AC30-ish build issues

[WRC34]

Hello all,

First I would like to thank the kind folks of this forum for helping me to solve an issue on a 50 watt EL34 based amp I made recently, THANKS!

I have been accumulating parts since last summer to simultaneously build that 50 watt amp along with a 30 watt amp based primarily on the AC30. They each have slight tweaks to the circuits, components, values and tubes used. I am not going for clones so much as for simple versions of these amps (with minor changes).

The big parts:
-PT is gigantic, from a Baldwin organ. It powered eight 6BQ5s, two 5U4s, and a nest of 12AX7s. I am using a GZ34 rectifier, four EL84s, one 6SL7 and one 12AX7. Theoretically it should be able to handle these tubes without breaking a sweat. It's HV winding is 290-0-290. It has a 6.3V tap for heaters, a 5V tap for rectifiers, and unused 16V and 15.1V taps that are heat shrinked and tucked away.
-OT is an unknown brand I bought from a guy on eBay. He listed it as having a 6K primary and being rated for 40 or so watts. It arrived with all leads labeled neatly and with long leads, with no visible signs of abuse.
-Choke is NOS Stancor, 10.5H, 110mA and 215 Ohms. It's leads were uncut and unused.


The issue is B+ voltage falling drastically, I think due to excessive current draw.

All tests have been conducted with a dim bulb tester/current limiter using a 300 watt incandescent bulb. This tester has been used successfully with my other builds and has been checked during testing of the 30 watt amp in question to ensure it is working properly. The first filter cap is before the Standby switch. Upon power up B+ reads around 330VDC, with wall power at around 116VAC. After flipping Standby switch with EL84s installed the bulb glows brightly and plate voltage/B+ falls to around 215VDC. With 6SL7 and 12AX7 installed, but EL84s NOT installed, the B+ stays at a healthy level, dropping only a few volts. I have conducted exhaustive tests and am starting to fear either the PT or OT is faulty.

Here's what I have done so far to no avail:

-changed first filter cap from 47uf to 22uf
-swapped out each filter cap individually in the amp with a known working/healthy cap
-replaced the coupling caps with equal value caps
-removed choke from B+ line
-swapped 300K grid leak resistors for 'stock' value 220Ks
-gone over/checked all ground points in the amp numerous times
-swapped cathode bias resistor for higher value (150 ohms shared between 4 tubes - I know this value is high but I figured if current draw being high was due to the cathode resistor being too low a value the 150 ohm would certainly help me identify this as being the problem. No change with the 150 ohm resistor in place)
-swapped the cathode bypass cap on the power tubes with a know working cap of equal value
-triple checked speaker jack/out wiring and ground
-tried three separate quads of EL84s. A Russian set of 6P14Ps, a set of JJs, and a set of Dutch made Philips EL84s.
-switched the 6SL7 for a known working tube
-switched 12AX7 for known working tube

The amp actually sounds nice & strong when being played, albeit with slightly earlier-than-desired breakup probably due to the low voltages. There is no noise or hum. I have not plugged the amp straight into the wall/without the current limiter for fear of destroying something (tubes, transformer, etc).

Attached are photos of the amp and a schematic I drew up. The first filter cap in the schematic has been changed to 22uf, and I've also added a 1K 5W resistor in front of the screen supply to drop screen voltage safely below plate voltage (it was already hanging out a volt or two under plate voltage, but I wanted a bigger margin. And yes the screen voltage drops right along with the plate voltage). I did conduct one test with the screen supply completely disconnected and the plate voltages stayed up at a healthy level with no glow from the bulb but obviously the amp is un-useable that way.

I will be in the great debt of ye who are wiser and generous enough to suggest anything else I might try.

Thanks!

[xtian]

I think you have a non problem. You're trying to run the amp with a 300 watt bulb limiter. At 120vac wall voltage, that allows you 2.5 amps max current. You're probably starving the amp, which is why the voltages drop when you're all warmed up and playing.

The bulb limiter is a good tool to detect if you have a runaway current problem. If you do, the bulb will go bright and stay bright, indicating a fault. But trying to play an amp with bulb limiter in-line? Never heard of anyone doing that.

[ampgeek]

Sweet build!

What is the voltage drop across your power tube cathode resistor?

With that, you can calculate the power tubes' B+ current draw (plates + screens) using ohms law. Compare results with typical EL-84 curves to see if you are experiencing excessive draw for the conditions you are operating at.

I suspect that you would have seen red plating of the power tubes if this were the culprit area.

Of course, you may just be seeing the "natural" voltage drop across the light bulb limiter filament from a fine and dandy power section.

Does the 6.3 VAC filament supply voltage drop as well? Likewise, how about the 120 VAC'ish supply at the amps fuse?

Good luck,
Dave O.

[ampgeek]

LOL! Looks like Xitan had the same idea and punched it along while I was composing.

My sense is that you are probably good to go with a test sans light bulb limiter.

Cheers,
Dave O.

[WRC34]

I think you have a non problem...trying to play an amp with bulb limiter in-line? Never heard of anyone doing that.

xtian, thanks for this info. In my other builds, including the 50 watt amp which I perhaps wrongly presumed would draw more current than the 30 watter, the bulb stays dim even when playing. I have used to 50 watt amp at rehearsal without a bulb limiter for 3 or so hours of near-continuous play with no problems, but only felt safe doing this after it 'passed' the bulb test.

I suppose it's better to err on the side of caution, but I'm about to go for broke and plug this thing 'straight into the wall'. Will report back momentarily with results!

[cbass]

A 300watt bulB is like almost not having a limiter try a 60 watt before further troubleshooting. If that 300 watt bulb is glowing bright you definatley have problem

[cbass]

trying to play an amp with bulb limiter in-line? Never heard of anyone doing that.

It works on single ended amps

[WRC34]

You guys were right - without the current limiter the voltages are healthy as a horse! 330VDC on the plates and 300VDC on the screens (maybe a bit too low screen voltage?) The amp is SUPER loud. A nice big G chord cranked up has excellent breakup going on BUT the sustain/decay of notes has a crappy soundng crackle to it...could this be a leaky cap somewhere or a result of the screen voltage being too low?

[xtian]

A 300watt bulB is like almost not having a limiter try a 60 watt before further troubleshooting. If that 300 watt bulb is glowing bright you definatley have problem

Wanted to mention this but forgot. CB wins. If you DO have a fault, a 300 watt bulb will let 2.5 amps burn through whatever component is in the way.

[WRC34]

Sweet build!

Thanks Dave!

A 300watt bulB is like almost not having a limiter try a 60 watt before further troubleshooting. If that 300 watt bulb is glowing bright you definatley have problem

cbass, thanks! You are correct. There IS a problem. Cathode bias resistor measures at 49.8 ohms, voltage drop across that resistor measures 10.25VDC. This equates to a whopping 51.4 mA per tube average, with 330VDC plate voltage this works out to be 16.9 watts dissipation per tube, ouch! I'm going to try a 62 ohm cathode resistor and hope for around 13 watts dissipation (which should be on the safer end of the EL84 abuse spectrum, ha ha). Will post results shortly!

[cbass]

30 volts seems to be quite a drop to the screens at idle. Does it drop that much after the choke or after the 1k screen resistors ?

[WRC34]

After the choke, with just the 1.1K screen resistors in place it was hanging out at just a volt or so below the plate voltage. I'm getting the 30V drop with the big 1K 5W resistor placed in front of the four individual 1.1K screen resistors.

[ampgeek]

Remember: ~10% of the total tube current draw is coming from the screen contribution and can be eliminated from the plate power dissipation calculation.

So....you are not quite as high on the abuse scale as you might think!

Still...yes..a bump up in the cathode resistor value will be heading in the right direction for sure.

http://tdsl.duncanamps.com/show.php?des=EL84

Rock on!
Dave O.

[WRC34]

Remember: ~10% of the total tube current draw is coming from the screen contribution and can be eliminated from the plate power dissipation calculation.

Thanks Dave, I tried to implement this factor into the calculations below, I hope I interpreted what you were saying correctly!

I ended up going with two separate 150 ohm bias resistors for each pair of EL84, individually bypassed by 100uf 100V caps. The readings I got:

pair 1
146.5 ohm resistor
-11.85 voltage drop across resistor
40.4mA per tube, multiplied by .9 for screen displacement = 36.3 mA per tube
11.9 watts dissipation average between both tubes

pair 2
146.7 ohm resistor
-11.70 voltage drop across resistor
39.8 mA per tube, x.9 = 35.8 mA per tube
11.82 watts average between both tubes

If I am right, this seems exactly where I want to be.

I think that crappy sounding note decay/sustain was simply from extremely overbiased output tubes saying "dude, you are f**kin killing us" because it has disappeared.

[ampgeek]

Purrrr..fect!

That buzzy, ugly decay at the end of notes might also have been/be crossover distortion. It is very common in these plans and is exascerbated by pushing the tubes hard as you imply. This type of distortion also has a swirly characteristic which, in the right doseage, is kind of pleasant.

Play it for a while and listen closely to the character of that decay distortion as the amp breaks in. There are easy/inexpensive ways to reduce that character if you so desire.

Well done mate!
Dave O.