• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Mono push-pull driver PCB

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm planning on building a pair of monoblocks based on this PCB, and the power supply PCB for a friend. I'm aiming for 30W-40W using Edcor 60W OPT's and a pair of EL34. Using the 4.2K Edcor and the 40% screen taps, to reach my target of ~35W at <1% THD I believe I'll need a B+ of 400V.

Antek has a 2x150V transformer (AS-1T150) which with Pmillett's power supply design will give me a B+ of around 400-420Vdc. I'd also need to use an AN-0240 to supply the 80Vac for the bias circuit. I like this idea, as it would be under 50$ per monoblock. The problem is that Antek is out of commission currently. They're piecing the operation back together, but it looks like it could be months before they get to the high voltage units. We'll see when the next shipment is posted this month.

The other options are using an Edcor XPWR220-120 (295V@350mA, 6.3V@8A) and another transformer for the 80Vac bias supply, having Edcor do a custom job, or using the Classictone transformer Pete used and dropping the B+ some with a bigger choke or resistors...

Any other ideas?
 
I finally got one monoblock completed and in its chassis. I've tested everything, and have found one problem.
I measure across the 10k resistors on the 12hl7 anodes to set the balance, and only measure about 4v on each resistor. In the other channel that's still just tacked together on my workbench, I measure about 9v across those resistors.
Everything else checks out fine.. the voltage at every pin of the hl7's match in both channels.

The tubes should be burning about 8ma. If I didn't have a 2nd amp to compare with, I'd assume (maybe incorrectly) that the extra power I'm not measuring across the 10k resistor is being taken across the feedback resistor. But, both channels should be the same..

I measured the V drop across r17 (set resistor for the 10m45), and get 37ma as I should, so the CCS is correct.

Here's the schematic
http://www.pmillett.com/file_downloads/ppdrv2.pdf

Any ideas?
 

Attachments

  • 829ppp2.JPG
    829ppp2.JPG
    261.1 KB · Views: 397
I finally got one monoblock completed and in its chassis. I've tested everything, and have found one problem.
I measure across the 10k resistors on the 12hl7 anodes to set the balance, and only measure about 4v on each resistor. In the other channel that's still just tacked together on my workbench, I measure about 9v across those resistors.
Everything else checks out fine.. the voltage at every pin of the hl7's match in both channels.

The tubes should be burning about 8ma. If I didn't have a 2nd amp to compare with, I'd assume (maybe incorrectly) that the extra power I'm not measuring across the 10k resistor is being taken across the feedback resistor. But, both channels should be the same..

I measured the V drop across r17 (set resistor for the 10m45), and get 37ma as I should, so the CCS is correct.

Here's the schematic
http://www.pmillett.com/file_downloads/ppdrv2.pdf

Any ideas?

did you adjust the driver balance pot? that pot is for compensating for unequal transconductance of the 12hl7's...
at setup, this pot is set at midpoint, then adjusted towards to the tube that has lower current...

if you have several 12hl7's you can test them for gm matching...
 
Tomlang, thanks! And like AJT says, 829b's. I regulated the screens and bias. Otherwise, it's mostly the same as Pete's schematic.
I had a bunch to choose from so these are matched fairly well. Few ma difference between sections at worst. I might be adding a bias adjustment for each section. I've got the parts but am not sure I want to add more caps (or unnecessary complexity). So.. I'll frequently test each section for a while and decide later..

AJT
Yes, that's in balance, each at 4v drop. Other monoblock is about 9v drop across each 10k resistor.
 
I disconnected the ul taps local feedback (e-linear?), and it all biased up normally.
And I realized my comparison w/ my workbench copy was flawed.. 9v across 10000 ohms is only .9ma. So, there's more going on w/ the feedback than I understand right now, but both amps are functioning the same, w/in tolerances.
I think everything is ok.
 
I disconnected the ul taps local feedback (e-linear?), and it all biased up normally.
And I realized my comparison w/ my workbench copy was flawed.. 9v across 10000 ohms is only .9ma. So, there's more going on w/ the feedback than I understand right now, but both amps are functioning the same, w/in tolerances.
I think everything is ok.

you can series connect a 1k resistor on your CCS and then you can verify that your CCS is feeding the correct current...
 
Well, I've pretty much decided to build a 125wpc stereo amp using two of these boards, a pair of 300v(ish) power supplies stacked on top of each other, and 6hj5 sweep tubes.

Similar to what people were doing with the big red board. I read that entire thread. And this one.

I understand that I'll need to add an active regulated supply for the screens. I'm debating laying out and etching my own power supply boards or just buying ready-made boards for that.

Would pete's 'maida' style board be a good pick for the screen supply or is something like the zener referenced mosfet on the DCPP boards going to work better?
 
Bought the boards for this several weeks ago. I've been meaning to work on a tube project for some time. The A-431 output transformers for this should arrive tomorrow.


Tubelab, have you ever melted any of the Philips ECG 6BG6GA tubes? Supposedly, they have 35 watt plates and can be used in place of 7027A or a 7581A tubes. They have a plate cap and can handle 700 volts.


I bought a whole case in 2003 and the cost worked out to $3.25 each. I've used them in a guitar amp and was wanting to try them out connected for ultra-linear in a hi-fi amplifier. I think they can handle 450-500 in ultra-linear and they are cheap enough to pop a couple and not feel too bad about it.


I worked on a Fender 140 back in the late 90s, early 2000s. If a quad of Russian 6L6GC tubes can handle ultra-linear at 515 volts, I should be good as long as I have a screen resistor in place.


Classic Tone has the interesting 40-18069 transformer that can be configured for 300-410 volts at 500mA then run through a FWB for a good range of B+ voltages and still have 300mA to work with. The two extra 15 volt 2 amp windings look like they could be converted to 12 volts DC and used for 12SN7 tubes or what have you.
 
Last edited:
Tubelab, have you ever melted any of the Philips ECG 6BG6GA tubes?.......I bought a whole case in 2003 and the cost worked out to $3.25 each.

I remember when SND first started selling those......I should have bought a box full, but didn't. I bought a few, made the adapter and wound up sticking all of them into guitar amps that were notorious tube eaters......didn't eat those.

Supposedly, they have 35 watt plates and can be used in place of 7027A

They ARE 7027A's stuck in 6BG6GA bottles. Sylvania did quite a bit of glass stuffing near the end of the vacuum tube era to fulfill government contracts. I have seen a few 807's with the same guts too. As lines were getting shut down they made "one size fits all" tubes. A 7027A should work in any 6L6 type application, so why make 807's, 6L6GC's, 6BG6's and 7027A's when you can just stuff each envelope with the 7027A guts.

I never torture tested any of these since I never got any more while they were still cheap, but they should be good for some serious power.
 
I just so happened to see SND's site, shrugged, sent him an email and asked how much for a case? Times must have been tight. I was expecting double what I paid.

The guitar amp I used them in went to this kid I helped with his high school graduation project. I biased them just to see what would happen. After one of his shows, I checked the amp. Tubes where happy as a clam but I was afraid I was going to have to add a vent slot in the top of the cabinet or the tolex might melt. I decided that turning the bias down a bit would suffice.

I imagine in pentode they could do some damage. Not sure what four of them and say 470 volts along with the 4300 ohm primary can acheive. Triode and/or ultra-linear may or not be optimal with that load. My ears are already ringing.

When I first saw your thumbnail profile picture, I thought it was Rutger Hauer with a guitar.
 
I imagine in pentode they could do some damage. Not sure what four of them and say 470 volts along with the 4300 ohm primary can acheive.

All of my earlier posts in this thread occurred when I lived in Florida. In 2014 I left my 41 year engineering career at Motorola and moved 1200 miles north. Having to move 50+ years worth of accumulated "stuff" wasn't possible, so the "stuff"reduction plan went into effect long before the actual move. I sold, gave away, of trashed most of it.

We lived in a small house while our "retirement house" was being built, with all the remaining "stuff" stored in 4 different places. The new house was finished in 2016 and we moved in. There was no glowing tube activity during that time. I did finalize one of my driver board designs and sent it off to a PC board house for fabrication. It was the 9 pin miniature tube version of the driver circuitry designed in the 6L6GC in AB2 thread dating back 10 years. It is capable of driving nearly anything well into the positive grid region.

6L6GC AB2 Amp

The boards came back from the fab and sat on the shelf for a couple years before I stuck a pair in the breadboard I built during the 6L6GC in AB2 thread. They worked well enough for me to trash the old masonite breadboard that was falling apart and make a new one with plywood.....an upgrade. It got some occasional use until I got a workbench set up for tube stuff, early this year. I stuffed a bunch of different tubes in it and started a new thread here:

Tubelab Universal Driver Board, 2015 version

I still have tons of stuff stored in a derelict 850 sq ft house trailer next door, so I had been fetching boxes of tubes and "testing" them. The previous experiments were done with several flavors of 6550's and KT88's in triode, pentode, and UL with 100 watt 3300 ohm Edcor OPT's because I have them. YouTube video of 75 WPC flowing from GE6550's in triode here:

YouTube

I happened across a large box of tubes labeled "6L6" containing everything from old metal tubes to RCA black plate 6L6GC's. I had a lot of 6L6G's and 6L6GA's, so I decided to "test" some in the breadboard with the 3300 ohm OPT.

The results were surprising. I could push a single pair of 70+ year old tubes beyond 100 watts without making any of the internals glow......several attempts to venture beyond 100 watts resulted in exploding the cathode resistor, yet the tube still worked.

Later experiments pushing medium sized sweep tubes beyond 160 watts per pair resulted in a couple of shattered tubes. That's when a tube autopsy revealed the cause was a tube arc.

Beating on 6L6GA's at 525 volts into a 3300 ohm load is way past where they were ever dreamed of operating them, so a pair or a quad of newer bigger tubes into 4300 ohms should be no problem. The plate voltage spec can be bent, but keep the screen grid down to about 300 volts if you plan to lean on them.
 
I have to say you made reading these 24 pages interesting. Now I have two more threads to pore over. I just received the A-431 transformers today. At a later date, I need to purchase a few Edcore transformers for further tinkering. Nothing pushed to explosive limits I'm afraid. I need to rearrange my shop and set up some test speakers.

I almost forgot there's an ancient pair of Pioneer SX-88 speakers stashed away down there. The tweeters had those foam surrounds that need to be replaced. Another project.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.