D-AOC PCBs - the building and help Thread

[[silent:arts]]

I can't find your voltage readings in the thread Volker, I think your page is still down ?

fixed the link:
http://www.groupdiy.com/index.php?topic=26585.msg330048#msg330048

 

[MartyMart]

Thanks Volker, so my "0" is good and all the rest I have about 3 to 4 volts below
your readings, so that all looks fine.

Perhaps R1 just needs to be bigger in my case ... 5W !
And probably need to move the 783 to the case bottom or side to cool it a bit !!

What do you think ? ....  is 300v AC from the pwr traffo a bit much ?

MM.

 

[lolo-m]

Thanks Volker, so my "0" is good and all the rest I have about 3 to 4 volts below
your readings, so that all looks fine.

Perhaps R1 just needs to be bigger in my case ... 5W !
And probably need to move the 783 to the case bottom or side to cool it a bit !!

What do you think ? ....  is 300v AC from the pwr traffo a bit much ?

MM.

250V AC means approx 353V DC... 300VAC means approx 424V DC... This is a lot of lowering voltage for the 783... Maybe you'd better put a serial resistor to lower the AC... I'm sure your 783 will feel better with less voltage than with a bigger heater...

 

[MartyMart]

Do I just need one on one of the AC lines just before entering the board, if so what
would you suggest ?
I have a few 25W rated @ 22 ohms and some 10W @ 10 ohms
The 25W is metal cased with bolt holes and about 5 cm long

MM

 

[lolo-m]

Do I just need one on one of the AC lines just before entering the board, if so what
would you suggest ?
I have a few 25W rated @ 22 ohms and some 10W @ 10 ohms
The 25W is metal cased with bolt holes and about 5 cm long

MM

Yes... on the HT AC line. I don't know the consumption of the DAOC tubes, so I can't tell you the value... I bet it will be more than 22 ohms to notice a real change in the HT voltage...
Test it with one 22 ohm, measure the voltage, calculate I = U/R and you'll know the consumption in static mode... Leave a margin for compression and you'll know the resistor you need. 

 

[MartyMart]

Thanks , After some thought I calculate a 1k1 resister needed.
Tried a parallel pair of 2W 2k2's and the voltage at the board comes to a nice 249v before
the rectification .... more like it

They are cooking a bit so my calculation using 0.1 A was off, I thought that 1W resistors
would work, anyhow dropping 50v is a strain, so I'll order a much bigger wattage 1k for the job.

This resulted in the 783 running quite cool .... the new resistor was creating all the heat !

MM.

 

[Gold]

After playing around with the DAOC today I noticed a high frequency roll off. According to Volker's graphs the frequency response should be pretty flat. I'm not sure my testing methods are correct, but I do hear it.

When I have it in the studio and I feed tones from the DAW through the DAOC with its output feeding an electronically balanced input. The DAOC is down .5dB at 10K and down 1dB at 15k. This agrees with what I am hearing. I'm using Jensen JT-11P-1 input transformers and  Jensen JT-10k61-1M output transformers.

When I test in on the bench the frequency response tests way worse. I feed the input transformer from an oscillator and have measured the output after the transformer with both  a battery powered Fluke 8060A and an HP400EL. When I do it this way it measures almost 6dB down at 15k. Something doesn't seem right.

I measured on the secondary of the input transformer with the secondary attached to the PCB. It measured flat. I also measured at the output before the transformer primary with the transformer attached to the PCB. I saw the high frequency roll off. This leads me to believe it's something in the circuit but I'm not sure.

Are there any tests I should do or things I should look for?

 

[Gold]

I did some more tests today. I made some stupid mistakes in testing on the bench but I figured those out. Unfortunately the problem still exists. I have eliminated the transformers from the equation. It tests the same both with and without them.

The circuit is -.5db @10k, -1dB@16k and -3dB@32k. Is there a pole at 32k anywhere in the circuit? Am I the only one who is seeing this? On Volker's graph it looks like it is down .5dB@ 20k. I'd be happy with that. I sure hope I can get it flat at least to 15k. Yikes!

 

[Lowfreq]

Am I the only one who is seeing this?

Nope my DOAC is also a little dark sounding. I haven't had time to really investigate tho.

I've used the same Jensen on the input as you, but with edcor out.

When I did some RMAA tests it seemed that it was the edcor that was  dullin things for me as when tested unblalnced out it had a high freq rise. Yet when I listened to some audio files run through the comp. the unbalanced out also sounded dark.

I'm mainly using mine for vocal tracks at the moment, and a little eq up top seems to fix everything, so I haven't worried too much.

I have most of the Kubi mods on mine, except the input transformer, and I've still got the solen 10uf on the output.
I tried adding resisitors to the primarys and secondarys of the output transformer, but still got better results with just the transformer itself.

So yeah, mine is kinda dark too, but still works well..................... however if you have any luck improving it, I'd be interested to know what worked for you.

 

[MartyMart]

Mine seems to be unchanged in / out but I'm not measuring, just checking some bright
vocal takes - I have the ears of a "Bat" though

I've used Lundahls in and out, inputs are a 1:5 LL3815's and outputs are 3:2 LL3814's

It sounds absolutely gorgeous, even NOT compressing, stuff just sounds better going through
this thing !!

My Sifams seems to be off though, display almost nothing and then peg left when I can just here
compression coming on .... too sensitive, but come on too late and too fast.

MM.

 

[Gold]

Mine seems to be unchanged in / out but I'm not measuring, just checking some bright
vocal takes

It sounds absolutely gorgeous, even NOT compressing, stuff just sounds better going through
this thing !!

Maybe because the vocals are too bright

My Sifams seems to be off though, display almost nothing and then peg left when I can just here
compression coming on .... too sensitive, but come on too late and too fast.

I replaced R126 with a trimmer and followed the link for instructions on calibrating the LA2A meter. I was able to get the meter dead accurate with good looking ballistics from 1-3dB which is what I'm concerned about. After that it's not that accurate. You could spread out the error over a wider range if you want.

 

[MartyMart]

Maybe because the vocals are too bright


I replaced R126 with a trimmer and followed the link for instructions on calibrating the LA2A meter. I was able to get the meter dead accurate with good looking ballistics from 1-3dB which is what I'm concerned about. After that it's not that accurate. You could spread out the error over a wider range if you want.

Na - I just mean that they are as bright on the way out as on the way in ... not "uber bright"  ! 

Thanks, I have to read that link again as it didn't make sense that last time I looked ... or was broken link.

MM

 

[Gold]

Na - I just mean that they are as bright on the way out as on the way in ... not "uber bright"  ! 

If you get a chance to run some tones I would be interested in the results.

 

[MartyMart]

Sorry Paul, I know it's not "scientific" at all but I'm trying stuff that I know well through it and
it's coming up sounding very good, ... I'm more of a "listen to it" than a numbers guy 

MM

 

[diyfanatic]

just a little question!

since I don't want to use a sifam meter and I've seen that a 250µa meter working nicely...

I found some really nice vintage meter for cheap that are 200µa, is these are too sensitive or like I saw the compression is subtile so it should work?

thanks

 

[Gold]

No apologies necessary. Mine sounds good too. Just not entirely flat. It doesn't match the graph that Volker posted but it's not way off either. I'd just like to figure out whether this is typical.

If you could run a 1k, 10k and 15k tone though it with no compression and let me know what you get I would appreciate it. No need for disassembly or anything.

There are a distinct lack of capacitors in the circuit. I'm going to try to measure at different points in the circuit and see if I can figure out where it is happening. I not much of a schematic reader but C102 and C105 seem worthy of investigation.

 

[mikeyB]

Got the same response on 2 builds - they just don't seem happy with either the 10k/10k or the 600/10k Edcor input transformers! Not had time of late for any DIY (getting the sh-sh-shakes!!!) but i did briefly swap input's for sowter 10k/10k - still missing some top, some other sowter 1:3 step up and oep 1:1 600ohm and 1:2 600:2k4 - noticable top and bottom roll-off. The input circuit seems really picky as to what you put in it! In the end, omitted the input transformers and fed it unbalanced!!
The output is no trouble at all. Try it without in tranny - very open sound!!
Would eventually like input tranny, maybe option for both - any thoughts from the gurus?? - are we doing something wrong with the wiring etc, etc????

 

[MartyMart]

My scope has died on me Paul !
I have no way to measure that stuff at the 'mo , nor software either :-(

GOOD NEWS THOUGH  ......

I installed the 1k 25W between secondary 250v rail and the board, it's a metal body wirewound
that bolts to the case.
My voltages all look "spot on" now and the heatsink runs totally cool, the 1k 25W gets about as warm as
a cup of coffee - so touchable, it's directly on the side of the case so has a HUGE heatsink.
I think this will all run cool and happy from now on.
I bought a few more as the Dual LA2a / PM670 have "european" traffo's expecting 220v an I'm at 249v here !!

Peace,
Marty.

 

[MartyMart]

Got the same response on 2 builds - they just don't seem happy with either the 10k/10k or the 600/10k Edcor input transformers! Not had time of late for any DIY (getting the sh-sh-shakes!!!) but i did briefly swap input's for sowter 10k/10k - still missing some top, some other sowter 1:3 step up and oep 1:1 600ohm and 1:2 600:2k4 - noticable top and bottom roll-off. The input circuit seems really picky as to what you put in it! In the end, omitted the input transformers and fed it unbalanced!!
The output is no trouble at all. Try it without in tranny - very open sound!!
Would eventually like input tranny, maybe option for both - any thoughts from the gurus?? - are we doing something wrong with the wiring etc, etc????

I have edcors for this but went with the Lundahl's, as I needed a "step up" at the front end. ( 1:5 at present )
Of course the Lundahl's cost about 4 x what the edcors did !!
No zobel nework, just direct in/out of the Lundahl's , I expect them to be almost flat, and I think there's a graf
showing some somewhere IIRC.

There are a distinct lack of capacitors in the circuit. I'm going to try to measure at different points in the circuit and see if I can figure out where it is happening. I not much of a schematic reader but C102 and C105 seem worthy of investigation.

In all those positions 102/202 105/205 I have MKT 1uf's if that's any help ?

MM

 

[Gold]

A more detailed description of meter adjustment:

1. Install a 100k trimmer as a variable resistor. Before putting it in adjust it to 33k which is the value of R126. The DAOC should behave as it did before you did this. The meter should be resting at 0 assuming you previously adjusted it.

2. Attach a level meter to the output of the DAOC.

3. Feed a tone to the input.

4. Turn the compression knob until you see (let's say) 2dB of gain reduction on the level meter at the output.

5. Adjust the R126 trimmer so the meter on the DAOC matches.

6. Remove the tone. The 0 position on the DAOC will be off.

7. Adjust the meter 0 trimmer until it's right.

8. Apply the tone again and adjust.

9. Rinse and repeat until it's good.

10. Adjust for more compression and repeat the previous steps. I would make the second point the maximum amount of compression you think you will use and the first point halfway between that and no compression.