Do schematics of the API 2520 over the years even exist or is it a secret?

[vdr20]

Specifically from the late 60s. Since they are no longer manufactured that way, is the schematic of the 2520 from that time known, but kept a secret? Or has no one been able to figure out the schematic due to the construction techniques used.

 

[gyraf]

API always kept this secret - which makes sense, as this is more or less the core of their business. Then and now.

There has been a couple of reverse-engineered schematics around here, precision of which may differ.

/Jakob E.

 

[JohnRoberts]

I don't know about back to the 60's but Paul Wolff was a service tech or something like that working at the old API (80s, 70s?) so may have some insight into their evolution. Documentation or willingness to share is another story.

JR

 

[vdr20]

API always kept this secret - which makes sense, as this is more or less the core of their business. Then and now.

There has been a couple of reverse-engineered schematics around here, precision of which may differ.

/Jakob E.

Out of all the reverse-engineered schematics, which one is the closest to any of the 60s versions?

 

[jensenmann]

IIRC I read in a Paul Wolff interview that the 2520 had 20 revisions over the years. Of course all sound magical and have THE API sound ;-) Decide yourself if it´s worth the effort to hunt down a certain "mojo" (=distortion) of a special version of this DOA.
From a manufacturers POV what happened was that transistors have been replaced because of unavailability and the circuit has been improved to bring certain parameters to better specs. On top of that it surely happened that the ladies in the assembly line installed other transistors than required because they were too lazy to go to the warehouse and pick a box of the original type... Now which one is the best 2520?

In 2020 it makes barely no sense to recreate something that old with unobtainum parts.
YMMV

 

[vdr20]

Okay thanks. Do you know the overlap between when the Melcor 1731 was used in API products i.e. what years the Melcor 1731 was used for API? I assume it was during the very early months of the company.

Considering (according to Paul Wolff), the commonly circulating schematic for the API 2520 is incorrect, do you think it's best to make a Jensen 990 or John Hardy 990 instead? I assume the schematic for the Melcor 1731 circulating is correct however I'm not to sure if it's too coloured for my purposes.

My plans are to build an Pultec EQP1-A with a solid state amplifier section, however I'm thinking that maybe it's best to have a non-coloured amplifier section so that I can choose the colour to be from the preamp driving the eq.

 

[jensenmann]

I made a bunch of 2520s over the years and used whatever transistors I had available. Some were done with original transistors when I could get my hands on them, others were done with equivalent types. I could not hear any difference. I even installed some as replacements in a 2500 and 5500 when I had them on the bench for repair. The owner, a high profile mastering guy, who´s name shall not be mentioned, didn´t notice any difference. If he doesn´t I surely won´t either.

What I can tell, though, is that Melcors are not a drop-in replacement for 2520s. I tried that several times and it didn´t work.

Jensen 990s are a very different beast. They want to see a very low source impedance at their inputs. If not they will fry. That means they are in no way a replacement for 2520s. This has been discussed here before. If you´re interested in details search here: https://groupdiy.com/index.php?topic=281

Solid state Pultecs are available here as a kit from our member Mitsos: https://groupdiy.com/index.php?topic=46920.0
They sound great. I have one of these and it´s the Pultec with the most solid low end I ever heard. I shot it out with some big $ brand Pultecs and it won hands down.
I´ve seen other products or DIY Pultecs which use a simple chip-based gainstage. A NE5534 is fine as well as some über-chip with spectacular specs. The good thing about DIY is: try and use whatever you like.

 

[JohnRoberts]

If an amplifier in a NF circuit is serviceable with adequate loop gain margin for the application, the transfer function will be dominated by the NF network, not the amplifier....

JR

 

[vdr20]

Thanks. What schematic did you use for the API 2520 as I'm struggling to find any accurate information. From the link you provided, that Pultec clone from Mistos uses a 1970s Huntington era 2520. Does that mean that there is a schematic circulating for that era of opamp?

 

[jensenmann]

Pm me your email, I can send you what I have.

 

[vdr20]

Thanks, I just sent you a PM.

 

[vdr20]

Your first e-mail didn't seem to go through. I just PM'd you another e-mail address. Hopefully that will work. Thanks

 

[jensenmann]

Tried it with the new email.

 

[vdr20]

Thanks I received them.  I might just go with the version of the 2520 containing the BC series of transistors. Do you know any information about the era of that version or it's origin?

 

[john12ax7]

Do you specifically want to do your own pcbs? If not there are plenty of diy kits available. The gar2520 would be a good choice.

 

[jensenmann]

I have no first hand information, but in that Paul Wolff interview I think he said that the version with TIS transistors is an early revision. The schematic with BC transistors is the same as the other one. The obsolete transistors have simply been replaced with currently available ones.

 

[vdr20]

I'm building it on veroboard. I have a bunch of Mullard capacitors that I would like to use with a larger footprint.

Also thanks for the information jensenmann.

 

[pucho812]

Paul Wolf once told me the main schematic that was floating around of the 2520 was incorrect. He never told me what was incorrect about it and I never confirmed if that was correct. I have no reason to doubt him.