RCA ba-25 6386 vari-mu compressor schematic / PRR synopsis

[wilebee]

I think a vf14's a bit more

 

[ToobieSnack]

yah those collectors tubes can go sky high ...
i was just exaggerating for effect .. lol

anyway i guess i can fire it up and check the voltages at the socket before i install the tube..

i am about to do that in the next hour or so ...
later
ts

 

[wilebee]

well with out the tube in place you won't get the appropriate voltage drops where they are suppose to be. I shold have sent you a 5670 with that 6386 then you could have at least seen if you voltages were correct. I know the tube is good ss do what you have to do but I wouldn't hessitate to install it. But the unit is not sitting right in fromt of me either so I can look for any potential problems . Wil

Wilebee

 

[ulysses]

Hey 2B,
Any chance you could put the schematic back up? I'm late to the party and would love to see that schematic. I had a pair of BA-25's a while back, maybe 5 years ago, that I bought on eBay for cheap. They weren't really useful as line-level compressors, because they had so much gain and a 40-year-old volume pot. But I used them as mike preamps with compression built in. They were painfully slow compressors, and I ended up selling them on eBay for a HUGE profit. But right before I sold them, I got around to cleaning them up and replacing ONE capacitor in the time constant part of the circuit. This made them MUCH faster and infinitely more useful. They also somehow became less noisy, Idunno. This was done of course with the consent of the buyer, who was very happy with them. At that point I hated to see them go, but I was very happy to see the big bag of money I got in their stead. I never did find a full schematic though, which is why I didn't clean them up sooner.
I like the idea of reversing the input transformer for lower gain.

 

[GregL]

I have one of these and also missed the schematic. If there's any way I could get a copy it would be greatly appreciated.

 

[ToobieSnack]

yah guys np i can re-up the schemo ..
i have even taken the schemo and used photoshop to color trace some of the paths (looks like highlighter marker or different colors)

in an attempt to try to better understand the unit.

i am glad to hear someone else is intereste in these beasts .. i have yet to get mine to work .... :?

i did check my 6386 voltages and the filament was right on and the plates were very close to the limit (375v...400v limit)

i thought these should have enough gain to be used as a mic pre (thanks for verifying that)

i plugged a ribbon mic in to see if i could get needle deflection ... nope ..
i was trying this with the output un-terminated .... (unloaded) would this make a huge differednce ... i mean i wouild still get needle deflection right?

the meter does work in GR mode but i get nothing in I/O modes ..
i guess i am off under the cover to see what my votages are ...
could this be a bad trafo .... ?
can io measure the trafo IN circuit?

my signal generator and scope are here somewhere but finding the leads too will take me some time .(l never really got straighted out from moving to this tiny apt)

ALSO: i have a whole suite of mods for these ... i have a wonderful friend from ebay and lives close to me here in NC .. this guy is genious ...:
i had hopes that he might join our forum ...... i'll try again soon ... :wink:

anyway i have my ftp stuffed full .. and i have a web account that will activate in a few days .. but i will try to find a host sooner than that ..
if you have andy free pic hosts lemmme know and i can upload the schemo sooner.
later
ts

 

[Butterylicious]

I'd like to read this schematic too. I could host it on my site if you like. eMail it to me or PM me or something.

[email protected]

BTW: If you do mail it to me, make the subject like clear and on topic. Tis address gets tons of SPAM and I can be a bit heavy-handed with the delete key.

 

[ToobieSnack]

hey all .. i didn't forget about you all .. hehe ... sorry for the delay ..
the colors will aid in discussing this schmatic ..
it was an attept to understand how the rectifier tube was controlling gain for compression ...
maybe someone can elaborate on this...
the purple sectiona are a mytery to me????
i hope this color code format proves to be helpful for the purposes of this discussion ..
here is the schemo:

looking forward to a great discussion of this ancient beast!!!! :grin:
later
ts

 

[PRR]

> the purple sectiona are a mytery to me????

The purple from the 6AL5 plates to the 6386 grids is the Control voltage.

12AU7 swings big push-pull voltage to the output transformer to drive the load. C14 C15 tap this off. 6AL5 allows only the negative peaks to pass. R40 and C16 slow the attack time. C16 and R39 set the release time. You can watch (or modify) this on P1 pin 6, but that isn't essential. The key thing is that the voltage on C39 goes to the 6386 grids.

At no or small output signal (12AU7 plate swing less than the voltage on the Threshold pot), the 6AL5 is dead. R39 and R11 R12 hold 6386 grids at ground. Gain is max.

For large swings on 12AU7 plates, 6AL5 starts conducting peaks. This drives C39 negative. This voltage passes to 6386 grids. 6386 conducts less, gain is reduced.

One problem is that, in addition to R40 and C39, R11 C5 and R12 C6 also slow the attack time. 56K/0.22uFd is 12mS, 680K/0.047uFd is 32mS. So the attack is really set at R11/12 C5/6, though R40 C39 have a strong effect. And the double-pole response can lead to limiting instability. (It gets worse when you look at C3 C4.) We might like to make C5 C6 smaller for faster attack, but that would cut bass response. That is a key detail in fast audio limiting: the GR rise time is within the audio band. Things are easier for slow limiters, or in radio where AVC is applied to stages that don't pass below 440KHz.

The 5Y3 rectifier is, of course, just a power rectifier. You could replace that section with a 300V battery and it would work the same.

The 0B2 is there because the rate of gain reduction depends on the voltages applied to the 6386. The 0B2 holds a constant supply voltage on the 6386 stage. (It is however a bit odd that the Threshold pot is not also given regulated voltage. I suspect the limiter is very soft-knee, the Threshold shifts the knee a bit but you don't get real GR action until the 6386 grid voltage drops a lot. In that range the 6386 supply voltage is the main reference for the output level, the threshold hardly matters. Anyway the places that this would be used would take care to have reasonably constant line voltages.)

 

[synthi]

Is going to happen the PCB making for this comp?
I think it would be a great adition to thegroup projects available :grin:

Synthi

 

[ToobieSnack]

synthi... YEAH! PCB ! :grin: :thumb:!
and the cool thing is:
this is ALREADY ON PCB in it's original form! :grin:

one could "almost" make the PCB from a pic ... :grin:

i have gently removed the bottom cover leaving her "naked"..
here is the "gear porn" ... lol ... and a head start on the pcb making for this .:grin:

I need to remove one transformer to get a really good pic if anyone is serious ... let me know ... here's the pic ...

later
ts

 

[synthi]

Great Toobie! :razz:

 

[ToobieSnack]

Well actually even though i did have a super hard time finding this schematic ..( i had an actual paper fairchild 670 schemo before this)
and i did some tracing ...in photoshop
the real cool part ... and thanks ... goes to PRR who provides such concise info ... THNX PRR :grin: :thumb:
later
ts

 

[wilebee]

hey toobie is that thing up and running yet ? how does it sound ? Wil

Wilebee

 

[ulysses]

I don't think I would bother making an exact clone of this device. It's not exactly famous, and although original units sell for a lot of money we're not talking about holy grail dollars here. The cool parts of this circuit are essentially identical to the Gates Sta-Level. The only real differences are the BA-25's weaknesses, in my opinion. I would be much more interested in a project that looks at several different designs. There are many vari-mu compressors out there, both old and new. There's the Altec stuff, for example. The biggest expense for a Sta-Level or BA-25 clone would be the 6386 tubes, which are practically unavailable. There are other tubes which would work every bit as well, while costing dramatically less. The transformers will still cost real money, but there's no reason the tubes have to be a deal breaker.
The BA-25 has two extra gain stages it really doesn't need, in addition to a high-ratio input transformer. It has far more output drive than a recording studio will ever need. And it's needlessly complex because it has a phase splitter circuit instead of a center-tapped input transformer secondary. It's a cool compressor, I wouldn't kick one out of my bed for eating crackers, but I wouldn't go to great lengths to recreate that exact circuit either.
And rather than copying a cluttered, rather sketchy layout from 50 years ago, why not do a new layout so you can make useful changes, such as altering the form factor of various components, so you can use newer, better capacitors, switches, pots, tube sockets, and so forth. I don't see the point of DIY if you're going to be stuck with the same mediocre options engineers struggled with in the past.

 

[emrr]

I think you guys must have been referencing one of the two I refurbished and modded for ebay sale last year. I had at least one buyer who thought it killed his Sta-level; I can't speak to the condition of that Sta-level though. I agree that there are simpler/better/lower gain circuits that will sound almost the same that can be built from scratch, unless you really need to have an exact clone. In which case you really need the real thing.

Attack and release are quoted in the manual as 12.5 ms and 1 sec. Gain is quoted at 70 db. You can use it as a mic pre, but that 100K input attenuator is going to reflect too low an input impedance back to the mic.

The ratios are in the 2:1 to 6:1 range, which will make it softer. The R38 / R47 relationship determines the ratio and threshold, and can easily be changed to get higher ratios since it is a low current voltage reference.

The quoted outputs for each ratio are:

3:1 5 dbm
4:1 14 dbm
5:1 18 dbm
6:1 21 dbm

...so the ratio does make a real difference in actual level.

These really need an output attenuator like a Daven type ladder so you can adjust the ratio (which is tied to threshold in this type of limiter) independently of output level.

Here's one I'm probably missing something about; seems to me the 6AL5 is working with the same PP inverted/non-inverted phase relationships that the amp stage is, so each section will be letting the relative negative wave pass. With the PP action this should add up to both the actual positive and negative waves being rectified on opposite sides, and likewise the same information also existing in a non-rectified manner (due to the half-wave action on each side), all mixed together at R40 where the 6AL5 plates join. Not sure that's very clear reading.

I think in many ways the Sta-Level is as popular as it is because it's knobs make sense to the modern user who tends to get scared of high gain devices with knobs turned way down to achieve unity gain. The Sta-Level is already set up with a 20 db input pad and an output pad for a 35 db total gain range. It has a very low output adjustment range of only 6db. This seems to feel safer to use to many folks.

The Collins 356E and 26J along with the GE Unilevel are pretty much the same circuit, though lacking the 12AT7 driver tube, and having no input or output pads, or attached meter.

 

[skipwave]

[quote author="emrr"]The Collins 356E and 26J along with the GE Unilevel are pretty much the same circuit, though lacking the 12AT7 driver tube, and having no input or output pads, or attached meter.[/quote]

It is so great to have somebody who has actual had his head inside these various limiters to discuss them! :thumb:

I'm glad you brought up the Collins. In another thread I think you mentioned that the Collins side-chain caps allow bass freq to pass, but the Sta-level does not. I've been looking at the schematics and I see the same cap values in the side chain. Did that only apply to the Collins 26U, and not the 356E/Unilevel?Thread with Collins 356E/GE Uni-level schematics. Also, PRR said the 356E/Unilevel used a weird feedback scheme that only has an affect near the end of the 6386's range. What's that about?

Sorry to hijack, but while we are comparing... might as well ask. Many thanks!!! :sam: :sam: :sam:

 

[emrr]

Yes, only the 26U. 26U has 1 mfd caps feeding the rectifiers, as opposed to the more normally seen 0.1 mfd in most other units. This will equate to about a 3 octave difference in low end hitting the sidechain. Does not apply to the 356E or 26J which match the Unilevel. All that low end presents an enormous amount of extra energy to the sidechain, so it can be critical to issues like thumping. It probably does change the knee, now that I think about it, but most of us are using these circuits for color rather than accuracy, right? It doesn't really affect the usefulness that I can see, rather it extends the useful range if you really want the sound that comes with a lot of meter movement without obvious pumping. Of course you can always put in a switch for various cap values.....

I recall the Collins 356E and GE Unilevel being identical circuits, made with slightly different iron. Both iron from ADC, though, so it could really be identical other than casing and labeling.

These and the Sta-Level all have pretty much the same compression sidechain bias network feeding the 6386's, so not sure what PRR means. Not too different from many many others, either. Any and all are easily changed to different operating points. I think the Sta-Level manual has resistor values listed for permanent change to different operating points, though a pot or stepped switch makes the most sense.

Makes the price of Sta-levels look pretty silly.

 

[soundguy]

[quote author="emrr"]Makes the price of Sta-levels look pretty silly.[/quote]

like anyone cares, but I remember when sta-levels were $350 I thought they were too expensive...

dave

 

[NewYorkDave]

Dude, I remember when Sta-Levels were $25 to $50, or free if you happened to be in the right place at the right time.

"If I knew then what I know now..."