input impedance of Altran C-3837-1

[orangerec]

I'm trying to build an H-pad for my 1176 Rev-A following the 600ohm t-pad attenuator.  I can barely turn up my input pot without getting instant compression. I had a tech look at it an we both agree the thing sounds amazing but we don't have much resolution on the pot. All the parts follow mnats schems and like I said, I love the way it sounds.
Anyway, my tech recommended installing the H-pad witch there are there's a spot on the mnats x-frmr board. So, does anyone know the input impedance of the Altran C-3837-1 (purchased from Hairball audio)? The output of the t-pad is 600ohms but in need the input of the Altran C-3837-1.

Thanks in advance. 

 

[orangerec]

I measured the input of my Altran at 50ohms. Can anyone else verify? This doesn't seem right?

 

[PRR]

> I measured the input of my Altran at 50ohms.

How? With an ohm-meter?

An ideal transformer has zero DC resistance (what the ohm-meter reads) and either infinite or reflected AC (audio) impedance.

In real life, the DC resistance is just "small", like 5%-10% of the nominal audio impedance. 50 ohms DC is not out of line with 600 ohms nominal audio impedance.

If you must know: Build a 10K:100 voltage divider, to feed with 100 ohm impedance, at VERY low level (obviously the thing does not like large level) from your audio signal generator. Sweep 10Hz-100KHz (or 20cps-40KC if that's what you got). Measure voltage at the 100 ohm, un-loaded and with the compressor attached. If the voltage drops to half (from 4mV to 2mV) then the impedance is 100 ohms, if it only drops a tenth (4mV to 3.6mV) then impedance is roughly 10 times higher. You will probably find a low impedance at 10Hz, probably going to 600 (or 150) by 50Hz. It may stay steady, or it could rise and rise to midrange.

 

[orangerec]

Yes, I used an ohm-meter. It's all I have.
Wow, thanks for your answer. It's a bit above me, I need to do some research and thinking. I found an H-pad calculator online and it just asked for the in/out impedance and desired attenuation.
http://www.nu9n.com/tpad-calculator.html
I figured, I know the t-pad is 600ohms, just had to figure out the Altran's, plug in the numbers and build it. Or, is anyone else using the h-pad with the resistor values on a rev-d or rev-a with the Altran input x-frmr?

Thank you, I will figure this out.

 

[emrr]

I don't know what's up with the H pad on the Mnats board, but an H and a T shouldn't be connected directly together.  You have one T already, one generally uses another T or L.  Besides, an H is two more resistors.   I'd probably try a switchable L in front of the variable T, with an eye towards raising actual input Z to a modern standard.  A 680 shunt and whatever series input Z gives the desired attenuation.  There are plenty of attenuator calculators out there; many times physical experimentation is faster.   The Altran is supposed(?) to be suitable replacement for the UTC O-12, which we know to be 500/600 ohms, so there's your probable answer.

 

[PRR]

> I'd probably try a ...

I'd try just a 1K pot. We aren't trimming fine cabinets. We don't need a precision fit. Good transformers are tolerant of large impedance "mismatch". And most will be quite happy with a low-Z source. The 1K pot output is zero to ~~250 ohm impedance, maybe 400 ohm if there is a true 600 source in front. Its all "500-600", the probable transformer impedance.

Yes, one simple pot. Yes, "unbalanced" drive to the transformer. If the tranny is good, and the wires are short, this is not a sin.

 

[PRR]

Isn't an 1176 an FET limiter? Then the path -after- the transformer is surely unbalanced. Unless you are feeding whopping levels into the transformer, an appropriate pot -after- the transformer is elegant.

 

[orangerec]

One simple pot. Thanks PRR. I'll try a 1k pot post the x-frmr, pre the amp.

 

[emrr]

pot would replace R4, the 270.  So, a (perfect) 250 audio pot on top of a 20 ohm would be an 'exact' match for the transformer.  Maybe 500 ohm audio is all you can find, and works just fine too.  But, now you have two variable attenuators, which does nothing but make knob(s) position 'look' more aesthetically pleasing to the eye, while adding an additional stage to share the load, and possibly introduce other problems.  I say the existing attenuator is already fine.  Now that I think about it, simply adding a series resistor to the input before the T may in of itself create a better range.  It probably changes the taper too, as downside.  But, an L made of one (value to be determined) series and one 680 shunt before the T would both fix the load on the T and give additional fixed attenuation.  I think far better than two pots.  You could determine the value of the series with a pot wired as rheostat, then measure, then insert nearest fixed value resistor.  And I'd still make it switchable, for those times you want to really blow something up or need more gain than usual.  YMMV

 

[orangerec]

Emrr, excuse my technical ignorance but going off of what you mentioned, why can't I just change R4 to a higher value of 500ohms? Is it because it won't match the x-frmr's output? I think I just answered my own question. I'm going to have to wait for the weekend to try all this out. Also, going to try PRR's suggestion with the simple pot after the x-frmr.

thanks again.

 

[emrr]

PRR can correct me on this, but I think he's shooting from the hip without looking at the actual circuit, and realizing there would then be two variable series attenuators.  My point, is another fixed pad up front is technically simpler and easier to implement.  It's only harder in the thinking and math part.  And experimentation avoids both. 

You did just answer your own question.  Whether or not it's a problem is a different question. 

 

[orangerec]

For anyone interested, I tried an L-pad before the T-pad with a 9dB drop and it balanced out the potentiometer resolution nicely. It matches my other Urei 1176s more with the 10 and 2 o'clock positions on the input/output pots. Thanks for your help. 

 

[emrr]

Glad it's working for ya!

 

[edanderson]

just a note to say that part of the behavior you are seeing is "how it is supposed to be" -- the rev A had IIRC about 6db lower threshhold for onset of compression than the later revisions, due to higher gain in the first amp stage.  this is part of why people think they sound so different from the rev D, etc.  people who know the rev D would set the knobs on the rev A the same, and the rev A would compress at lower input levels.

your solution will work, but you might consider instead looking at the schematic where the pickoff to the sidechain starts (at the top of the output pot, R23) and putting an attenuator there or possibly right at the input to the sidechain (C17).  that way you don't have to attenuate your audio all the time to get the threshold in the range you want.

ed

btw -- the C-3837-1 is a replica of the UTC O-12 used in reverse, so it is a nominal 500ohm to 200ohm stepdown.

 

[orangerec]

Ed, thank you so much for the input. I have a couple sessions where I need all my 1176s, original and Diy. I'll see how they do and then experiment with your recommendations.
Isn't it great that we can build gear as good as the originals? Thank you to all of you who make all this possible, I love learning all this stuff.

Cheers. 

 

[mpaavola]

The datasheet, is your best source for such information.

https://library.hairballaudio.com/datasheets/3837.pdf