Opinions on Push-pull, class-A, transistor preamps

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

MaxDM

Well-known member
Joined
Oct 23, 2018
Messages
206
I've noticed that a lot of early transistor designs are differential circuits.

I suppose that, in a way, given the nature of the transistor, there is good reason to try and cancel distortion with such a circuit.

Most of the classic preamps are asymmetrical class-a gain blocks, though.

Any reason why they didn't catch on? How would you compare the sound to the more typical 3-transistor circuits?
 
Of the early transistor circuits, I really like the AM-16 and the slightly later Altec 9475, both designed by John Hall.

Class A, push-pull, DC coupled, what's not to like :)
 

Attachments

  • Altec9475a1.pdf
    73.6 KB · Views: 104
John Hall had test equipment too. Granted it wasn't all in one box or as easy to get the results as it is now, but he had the means and used them.

From both an engineering specifications and user's standpoint, I personally think those Langevin's and Altecs beat the pants off the 3 transistor Neve-type pre's of a few years later. Not as flexible of course due to the fixed gain, but apples to apples at 40dB or whatever, an easy call for me.

That last bit might get me thrown in The Brig for heresy against King Neve 😳
 
I found an interesting old thread online from 2005 where John contributes and talks about the Altec 9475 design.
His first post:

"I designed the 9475. It is an updated Langevin AM-16. Has wider freq. resp and less distortion. Also same ckt. was put in special package for MA BELL to use in repeater units. Had octal base like vacuum tubes. This was in 63 when ALTEC had the AUDIO CONTROL DIV. Schematics same as AM-16, just better xfmrs. ALTEC's Peerless xfmr div would not make the xfmrs. The eng dept said that it would not work. A company in Santa Ana , ROBOT COIL built them."


The rest here:
http://www.3daudioinc.com/3db/showthread.php?53-Altec-9475a-
 
I suspect there are two reasons. One would be cost because you pretty much need at least two transformers with center taps to source and sink the anti-phase signals. The other would be that those early designs used fairly low open loop gain with less feedback but with even two or three transistors you can achieve very high open loop gain and a high degree of feedback such that distortion is reduced to probably less than you would get from using two anti-phase circuits (which use twice as many parts and space and, again, $).
 
I probably got my numbers mixed up then, it's been years since I owned and used the Langevin's and Altecs. I read that thread at the time he wrote in it too and probably got the numbers mixed and stuck in my mind from then also.

Regardless, whichever is the AM-16 type of Altec, they're both great mic amps.
 
Are they? OK.
I don't have a 9470 schematic and the places I can find it online want money so, I'll post the AM-16 for folks that want to see that.
One difference I see in transformers between the 9475 and AM-16 is the output, which on the 9475 is a much simpler affair.
If John had left Altec by then, is it possible the Altec transformer division got their way in winding what they thought it should be, as he alludes to in that thread?
 

Attachments

  • Langevin AM-16.pdf
    1.5 MB · Views: 52
Cool, I did not know that. I had the rack mount Altecs, similar in appearance to the AM-16.

Good info from you too in that older thread you linked to.
I noticed that Port Barlow replied to you - I must have bought maybe two dozen AM-16's off him in the '90's, he ended up getting a lot of the old Langevin stock and parts off Manley after they bought the name and rights. Although at least half the AM-16's I got from him were labelled "Capitol Records' (printed in black on the input transformer) so, not sure what that was about.

Anyway, I agree with your observations from the time on the differences. I can't say I had a favourite of the two myself, although I certainly owned far fewer Altecs,. They just didn't come up that often, at least in the circle of folks I knew.

👍

P.S. I just remembered that the going rate for AM-16 then was about $150 each, including connector. I'll take a dozen now at that price please :D
 
Last edited:
P.S. I just remembered that the going rate for AM-16 then was about $150 each, including connector. I'll take a dozen now at that price please :D

Yeah back then no one had heard of the Altec, somehow. Yet the Nashville Bradley's Barn console was the Altec at one point, now lives at Niles City Sound after being at the former Sound City awhile.

When AM-16's where $150, Altecs were $50. And RCA tube pre's were $100....unless they were OP-6's, then $200. etc etc etc
 
Last edited:
As often, I just love to lean back and "listen" to your informed conversations.
But one thing I have to throw in, how to replace these unobtainable center tapped sec. (EDIT: they are 2 independent windings) transformers?
These Sowters are not inexpensive, but acceptable, but are they still available at all (hearing Mr.Sowter retired)?

Sowter catalog
from which also comes
AM-16 info sheet
 
Last edited:
Last time I spoke with Mr. Sowter (January)he was consulting for the company but had nothing to do with sales. Their web site says they are concentrating production an s small number of transformers. Cinemag is also experiencing delays. I am still waiting for a couple of inductors I ordered from them in Jamuary.


Cheers

ian
 
For the Altec scheme, there would be any number of output transformers you could try.
For the Langevin, the Sowters certainly worked when I tried them (they were in the Mercury Recording Equip. rack mount AM-16 which I did pcb and design work for, no idea about their 500 series version) - and hopefully Carnhill get up to speed and we start seeing more of the various transformers available.

But check it - I haven't tried these but Cinemag now have transformers suitable for the AM-16 circuit so that's a possible optipn to try too
 

Attachments

  • AN-107.pdf
    156.4 KB · Views: 31
Last edited:
Would modern matched transistor pairs in IC packages make building these amps easier than in the past?
 
Define easier. Are we talking about making 1000's, or just one or two for DIY?

On old Langevin modules I often replaced the smaller transistors and never went beyond nominal matching of β. There's a trimmer for balancing the push with the pull and it did the trick.

For what it's worth, those old transistors were very rarely defective, I was just putting in a slightly quieter and faster modern part. Not much difference when it was done but I felt better, that I'd actually done some work.

On the Mercury commercial box, I think I might have used 2SC2240's with decent current gain, but it's been ages so I might be wrong on the part.

If you were building 1,000's then I think the hardest part about it would be convincing anyone to buy all those transformers! :D
 
Would modern matched transistor pairs in IC packages make building these amps easier than in the past?

After repairing many Altecs (I've probably gone through around 100 of them) I found it's tough to find modern transistors with low enough hfe to avoid oscillation. The manual specs a minimum number, and almost everything new I tested was 10x that. I recall I could get away with about 4x max.
 
I imagine then, that a re-designed version with less transistors might work better?
 

Latest posts

Back
Top