Several years ago, I acquired an Epiphone Genesis electric guitar. It was free, but needed repair. The neck had snapped, but the amazing people at Specimen Products did an even more amazing job restoring it for far less than what that guitar was worth. So I figured it was a sign that I need to learn to play. (Now the cats knocking over my guitar stand and snapping the neck a second time, that might have been a sign too….)
Being an electronics geek, and having an interest in retro technology, of course the only way to go would be a tube amp. And a friend moving out of town conveniently had a tube amp kit she built, but it didn’t work. She’d part with it for the sum of $20. Sold! The chassis was clearly homemade, but I figured that if I could get it to work, I could always rebuild the amp on a new chassis at a later date. And Pumping Station: One recently acquired a finger brake for bending sheet metal, which will be ideal.
First, about the circuit design. Most normal single-ended amps work by having a power transformer that takes 120V on the primary winding, and at least two secondary windings. One provides a low AC voltage to powers the tube heater filaments. The other steps the voltage up so it can be rectified into DC, which becomes the B+ voltage that the tubes use to amplify signals.
This amp isn’t designed like that. Rather, it is similar to the All American Five AM radio. A while back, someone realized that if they were willing to compromise on safety, they could cut the cost of radios substantially. Instead of using a power transformer to step the voltage down to heat the tube filaments, they could use carefully chosen tubes whose filaments could be wired in series and connected directly to the 120V AC main, resulting in each tube getting the right voltage. The B+ power to amplify the signal would come by taking 120V AC and rectifying it to DC without stepping it up first. This means a much lower B+ voltage, so less amplification. But no power transformer is needed, which cuts the cost.
Now the safety issue:
These early radios had no isolation from the 120V AC coming out of the wall. Often, the chassis was directly connected to one side of the AC. So they posed a shock hazard. With a guitar amp, it would be worse because the guitar cable, and hence the guitar, would be connected directly to the wall. Not a smart idea!
Back to the guitar amplifier…. This circuit design comes from a book by Tino Zottola and uses 4 tubes: two 12AV6 pre-amps, a 50C5 final amp, and a 35W4 rectifier. It should provide 3 watts power output. The heaters are wired in series, along with a large 5W 75 ohm resistor to absorb the current that a 5th tube would have used. But, to satisfy modern safety concerns, the amp uses a 1:1 120V isolation transformer. So the circuit itself is entirely isolated from the mains.
A while back, I decided to take the plunge and turn it on. Nothing happened. Not even the tube filaments lit up. A multimeter check showed no voltage on the filament circuit. Why? Well, I checked continuity, and sure enough, the rectifier tube had a burnt out filament. Fortunately, getting a replacement from eBay was trivial.
This weekend at PS:One, I figured I’d do more troubleshooting. As I didn’t originally build the amp, I wanted to double-check every wire against the schematic, which took a while, but at least assured me that everything matched. Well, except for a 180 ohm cathode bias resistor on the final tube which looks like it might really be 150 ohm. Note to self: I should check the current through the cathode to make sure I’m not frying the final amp tube with excess plate current.
So, the moment of truth…. I wire it up to a guitar speaker and power it up! And…. Nothing. Well, something. The tubes feel warm, but they’re certainly not glowing. And the two power transformers definitely feel warm. The voltage across each filament is about 1/3 of what it should be! Then I think back to something Ed Bennett noticed. The original builder had, instead of using a 1:1 isolation transformer, instead used a pair of 120V to 24V step down transformers. The secondaries were wired together. So in theory it would drop 120V to 24V and raise it back up to 120V again. Ed’s concern: this was really inefficient, and there wouldn’t be much power. Then I noticed the markings on the secondary side saying 24V 0.2A. So this transformer could only provide 5 watts! Worse, running that 5 watts backwards through a second transformer probably wastes more power. And the designer of the circuit specified a 30W power transformer. No wonder the tube filaments didn’t have enough voltage!
Well, I need a new power transformer. But is that the only thing wrong with the amp? Then an idea hits me. A dangerous, stupid idea, I admit…. What if I bypass the isolation transformer and hotwire the mains straight into the circuit? Hmmm…. The chassis is sitting on a non-conductive table. I’ve got rubber soled boots. Worth a try. I disconnect the two 24V transformers from each other, bypass them with jumpers (although, to my credit, I made sure the neutral power phase went to the chassis), and, feeling a little like Egon switching on Ray’s pack for the first time in the elevator, hit the switch.
And… the tubes light up! Next, I turn up the gain (making sure I’m not touching anything grounded) and I can start to hear hum coming from the speaker. Good signs! So I plug in the guitar and strum the strings. Nothing. Hmmm. Wait a second, the guitar’s volume control is turned down all the way. I turn it up, and success!
The amp has a little bit more AC hum than I’d like at high gain values. (Ed suggests using shielded audio cable for the signals, which I’ll do when I completely rebuild it. Also, shielding the pre-amp tubes might be helpful, too.) I’m getting a little crackling with a hot signal from the guitar; I’m not sure where that is coming from. Is this an audio transformer problem? Am I driving the final tube too hard because of the wrong sized cathode resistor? I should probably break out the signal generator and oscilloscope. But all in all, it sounds pretty darn good.
So it looks like all I need is a transformer ($45 for a nice Hammond transformer from element14), and possibly replacing the cathode bias resistor, and I’ll have a really nice practice amp! Maybe I should also get around to building a cab for my speaker, too….