Effectrode LA-1A tear-down and trace adventure

[rankot]

Well I figured out where the low output level is coming from, and it looks like I have some homework (and probably asking for help ) to do. I have a B+ of 220V, and with V1 and V2 in parallel I have essentially 25k as the equivalent plate resistor (4x 100k in parallel). But for whatever reason the original LA-1A is measuring 136V on the plate but mine is dropping to 87V. The line itself isn't sagging at all, V3 sill gets 220V on the plates. I removed the coupling cap to make sure nothing downstream is messing things up, but no change.

Any thoughts anyone?

Is your SMPS capable enough?

[willienillie]

Easy check, swap tubes from the original to your clone.

[phasetrans]

Nathan,

Really dig the boards, it looks sort of like the Matte Black solder mask from PCBway.

Do you mind sharing your right angle double row connector, and tube socket?

[vigilante397]

If all the extra voltage drop is across the plate resistor then one or more triodes are conducting too much? What are the cathodes & grids? Tried tube swap?

I did try swapping tubes from the original, same result. I've used this SMPS on three-tube designs before, but I dind't even think about the fact that 12AU7 has a higher plate current than 12AX7. My big question there would be, if I was overloading the SMPS shouldn't the whole B+ line sag? I'm still getting 220V to my 12AT7, it's only the parallel 12AU7 plates that are dropping.

I posted the schematic a while back, but I'll post the updated one the PCB is based on. The plates are all in parallel with an equivalent 25k resistance, but each stage has its own 27k grid stopper and 4.7k cathode resistor, and each cathode is bypassed with 32uF caps (10 + 22 in parallel).

I'll do some more playing and see if I can figure out any more clues.

Really dig the boards, it looks sort of like the Matte Black solder mask from PCBway.

Do you mind sharing your right angle double row connector, and tube socket?

Thanks, boards are from JLCPCB. The connector is just a double row right angle 100mil header and the tube socket and header are soldered to a daughter board with a bi-color LED under the socket. It's not the setup I originally designed, but it works well and gives a nice solid connection.

[vigilante397]

Here's how the 12AU7s are set up:

[phasetrans]

Really dig the boards, it looks sort of like the Matte Black solder mask from PCBway.

Do you mind sharing your right angle double row connector, and tube socket?

Thanks, boards are from JLCPCB. The connector is just a double row right angle 100mil header and the tube socket and header are soldered to a daughter board with a bi-color LED under the socket. It's not the setup I originally designed, but it works well and gives a nice solid connection.

Nathan,

Thanks regarding JLC! You've just inadvertently helped me on another project, where the test units are coming from OSHPark, but I wanted to try the JLCPCB ENIG for the FX board and Graphic. I've got a half dozen guitar players building alongside, so it would be nice to get ENIG for less.

[marcelomd]

Thanks, boards are from JLCPCB. The connector is just a double row right angle 100mil header and the tube socket and header are soldered to a daughter board with a bi-color LED under the socket. It's not the setup I originally designed, but it works well and gives a nice solid connection.

You are running them at 220V, right? Any comments on high voltages and 100mil headers?
Thanks!

[vigilante397]

You are running them at 220V, right? Any comments on high voltages and 100mil headers?
Thanks!

I've been using 100mil headers to breadboard tube circuits for years and haven't had any problems with them. According to the datasheet each pin is rated for 3A and I'm obviously nowhere near that, and in order to reach the dielectric breakdown of air and arc across 100mil you would need thousands of volts, so 220V isn't an issue.

[marcelomd]

I've been using 100mil headers to breadboard tube circuits for years and haven't had any problems with them. According to the datasheet each pin is rated for 3A and I'm obviously nowhere near that, and in order to reach the dielectric breakdown of air and arc across 100mil you would need thousands of volts, so 220V isn't an issue.

Sweet!

[vigilante397]

Alright, so I did a little more digging, and I think anotherjim and rankot nailed it, I think the tubes are pulling more than the SMPS can handle. 12AX7 plates pull a max of 1.2mA per stage, which I know my SMPS can handle piece of cake, but I overlooked that 12AU7 can pull a max of 20mA per stage, and I doubt the SMSP can handle anywhere near 80mA. I have an SMPS daughter board I designed a million years ago that I can tack on top of it, see if feeding V1 and V2 off separate high voltage supplies takes care of anything.

[iainpunk]

It seems to me like it's a tube preamp with a compressor tacked on.

At the end of the day all compressors have some sort of "preamp" so a tube compressor could only be expected to have a tube preamp.

but controlling the gain of a tube isn't even that hard tho... especially when its optically controlled, is it so hard to implement tubes in the actual compressor circuit???

cheers, Iain

[PRR]

...12AU7 can pull a max of 20mA per stage....

I thot you had checked for power rail sag?

And use your Ohms Law. 220V toward 12AU7 with 100k and 4.7k won't pass 20mA.

I make it 90V across the tube (96V plate to ground). About 1.25mA per section or 5mA per quad.


Wiring error or wrong part value.
_________________________

> controlling the gain of a tube isn't even that hard

By varying DC current, yes. But when you vary the DC condition at program-dynamic rate, it thumps. It can be cancelled. But not easy. The real LA2a was notable as a limiter which did not balance-out thump, is essentially thumpless (except the unavoidable distortion of rapidly varying gain on loud transient).

[vigilante397]

I thot you had checked for power rail sag? 

Well that's the frustrating bit, it's been a long day The main B+ rail does not sag, just "HV_A." So maybe I still haven't completely convinced myself, but when I pull out one of the 12AU7 (doesn't matter which one) then I get great voltages, and when I have all three tubes installed HV_A drops. Power rail sag is the simple answer my brain wants I haven't gotten around to testing with the external HV supply yet, but I will tonight and post results.

[PRR]

.....Any thoughts anyone?

You are keeping key details to yourself. Even your "complete updated schematic" lacks a complete and clear power rail. There's 11k(??) from HV_A to HV_B, but which is primary? And does 220V just appear by magic? Which end of the 11k?

I'm sure you know, but if you won't tell us, we can't play.

[vigilante397]

You are keeping key details to yourself.

That was not my intention, you're right. In the interest of readability I have the schematic as two pages in my software, one with the preamp/compressor section, the other with the power supply and gain reduction indicator, I only uploaded the first initially as I felt the second was less important for "design review," but I agree both are needed for debugging. I'll get both uploaded tomorrow so you can see the whole picture of what's going on.

This is obviously not an original design and while I do plan to make money with the finished product, it's not pay-the-bills-and-feed-my-family money, so keeping the design a secret was never my plan.

[vigilante397]

I meant to get to this earlier, weekends are always busy. I made the images as big as I could get them (click to enlarge, you know the drill), but if they're still too hard to read I can get PDFs. So here's the updated schematic of the preamp/compressor section:


And this is the power section. Typical nixie-style SMPS, schottky bridge rectifier for polarity protection, 78M05 for 5V rail to the op-amps, LM3914V is the LED driver and LTA-1000G is the LED bar for the gain reduction indicator.


You'll notice my pin numbers are weird for the tubes and don't match up with typical 12A*7 pinout, that's because I created my "daughter board" as a component so the pin numbers line up with the 10-pin connector I use between the boards.

[diffeq]

Just a guess, but that HV_A is separated from SMPS supply with R13, which is 11k. Wouldn't that form a voltage divider with 25k load? I'd try to drop 11k to 2k-3k. It would need to bump up the cap value to keep filtering going, however. *

* - would it even make a difference, with oscillator running around 45kHz.

[anotherjim]

I would suggest lowering that final dropper resistor too. I was going to suggest temporarily fitting a 100k pot across it -  but you have shock danger with that supply.

[PRR]

...Wouldn't that form a voltage divider with 25k load?...

It's not 25k. The tubes are not shorted. They are in a "fair fight" with the load resistor. The tube may run at half the supply. So say 200k each, 50k for all four. Yes, that is voltage drop. We can compute it but what's the point? Nathan has not said how much drop he has, and he is sure smart enough to work it out. I think he is just talking to himself in public, a VERY useful analysis technique. Note how many other problems here are solved once the problem is fully expressed.

[merlinb]

Nathan has not said how much drop he has, and he is sure smart enough to work it out. I think he is just talking to himself in public

+1.  It's pretty obvious the PSU dropper R13 is causing the loss. Try 1k or something. 11 is a silly number anyway!