Building the WHAT compressor

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
[quote author="AnalogPackrat"]I should have also said that I plan to use rotary switches for gain, threshold, and ratio. That's why I'm being a bit more careful with the vactrol matching. I suppose I could adjust the R ladders on the each channel to make up the difference but that sounds like a royal pain! I have the, perhaps silly, notion of using this thing for a stereo bus compressor and maybe even for trying my hand at mastering (really silly).

I also got some NSL parts--dang--should have checked those as they might have matched a bit better.

So, Stephen, how do you like your WHATs? I haven't heard anyone here talk about using them. Which options did you go with? What iron? 1731s? How have you been using them? Anything you would have done differently? Any photos?

Well, I´m in Madeira on holiday at the moment! I use them as in the circuit diagram except for a larger resistor somewhere to increase the compression ratio. I use a condenser mic with a Behringer mixer preamp - what do do mean by iron?

Cheers,
A P[/quote]
 
[quote author="StephenGiles"]

Well, I´m in Madeira on holiday at the moment! I use them as in the circuit diagram except for a larger resistor somewhere to increase the compression ratio. I use a condenser mic with a Behringer mixer preamp - what do do mean by iron?
[/quote]

That was a bit confusing as my quote morphed into your answer...

Anyway iron = transformers. I'm getting the feeling that you built yours per the original Dee Tee design and not using Fabio's souped up version with DOAs and transformer coupling options...

Have a good holiday,
A P
 
i built my pair with OEP input trannies and one with IC opamps and one with fabio's DOA's (were they 1731's? i must'a remembered wrong - i started thinking they were 2520's)

rotary pots for threashold, ratio and output. (on one unit, the threashold pot makes no difference, and i still haven't figured it out)

the 'compression' LED's don't light up, but the unit sounds great and compresses like a big gooey blob of goodness.

my main complaint? the units are just swimming in a 2RU case. i think i need two more boards and i'll fit the case as a quad-pack of opto comps!

chris.
 
Just make it like a stepped attenuator. Trim in with a potmeter, measure it's resistance at the desired ratio (lift the pins first!) and fit the switch with the closest values.
 
I've searched for info on this subject but I can't find any clear answers.

VTL5C9 is specified in the schemo. I assume that it's been chosen for it's dynamic range of 112dB, but since the VTL5C1 with it's 100dB was chosen for the Forssell opto comp it seems that 100dB is sufficient.

But I really am in the dark here, so I'd like to hear if anyone can tell me if I'm looking at the right kind of specs here for a comparison and if the c1 can be used as a replacement for the c9 in the What comp??:

9:
vtl5c9.JPG


1:
vtl5c1.JPG


The problem is that the c9 is getting increasingly harder to find whereas the c1 is available everywhere I've looked.
 
I think the C1 will work. The difference between it and the C9 come down to the classic CdS tradeoff of switching speed vs. on resistance. You may have to drive the C1 harder to get the same level of compression as the C9. Fiddle with the series R going to the opto's LED (reduce it for more drive). In the WHAT, unlike the LA-2a (for instance), external components are used to modify the time constants with the opto only really affecting the lower bound for attack/release. The C1 is faster, so you can catch somewhat quicker transients using it.

BTW, which version of the WHAT are you building? Fabio's? mnats? The original D-T design? Just curious.

:guinness:
A P
 
Thanks a bundle AP! That's very very helpful. Perfect. I'm gonna give it a shot with C1's. I'm building 8 of these in one box and since I have the opportunity to test them up against C9's (my buddy's building 8 of those and he already have the Vactrols he needs) it's really not that much money wasted if they don't work the way I'm hoping. Then I'll just have to find 8 C9's to replace them with. It's not like it's impossibe, just hard.

It's based on the DT design of course and with another diy'er we're adapting Peter Cornell's layout to Gerber files and changing it here and there to our own liking. For instance a 680p cap is suggested as a feedback cap on one of the TL074 but it seems to take a little too much off the high end so we're getting rid of that, and my buddy's added a "meter LED" in parallel with the LED in the Vactrol and that works really well too. Furthermore I'm making a version for myself with resistor values "converted" to nearest regular 5% resistor value.

The files and changes we make will be available for everybody out here when we're done testing thoroughly of course. We're giving Peter Cornell the Gerber file version of his layout so he can post it at his site.

We've used the Sprint Layout 5.0 program to import the graphics file, redraw traces, all details, silkscreen and what have you. Cool program that we can recommend! Easy to use. I have never done that before and I was done with the main layout after about an hour! There's a downloadable test version on their site. http://www.abacom-online.de/uk/
 
Hehe Rasmus here in Denmark found that a LED in parallel with the LED in the vactrol worked quite nice as a "meter". The LED marked "Compression" in the schematic didn't light up at all. It turned out that the LED has been turned around in Peter's layout. Mounted correctly it works fine as a meter LED. So be aware that D11 should be turned around if Peter's Layout V2 31-10-1 is used. I believe Peter has been informed.
 
Alright. I was looking at that opto too. It's from Silonex and they have one model called NSL-37V82 with specs that looks a lot like the VTL5C9. I've decided to go with VTL5C1's though.

Are you able to tell the different opto's from eachother in the comp? I assume you've tried the C9 as well...
 
Got the first What (of 9 planned) up and running. Really nice little machine, I must say.

Alterations to the design:

VTL5C1 instead of VTL5C9 vactrol. This means that R41 has to be changed from 1KΩ to 200Ω. I just parallelled a 270Ω resistor over the 1K, giving me 212Ω, for my tests and that works just fine. 220Ω will of course work fine too if you'd like to stick with standard E12 values.

The schematic specifies 10K linear potmeters for Compression and Gain, but in my opinion it works most correctly with a log pot instead for Compression. That provides good, even and smooth control.

The 680p cap C38 has been removed completely. It took off a bit too much treble. Darkening the sound a bit. To me/us it just sounds better without it.

The C1 vactrol gives th comp a slightly faster attack and thus shaves off more transients. It's a quite transparent comp and since it's character remains the same when replacing the C9 with a C1 it suits me fine that it becomes a little faster. I have been wondering how Bob Rock got that very powerful and punchy yet quite soft sounding bass drum on the self titled Motley Crue album (the one with Hooligan's Holiday) and now after testing it I'm very confident that it's with an opto comp of some sort. It has that exact character.

I also tried it on:
Distorted guitar (boogie hard rock, Humbucker, Soldano Stack) - Nice and tight bottom, transparent while levelling quite nicely, good influence on the attack for this type of playing.

Vocal - Not the one I would choose for the job if I could choose only one BUT if you've tracked with an 1176 and just need an extra comp in the mix this comp will be spot on in my opinion! Not muddying anything up at all.

Snare Drum - For that snappy "rubber band" snare you should pick something else that you can control more or in my case I would use one of my dbx163 comps. But for a transparent but well levelling snare compression this one is just fine.

Bass guitar - So far I have a very good impression, but the bass I tested with is poorly recorded. Again nice transparency and levelling.
 
Luny,

Glad you got one up and running. Thanks for the detailed update. Have you looked at the other versions of the WHAT which have the additional attack, release, and ratio controls? mnats has the "Weak Joe" and Fabio Bauman's version has the same options plus DOAs and input and output transformers. Anyway, before you build your other 8, maybe you can hack in these extra controls for a test. FWIW, I'm planning a stereo pair with all controls brought out to the panel. Since you mentioned the 680pF cap issue, I might also add a "dark" switch.

A P

<edit--mis-attributed the Weak Joe>
 
Yeah, i wanted to test the thing before multiplying it. :grin: Also I'm giving this first test model to a friend afterwards.

I specifically want a comp with as few controls as possible. I'm quite fond of the dbx163 and decided some time ago I wanted to replicate that one or find another design that gives me the same/a similar tool. The What fits the bill very very nicely and for some applications I like the What better than the dbx. So I'm sticking with this design for this Box-O-Whats I'm building.

But hey, it's diy...I may change my mind, build more or Whatever I want...! mMUAHAHAhahaaa :twisted:

Thanks to David B. Thomas for the design, Peter Cornell for the layout and, on a personal note, Rasmus Friis for great help!

Diy ROCKS! :green:
 
And a photo or two before I shove it in the rack. :grin:
I know it's been a while, but quite a lot of work halted the process for me...
Anyway, it's done and all that's missing now is an engraved front panel from Schaeffer. For now I'm settling for this self made panel and if I can get my hands on some white Letraset rub sheets, and if that looks ok, then I might just keep it that way.

Please bare with the poor quality of pics.

OctoOptoKnobs.jpg


DoneTop.JPG
 
Thanks, AP! :grin:

It's not much to look at, but it sure comes in handy with eight of those babies when I'm mixing. just a little control and nice flavor here and there. I'm in love with them on bass drum, acoustic guitar and distortion guitar...

I've been testing a bit with C1 vs. C9 and I'm quite fond of the C1. Maybe it's because I've got plenty of snappy comps so I'm very pleased with having the faster C1 in there.
 
Back
Top