Stereo Amp build based on Hammond organ circuit

Choke input power supplies will have a lot less voltage than a cap input supply will. Its roughly 0.9x the AC voltage from center tap to end vs 1.4x, and then there is the 20 odd volts the rectifier drops.

You might get by with the large choke and a small cap, maybe 10uf, right off the rectifier. Adjust the cap as needed to get the voltage where it needs to be. That will cut the ripple through the choke considerably.
 
Will that quiet down the PT though?

I want to try something tonight that didn't think of last night. I will solder original series chokes back and pull tubes out of one channel. Then I'll watch all voltages through variac and set tubes to draw somewhere between 35-40ma and see how things are. Then I'll repeat with another channel alone and compare results. It seems that one channel hums more than another. Just want to eliminate circuit mistake possibility.
Thanks for your help!
 
It should to some extent. Removing half of the power amp would definitely unload it.

Is the hum mechanical or coming through the speakers? Hum through speakers could be poorly matched output tubes, or having the balance control not set right. Power amps want reasonably equal current draw on each output tube in order to cancel out hum.
 
Hum is both, mechanical through screwdriver between transformer and my ear and it is coming through speakers. I can make it better or worse by adjusting individual bias but unable to make it go away.
Tubes are new Tung Sol from Jim McShane. I selected closest pairs from his matched quad.
 
So I tried eliminating one channel at the time. Hum is still there in each channel.
Added 40/500 cap across rectifier output/choke input. Variac to needed voltage, set current to 23ma. Hum is lower but still there. Then I took a scope pattern of ac ripple on bias line out of IBAM. Picture below.
All my scope settings are visible in the picture. Is this ripple expected or I need to bump up IBAM cap values?
 

Attachments

  • IMG_2215.JPG
    IMG_2215.JPG
    87.4 KB · Views: 30
The bias can actually be kind of dirty, though thats probably worse than I'd personally prefer to see. The noise gets injected common mode so it cancels. What exactly is your bias supply setup?
 
Here is a AC ripple on 350vdc to 7247 driver. 200mv/div
 

Attachments

  • image.jpg
    image.jpg
    74.4 KB · Views: 20
Filtering on the bias supply seems sufficient. Are you sure its ripple coming from the bias supply and not from the plate circuit of the 7247? Bias supply noise should be at 60hz, ripple from the 7247 should be 120.

Either way, that does seem like a lot of noise. How many filter stages are between the output of the rectifier and the 7247? I'd figure it needs at least 2 caps and a resistor to get it decently quiet. That one doesn't cancel any noise so it has to be quiet.

10v of ripple on the main B+ also seems like a lot to me, especially if its being measured after two chokes. Does it knock down with a 10uf cap right off the rectifier before the chokes?
 
I will take a sample out of speaker terminal and compare frequency. Can feedback wire pick up interference? It is not shielded. Just a thought. I'll check that tonight.

7247 power supply has 64ohm resistor from rectifier and 47/500 cap after it.

The only high voltage cap I had laying around was 40/450 and it bumped up voltage quite a lot and noise did not change much.
I did not think to take scope shot before disconnecting it.
 
It really shouldn't. There is enough of a path to ground that unless you've got it wound around the AC power lines its not liable to pick up enough to bother anything.

Might be worth lifting the coupling caps between the 7247 and the output tubes to see if you can figure out where the noise is coming in at. if it gets quiet with the coupling caps out you know its from the driver stage. You may need to use a larger resistor between the first and second cap on that power supply stage, or sub the resistor with a choke. That stage shouldn't be drawing much current so a larger resistor won't cost much in terms of voltage but it will drop the AC ripple a fair bit. The cap after the resistor can also be fairly large if needed for filtering reasons. The one right off the rectifier has limits though, not sure what a 6CA4 is good for off the top of my head but probably not bigger than 47uf.
 
Strong possibility on feedback wires. They run by power transformer, chokes and filament supply (6.3vac in this case)
I'll scope that first against power tube input ripple and compare. Any reason feedback wires can not be disconnected for test purpose? I'm not putting any signal through and my input jacks are grounded for now. If that isn't the source of the noise I will pull couplings off on one channel and see.
 
You should have a cap between the chokes

What value?
Original design did not have one. And I don't remember having this hum with original components. PT, OPT and chokes are original Hammond parts. Everything else is new on both channels.
 
20-40 mfd would be good. You could pull the 6CA4 to see if the hum is in the driver or feedback circuit.
Can you post a picture of the underside?
 
Last edited:
I think I got it figured out. I forgot that 6CA4 was not part of original power amplifier PS circuit. It was powering Reverb Power Amp and it had a problem. Upon initial power-up it quickly burned one of dropping resistors at can cap. I pulled 6CA4 rectifier tube so I could focus on main power amplifier. It was lacking performance due to several resistors in driver stage nearly open, but it did not hum.

Scott, I remembered all of that as soon as I saw your post. Thank you for an eye opener.
I pulled small rectifier and amp is quiet. Some faint noise is there but main hum/buzz is gone!

Next I replaced 64ohm from 6CA4 with 900ohm and added 40uf cap per Gadget's suggestion. Also added 40uf cap between chokes as was suggested by Scott.
Ripple is significantly reduced but is still visible and audible as well.
Picture below is after all those modifications.

Lower wave is taken at between coupling cap and 1000ohm.
Upper one is at 300vdc (CORRECTION 350VDC) 7247 B+

Does it look and sound like I have failing 6CA4 rectifier?



IMG_2230.JPG
 
Last edited:
20-40 mfd would be good. You could pull the 6CA4 to see if the hum is in the driver or feedback circuit.
Can you post a picture of the underside?

Underside does not look so pretty right now with all mods done within last few days. 2 can caps added with extension wiring.
Some dropping resistors are in series to obtain needed (or close to needed) values. IBAM board is suspended in the air because I was thinking that it was picking up interference from caps next to it. Feedback wires are lifted. Several additional grounds soldered. Top plate is made of anodized aluminum so I had to run heavy gauge ground bus between all points on terminal strips and tube sockets. Several ground points are bolted to the chassis.
I will post pictures when it actually can be evaluated. It is crowded and looks like big bang experiment.

Thanks for your help!
Dmitriy
 
A failing rectifier would not be likely, unless one of the sections was completely dead (1/2 wave rect vs full) and your voltages would be quite low.
Your waves look strange, not like a sine wave I would expect.does a filament wave look similar?
Is your scope grounded to the chassis?
 
Filament frequency does not match.
Bias, 7247 B+ and speaker output frequency shows same.
Scope is grounded.
 
Back
Top Bottom