Using Alternate Transformers in ClassicAPI Preamps

[JuniusRecordingCo]

Years ago, I purchased four Old School Audio (OSA) preamps and a 500-series rack and PSU. Since then, every single one has had to have been serviced or replaced, including the power supply. Buttons stick, pots turn uselessly, phantom stops working, the PSU popped and smoked - the point is, I've not been happy with them.

Since then, I've been building Jeff Steiger's Classic Audio Products of Illinois (CAPI) VP26 modules to great success. And I want to continue adding to my collection - but wondered if I couldn't harvest the OSA preamps for parts. Specifically, the transformers and DOAs.

I have two each of two OSA models: MP1-C and MP1-L3. The C model has a "custom Crimson" input transformer and the L3 has a Lundahl LL7903. I have no idea what the output transformers are (don't have my hands on the units at the moment; relying on Google for information).

My electronics ability has been limited to "paint-by-numbers" until now, and Jeff's kits are fantastic for getting a start.

I was hoping someone could help guide me in the process of using these parts in Jeff's circuit? I'm aware the turn ratio on the LL7903 is different; I have no idea about the Crimson. I understand, too, that I'd clearly need to "deadbug" these to the board, if they'll fit at all.

How "plug-and-play" are input transformers? Without pin-out diagrams, how can I measure the Crimson? Should I assume the DOA is a drop-in replacement?

Any and all help is appreciated.

 

[desol]

I don't know if you'd want to do that. Probably best keep all the new preamps using the same components...rather than creating a kitchen sink kind of thing.

OTOH, you could probably use those other components for an entirely different DIY project.

 

[okgb]

the classicapi's don't have multiple footprints for input transformers , the crimson may fit but unlikely about the lunhdal
you may be able to save some money and have alternate pres ,  it's a matter of fitting the parts .
Don't forget the VP28 AND Hairball Lola  ,  those are a couple of nice sounding 500 type pres

 

[JuniusRecordingCo]

I don't quite get the "kitchen sink" comment - I have now four API-ish 500-series preamps, and would end up with four API-ish 500-series preamps with exactly the same transformers and DOAs - but hopefully a better backbone and build quality. The only real problem I foresee is the size of the output transformer - it might not fit without modification to the PCB.

Though these OSA preamps are troublesome - they are in service. I'd like to leverage these parts, if I can: they're the bulk of the expense of the preamps.

I'm reading up on the transformer sticky - but guidance is still appreciated!

@okgb: yeah, which is why I mention "deadbugging" the transformers. Never done it, but it seems simple enough: flip the things over, adhere it to the board and run leads to where they need to go. Apart from the size and fit - I am most interested in learning more about the interchangeability of the parts with the circuit.

 

[plumsolly]

If the DOAs from the OSA preamps have the same footprint (and I can't imagine they wouldn't) then they should work fine with no modifications. I would guess they are 2520's or very close.

The transformers should also work fine if you can get them to fit. Exact ratio is not imperative. There are Zobel networks that are specific to each transformer, but that should be easy enough to deal with. See what OSA used (if anything) as a Zobel for the Crimson, and look at the data sheet for the Lundahl.

Output transformers will also likely be fine to replace with no modifications.

Best,

Ben

 

[desol]

I don't quite get the "kitchen sink" comment

For me personally, i don't like mixing up components, preamps, opamps, etc. I believe in keeping my preamp(s) the same...across all the channels being tracked.

If it were me, i'd probably refurb the osa stuff(with better parts) so it's working good...sell them as good sounding, good working units(which they most likely are) and then buy more of the classic api preamps that i already had. That's just the way i'd do it....to keep my preamps identical. ie: 10 channels of x preamp...and call it a day

 

[nielsk]

Go for it, you can usually secure & wire the transformers if they do not fit the footprint, and their clone of the 2520 will work no problem.
It is expensive & a PIA to change the switches in the OSA's, I say your idea has merit.

 

[EisenAudio]

If I recall correctly...

I believe OSA used Crimson Audio CT-2385 output transformers (see attached). And their "output" pot was merely a 1k voltage divider across the output transformer secondary. Granted, the overall circuit is really not much different than any API, especially Classic's. A brief examination of the OSA PCB should tell you what to use for input termination and op amp feedback/compensation RC values. Probably 100k and 20k || 100pF, or close to it.

 

[okgb]

Hey Jens , thanks for stopping by & Nice to see you

 

[JuniusRecordingCo]

OK. I successfully de-soldered and removed the op-amps and transformers, with one unfortunate casualty: the Lundahl. The pins on one side ripped right away from the body. I can't see how it could be repaired.

Sad.

But, moving on, I've populated the CAPI VP26 board and cut away a little bit of the board to make the output transformer fit. My problem is how to wire it up. I've got the two schematics, one that Jens provided above, and the one from the EA-2623 (http://www.classicapi.com/catalog/images/gallery/2623/2623-1-specs.pdf). I'm going to have to ignore two of the Crimson's leads, but can't figure out how to determine which ones. How should I wire this up - or better, how would I know which ones to ignore from the schematics?

 

[JuniusRecordingCo]

The damage.

 

[JuniusRecordingCo]

Looking at the OSA board, it looks like four of the leads were terminated to themselves: Yellow and Green were connected to each other, and Blue and Violet were connected, too. Orange ran to the "Output" pot, Brown and Red to the "Phase" switch, and Gray to pin 4 (Out-).

On the CAPI board, Green and Brown are connected to themselves, Black to pin 4 (Out-), Orange and Red go to the Mute and then Phase switch, Blue runs to the Output Attenuate pot.

Using only logic, and not much real electronics experience, I get:

OSA      CAPI      Function
-----      ------      -------------
Gray      Black      Signal Out (-)
Violet    n/a        Interconnection
Blue      n/a        Interconnection
Green    Brown    Interconnection
Yellow  Green    Interconnection
Orange  Blue      Signal Out (+)
Red        Orange  Signal In (+/- phase select)
Brown    Red      Signal In (-/+ phase select)

How completely wrong am I?

 

[MatthisD]

For 1:2 connection, paralleled primaries try:

OSA                        CAPI      Function
-----                        ------      -------------
Blue                      Black      Secondary start
Red                        Brown   
Green                    Green   
Brown                    Blue     
Gray & yellow        Orange  Primary starts
Violet & orange    Red       

 

[CJ]

the 7903 can easily be fixed as the wire is fairly big for an input,

if not, don't worry, it is not the best LL coil out there, it is a problem solver type that tales huge input levels, possibly with a sonic trad3eoff,

i might have a crimson audio output here from somebody who wants to kill me, i guess i should do a " take apart before" the car bomb goes off, we will fix this situation shortly,



i wound a 2503 on a nickel core, did not do anything to improve sound, hard to improve stock api,

or melcor, think one of my melcors is melting, once in a while it crackles a bit,

 

[madshamanstudios]

Light years late to this thread but I'm wondering about this mod idea:

Replace a VP28 input transformer with the Reichenbach RE-0887 by Cinemag used in the CAPI Heider FD312

REASONING
I found this language interesting from the CAPI FD312 page:
"The original Heider/DeMedio preamp cards used stock AP2503 output transformers..."
But the CAPI Heider model uses newly designed Cinemag output transformers.
So how about an RE-0887 Heider input transformer in combination with more classic API output transformers?

Now, the VP28 doesn't use 2503s, it uses 2623s. But based on Jeff's gearslutz comment, the 2623 "was typically the only output transformer used anywhere in the old 2488 recording console." So that would still be a classic API style output transformer.

I know the AC coupling capacitors were removed in the Heider/DeMedio console, I'm not trying to get that deep. I'm not experienced in the concepts and maths of audio electronics. More of a "switch the opamp" level kind of tweaker. But I can solder.

QUESTION
Just wondering if this input transformer change would be:
- possible
- easy
- worth trying

Any of you kind folks have thoughts on this?

 

[FIX]

Their is no reason to remove the coupling caps from these modules of consoles. If the op-amp has any DC offset, you will slowly bake the transformers. If one fails you will bake the transformer. Just replace them with new, and going up a value will accomplish more than removing them.

Jeff is correct, the 2623 was the standard transformer used in just about every older API console, unless they used the 300 series cards. Switching op-amps or transformers is not the best thing to do, as the transformer was designed with the op-amp, and working together, they could a certain way. Replacing one of the other may not have a positive effect.

The different choices the Jeff offers are all in the same ball park, so they all have slightly different tones. The original 2503 was a quadfilar transformer where the 2623 was trifilar, meaning that all the windings are wound at the same time with the same ratio.

Nickel has a smoother low end, and a mix of steel and nickel can be tailored from cleaner to dirtier.

I would buy anything that Jeff makes, but I would focus on the song and lyrics more than the minutia of swapping out transformers...

 

[madshamanstudios]

Their is no reason to remove the coupling caps from these modules of consoles. If the op-amp has any DC offset, you will slowly bake the transformers. If one fails you will bake the transformer. Just replace them with new, and going up a value will accomplish more than removing them.

Jeff is correct, the 2623 was the standard transformer used in just about every older API console, unless they used the 300 series cards. Switching op-amps or transformers is not the best thing to do, as the transformer was designed with the op-amp, and working together, they could a certain way. Replacing one of the other may not have a positive effect.

The different choices the Jeff offers are all in the same ball park, so they all have slightly different tones. The original 2503 was a quadfilar transformer where the 2623 was trifilar, meaning that all the windings are wound at the same time with the same ratio.

Nickel has a smoother low end, and a mix of steel and nickel can be tailored from cleaner to dirtier.

I would buy anything that Jeff makes, but I would focus on the song and lyrics more than the minutia of swapping out transformers...

Great advice, thank you! But...I bought a RE-0887 before reading this! I'd like to use it since I bought it. This RE-0887 would follow into a 2520, as it does on the Heider. Since both my CAPI and the RE-0887 input transformer were designed with the same op amp, they should be switchable, no? Even if it is affirmatively minute and theoretically pointless. OK! I'll get back to songwriting.

 

[FIX]

Well the main rule in audio is that their is no rule... Try it and see if you like it!

 

[MisterCMRR]

There's a lot more to DC offset and transformers than just "baking" the transformer. Even a few mV of DC offset can change the distortion profile (mix of odd and even harmonics) and especially so at low to medium signal levels! This is true for any core material that has a magnetic hysteresis loop - which is any used in audio applications. Ignoring DC offsets may work in a "no rules" world, but not in real engineering. In any case, with an output transformer (low winding DCR), if enough current to "bake" the transformer is present, it's almost certain the line driver stage will have already melted down.