D-AOC PCBs - the building and help Thread

[jensenmann]

Regarding matching vactrols you should have a look at the datasheet. There are three interesting plots:
Uf of the LED needs to be matched for the same threshold (well, kind of threshold)
input current vs resistance is interesting for matching the slope of compression. Indeed, as suggested by flintan before, three measuring points should be enough for matching. The manufacturer offers three values in his datasheet, too, for comparing the characteristics of his different materials.
Measuring response time would be necessary to have the same time constants for both channels. Chances are good, that if you get vactrols from the same batch, the response is very similar. This would be the least important part of matching from my understanding.

good luck with all that. It´s an expensive sport to do all that.

dripelectronics did a lot of investigation in this direction but removed the most interesting parts of his posts because he started a business. Do a search for his t4b posts. CJ worked in the same direction even before Gregory, but his page is down and he is not very often here, not to say he disappeared from my radar without sending my V72s back - but that´s another story.

 

[flintan]

Interesting. Wouldn't Uf variations also reflect in different output R curves? It would not be too big a deal to do a first run of Uf matching and then match the matching ones in regards of I -> R, and then out of those there's the response time.. Well, realize i have quite some to read on this before/if i'm digging in to it.. I'll finish and try my unit first and see, maybe it's good enough as is.

I also found this in the sheet:

Consult factory if closely controlled characteristics are required at low input currents.

 

 

[jensenmann]

For matching the optos Fred Forssell did wire a 500 Ohm trimmer in front of the vactrol-LED in his sidechain design. I just realized that by crosschecking the schematic. That could be done with the DAOC, too, by lifting one LED-leg of both vactrols (if you have the kubi-mod) and wire them in series with a trimmer.

 

[ilfungo]

GREAT
Interesting!!!!
it does not matter which leg of the LED?
Thaks

 

[flintan]

GREAT
Interesting!!!!
it does not matter which leg of the LED?
Thaks

My guess would be no, but rather lift the side that is not connected to ground.

I talked to Rowan at Banzai Music who said:

We can offer matched optocouplers. There is a service charge of EUR 5,00 per item on top of the normal product price. Let me know if you would like to order some and i will make the necessary arrangements for you.

Too bad most of us already bought unmatched vactrols..

 

[ilfungo]

Thanks flintan
I also already bought several vactrol,
regarding my DAOC I noticed that when you just turned on the two channels are matched ...
After about 10 minutes no more ...
could be a valve problem and not Optos?
Thanks

 

[flintan]

Soon finished. One of my 6e1p had a dead heater so i'm waiting for a new one from russia.

The musikding transformer gives too high voltage (for B+) in my opinion. The regulator has some 100V to eat. After a while being powered for measuring etc it started eating fuses and i realized R4 had desoldered itself. R1 is also getting quite warm under these conditions. So anyone planning on using the musikding transformen, don't! Save it for something else and get one with 220V secondary instead.

 

[jensenmann]

I talked to Rowan at Banzai Music who said:

We can offer matched optocouplers. There is a service charge of EUR 5,00 per item on top of the normal product price. Let me know if you would like to order some and i will make the necessary arrangements for you.

Who knows which parameter will be matched?

 

[flintan]

I asked if they could match Uf, I vs out R and response time and this was the reply. But he did not say down to what tolerance and i guess we don't know how good it gets til someone try and report the results.

 

[Holger]

flintan, your mono-meter design reminds me on another build... 
Very good looking. IIRC I used the Musikding transformer, maybe I need to check the B+ again.

 

[flintan]

Yes it is a bit inspired by your (awsome) build with the big compression knobs and single meter.

 

[flintan]

I put a 1k resistor in series with the 250V AC (which is closer to 270V). It draws 47mA at idle so 1K=47V. Now it's closer to 220V and maybe a bit low on the DC side but this will do for now. It goes into the rack for a while before any more changes.

There's a tiny HF roll of by 1dB at 20kHz. I tried with the output pot+wire disconnected with a resistor+shortcut for max gain which gave a 1dB boost at 20kHz. So i guess the capacitance of my cable is eating 2dB at 20kHz. I use aluminum shielded wire with a capacitance of 140pF/m.

The meter tracking is good. Stereo tracking is ok. But as others i've observed that it's less consistent when stereo-linked. 1dB off at -7dB reduction.

Some more pictures (sorry about the poor quality, iphone):

 

 

 

 

[[silent:arts]]

very nice ;D

 

[j.frad]

Hi there folks!
I just powered mine and the heater supply drops dead with the tubes in!
With no tubes i can adjust to 6,3V, and as soon as I put in the tubes, I got nothing left...

Regulator is a LM338k but I don't think it could be an issue (or could it?), Transformer secondary measures 9.97VAC.
With tubes, I get some adjustment range but I can't get the output higher than 1,18VDC.
Anyone has a clue? I might try with the recommended regulator, but datasheets seem to indicate that they are interchangable...
http://www.national.com/profile/snip.cgi/openDS=LM338
http://www.national.com/profile/snip.cgi/openDS=LM350
No tube glow for me!!!

 

[seavote]

i would check the tubes first. make sure each tube has continuity from pin 4 to pin 5. if thats ok check the heater voltage trace for gaps or bad connections to the tube sockets.  good luck

 

[j.frad]

Hi, thanks, I checked and they are all OK.
But I think it wouldn't matter because broken tube/trace or no tube at all is pretty much the same, so I ruled that out.
I started taking parts out of the Heater supply, found the trimpot to make my tester beep as if shorted which is wierd, I also didn't know electrolytic caps could make a continuity tester beep but it seems they do... So I guess I have a bunch of stuff to change.
If anyone has experience in this kin d of stuff... I've had regulators shut down because input voltage was too high (on gssl builds), but it doesn't seem to be the case here..

 

[Michael Tibes]

I also didn't know electrolytic caps could make a continuity tester beep but it seems they do...

You charge the cap with the meter, it will beep until the cap is charged - which can take a while with a big cap.

Michael

 

[flintan]

My own daoc banner Thanks! And a big thanks for the great pcb's!

Capacitors do make you multimeter beep when they are charged or get charged by your mm. They also make resistance measuring in paralell behave strangely.

As long as there is no smoke or signs of things being really bad there's usually easier to track down the problem by measuring voltage with power on than by desoldering parts. They don't like being soldered in and out..

How does the voltages compare in the working psu (no tubes) to when tubes are in? What happens with only one tube in? Two, three etc?

 

[j.frad]

You charge the cap with the meter, it will beep until the cap is charged

Ah, that makes sense...

The psu works with no tubes in, I can adjust it to 6.3V, but as soon as I put tubes in, it drops to less than 1V. I tried with alll tubes, just two, anoher pair, etc. With 2 tubes it goes to something like 0.6V, 4 tubes makes it go around 0.3V.

I still took the caps out to test the pcb tracks and they seem to be fine. But there must be somehing that the tube connects the power supply to, that makes it shut down or something like that.

Is it possible that it happens because I have lm338 instead of lm350? I know lm350 needs a 3V offset between in and out but I couldn't find that info for the lm338. I don't see a reason why it would be different, but...who knows...

Thanks again guys

 

[[silent:arts]]

could you give us the specs of the PSU transformer ?