New Project -> Hammond AO-63 to Modded 5F6 Bassman

[turbofeedus]

Picked up a used Hammond AO-63 chassis, this is the one with two amps in one.

sluckey AO-63.pdf

Sluckey did a very cool conversion of this amp into an Ampeg mashup, taking advantage of the 7591 output tubes.
I'm going a different direction; gutting the amp and build a modded 5F6, keeping the 7591 and adding switchable cathode or fixed bias.
Several other mods as well (as suggested by Robrob); channel jumping switch, defeatable NFB, "Raw" control instead of a mid, frondelli MV. Also adding DC heater elevation for the cathode follower.

Here's what I've got so far:

conversion schem.pdf

DIYLC layout 1.pdf

ao63-1.JPG

ao63-2.JPG

A couple questions;

1) Does it matter from what node on the power supply I build the voltage divider for the heater DC? Aside from getting the resistances correct that is...
2) Is the artificial center tap still necessary after the heaters are elevated? Should I skip the 100ohm resistors and just tie the transformer center tap to the voltage divider instead?
3) The OT secondary is confusing with the center tap. Is there a way I would know just looking at the schematic that the secondary is 8ohm, or would I just have to measure it?

Thanks y'all. I'll keep updating as the amp get's built.

[martin manning]

1) Does it matter from what node on the power supply I build the voltage divider for the heater DC? Aside from getting the resistances correct that is...

Screen node is the typical source.

2) Is the artificial center tap still necessary after the heaters are elevated? Should I skip the 100ohm resistors and just tie the transformer center tap to the voltage divider instead?

You could reference the artificial CT to the elevated voltage instead of ground, but I don't think it would be any better than tying one side of the heater winding to the elevated voltage.

3) The OT secondary is confusing with the center tap. Is there a way I would know just looking at the schematic that the secondary is 8ohm, or would I just have to measure it?

I would measure it, but from Steve's schematic it looks like using the whole secondary goes with an 8Ω speaker. You could ground one side as common, and double the NFB resistor to get 5F6-A NFB level.

[turbofeedus]

Screen node is the typical source.

Gotcha, I'll make that adjustment to my schematic.

You could reference the artificial CT to the elevated voltage instead of ground, but I don't think it would be any better than tying one side of the heater winding to the elevated voltage.

I could also just tie the heat center tap to the DC correct? That would probably be easiest and makes the most sense in my noggin.

I would measure it, but from Steve's schematic it looks like using the whole secondary goes with an 8Ω speaker. You could ground one side as common, and double the NFB resistor to get 5F6-A NFB level.

Yep that's the plan. Already got the correct NFB resistor in the layout.
Thanks Martin.

[sluckey]

I struggled with that OT. I kept the same PA circuit and found that the OT is an 8Ω output with a center tap. It works best in that circuit with NFB only when wired exactly as the orig. schematic shows. I never did any tests on the OT to see exactly what was going on with the secondary taps. I wanted to keep the PA original.

[martin manning]

You could reference the artificial CT to the elevated voltage instead of ground, but I don't think it would be any better than tying one side of the heater winding to the elevated voltage.

I could also just tie the heat center tap to the DC correct? That would probably be easiest and makes the most sense in my noggin.

If you have an actual CT, sure. Since you mentioned the 2x 100Ω, I assumed you didn't.

[turbofeedus]

If you have an actual CT, sure. Since you mentioned the 2x 100Ω, I assumed you didn't.

I'm glad you made me go check, I thought there was but there's actually not. They used a hum balance pot.

I struggled with that OT. I kept the same PA circuit and found that the OT is an 8Ω output with a center tap. It works best in that circuit with NFB only when wired exactly as the orig. schematic shows. I never did any tests on the OT to see exactly what was going on with the secondary taps. I wanted to keep the PA original.

I see, so you did keep the center tap. I guess then the output jack would be isolated from the chassis?

Are you saying the NFB has to be injected at that exact spot for best results? I was planning to do more original to the 5F6/fender, connected into the phase inverter.

[sluckey]

Keep in mind that as soon as I heard the original power amp that my goal was to keep the PA all original. That 7591 power amp sounds very clean and punchy when driving an Eminence 15" Big Ben or a a pair of Celestion Heritage G12H speakers. Sounds good with a EVM-12L too. Yes, I had to isolate the speaker jack from chassis. I used a switchcraft with two fiber bushings.

But your goal is quite different. While you have the OT totally disconnected I suggest you check the impedance ratios. (I never did this on my conversion.) Then you'll know how to best adapt that OT to your 5F6 circuit.

And don't forget that you have another high quality OT that works very well with PP EL84s or PP 6V6s.

[turbofeedus]

Keep in mind that as soon as I heard the original power amp that my goal was to keep the PA all original. That 7591 power amp sounds very clean and punchy when driving an Eminence 15" Big Ben or a a pair of Celestion Heritage G12H speakers. Sounds good with a EVM-12L too. Yes, I had to isolate the speaker jack from chassis. I used a switchcraft with two fiber bushings.

But your goal is quite different. While you have the OT totally disconnected I suggest you check the impedance ratios. (I never did this on my conversion.) Then you'll know how to best adapt that OT to your 5F6 circuit.

And don't forget that you have another high quality OT that works very well with PP EL84s or PP 6V6s.

Understood, thanks Sluckey. Your conversion has been a big help.
One of the things I was considering was keeping this a 2-in-1 amp, sorta how people will do a 2 amp rig. Those 6GW8 tubes are cool though, I love the triode-pentode in the same package; I'm gonna save that for another build.
I'll try and measure the OT and see what happens.

[turbofeedus]

Alright, here's some measurements on the OT.

With 4.972V applied to the secondary winding:
Voltage across primary: 138.5V
138.5/4.972 = 27.9, so probably spec'd as a 28:1 turn ratio, or 784:1 impedance ratio.

So for an 8Ω load on the secondary, the primary with reflect 784 * 8 = 6272Ω
Looking at this datasheet for 7591, sorta seems like the Hammond designers followed the last column in AB1 fixed bias.
That shows with an effective load resistance of 6600Ω, a push pull pair will make 45W at 450V plate voltage under fixed bias.

So am I to take this to mean I could just lift the OT secondary center tap, and ground the one side of the secondary like a "normal" setup?
I'm also trying to wrap my head around what it will mean for bias when switching to cathode bias? Datasheet shows a certain output wattage at 9000Ω, but what about at ~6300Ω?

[sluckey]

So am I to take this to mean I could just lift the OT secondary center tap, and ground the one side of the secondary like a "normal" setup?

Probably.

As for fixed bias... be aware that the PT bias winding only puts out about 33VAC, which is not enough for 6L6s unless you use a voltage doubler. Simpler to just feed the bias circuit from the HT winding (like a Princeton Reverb or a Marshall).

[turbofeedus]

Probably.

As for fixed bias... be aware that the PT bias winding only puts out about 33VAC, which is not enough for 6L6s unless you use a voltage doubler. Simpler to just feed the bias circuit from the HT winding (like a Princeton Reverb or a Marshall).

I intend to use the 7591 tubes. My power amp is essentially the same as yours, with the exception of an LTP phase inverter and the NFB connecting to that rather than a preceding gain stage.

[martin manning]

So am I to take this to mean I could just lift the OT secondary center tap, and ground the one side of the secondary like a "normal" setup?

Yes.

I'm also trying to wrap my head around what it will mean for bias when switching to cathode bias? Datasheet shows a certain output wattage at 9000Ω, but what about at ~6300Ω?

No problem, just ground the grids and bring in a suitable cathode resistor or two. The effective plate and screen voltages will go down, of course, so you'll lose some output power.

Why not keep the existing HV supply up to the screen node, and the existing bias supply? Do you know what AC voltage is on the HV secondary feeding the 5AR4? It's probably too high for a 5F6-A CLC arrangement.

[turbofeedus]

Why not keep the existing HV supply up to the screen node, and the existing bias supply? Do you know what AC voltage is on the HV secondary feeding the 5AR4? It's probably too high for a 5F6-A CLC arrangement.

My power supply is basically identical to sluckey's, and I was going off the voltage they listed on their doc (475 on the plates, 388 on the screens, the 350 for LTP and 310 for preamp). I should have really called this a 5F6 preamp, with a 7591 power amp. Anyway, those voltages seemed to match up pretty well to this 5F6 schematic, so I just stuck with it. Maybe I'm mistaken and I should strive to get the voltages even closer?

The bias supply is also from sluckey, which is pretty much also original. I just dropped the filter cap value, just for what I have on hand.

[sluckey]

This PT puts out 430-0-430 so you must stay with the choke input filter. If you try to use a more common CLC filter you'll end up with almost 600VDC for B+.

Notice that the original used a bias balance circuit. I converted that to Marshall style bias adjust circuit. Also notice that the cathode bias sense resistors are 10Ω. So don't freak out when you see 250mV or 400mV on the cathodes. I set the bias at a cool 250mV (v. 400mV on the orig). That accounts for my plates running at 475V rather than the orig 450v. The clean PA and warm Ampeg preamp gave exactly the sound I was hoping for. But don't be afraid to run the 7591s hotter than I did for your Bassman. They can take it.

Looking forward to see how your project turns out.

[turbofeedus]

This PT puts out 430-0-430 so you must stay with the choke input filter. If you try to use a more common CLC filter you'll end up with almost 600VDC for B+.

Notice that the original used a bias balance circuit. I converted that to Marshall style bias adjust circuit. Also notice that the cathode bias sense resistors are 10Ω. So don't freak out when you see 250mV or 400mV on the cathodes. I set the bias at a cool 250mV (v. 400mV on the orig). That accounts for my plates running at 475V rather than the orig 450v. The clean PA and warm Ampeg preamp gave exactly the sound I was hoping for. But don't be afraid to run the 7591s hotter than I did for your Bassman. They can take it.

Looking forward to see how your project turns out.

Yeah like I said I should have really called this a 5F6 preamp and LTP, and the sluckey/hammond power amp.
I pretty much copied your power amp(except the phase inverter) and power supply (except the extra nodes for the trem, and added the HT DC elevation).
As you can see, I also gutted the amp, so I'm rebuilding from the ground up. I found that a lot of the CC resistors had drifted anyway.
I ended up having to move the bias switch off to the side to clear the OT, so I'm not sure if I will reinstall bias sense resistors.