Putting Harmony In Your Life: Reviving A Harmony H-322 Amp Part I.

The irrepressible Dirk Newton brought me this barn find amp he'd discovered on one of his sojourns, and after a while he decided he didn't want it as much as he wanted some folding money so it came my way. It's an interesting piece of single channel gear that sports volume, treble and bass controls,  two fixed bias 7591 power tubes and four 8" Jensen 5w speakers. I'm always a sucker for oddballs and orphans so I was all in on this. The first task was to see if it passed signal and it did-barely-with a raft of junk tubes scrounged up around mi ranchito.

That all happened a while ago and today I decided to see what I'd acquired. The Mullard GZ34 rectifier tube will just about pay for this rascal.

Here are some images.The before and after stuff should be pretty clear, and the amp had had some significant bodging before I got there. The first task was to replace the electrolytic capacitors, rationalize the sorry excuse for a bias system, and replace whatever else came to mind. Constructing a bias board on some vulcanized fiberboard with eyelets and stuffing it with a diode, resistor and Fender style 10 k bias pot was no great task. It allowed me to dump the old fashioned selenium rectifier which was pretty much dead anyway. That's the funny looking gray thing next to the old pot.

There's also a mystery capacitor of 1.0 uf 100v value and I've got nothing like that in the shop and it doesn't show on the schematic so I may eliminate it and just continue building to print. The wax paper caps are gone and the amp's loaded with disc ceramic caps which are usually pretty reliable.

The power supply electrolytics were replaced with a single JJ 40-20-20-20 can and bracket mount. Thus required opening up the mounting hole with my $10 set of chassis punches from Harbor Freight which are now $25. They're still a bargain if you've priced Greenlee punches lately. They work fine on aluminum and mild steel.

After all this I noticed on the schematic that it has a provision for a 33 volt supply and that's something I have to add tomorrow. It dumps 33v DC into the tube filament section which I suppose is for noise reduction. Another oddity about this amp is that it had not one but two so called "death caps" and as I was installing a grounded power cord these went in the trash.

Powering up the chassis yielded an adjustable bias that got me the required 19 volts. After I got done rethinking the power supply, I realized last night that I hadn't grounded the JJ can, so doing that yielded 450v on the plates-much more like it. I hooked up the 33v power source but there's still some hum. I think some grounded inputs and 100 ohm resistors may take care a lot of it.

The power and output transformers are marked Transco, a brand I've never heard of  unless it's the same Transco that makes neon sign transformers and sign ballasts. 

I also incorporated a pair of 1 ohm resistors from the cathodes to ground so I'll be able to measure how much current the power tubes draw.

According to the Tung Sol data sheet a class AB pair of 7591s, fixed bias of about 20 volts and a plate voltage of about 350 volts should yield about 30w of power although I'm reserving judgment until power up time- I would be pleasantly surprised if this amp was capable of that kind of power.

The goal is to evaluate the amp before I spend too much time on the cabinet, which is pretty shabby looking at this time.

EDIT: I finished the power supply this morning and got a good solid set of voltages and a usable bias supply. A pair of matched 7591s and some tweaks and I'll be in business-it's a good sounding amp.