Getting back to an old Hammond AO-29 project

[AmberB]

Hello, all!
I started on this project back in 2018 and basically ran into a wall with it so I put it away.  I recently dug it out again to work on it.  I was building a more or less copy of a Fender 5F6A with it, as far as the preamp goes.  When I put it away, it had a major buzz that I couldn't figure out at the time.  In the past couple of years, I've learned a lot about grounding, so I decided to pull it out and see if I could get it working.  I rewired all of the preamp grounds so that they terminate to the chassis at the input jack.  Lo and behold, that fixed the ground loop buzz.  Before that, I had grounds going everywhere...  Now I need to dial down the preamp, it's way too strong right now, it distorts much too easily.
I also found a real rookie mistake in the preamp circuit after redoing all the grounds.  I had a couple of the pins on the preamp tube connected to the wrong components...
Here's a picture of the chassis from back when I put it away.  When I built it, I pretty much left the power amp section the way it was on the original board.

[shooter]

don't have your schematic so..;
This should drop gain by converting a gain stage to CF.  It might even add tonal "flavor"


The Valve Wizard

[AmberB]

Upon looking at my circuits in the amp, I realized that there's an extra gain stage in this amp as compared to the 5F6A circuits.  Where the 5F6A uses the first stage tub for 2 inputs, one on each triode, I only have 1 input on this amp.  I used the other triode in the first tube as an extra gain stage, much like the Hoffman tube pedal 2 schematic.  I don't seem to have the schematic on my computer, so I'll have to find it and drop it here.  I usually work from schematics on paper, and I have a hand drawn schematic for the Hoffman Tube pedal.
Anyway, I'm sure that the extra gain stage isn't helping me here.  I wonder if I can run a cathode follower into the gain stage of a cathode follower...I have 1 too many triodes in this circuit...

[AmberB]

It took me a while to find this schematic, I had to figure out where to look...
This is the schematic for the preamp that I put into the AO-29.  The output of this goes to the PI tube for the power amp.

[AmberB]

It would seem that I didn't install the feedback loop on the gain stage of the cathode follower that drives the tone section...that may solve some of the problem...

[AnalogDok]

I'm using AO-29 iron to build 3 different amps.

I added a Bias adjustment trim pot for the AA1164 Princeton Reverb. I had to move the 22K to 27.6K to get it close but I wanted it closer and small changes in value really affected the bias. I basically didn't have an exact resister value I needed.

I had to change the fixed cathode bias resistor in my 5E3 Tweed Deluxe from 250Ω to 330Ω to get it about 85%.

The higher voltage on the trannys can cause the Bias to be too high and there is noise etc. The 5E3 with the normal/stock 250Ω was pushing 45mA with max at 41.7 at that voltage.

My next build is a AA763 Deluxe Reverb with the same iron. I bought 3, AO-29-13 amps off the bay. (non-field coil)

[AmberB]

The AO-29 iron does seem to be pretty good to build with.  There's a lot of heater power available for extra tubes such as tremolo or reverb.
I built an adjustable negative bias control for the power tubes in this amp build so that I could adjust the bias to suit the tubes being used, and I also installed a 1 ohm-5 watt resistor between the cathode and ground of each power tube so that I could measure the milliamps going through each tube for setting the bias.  I don't know if all of the AO-29 chassis used the 6V6 power tubes, but the one I have does.

[AnalogDok]

The Hammond AO-29 all use dual 6V6. The early ones had higher voltage transformers for the field coil speakers.
All of my donor amps have been AO-29-13. Stamped on the chassis, non-field coil versions.

The only issue is that they were made for 117VAC and the 125VAC of today causes a higher voltage in the 5V, 6.3V and 315V outputs.

[AmberB]

The chassis that I have is stamped "H-AO-29-7"  I would imagine the H is just for Hammond.  I'm not sure what the 7 designates.

[PRR]

....I'm not sure what the 7 designates.

https://organforum.com/forums/forum/electronic-organs-midi/hammond-organs/9180-calling-m3-enthusiasts-for-a-study-on-speaker-amp-changes

"...early field coil M3's ...AO-29-1 amp .... with permanent magnet would be around AO-29-7 and later AO-29-13..."

[tubenit]

Given the chassis and wiring layout,  I'd sure be inclined to remove that extra gain stage to cut down on noise.  The 5F6-A is a great design in it's own right.  IF you still want more oomph, what you could do is parallel 2 triodes the first gain stage for 30% more gain with almost no increase in floor noise, then go into the second gain stage with the cathode follower.


With respect,  Tubenit

[Stuff4bikes]

Given the chassis and wiring layout,  I'd sure be inclined to remove that extra gain stage to cut down on noise.  The 5F6-A is a great design in it's own right.  IF you still want more oomph, what you could do is parallel 2 triodes the first gain stage for 30% more gain with almost no increase in floor noise, then go into the second gain stage with the cathode follower.


With respect,  Tubenit

Don't want to high Jack this thread, is there recommended reading that explains parallel use of pre amp tubes?

[SILVERGUN]

Designing a Preamp with Parallel Triodes (ampbooks.com)

[AmberB]

Given the chassis and wiring layout,  I'd sure be inclined to remove that extra gain stage to cut down on noise.  The 5F6-A is a great design in it's own right.  IF you still want more oomph, what you could do is parallel 2 triodes the first gain stage for 30% more gain with almost no increase in floor noise, then go into the second gain stage with the cathode follower.


With respect,  Tubenit

That looks like it might be my best bet here.  There's a switched cap on the cathode of V1, what is the function of that, high pass or low pass?

[AmberB]

I had to replace the tube socket for the PI tube, the socket I replaced was an original socket that Hammond put in.  I couldn't get the thing to work and probing around with my nylon stick told me that there was a connection problem, most likely with the socket.  I just tested the amp, it works now, but holy crap, the distortion this thing is capable of!  All preamp distortion.  I understand why that Hoffman tube pedal was designed the way it wes, you can get some serious distortion from it.  The problem in this amp is that there's almost no room for clean.  When you get the gain up past about 2, you're into distortion, no matter the master volume setting.  If you back way off on the guitar's volume, it's easier to control, but that makes for a poor design. 
It's possible that the B+ voltage is a fair amount lower in the tube pedal so it's easier to control that thing.
I guess the next step is to parallel the first tube and see what happens.

[SILVERGUN]

If you are using that Hoffman pedal schematic, try eliminating the 1uf bypass cap on the 3rd stage. Or put it on a switch.
Just one of a few steps to try.

[AmberB]

I wired the first tube in the preamp, V-1, in parallel as was suggested, and that helped to tame the distortion quite a bit.  Now, of course, I have another problem with the amp.  Apparently the PI tube socket was not the problem.  I've switched out all of the tubes with no effect on the problem.  What happens is that if I leave it sit for a day and then try the amp, it works for 2-3 minutes, and then it just cuts out with intermittent bits of sound from the guitar.  Since the capacitors in the PI circuit section were all original ones, I decided to replace them all with some Sprague Orange Drops that were in my stash.  The Sprague caps are 600 volt caps and the original caps are 400 volt tubular caps with the ends sealed with some kind of epoxy resin.  It's possible that one of those caps is bad.
I can't test it until tomorrow to find out if that solves the problem.  I guess I'll know soon enough.
PS, when the amp loses sound, none of the B+ voltages seem to be affected, I checked with my meter...

[shooter]

it just cuts out with intermittent bits of sound

make sure you didn't sneak a 470K in place of a 470ohm in the PI circuit.  should be a quick check

[AmberB]

it just cuts out with intermittent bits of sound

make sure you didn't sneak a 470K in place of a 470ohm in the PI circuit.  should be a quick check

The Hammond AO-29 uses a split load phase inverter circuit.  I didn't change that, I just replaced the resistors in the circuit with the values shown in the Fender Princeton AA964 schematic.  Since the amp uses 6V6s, I figured that was a good way to rebuild the circuit.  I did check all the resistors to make sure that they were the proper values when I replaced the original capacitors in the circuit a couple of days ago

[AmberB]

I did some testing of the amp this evening, and it seemed to be working well for a while, but it eventually started having problems again with the sound dropping out and getting super distorted.  It occurred to me that I had not checked the bias voltage for the power tubes since starting to work on the amp again.  When I had first started working on rebuilding the amp, I had built an adjustable negative bias circuit for the power tubes, so I figured I should check it just to see where it was at.  I had put a 1 ohm 2 watt resistor between the cathode and ground of both power tubes to make biasing easier, so that's where I checked.  Basically, whatever the voltage reads on the side of the resistor hooked up to the cathode, that's the amperage flowing through the cathode.
The AO-29 puts about 389 volts on the plates of the 6V6s, at least this one does...
When I tested the bias voltage, the idle current was 30ma, rather too hot!
I adjusted the bias current down to about 20ma, which is supposed to be somewhere around 60% at idle.  Now I have to test it again...

I think that changing out the original caps in the PI did help, because the amp didn't just cut out this time, it was more of a fade away.  I was probably overheating the poor 6V6s with that hot bias and playing loud...

[shooter]

I had put a 1 ohm 2 watt resistor between the cathode and ground of both power tubes

the idle current was 30ma, rather too hot!

just to clarify;
you took the reading you got and divided by 2 to get the 30mA?, since the 1 resistor supplies 2 tubes

[PRR]

> about 389 volts on the plates of the 6V6s...the idle current was 30ma, rather too hot!

Whether one tube or two, that's 11.7 Watts, perfectly fine for 6V6. 20mA sounds icy-cold.

[AmberB]

I had put a 1 ohm 2 watt resistor between the cathode and ground of both power tubes

the idle current was 30ma, rather too hot!

just to clarify;
you took the reading you got and divided by 2 to get the 30mA?, since the 1 resistor supplies 2 tubes

That was each tube.  I have a 1 ohm resistor for each tube's cathode.  That way I can test each tube separately and get some idea if they're fairly well balanced or not.

[AmberB]

> about 389 volts on the plates of the 6V6s...the idle current was 30ma, rather too hot!

Whether one tube or two, that's 11.7 Watts, perfectly fine for 6V6. 20mA sounds icy-cold.

11.7 watts per tube?  That does seem a bit much for a 6V6, although the RCA book does say 14 watts max plate dissipation...I suppose it could be these old and tired tubes in the amp...

[AmberB]

Of course, I'm talking about idle current with negative bias on a push pull setup.  The current will increase when I start playing the guitar through the amp.

[shooter]

If you have a record feature or MIN/MAX feature on your meter
clip the meter across one of the resistors, play soft-to-hard-to-soft for 5 minutes, you'll be surprised at the total current swing

[PRR]

> The current will increase when I start playing

And most or all that increased power input gets dumped to the *speaker*, not the tubes.

12W is fine forever on any semi-healthy 6V6. Fender sold a recent Champ idled at 17 watts or more.

[AmberB]

Well, I still have no clue as to why the amp fades out after you've been playing it for a few minutes...adjusting the bias didn't change anything...

[PRR]

> the amp fades out after you've been playing it for a few minutes...

With amp off and cold, measure resistance from each grid to ground. Is any of them "infinity"? (Open circuit)

[AmberB]

> the amp fades out after you've been playing it for a few minutes...

With amp off and cold, measure resistance from each grid to ground. Is any of them "infinity"? (Open circuit)

I measured all the grids for the preamp tubes, and I found that the grid for the input of the PI tube reads infinity.  I can see why now that I look.  I have a capacitor between the volume control and the grid of the amp triode for the split load PI that I'm using.  The Hammond was using a split load phase inverter and I didn't change that configuration when I rebuilt the amp.  I'm basically using a Fender 5F6A preamp circuit into a Fender Princeton AA964 split load invertor circuit.  The original Hammond had a capacitor on the input of the original PI circuit and apparently I left it in when I rebuilt the preamp section.
Basically, the triode has no path to ground for that grid because it's isolated from the volume control.  I guess I need to remove that cap and wire the volume control direct to the grid and test it again.
I probably would not have found that basic mistake if you wouldn't have asked, so thanks for that!

[AmberB]

I finally got around to testing this amp today, and it seems to work now that I gave the PI tube grid a path to ground through a 470kohm resistor.  I like the way this amp sounds...