap then apart from just being another connection to ground?
What's the use of 100ohm resistors when creating an artificial center tap?
Electrons will always take the shortest and less restrictive path, won't they?

Regarding the voltages at the standby switch. Shouldn't it be higher (400ish?)?
Does it indicate a faulty rectifier? It is a JJ GZ34S.
(Edit: I guess the value is not the same when disconnected from the main board and that Ohm's law is my friend)

Thank you.
Have a nice day.
Wil

[d95err]

If one side of the heater winding is connected to ground, then your multimeter may see the other side as continuity, since the resistance of the winding is so low (something like 0.0x ohms).

You can tie one end of the heaters to ground, but you probably get less hum if you reference it to ground via two 100 ohm resistors.

I would go through the heater wiring again to find where it is connected to ground.

[Wil]

Hi d95err,
The measurements have been taken while only the center tap was connected to ground, both wires of the winding were disconnected.
That's maybe why I'm getting connectivity to ground (center tap). Sometimes, my brain starts to figure simple things out when I'm writing here...

[sluckey]

Can you please explain to me why there is connectivity to ground from the heaters (when disconnected)?

See that green (with yellow stripe) wire coming out of the PT? It's connected directly to chassis. That's the center tap for the filament winding so that means it is connected to the filament winding. The filament winding is just a few turns of heavy wire and the resistance is probably less than 1Ω from end to end (green to green wire). So there is probably only about 0.5Ω resistance from the green/yellow center tap to either of the green wires. If you disconnect the center tap from chassis you will no longer measure any resistance from the green wires to chassis.

The purpose of the center tap is to provide a ground reference to the filament circuit. This will reduce 60Hz hum in the amp. You can prove this by listening to the amount of hum coming from the speaker. Now disconnect that center tap and you should hear an increase in hum. You can use the real center tap or you can use the 100Ω resistors artificial center tap to reduce filament hum, but you should not use both. Since you have a real center tap connected to ground I suggest you remove those 100Ω resistors on the board or at least disconnect any wires going to them.

When the standby switch is in standby position, the rectifier is disconnected from the filter caps. The only voltage at the switch is the rectified pulses from the rectifier tube. Your meter cannot measure the peak value of these pulses. Instead, it averages the pulses and the voltage in the meter display will be much lower than what you may expect. Switching to operate connects the filter caps. Now they will charge up to the peak value of the rectified pulses and the resulting smooth dc voltage will be much higher.

[Wil]

Thank you for your answer.
Unfortunately, my problems  persist !
All ground connections seem good but will double check tomorrow (will check all the connections as well with the schematic).
Now when I power up the amp, the heater circuitry looks fine, PT voltages are fine, then, when I turn on the standby switch, I get a long increasing hum. See video:

VIDEO : https://tinyurl.com/yd7yo2ee
Sorry for the vertical orientation...

I'll try going upstream and adding the tubes one by one to locate the issue.
It's not 50hz, not 100hz... sounds like bad grounding but can't seem to locate the issue.

Any idea?

[sluckey]

Everything went fine for a while, then I started modifying things...

You've messed up something with your mods. I suggest you remove all the mods and get back to the original Hoffman circuit. I bet you will find what went wrong in the process.

[pdf64]

Is there adequate bias voltage at the power tube control grids, socket terminal #5?
It’s bad practice for a standby switch to cause surge current in the rectifier plates circuits (hot switching), eg by disconnecting the rectifier from its reservoir cap.

[Wil]

Hello,
I put numbers on the issue.

Issues are the following:
- No tremolo
- Lack of headroom
- Lack of power
- Lot of heat
- some buzz, hum (might raise the first filter caps value)

- pretty cool mellow vintage sound for a Danny Kirwan fan...

Measurement were taken with a Fluke 101 multimeter.


Everything seems to be pretty high.
Could the dropping resistors alone cause this situation?
Filter caps are brand new 22µF 600V.

What would you recommend? What should I look after?

Thank you.

[sluckey]

Divide and conquer. Tackle the redplating issue first.

Step 1.
Pull all little tubes and disconnect the feedback wire from the speaker jack (or board, whichever is easier). Adjust the bias pot for maximum negative voltage on pin 5 of the output tubes. What voltage do you read on pin 5 of each output tube? What millivolt readings do you have for pin 8 of each output tube? Is it still redplating?

Step 2. (DO NOT PROCEED IF AMP IS STILL REDPLATING!)
Insert a 12AT7 into V6 socket. Check voltages. Pins 6, 7, and 8 should be similar to pins 1, 2, and 3. (Your chart looks like you just copied the numbers from V5 for pins 6, 7, and 8.)Post your new voltage readings. Does the amp make any abnormal sounds? Reconnect the feedback wire. Any abnormal sounds or redplating?

Post your findings.

[Wil]

Divide and conquer. Tackle the redplating issue first.

Step 1.
Pull all little tubes and disconnect the feedback wire from the speaker jack (or board, whichever is easier). Adjust the bias pot for maximum negative voltage on pin 5 of the output tubes. What voltage do you read on pin 5 of each output tube? What millivolt readings do you have for pin 8 of each output tube? Is it still redplating?

Step 2. (DO NOT PROCEED IF AMP IS STILL REDPLATING!)
Insert a 12AT7 into V6 socket. Check voltages. Pins 6, 7, and 8 should be similar to pins 1, 2, and 3. (Your chart looks like you just copied the numbers from V5 for pins 6, 7, and 8.)Post your new voltage readings. Does the amp make any abnormal sounds? Reconnect the feedback wire. Any abnormal sounds or redplating?

Post your findings.

I imagine by little tubes you mean 12a*7. So, I should only have the power tubes and rectifier in place.
I will try that and post my results.

[Wil]

Hi Sluckey,
Here are the measurements:

With only the power tubes and rectifier plugged in:
Pin 5 of each power tube measures at -55.2 VDC
Cathode to ground measures at 5mV (V8), 8mV (V7).


(maybe my title is misleading, I don't know but I get some red glow at the end of each side of the plate when biased at 22mA)

With V6 plugged in, I get:
Pin 1: 299VDC
Pin 2: 84VDC (sounds like a big ground loop when probed)
Pin 3: 130.3VDC
Pin 6: 308.5 VDC
Pin 7: 85.4 VDC (sounds like a smaller ground loop)
Pin 8: 130.6 VDC

Anything abnormal?
Please note the PI resistor are 47k/47k and not the usual 82k/100k.

Have a nice day,
Wil

[shooter]

can you post/link to the schematic used to build this, would be a great help
pin 2 and 7 are typically zero volts DC, IF V6 is a 12A?7

[sluckey]

V6 voltages look OK. Why do you have 47Ks for plate resistors?

[Wil]

V6 voltages look OK. Why do you have 47Ks for plate resistors?

I read here:
http://music-electronics-forum.com/showthread.php?t=23490
that 47k PI resistors better fit the 12AT7 tube (while 82/100k better fit the 12AX7) but could drive the power tubes a little harder.
What's your opinion on this?

can you post/link to the schematic used to build this, would be a great help
pin 2 and 7 are typically zero volts DC, IF V6 is a 12A?7

This is the basic Hoffman ab763 schematic with the 6g16 tremolo unit and a couple of Rob Robinette mods (tone cut, fat switch, tremolo controls at the back of the amp, NFB resistor switch, bright switch, mid pot)

Find below a picture of my build (still in prototype form, though):

[mresistor]

recommendation -  side note - is that an automotive crimp connector on your HV wire of the PT? I would recommend to not use those in a guitar amplifier. It would be better to strip the sheath and then twist the bare wires together in line parallel so to speak not at a right angle, solder the splice and then insulate with heat shrink tubing.

[sluckey]

The barrel crimp is on the bias lead, not the HV wire.

[mresistor]

I just figured that out..  thanks  I still wouldn't use those crimp connectors in an amp.

[Wil]

Regarding the bias lead wire from the PT, it was like this when I bought the amp and it never failed so far so I'll leave it like that, I reckon.

Never managed to have the tremolo work as long as I can remember.
Every resistor is within spec (can't say for the caps, yet), the wiring is correct, V5 is a known good 12AX7, V4 as well and I don't think the dropping resistors (1,5k and 5k instead of 10k/10k - need to be changed) have any effect on this since since it's sourcing B+ for plate 2 before the dropping resistors.
The only different thing in this area is the 0.001 cap instead of 0.01 between the two 220k resistors that bring the signal of each channel and V6 pin 2 and that it's not grounded with a footswitch but with a DPDT switch (it wouldn't work either when it was supposed to be footswitch-activated).
Any idea on the tests I could run and things I could measure?
I think I will measure the pots, they are the only resistors I haven't checked.

Have a nice day.
Wil

[Wil]

The only abnormal thing I found so far is that I can't get accurate measurements of the 1M resistor going from the V5 cathode (pin 3) to ground. But I think that's because it's wired in parallel with a couple of caps.
I'll lift one end to see if it gets any better.
I thought I had found the problem with two missing wiring going from B+ (after choke) to screen grid resistors to screen grids but they're actually very well there. I'm using an older Hoffman schematic and the newer Hoffman design puts the screen grids of the tubes sockets.
On this older schematic there was a missing 22uF going from V5pin3 to ground in parallel with a 4.7k resistor. In my amp, it's in (see the tantalum caps of the hi-res picture).
The search continues...

[sluckey]

The only abnormal thing I found so far is that I can't get accurate measurements of the 1M resistor going from the V5 cathode (pin 3) to ground.

There is no 1M from V5-3 to ground. If you have one, then you have a wiring error and that could very well be why you have no tremolo. Nothing else should be affected.

V5-3 should have a 4.7K and a parallel cap to ground. There is a 1M from pin 2 to ground. Maybe you have those two wires swapped? Can't say from looking at the pic, but if this is the case, the tremolo cannot work.

[Wil]

Pin3 is not directly going to ground but through the aforementioned 1M resistor + the DPDT switch but also through the 4.7k/22uF you mentioned.

Could this be a mistake in the schematic?

I've checked a vibroverb schematic and it's there as well.

[sluckey]

...aforementioned 1M resistor + the DPDT switch

The schematic is correct. Be aware that when the 1M is grounded through the switch, the tremolo will be disabled. Switch must be off (contacts open) to enable tremolo.

You should be able to get an accurate measurement of that 1M when the switch is open. When the switch is closed you should read 4.7K

[Wil]

...aforementioned 1M resistor + the DPDT switch

The schematic is correct. Be aware that when the 1M is grounded through the switch, the tremolo will be disabled. Switch must be off (contacts open) to enable tremolo.

You should be able to get an accurate measurement of that 1M when the switch is open. When the switch is closed you should read 4.7K

Yes now I get 0.96M when the switch is closed as well as open. It takes some time to get there, which makes me thing they're might be some capacitance somewhere in the path.
The switch has absolutely no effect and I'm therefore not getting 4.7k across this resistor when the switch is closed. I'm gonna go upstream towards V4 to see if they're could be a short somewhere that's bypassing the tremolo circuitry which is exactly like the Hoffman schematics. Every cap and resistor (pots included) is within spec and I followed the layout and wiring of the Hoffman schematic.

If you have any idea, I'm all ears.

[sluckey]

Measure resistance from V5 pin 3 to chassis (one probe on chassis, the other probe directly on pin 3 at the socket). What have you?

No need to look upstream at V4. V4 has nothing to do with the trem.

[Wil]

Measure resistance from V5 pin 3 to chassis (one probe on chassis, the other probe directly on pin 3 at the socket). What have you?

No need to look upstream at V4. V4 has nothing to do with the trem.

What have you? Hm... well. Some crazy readings. A video speaks better than words:
https://drive.google.com/open?id=1pVcFfZyolx_8IVcfTOniTgq2yoRyDQ8T
(If you don't have access via this link, I'll upload it somewhere else)

I'd love to know what this variable value means.
Same results when the DPDT switch is closed or open.

Thank you for your time.

[sluckey]

Assuming the power is off for the amp... It means the 4.7K resistor is open or there is a bad connection to ground. That cap is causing the crazy changing readings.

… the other probe directly on pin 3 at the socket

There is a reason I specified this rather than just check the resistor.

[Wil]

Assuming the power is off for the amp... It means the 4.7K resistor is open or there is a bad connection to ground. That cap is causing the crazy changing readings.

[quote author-me]… the other probe directly on pin 3 at the socketThere is a reason I specified this rather than just check the resistor.

Took some measurements before going to work. The 4.7k resistor is within spec. But, because there's a but and that's what interesting. When I measured from V3 to ground, I would get what you saw. And then tried to clip the lead going to ground from that resistor and, to my surprise, that end of the resistor came off of the solder joint! A super shiny solder that, apparently, is at fault.
I'm on a business trip for the next two days but can't wait to give this mini ground buss a little heat!

If this solves the problem, you'll have to give me your address so I can send you a good case of Belgian beers!

[sluckey]

Hopefully that will fix the tremolo. The low power is probably a separate issue. We'll tackle that after the tremolo is fixed.

[Wil]

Sluckey,
The tremolo is working! Thank you. It's a bit weak but I'm sure this can be tweaked (from weak to tweaked, I'm making play words now...).

Now I need to tackle the noise/hiss/buzz/hum issue which has gotten worse somehow.
Please note I'm living in Europe, so my mains are in 50hz.

I'm getting a lot of hum of both channel. Even some squealing on the normal channel (seems like there's a mockingbird living in my normal channel...).
The reverb is gorgeous but amplifies another type of hum (not the same that affects the normal and vibrato channels).
The hum is not linear though. I'm not sure if you can hear it but there are two level of hum that go on alternatively every second or two.
There's also a ticking sound, just like pendulum clock that is not related to the tremolo speed (was there already when the tremolo wasn't functional)

Once again, a video speaks better than my words
https://drive.google.com/file/d/1t9xAfucBsV82H6uQRA_rFGnOOzM2N1im/view?usp=sharing

I guess there has to be a ground loop somewhere and I might look into moving the filament wires. That's the first next step.
Once again, thank you for your help.

[Wil]

The reverb was noisy, it's not anymore!
- I raised the red wire from the reverb transformer
- Isolated the reverb jacks and grounded them to the preamp buss bar
- improved the lead dress here and there

What a difference! It's now almost dead quiet even on 10. Never wanted to believe how lead dress could actually impact the amp overall background noise, now I do.

I noticed something funny, you tell me if that normal of not.
When I measure the resistance between lug 2 to ground on the normal channel pot, it goes from 0 to 1M. When I do the same on the vibrato channel volume control, it goes from 0 to approx. 250k then goes back to 0 when on ten. I frankly don't get it (I'm a little tired tonight so I might want to check again tomorrow).
The operation of this volume pot is way faster, I mean, way less resistance than the other pots. I might have changed it at some point, I can't remember. Maybe I picked the wrong value and that's where the lack of power could come from (will measure tomorrow). I remember my walls shaking when I first completed this build and cranked this amp for the first time and it's not the case anymore. Can't remember either if I was using an Eminence Swamp Thang (102dB sensitivity) or my WGS G12C already (97dB sensitivity - first generation G12C) - 5dB sensitivity delta is a lot.

I also replaced the 0.001 PI input cap (DRRI value) with a 0.01uF (AB763 value - Orange Drop 6PS) and, oh my god, it sounds way way fuller.

Some problems remain:
- Still have some background noise, too much to my liking (and I learnt to be tolerant toward background noise...)
- Still have some ground noise and squealing on the normal channel when cranked. I'll swap the tube and then investigate the circuit if the tube doesn't solve it
- Still have that lack of power (The first time I completed the build I remember how the amp would just scream and push each and every note is such an articulate manner, but then again he could just be the speaker)
- I have what I would call static ground noise (the kind of ground noise that vanishes when you touch some grounded component like the chassis of the bridge of your guitar. I will check each and every ground connection individually.

Still have to change the dropping resistors to 10k/10k and I might raised the first filter stage from 44uF to 60uF and see if there's any improvement on the background noise. But first, I'll improve the overall lead dress based on the wonderful article written by RG Keen (who I remember helped me with my first build! Will always be grateful for that).

Well, there was some real improvement today! And Sluckey's solution for the tremolo, this is pretty encouraging!
Thank you in advance for your advice.

[sluckey]

I noticed something funny, you tell me if that normal of not.
When I measure the resistance between lug 2 to ground on the normal channel pot, it goes from 0 to 1M. When I do the same on the vibrato channel volume control, it goes from 0 to approx. 250k then goes back to 0 when on ten. I frankly don't get it (I'm a little tired tonight so I might want to check again tomorrow).

I wouldn't worry about it. The settings of the tone controls greatly affect the resistance readings you're trying to do. If you want similar resistance readings you'll need to set the tone controls the same for both channels.

[Wil]

Hi Sluckey,
I managed to measure the volume pot before going to work and it's all fine. 1M. It's a linear pot, so the amp only starts breaking up at 8. That's pretty strange, though, that normal and vibrato volume pots don't read the same values (with all controls on 1 and only playing with the volume pot).
Since the normal channel is squealing on 10, I'd say there's something wrong with it and not the vibrato channel.
Maybe that lack of power is just psychological, because of that pot (we're all used to that big jump in volume on Fender amps between 1 and 2). But I'd rather put numbers on the matter.
Since all my tubes voltages look fine, is there a way to measure the output power of my amp?

[shooter]

is there a way to measure the output power of my amp?

inject a sine wave at input, put a scope on the speaker, adjust sine wave to just before clipping, measure sine wave Volts rms, square the reading then divide by speaker impedance

[Wil]

I was thinking of a more jerry-rigged way to do it 

I rewired the amp with the ground scheme recommended on this very website.
It sounds very good but I have some hiss. Too much to my liking.
I've read on forums about the SuperSonic 22W amp that using a 12AU/AY/7 on V4 could help with the problem as well as increasing V4's plate resistors value to 180/220k. Worth a try... But still, I'm just guessing here since I still need to replace the dropping resistors to 10k/10k and might raise the A node filter capacitance value to 60uF instead of the current 44uF.
Should be done this weekend hopefully.
Anyway, a big thank you already for your help.