D-AOC PCBs - the building and help Thread

[Script]

I find M/S processing so much more logical and easy to use on stereo signals and buses than L/R.

Same thinking here. Although on some, non-linked L/R can add something interesting -- even to electronic tracks.

As for 2W, I think they are underspec'ed. I measure temps the analogue way, so too hot to me means 'too hot to touch' for longer.

So I opted for 'high power line' approach. Not sure that was a good idea, but better than long-term prospect of burn marks or, even worse, slowly reflowing solder, I thought.

Tubes are still easily accessible. Also makes it easy to lift entire board like a brick out of the case. Standoff poles are permanently fixed to the board with another standoff underneath that tightens with a screw from case underside (so attached to case ground).

 

[fragletrollet]

wow

 

[Script]

Those 25W all speak Chinese

 

[Murdock]

I have a big 3.2:1 transformer here which seems to be gapped. How much ampere is going through the last output cap? And would it be possible to remove it if the transformer could handle it?

And about the input. Could I increase the 15k load resistor if I would use a 1:4 transformer?

 

[5v333]

Current does not Flow through capacitors.
AC current in circuit can be described with
i = dq/dt = C du/dt

 

[Murdock]

Yeah, isn't the capacitor there to block the current?
That's why I thought that I could omit the capacitor when the transformer could handle the current.

 

[5v333]

if I got it right then you have a defined current through the tubes of about 7 mA and a voltage, at the west side of the cap derived from the previous stage, of about 130 V (guessing here that B+ is 250 V).
If you remove the cap you will disturb the bias setup.
Your gapped transformer is prob better of for a common cathode stage but you prob need more info on how it likes to be treated. If you dont have info you could prob experemint with it to find a good setup .

 

[timtom]

I finished a D AOC Compressor with Haufe RK 310 Input transformers and edcor 10k/600 out transformers.
The unit passes audio and also compresses. Voltages are fine.
But as soon as it starts compressing the signal starts to distort.
Xvive VTL5C4/2 are the opto elements.
I listenend to the stereo link point at the pcb and the signal there is distorted as well.
At r110 before (sidechain tube) signal is clean.
Listening to the Stereo link when I remove the vactrols the signal is clean.
When I only add a LED into the LED Vactrol spot it starts to distort and the led starts to flicker.
The vactrols are brand new and the problem is at both channels....
I swaped the sidechain tubes with the output ones, same problem.
I measured the vactrol resistance and it works as well.

 

[Script]

Played around rolling tubes in this unit again. I had done this a long time ago, when I built it, and as far as I remember, back then it didn't lead anywhere. But still I had always wondered why tubes in this unit are hardly ever discussed. Some swear by expensive ones, others use cheap and are happy... Either way, I'm not a tube guy.

I used old-time RMAA for visualization of frequency response, THD, IMD and phase, and then did extensive blind listening tests. After several hours spread over several days I couldn't help but conclude for myself that just rolling tubes in this unit without additional changes in the circuitry (which I also tried) were a waste of time. And that's good enough for me as a result regarding tube rolling

Negative feedback is plenty in this unit -- some 27dB/26dB due to the 470K in the NFB loop -- linearizing the tube stages. I don't keep fancy tubes around, and probably I rolled not so good ones (including some NOS), but with unit stock, changes are rather subtle to say the least. Maybe very expensive ones add something, but I'd guess not that much after all. Tubes with a slightly different frequency response (+/-0.5dB) are meaningless to me, as program material goes through equalizers anyway. So it's down to noise floor, THD and/or IMD (7K). Eventually, I settled on the best cheap ones -- that is the best for me. This can be done in two hours.

 

[timtom]

After I got some original Vactrols from Volker the unit works. Sound very nice!
After using ist for some weeks I found out that there is sometimes a clicking sound with some material. Anyone has realized the same behaviour?

 

[dagoose]

Hey guys! Long time not hanging around here, still alive and well though.

I'm trouble shooting my d-aoc as we speak.
I think I might have a dead vactrol since it's passing audio fine but it doesn't compress on one channel.
It's been ages since I last looked into this compressor since I mainly just use it to pass audio through withoit compressing, which it does fine.
I have 4 vactrols in there, the reversed parallel trick with stepped switches and various leds mounted for various curves (which I might partly remove).
I do measure about 0.8v on the stereolink/c104 pin2 vtl pin 2 on the working side, but nothing on the broken side.
So it looks like a dead led in the vactrol (grr..) but before jigging the boards and vactrols around I was curious if someone else seen something this before.

 

[dagoose]

My unit stopped compressing on one channel. I switched sidechain tubes, made no difference.

The voltage at R221 won't change despite lowering the threshold, but I can see distortion going up while the amplitude stays the same. Can the vactrols have died (I've got two on each channel via the Kubi advanced mod)?

Did you happen to find a solution for this? Was it a broken vactrol?

 

[Script]

You swapped the sidechain tubes L / R aleady, I assume ? I had one channel not working at some point, thought it was the opto, but turned out it was a bad tube.

 

[dagoose]

You swapped the sidechain tubes L / R aleady, I assume ? I had one channel not working at some point, thought it was the opto, but turned out it was a bad tube.

Yes, that was the first thing I did, but forgot to mention that.

 

[Script]

THR pot / rotary connects fine ? And so does everything else ?

If only one opto of the combo is gone, unit should still compress, but threshold is not deep enough with just one opto LED. Might test with an of-the-shelf LED in reverse.

 

[dagoose]

THR pot / rotary connects fine ? And so does everything else ?

If only one opto of the combo is gone, unit should still compress, but threshold is not deep enough with just one opto LED. Might test with an of-the-shelf LED in reverse.

I checked and swapped the threshold attenuator (stepped switch) which didn't work, same thing happened.

I have (2 types of of) leds in parallel and anti parallel with the opto leds with a stepped switch to get different ratios.
There is one thing I noticed, normally when powering on the meters go full scale as they should and after a couple of seconds they both drop and come back more slowly right after that. The broken channel doesn't do that anymore.
I'm pretty sure it's the vactrol itself, that's why I ask for people that might have had the same thing.

Anyone familiar with the Xvive VTL5C4/2 vactrols? I assume they are not 'real' vactrols?

 

[Gold]

I've been dealing with this too. I've got some dead vactrols in three out of eight channels. A couple back to back were totally dead. I also found some that were very far off from when I installed them and not working well. I matched all of them at 15mA and 45mA when I originally installed them. The strangest failure was having a couple that weren't that far off but the release time was way way off. Like a few seconds to get to mega ohms with a lazy detour through Kohms.

 

[Script]

I'm pretty sure it's the vactrol itself

That's bad.

I skipped on the Xvive after reading pretty unfavorable reviews here and elsewhere.

Also briefly tried just photoresistors from Aliexpress to make own optocouplers in shrinktube. It works but one needs to make quite a few for matching (time consuming).
_____

So instead I had spent time experimenting and ended up with only one vactrol per channel, but added a network consisting of a few resistors, selectable LEDs, a diode, and a RATIO pot.

Two of the settings introduce a second knee, so that the unit can compress lower levels and limit high peaks at the same time, with the "curve" shape tweakable (via the pot) and going gradually from about 4:1 at lower level to about 10:1 at higher levels.

If you can't find original vactrols or other solution, and if interested, I could dig into my notes. (My own handwriting is a problem though.)

 

[dagoose]

I've been dealing with this too. I've got some dead vactrols in three out of eight channels. A couple back to back were totally dead. I also found some that were very far off from when I installed them and not working well. I matched all of them at 15mA and 45mA when I originally installed them. The strangest failure was having a couple that weren't that far off but the release time was way way off. Like a few seconds to get to mega ohms with a lazy detour through Kohms.

ah.. bonckers... this sounds like I have a dead vactrol indeed. I was already afraid of that. I will need to make it work with 1 vactrol per channel and some leds to switch then. Hoping that I can get a good decently matched pair out of the 3 (hopefully) working then.
Also I found one new in my own stock but it had a broken off pin (don't ask). Maybe I will try to get a leg soldered back by removing the housing a bit.

Now just wondering, why do they die? I did notice on the schematic and by meausuring that there is, I think something like 120v at the R116 to C104 joint. The other side of C104 is going to the vactrol led. Maybe the C104 100nf cap (using WIMA) is leaking some higher voltages that could blow the led?

The story continues.

 

[dagoose]

That's bad.

I skipped on the Xvive after reading pretty unfavorable reviews here and elsewhere.

Also briefly tried just photoresistors from Aliexpress to make own optocouplers in shrinktube. It works but one needs to make quite a few for matching (time consuming).
_____

So instead I had spent time experimenting and ended up with only one vactrol per channel, but added a network consisting of a few resistors, selectable LEDs, a diode, and a RATIO pot.

Two of the settings introduce a second knee, so that the unit can compress lower levels and limit high peaks at the same time, with the "curve" shape tweakable (via the pot) and going gradually from about 4:1 at lower level to about 10:1 at higher levels.

If you can't find original vactrols or other solution, and if interested, I could dig into my notes. (My own handwriting is a problem though.)

I will skip the xvive then indeed. I will hopefully get a working and matched pair out of the 3 (hopefully) working vactrols. I already have 2 different leds mounted on a switch, but maybe I could add some resistors as well to get more ratios. But first make it work again.