Some basic opto comp questions - LA-2a schematic

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kiira

Well-known member
Joined
Aug 2, 2004
Messages
536
Location
Baltimore - Blobsville USA
Hi I have started building a tube type channel strip in a large 3U box, using a passive EQ (NYDave design) and I have also designed the makeup gain stage myself too... just a very simple thing. I decided I would try and design the tube preamp as well since I've done one before.

I would like to include a tube type compressor as well but I am kinda stuck. The vari-mu comps are all a little complicated except for PRRs simple design but using that would complicate my power supply needs too much I think though it would be perfect otherwise.

The only tube opto type comp I know of is the LA-2a and that's too large and complicated too. Plus I already started the metalwork for a seperate one I have all the parts for.

Is there any reason why a tube opto compressor could not be designed using a vactrol like a VTL5C9 (just a part I remember, not based on design)? I am thinking opto because I have an idea how they work from studying the LA2 schematics... veri-mu is a complete mystery to me.

Designing compressors is way over my head I barely know enough to make a working preamp stage, but I was thinking I could use the LA2 as a basis and then put together simplied side chain and gain amp and a makeup stage and give it a try and try and learn more of the basic of compressor operation. I have in mind an opto comp that is as simple as PRRs veri-mu but using only tubes.. the simplest thing possible initially.

Just looking for opinions and maybe something to read or some schems I don't know about that could teach me how to design a sidechain to work with light emitting resistor vactrol.

thanks I hope!

Kiira
 
[quote author="kiira"] light emitting resistor vactrol. [/quote]
Light emitting DIODE in most vactrols that I know of.

Since the type of LEDs in vactrols will likely have a forward bias of about 2Volts, and LA-2a's tend to have a hundred-volt or so of sidechain drive, it'ws not a match made in heaven...

You should see if Rowan's initial design schematic is anywhere (or flutter your virtual eyelashes at him if he's taken it down, as I think he possibly did).

Personally, I feel more comfortable with a signal diode back-strapped across it in reverse-parallel, to ensure the vactrol's long life when subjected to AC. My intial suggestion that Rowan's design possibly wouldn't work was plainly wrong, but I still sleep bettter at night with the way I saw it working!

PS "Rowan" = 'Analag'

Keith
 
[quote author="SSLtech"][quote author="kiira"] light emitting resistor vactrol. [/quote]
Light emitting DIODE in most vactrols that I know of.
[/quote]

Yes that's a thinko, related to a typo.

Since the type of LEDs in vactrols will likely have a forward bias of about 2Volts, and LA-2a's tend to have a hundred-volt or so of sidechain drive, it'ws not a match made in heaven...

Well I just mean in principal not any actual voltages, components etc. I just meant a vactrol opto sidechain and the voltage amp and probably a cathode follower out to the xformer.

I was thinking 1 bottle for the sidechain drive, 2 bottles for the makeup gain part, the voltage amp to a cathode follower.. I'm not sure I have all the terms right. And I don't know if this idea is practical but it would be nice... 3 tubes and I/O transformers would be simple and nice for me to build.

Thanks for the help. No I haven't seen Rowan's design I don't know anything about it but he usually designs hybrid tube stuff, which I'm not into. In a way I'd rather NOT see it if it is similar to the ideas I have because I'd like some way to try and design/cargo cult my idea so I could learn some more. Problem is, there's lots of books and stuff on the web for designing an amp but I havn't seen anything about designing compressors that a math impaired beginner could understand.

Kiira
 
We've talked about this a little via back channels. From what you've told me about the size of the enclosure, etc.--and assuming you want to avoid using any solid-state amplifiers whatsoever--I really think it'd be better to build this "channel" as a preamp, EQ and line amp sharing a common power supply, bring the three "blocks" out to individual input/output jacks and use an external compressor as needed.
 
Actually, I guess what I really need is some help designing a tube sidechain using a vactrol. The rest is a line amp and I can muddle through with one of them, and if I can't then there are some awesome line amp designs right here I could use. I just don't know where to begin with the sidechain though... getting the specifications of various vactrols I guess would be a good start.

Kiira
 
[quote author="NewYorkDave"]We've talked about this a little via back channels. From what you've told me about the size of the enclosure, etc.--and assuming you want to avoid using any solid-state amplifiers whatsoever--I really think it'd be better to build this "channel" as a preamp, EQ and line amp sharing a common power supply, bring the three "blocks" out to individual input/output jacks and use an external compressor as needed.[/quote]

The huge 3U 16" deep one? I'm pretty sure I could fit a simple comp in between the pre at one end and the EQ at the other. I just wouldn't be able to fit an LA2 ar a 4 xformer veri-mu jam in there. At any rate I'm building the EQ and line stage, then the pre-amp and I can see what's happening with things then. :razz: I'll just try and make the pre and EQ compact, leave some room and go from there.

Oh forgot to add.. the case is deep enough to allow a generous size turret board to go front to back and that makes it a little bit easier to lay out nicely.



Thanks Dave!

Kiira
 
The LED-part of a vactrol likes low voltage, relatively high current - so if you want it driven by a tube, you should probably consider using a cathode follower.

Still, the easiest would be to use a simple solid-state sidechain - possibly fed off your heater voltage. Look at the Urei LA4 for a simple, but good setup.

Jakob E.
 
The rest [of the compressor] is a line amp

I thought about that, too. But what if you want to put EQ after compression, which (if you're like me) you'll find yourself wanting to do most of the time? The output from the passive EQ is pretty weak, so the line amp is pretty much married to the output of the EQ. So you'd have to add yet another line amp plus a sidechain circuit, at minimum, to implement a compressor.

Doing it all in one box is possible but, as I see it, would make not only for a daunting project but probably some undesirable compromises as well. But if you insist, you may want to consider a 436-type. It's only three tubes, two transformers (not counting power) and, although it has its shortcomings, it's a quite viable compressor if you don't require a fast attack (shorter than 27mSec or so).
 
Hi Jacob,

[quote author="gyraf"]The LED-part of a vactrol likes low voltage, relatively high current - so if you want it driven by a tube, you should probably consider using a cathode follower.
[/quote]

Oh thanks I was wondering how it should be driven and if a cathode follower was called for. I can start from there. yay.

The thing is I'm trying to learn something and it's been a while since I did anything original so the easiest or even the best solution isn't neccessarily what would be best for this project. This huge box was what I had my first tube pre in and I dismantled that so I should build something as original as possible in it again. Even if the compressor is a failure I can still use the pre and the EQ and I will have learned something.

Kiira
 
[quote author="NewYorkDave"]
The rest [of the compressor] is a line amp

I thought about that, too. But what if you want to put EQ after compression, which (if you're like me) you'll find yourself wanting to do most of the time? The output from the passive EQ is pretty weak, so the line amp is pretty much married to the output of the EQ. So you'd have to add yet another line amp plus a sidechain circuit, at minimum, to implement a compressor.
[/quote]

Well what I have sketched out so far is the makup gain for the EQ will actually fit at the rear of the turret board that the EQ is built on... in the back right side of the box after the EQ IOW. Ergonomically that works out pretty well. Yes I did want the EQ AFTER the compressor but I decided to bring each I/O out the back and patch them together as needed. It's not elegant but it keeps things flexible and simple. I don't want to be wiring up switches carrying signal or any relays or that stuff.

Doing it all in one box is possible but, as I see it, would make not only for a daunting project but probably some undesirable compromises as well. But if you insist, you may want to consider a 436-type. It's only three tubes, two transformers (not counting power) and, although it has its shortcomings, it's a quite viable compressor if you don't require a fast attack (shorter than 27mSec or so).

I was wondering about the 436. It is really pretty simple and the tubes are easy to get. If I abandon the opto idea then I will build a 436.

This will all be built on sturdy turret boards to. THis box was my prototyping box and I built and took apart several preamps on the boards that I had in there. If I want to change it, take something out just build something completely different if I don't like it then it won't be all that hard.

thanks again

Kiira
 
bringing everything out to i/o jacks (or even better, a patchbay) is definitely the way to go. Hardwiring the "blocks" together inside the box would substantially limit your options.

You could do a really simple opto comp with two twin-triodes. The output wouldn't be able to drive any seriously low impedances, though. I would seriously consider allowing yourself the use of one transistor to control the LED in the Vactrol.
 
[quote author="NewYorkDave"]bringing everything out to i/o jacks (or even better, a patchbay) is definitely the way to go. Hardwiring the "blocks" together inside the box would substantially limit your options.
[/quote]

Oh yes... patch bay it is.

You could do a really simple opto comp with two twin-triodes. The output wouldn't be able to drive any seriously low impedances, though. I would seriously consider allowing yourself the use of one transistor to control the LED in the Vactrol.

Ok I see. I am going to draw this as soon as I can... just a first try using the tubes I have on hand and see where it goes, As for the line stage... I could use two bottles, have a cathode follower and something like a 4:1 output transformer. The sidechain would operate the first triode acourse I guess, looking at an LA2. I'm thinking 6sn7s for most of this because I have a lot of good ones. Also if I don't peek at Paul's pre I'll have to figure it out by myself LOL.

Hi Wayne... thanks for that info. If I get this thing working at all I'll think about improving it. :razz:

thanks everyone. I'm having fun.

Kiira
 
Interesting project, but I think you are making a rod for your own back by trying to design a valve (tube) sidechain to drive one (or a pair) of Vactrol control elements. Since the sidechain is not in the audio signal path, and the limits on the attack and release times are governed by the characteristics of the Vactrol, save your valves for the audio path and design the sidechain with opamps, where the currents and voltages are much more compatible with the required LED drive.

We've covered this subject at some length in previous threads, but one of the design problems with Vactrols is repeatability between units. Getting matching over the range of compression between the two channels of a stereo compressor (single sidechain or two sidechains in parallel) is quite difficult. These difficulties become an acute problem when designing for a commercial product, when hand matching of elements is out of the question. The commercial designs I have done using Vactrols have usually had a servo scheme to normalise the resistances at specified points in their operating curves. You would probably get away with manual matching in a project such as this, but it should be taken into consideration when sketching out the design.
 
Use the T4 and the side chain, it's not that complicated.
Why re-invent the wheel?
All the work has been done for you.
Or, I could do some of the work, like 41 dollars worth, but I do not know your financial situation.

I have a couple of T4's going out to Kubi this week.

If you want to DIY it, I could send the pc board and can.

My favorite tube compliment for the LA2 is

V1 - 12AY7/6072
V2 - 6/12SN7
V3 - 12AX7z
V4 - 6V6GT

You can use the 6/12SN7 for V1 also.
Or 6SL7.
Been there, done that. They all sound poretty good.
 
"Why re-invent the wheel?
All the work has been done for you."

Maybe 'cause cloning is getting old and the LA2A in particular has been done to death?

But that's just my opinion, and it's not my project.

Boswell is right about tubes being far from ideal for driving an LED, considering the fact that several milliamps of drive are required for max brightness. To avoid the requirement for another set of power supply rails, high-voltage transistors could be used to good effect in the sidechain. It could all be built on a small PCB which mounts to the inside of the chassis with a couple of standoffs... Nobody needs to know the dirty little secret! :wink:
 
[quote author="NewYorkDave"]Boswell is right about tubes being far from ideal for driving an LED, considering the fact that several milliamps of drive are required for max brightness. [/quote]
Which was exactly my original thought.

In fact, you never need to send AL that much current through an LED in an opto-comp, but none the less, LEDs are Low-volt, medium current items, tubes are at their happiest with high-volt, low-current arrangements... which explains the T4B EL panel.... 100V or so, microamps of current... Bliss.

Anyhow... Either solid-state sidechain or step-down (voltage-reducing, current-multiplying) transformer will slightly better match to an LED.

Me I'm a little more familiar with low-voltage, high-current sand-state-stuff. If you like hollow state better, by all means go that route... -And you don't NEED super-high current... Look at Analag's solution: He loves the sound, and many of us have come to appreciate that he's no fool.

Even if you don't build that exact version, it'll show you the way.

Keith
 
A stepdown xfmr could definitely do it. But with any sort of reasonable (read: off-the-shelf) ratio, it would still take a fairly hot output stage to drive it. It could be done with, say, two of my One-Bottle preamp circuits, one as a signal amplifier and one as a sidechain amp. Add input, output and LED-driving transformers and that's two tubes and three transformers, getting a little bulky already...
 
[quote author="CJ"]Use the T4 and the side chain, it's not that complicated.
Why re-invent the wheel?
All the work has been done for you.
Or, I could do some of the work, like 41 dollars worth, but I do not know your financial situation.
[/quote]

Gee I have to explain this again. The thing is that I don't want all the work to be done for me because then I won't learn anything. Heck if it came to that I'd just find a nice schematic and build something I know works great or even just ask someone to build it for me. A lot of times in 3D guys just assume that I'd rather they did something for me and sometimes that's totally ok but other times it isn't. I don't care if it doesn't work awsomley just that it works ok for what it is and I don't care if I fail completely no big deal. But I'd like to try and mqke this as original as I can so I'll get better at making stuff and especially get more practice doing math because I suck at math. That's why I was thinking of building an tube opto comp because the LA2 is the only one I've seen so I'm kinda forced to do the work myself. There are so many incredible designers just right here on PPro that it's hard to think of something that someone else hasn't already done that is better than I could ever hope to do.

Anyway I drew out a schematic tonight but I have to do it in orcad to show it here because I don't have a scanner. It uses a cathode follower to drive a vactrol and the rest is just a line amp... it's kinda generic because I just wanted to get an idea in my head how things would go together I can do the real design later if it looks like this isn't just me being crazy.


I have a couple of T4's going out to Kubi this week.

If you want to DIY it, I could send the pc board and can.

That is wondeful CJ! The reason I didn't consider using a T4B is the price. But the thing about building a T4B though isn't the can or the PCB because I've learned how to make nice PCBs from art. It's the EL cell, I've read your instructions on building one but the EL seller wants $100 minimum order. foo. Do you have any of those? I could get some $$$ together 'specially since I got some new part time work yay. Oh, do I owe you money? I've been so scattered lately I forget.

I'd still want to design everything else as much as possible though and not use the LA2 schematic except for reference. For one thing an LA2 would again be too complicated for this project, I need something smaller and simpler - I'm hoping just three tubes - plus I still want to have this be as original as I can make it and just copy what I can't get past at all.

The 436 Dave suggested would be a good thing if I just can't make anything work. But using a T4B would be great I think because that is known to work with tubes just fine. Hokay, post too long.

much thanks CJ and everyone,

Kiira

EDIT: I just read the replies that were added while I was writing this and I'm kinda beginning to warm up a little on the solid state sidechain plan. I could probably power it off the heater voltage... or also the xformer I plan on using is 250 volts 100mA and has a tap for a bias supply which is around 23 volts I think... better check again. Maybe I could use that in some scheme. But I would like to get opinions on the schematic I did this evening especially Rowans since he's done this I guess? I don't think I'd like to see the schematic he did though for now.
 
Kiira,

Sidechain is just an amplifier. build in an LF rolloff if you like, but get it working first.

Basically, it sees the same post-attenuation signal that the makeup gain 'preamp' stage sees in the signal path (for a simple feed-back topology).

Then you stuff the output into a capacitor (NO DC please!) and then into the LED, then to ground. Strap a 1N4148 or IS44 diode reverse-parallel across the Vactrol LED, and it'll compress.

I promise you.

And in fact, while an LED may be rated at 20mA-30mA for full brightness, you'll NEVER drive it that hard... 4-5mA should be the tops... if you kill a tube I'll be surprised. Look at the LA2a topology and you'll see that the peak reduction control (feeding the sidechain) sees the same post-attenuator signal that the makeup gain (signal path) does.

You'll learn a bunch. Do it however you want. It will work. -And if it kills a tube, you probably have a spare -right? and I learn more from a single failure than a thousand easy sucesses.

Keith
 
Now I understand.
OK, go for it. There are some vactrol compressors out there, we know that.
Punk, Byron, those guys were into it pretty heavy.
Way back when.

You always have an out if the vactrol approach does not pan out.
So may as well do something different.
 
Back
Top