D-AOC PCBs - the building and help Thread

[Script]

Yes. When I was looking for NOS vactrols (for second in reverse) but couldn't find any, I tried with DIY opto cells (including the one in the picture).

This thing yielded compression alright but requires a lot of testing. Need to find type with good att/rel timing and then at least two that match. Meter is another two that better somewhat match, but not that critical.

Either way, I ended up with an LED instead -- after testing severals, I ended up with an orange one (but sheer coincidence, I'd think).

 

[dagoose]

ok.. the saga continues. Now I have another dead vactrol. So basically their life expectancy is over after a set time and 3 out of 4 are dead now within like 6 months. All mods, no spikes, just dead. :/
I think I need to go with the fake ones now and replace them and hope for the best.

 

[dagoose]

OK, I found a source (local store, which was a nice surprise) for the xvive vactrol's. They only had the VTL5C2/2 in stock which has a faster release time and actually, I really like that! A bit more grabby sound, suits better for mastering I think. I ordered 4x VTL5C2/2 and I had to match them a bit, I had one that was something like 1.5/2dB off with higher GR, but now I have a really good matched pair, actually better then with the VTL5C4/2 I had. So i'm pretty happy with how it turned out in the end.

 

[Gold]

So i'm pretty happy with how it turned out in the end.

So am I. I like the compression action much better wirh a single vactrol. I was able to get eight channels matched up pretty well. It took a while and both vactrols and tubes needed to be selected.

 

[scott2000]

Is there somewhere in this thread that describes how to try and match vactrols?

 

[Admiral-dk]

I'm glad I don't have to do it .... but you should put the two LED's in series - in order to have the excatly same current through both.

Maybe the two LDR's in series too - parallel to two Resistors of same value - on end to GND and the other to Vcc.
You now have a Bridge, where you connect your Meter between the two midpoints - IF the Vactrols match, there will be no Voltage over the Meter on All Current values going through the LED's .... 0-20mA

Per

 

[L091c]

Ok fixed my PCB with the safety resistors, hoping my 4 vactorls don't die on me:



Question regarding dual vactrol vs series LED: DaGoose mentions that with a single vactrol and a LED in series the DAOC is more grabby. Guess this means a higher compression ratio. How does that make sense, what does the LED do, it seems to be just half a vactrol (no LDR part).

I'd like to make my dual vactrol DAOC switchable to single vactrol (with LED in series) and trying to figure out if using the LED in the vactrol, just disabling the LDR part, would do the trick.

First test run yesterday of my DOAC went great, I love the sound, using stepped switches instead of pots so now experimenting with values to strike the best balance between reach and accuracy.

 

[boji]

I imagine a lot has happened since post-321415 in this thread, but I've recently resurrected an old ebay D-AOC. Some thoughts:

On using Edcor's, to my ears, they do in fact add a little HF roll off. Nothing +1 or +1.5 dB shelf can't restore.
I'm also hearing a 'flattering' bump in the lows, welcome for tracking LF instruments.

Stepped switches (100k all around) really helped dial-in comp engagement. Also sightly increased the range of use across the dial faces.
Variable pots gave me a bit of trouble in this respect. YMMV.

Edit: Now, on to read about grabby vactrol mods...

Few other minor alterations:
Current limit VU led's (100r)
Std. front panel needs holes drilled out 9+mm to fit stepped switches
Heater rework lowered noise

 

[Gold]

I said it in a previous post but after switching from dual Vactrol to single vactrol I wouldn’t bother with dual vactrols. Even if you are able to match four closely. The compressor action is way better with a single vactrol.

 

[L091c]

I said it in a previous post but after switching from dual Vactrol to single vactrol I wouldn’t bother with dual vactrols. Even if you are able to match four closely. The compressor action is way better with a single vactrol.

I read that but do you know why? What is 'better action' and what is technically the difference. Having just half of the waveform influence attenuation just does not completely explain the more grabby action.

 

[Gold]

I read that but do you know why? What is 'better action' and what is technically the difference. Having just half of the waveform influence attenuation just does not completely explain the more grabby action.

It’s not more or less grabby. It follows the waveform envelope better. I assume it’s because there aren’t slight differences in the vactrols trying to do the same thing but not at exactly the same time. Like a drum flam.

 

[L091c]

I imagine a lot has happened since post-321415 in this thread, but I've recently resurrected an old ebay D-AOC. Some thoughts:

On using Edcor's, to my ears, they do in fact add a little HF roll off. Nothing +1 or +1.5 dB shelf can't restore.
I'm also hearing a 'flattering' bump in the lows, welcome for tracking LF instruments.

Stepped switches (100k all around) really helped dial-in comp engagement. Also sightly increased the range of use across the dial faces.
Variable pots gave me a bit of trouble in this respect. YMMV.

Edit: Now, on to read about grabby vatrol mods...

Thanks, guess I'll try 100Ks first!

I noticed the DAOC, even without trannies, bumps the low end subtilely (with the threshold I can dial in the bump), to be expected with a slow opto compressor.

Have the original Edcors, a new set of Lundahls but also some old Triads laying around, hesitant to put them in (now running tranny-less and sounds great), but I'll give them a listen before deciding, see wether or not they bring too much coloration.

 

[Script]

Edit: Now, on to read about grabby vactrol mods...

I might have one for bending the compression slope, but won't be able to post details right now as I'm out and about ( would have to wait for two weeks. Sorry.)

I used one vactrol and a rotary feeding a network of one pot, resistors, diode, LEDs) for stepless soft to hard, including two settings with a second 'knee' for a curve gradually spanning from comp to limit, 4:1 to 10:1.

 

[shot]

Hey L091C, could you explain a bit more on those safety resistors?
I've tried searching through this thread but couldn't find anything on that topic.
Where are those resistors located in schematic and what are the values?
Thanks!

 

[scott2000]

Hey L091C, could you explain a bit more on those safety resistors?
I've tried searching through this thread but couldn't find anything on that topic.
Where are those resistors located in schematic and what are the values?
Thanks!

couple of pages back people were complaining about vactrols dying.......post 1633 analag had a fix........

 

[Script]

RATIO (Slope) MOD

As promised, here's some data on the ratio/slope mod that I came up with and implemented. It's the result of hours of testing, plotting, listening, swapping component values, and more listening and plotting. Nothing academic though and it's not the end-all of ratio mods for a more 'grabby' sound -- rather consider it inspiration for more experimentation.

It started with the "D-AOC Slope Mods" plot and me being indecisive in which order (slope or threshold ?) to arrange the settings on a rotary. (I also experimented with a second DIY vactrol, but that's another story.) So I came up with the idea to (1) make compression from 'Soft' to 'Hard' stepless and (2) have the knee sit at the same threshold, which it obviously doesn't here:



The result on my faceplate is a "THR" pot in the middle, a "SOFT to HARD" rotary on the left (POS 1 to POS 5), and a "RATIO" pot on the right that ranges from "Ceiling" through "mid position" to "Floor". (Some dust on faceplate.)

 

[Script]

Here's what the pot does on POS 1. In the end I settled on a C50K pot. (For difference btw A50k vs C50k @ mid position see blue and orange). I did this cos I wanted (1) a meaningful "mid" position on the pot for all soft to hard settings and (2) more resolution between CCW and mid.



Through positions 1 to 5, the pot yields the same curve when turned fully CW. This might seem useless at first, but ultimately I found it to be more intuitive in use. Fully CW is hard compression close to the knee ("Floor").

 

[Script]

And here are the rest of the plots for the pot's position through all soft-to-hard positions (POS) on the rotary.


@ R (3 o'clock)


@ MID


@ L (9 o'clock )


@ CCW

Note:
Orange line (labelled POS 5 here, but actually is POS 6) was dropped cos it sits at a different threshold. (Numbering of "POS" in graph deviates from front panel labelling. They are all one up: POS 1 is POS 2 on faceplate.)

 

[Script]

And more graphs. This time ordered by rotary setting (POS) @ different ratio pot positions. Note that the knee ultimately always sits at the same threshold in fully CW position.

POS 2


POS 3

Note: Here it gets interesting. The rotary sets the upper limit ("Ceiling"), while the pot moves a second knee around. Also obvious why I wanted more resolution between pot CCW to MID.

POS 4


POS 5

Note: This POS 5 almost reverses the function of the pot. It yields the hardest limiting when pot is set fully CCW.

POS 6

Note:
This POS 6 was dropped, as the pot has no effect. Also, effectively it is the same as POS 1 to POS 5 @ fully CW.

 

[Script]

SCHEMATICS

As for a BOM, it's three LEDs, three resistors, one diode, one pot, one rotary. That's it. I soldered the rotary and components onto a small board for easier wiring both to and from the DAOC main board and to and from the LEDs (see attachment). Note that POS 6 on the rotary is disabled. I might come back to it later to add a second vactrol or something else -- or nothing.

As for the schematics, see the attachment. Sorry, I am a pencil guy without motivation or time to create nice-looking schematics.


FACEPLATE

The three LEDs are brought out onto the faceplate (orange, white, clear red -- I tried other LEDs including blue, and also swapped their positions, but ultimately settled on these. They are by different makers, so please experiment here, orange at bottom, white in middle, red at top). The LEDs flicker in a musically meaningful way (i.e. they kind of indicate where a signal peak hits the curve, almost like a peak meter), and in so doing, they also sort of indicate the POS that has been selected.



Personally, I find POS 3 to be most interesting both musically (more so than POS 4) and as a starting point for further experimentation. I also played around adding a second pot to tweak the curve even more (as in "ratio" and "knee bend"), but my patience hit a dead end (also no space on the faceplate unless I dropped the "soft-to-hard" rotary, which I might do at some point.

[Note]
Please feel free to use for experiments and don't forget to let us know what you come up with. But mind you the schematics, circuit idea and basic mode of operation is patent-pending, I claim. So should my lawyer see this very idea implemented on a commercial unit, better get ready for the hulks

[Liability disclaimer]
Be careful to not blow up your unit !