Langevin AM-16 DIY?

[ytrehalf]

[quote author="soundguy"]no...

patch it in to a limiter on the out or build a variable pad for the input, which you are going to need anyway.

dave[/quote]

I patch my AM16s into a Distressor when recording electric guitar to knock down the AM16s +26db output to a level my converters can handle.

The AM16 strikes me as sounding pretty accurate, not very colored, but it yeilds nicer (smoother) sounding upper mids than what goes into it, which, I guess, is why they work so well for electric guitar.

As complicated as the design may be, the parts count is very low, about 10 transistors, I think there's only one capacitor and a few resistors. But that, in my opinion, would tend to back what others have said about the sound being in the iron in the case of the AM16.

 

[rafafredd]

I´m sure it will not sound the same with the sowter transformer, but probably very close, or maybe very nice in their own. Maybe you could even like the sowter version better, who knows?

A close friend of mine has a pair, so, if I ever buy the sowter trafos to try a DIY version, I will be able to compare.

 

[CJ]

There are two settings for gain if I remember correctly.
I don't know if they make that much difference.

For vocals, it's like having a compressor built right in to the preamp itself.
I thought I remember something about using them on kick drums also.
DC coupled, so good thud. Maybe wrong here.

BTW, the very first article I read on Vintage equipment was here:

http://www.mercenary.com/readguidtovi.html

There are a few errrors in it, but it is still one of the best articles I have seen.

Brian said he worked hard to get the balance right on his, so I think they will sound fine. Can't wait to see. No other options really, due to all those windings. Never seen another transformer like it. Might need some breaking in as the original iron was 30 or so years old.

 

[NewYorkDave]

Output attenuator, huh?

Sounds like a good time to build that switchable attenuator I posted a while ago! :wink:

 

[Rob Flinn]

There are two settings for gain if I remember correctly.
I don't know if they make that much difference.

It is worth noting that when the unit is strapped for LOW power the harmonic generation is 0.5% as opposed to 0.75% when strapped for high power. Maybe it sounds more useful strapped one way than the other !

I always thought it useful to have the attenuator on the output. The reason being that the AM16 has several different impedance inputs by changing the i/p transformer wiring. If you use low Z ribbon mics this is a useful feature to tap (excuse the pun) into ! I worked out a method to wire mine up using a rotary switch which gives 50, 150, 600 & 600 with 48v phantom.

 

[Rob Flinn]

One other thing !

All the small black transistors in both of my original "untampered with" AM16 are 2n5210. I can`t read what the other 2 are without removing them, because they`re mounted upside down.

 

[Sleeper]

One of these was the first DIY module rack-up that I ever did.
Boy this got me hooked they sound so amazing. I recently had to replace some transistors and used the 2n30... (crap- this page isn't showing up on topic review...) anyhow, they work great, I had to pull out one or two of them and replace with more evenly matched ones because the bias was too far off to get it centered using the bias pot. Speaking of that, plug a mic in when you adjust the bias, there is a spot where the noise drops to nothing, and then the gain starts to climb , I pushed mine just a bit, push em too hard and they will oscillate, but jeez, this just puts the sound so in your face.

The input pad is a good idea, these are really hot pres so it can be difficult to get mic input low enough especially if you are recording in tight quarters.

Sleeper

 

[PRR]

> Any way of varying the gain in this thing?

Change R1 until it screams. Lower R1 is higher gain, and it probably has more gain than you could want. However DC balance will suffer. (And anybody cloning this better copy the copper resistances, as they are part of the DC network.) Distortion will rise, though I bet an actual unit has far lower THD than they claimed, and pretty mellow distortion.

Traditionally, these are used in a broadcast studio where levels don't vary all that much, so you leave the amp at fixed gain and adjust the mix on the fader that feeds the mix-network.

> he just shook his head, saying that it probably took years to design and tweak. I still don't understand it.

McIntosh would understand: the output stage is his idea. Half cathode follower, with the other half of the load in the plate. Gives great linearity and damping without the excessive drive needed by a straight cathode follower. Still needs a lot of drive, so McIntosh (and Langevin) bootstrap the drivers from the output stage.

The last three stages are just a triple Darlington. Think of a giant super-Beta emitter follower. First stage has all the voltage gain between the transformers, about 1:34. The output ratio from one emitter to 600Ω is about 1:2. This implies the input step-up and impedance is pretty mild.

AC gain is set by the R2 R3 feedback pair against R1 and the R8 R9 R10 network.

> There are two settings for gain if I remember correctly.

No, two power-consumption settings depending how much power you need out of it: 64mW or 250mW. Both give 40dB gain.

> I guess you could use many options at the input.

No, it is very unusual to find well-balanced push-pull secondaries at mike level. Not to mention split-winding so you can insert feedback. However it sounds like Sleeper here prefers the unbalanced sound of pushing the bias pot too far to one side, so a poorly balanced input tranny might also be kewl.

> If you're someone who worries about capacitors directly in the signal path then this is the one for you!

Langevin was on a kick to eliminate ALL electrolytics from the amplifier. Transformers, resistors, and even transistors had become reliable, but electrolytics still failed on-air. This was a historic dead-end, electros got better. Also the Langevins were built to last forever, and probably would, but we threw out all that old junk in the 1970s and 1980s. (Wonder how Sleeper managed to smoke transistors?) Still an interesting design concept.

 

[Helsing]

This pre kicks ass. Ive used it on hip hop, rock, pop, country, drums vox, bass, guitar. Great overall use pre.

I use a dual 10k log pot attenuator for the ouput. +/- outputs into pin 3s of the pot, tie pin 1s together and take your output off the wiper. If the gain is reversed switch pin 1 for 3. I know its not the "right" way to do it, but if your concern is turning your current paperweight into something that sounds great for a pot, a few feet of balanced cable, two xlrs and a power supply, just give it a try.

I had a mod when I was racking these for folks where I could squeeze out an extra 5 db of gain. Ill dig out the mod and post it.

Jon Erickson
www.helsingaudio.com

ps This is also a quick and dirty attenuator solution for the tele V72 and V72a. Oliver Archut taught it to me years back so it couldnt be too wrong!

 

[Rob Flinn]

A dual 10K log pot seems like quite a high impedance for the 150 or 600 ohm transformer to feed. Most transformers like to be terminated at roughly their o/p impedance.

There was this thread a while back with some suggestions on how to tame the output of this one:

http://www.groupdiy.com/index.php?topic=483&start=0&sid=bfafbafbd40cb85c49c225c83089c692

 

[rafafredd]

And anybody cloning this better copy the copper resistances, as they are part of the DC network

WOW, PRR...

Never heard about something like this. So it sounds like a clone is just possible with a perfect or close enought PCB. I was thinking about doing my own single sided boards to use with the sowter transformers but it doesn´t looks like a good idea anymore.

 

[alk509]

[quote author="PRR"]And anybody cloning this better copy the copper resistances, as they are part of the DC network[/quote]

Damn... and I was already making a PCB for this

So what exactly happens if the copper resistances aren't right? Is there any way to DIY this thing without copying the PCB?

* * * EDIT * * *

Excuse my ignorance, but what exactly is the copper doing in this circuit? What's that DC network dealy?

Peace,
Al.

 

[NewYorkDave]

Yo, guys, I'm pretty sure he's talking about the copper resistance of the transformers, not the PCB traces.

What's the copper doing in the circuit? Conducting the electrons :wink:

 

[PRR]

> close enough PCB

The PCB hardly matters. Just keep it tight, about the same size as the original. You'd have to go to wild extremes before PCB resistance mattered.

The output windings have about 25Ω DC resistance. (Much more than any sane PCB trace.) That is also the DC current sensing to stabilize the DC bias. I do not know the exact value, but any clone winding needs to be close-to-right. I mention it only because I thought someone said a clone transformer was being prototyped. And if my estimate is right, the winding resistance is kind of high compared to the working impedance, while we usually aim for low DC resistance, so the usual rules of thumb for wire gauge have to be modified.

 

[rafafredd]

[quote author="NewYorkDave"]Yo, guys, I'm pretty sure he's talking about the copper resistance of the transformers, not the PCB traces.

What's the copper doing in the circuit? Conducting the electrons :wink:[/quote]

Ok, I´m a dumb :?

Now, would be nice to see that PCB you have been working on, ALK...

 

[alk509]

[quote author="rafafredd"]Ok, I´m a dumb :?

Now, would be nice to see that PCB you have been working on, ALK...[/quote]

Yeah, that whole copper resistance thing threw me off too!

As soon as the PCB is done I'll post it.

Peace,
Al.

 

[jrmintz]

As soon as the PCB is done I'll post it.

That would be very nice Al. Thanks.

:thumb:

 

[NewYorkDave]

Why not just remove the components from a real AM16 and throw the PCB on the scanner?

Then you'll have a layout that's guaranteed to work as well as the original.

 

[alk509]

[quote author="NewYorkDave"]Why not just remove the components from a real AM16 and throw the PCB on the scanner?[/quote]

Sounds great! Anybody want to go ahead and do it?

I just spent about two hours making a PCB for this, but my net list had an error and now there's a trace missing

Of course, you could jumper it off, but who wants that? :roll:

I need a break.

Peace,
Al.

 

[NewYorkDave]

I didn't mean to sound facetious. As a former full-time bench tech, the idea of desoldering those few components doesn't seem as daunting to me as it might to others. Problem is, I don't have an AM16 :roll: