THAT1200 vs THAT1240

Help Support GroupDIY:

Script

Well-known member
Joined
Nov 26, 2008
Messages
1,315
Location
Japan
I want to use a combo of line receiver (THAT12xx) plus line driver (THAT 1646) to achieve high precision gain of 6dB to drive a 10k:10k input transformer. (I am aware that a step-up transformer is the best solution.)

Browsing the datasheets, I can discern a different periphery for the THAT1200 and better CMR results with the 1200 when hit with 'unwanted' deviations in input impedance. Should not be a huge problem here cos my output resistors are all meticulously matched, stil...

Also, the 124x's can be configured for positive gains, which is kind of neat. So a combo of THAT1246 + THAT1646 can yield +12dB. That's more than I need though.

Anyway, is there any other reasons than better potentially CMR why to prefer the 120x over the 124x?
 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
12,347
Location
Marcelland
Script said:
I want to use a combo of line receiver (THAT12xx) plus line driver (THAT 1646) to achieve high precision gain of 6dB to drive a 10k:10k input transformer. (I am aware that a step-up transformer is the best solution.)
This is debatable.  ;)

Browsing the datasheets, I can discern a different periphery for the THAT1200 and better CMR results with the 1200 when hit with 'unwanted' deviations in input impedance. Should not be a huge problem here cos my output resistors are all meticulously matched, stil...
Indeed, teh 1200 and 1240 are based on a different principle. However, I have not found the difference in performance significant enough in practice to justify spending the extra $.

Also, the 124x's can be configured for positive gains,
AFAIK, both are available in 0dB, 3dB and 6dB versions, while the 1243 can also be used in 3dB attenuation and the 1246 in 6dB attenuation. which is kind of neat.

So a combo of THAT1246 + THAT1646 can yield +12dB.
that's right.

Anyway, is there any other reasons than better potentially CMR why to prefer the 120x over the 124x?
[/quote] On the contrary; I've always favoured the 1240 because it's pin-to-pin compatible with SSM and TI parts. With the current semiconductor shortage, it's not unsignificant.
 

Script

Well-known member
Joined
Nov 26, 2008
Messages
1,315
Location
Japan
Thanks. It confirms what I had thought. Yes, the 120x are almost double in price  :eek:

Saw the thread by Heikki on transformerless vari-mu. In his second one he uses INA(2)137.  Also interesting. Although I don't expect much of a difference btw.  2xLine receiver to get balanced (possibly at +6dB) compared to receiver//driver combo at+6dB or even a brutal  +12dB.... although the more I think about it, the more I like +12dB possibility. Would require another -6dB atten. before output.

P.S.
Will also try harder to understand the choke circuit proposed in that thread  ;) Not sure I'm up to it though.
 

Newmarket

Well-known member
Joined
Oct 10, 2016
Messages
578
Location
Brighton Sussex UK
abbey road d enfer said:
Quite. I'm using a mix of transformer and electronically balanced circuits myself atm to go from my unbalanced  ADC box outputs to 'balanced' mixer inputs and enjoying it but can't say that transformers are superior in technical terms.
Apart from that why the OP 'precison' wrt 6dB gain ?
Given that Gain of x2 (Signal Level) is actually 6.0206 dB... but we essentially use 6dB as a 'shorthand'.
 

Script

Well-known member
Joined
Nov 26, 2008
Messages
1,315
Location
Japan
Thanks for pointing that out. An excess of 0.0206dB wouldn't make much of a difference here. I read somewhere that the smallest difference some people, not all, can possibly hear is 0.1dB). Also, the input and output pots of the unit here don't have such fine resolution. Although the unity gain trimmers before the NE5532 balanced output line drivers do -- for overall unity gain calibration.

With 'precision' I was referring to CMR behaviour of balanced line receivers, which looks better for the THAT120x at high frequencies, at least on paper, as well as to the onboard laser-trimmed resistors for gain staging without having to match resistors.

I want to keep the input transformer precisely for its technical imperfection, cos I take it to impart some 'colour' or whatever ;) Also I understand have to keep it cos there'is CV via centretap to grid (although I find Heikki's transformerless approach interesting, but not willing to do major redesign on this one here. Maybe next time).

In comparison, my aim is rather modest. Instead of having to crank up the output volume on the unit in front of this one, I'm looking for in-unit extra gain in order to (1) drive this comp harder, (2) gain a bit on in-unit S/N ratio, and (3) be able to finally hardbypass the unit for easy A/B comparison without having to also bypass or gain-adjust whatever ihappens to be patched in for extra gain before it.

(For low[er/est] cost, THAT1250 might be an option, when sacrificing on CMRR. Well, even a simple instrumentation amplifier might do ithe job. But this is DIY so luckily I do not have to consider production cost.)
 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
12,347
Location
Marcelland
Script said:
(For low[er/est] cost, THAT1250 might be an option, when sacrificing on CMRR. Well, even a simple instrumentation amplifier might do ithe job. But this is DIY so luckily I do not have to consider production cost.)
Actually, 50dB CMRR is pretty good. In a semi-dicrets opamp diff amp, using 1% resistors of same value from the same batch would typically result in 40dB (with possible deviation down to 23dB).
In a typical studio environment, where ground loops are the main source of problems, 40dB CMRR is plenty.
 

Heikki

Well-known member
Joined
Jul 31, 2008
Messages
226
Location
Finland
Something came to my mind when reading this thread. Can anyone tell me why this circuit wouldn't work as a replacement for input transformer in vari-mu compressors (see attachment). Only limitation I see is the amount of control voltage but it should be plenty for many designs.
 

Attachments

  • trr.pdf
    15.4 KB · Views: 30

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
12,347
Location
Marcelland
Heikki said:
Something came to my mind when reading this thread. Can anyone tell me why this circuit wouldn't work as a replacement for input transformer in vari-mu compressors (see attachment). Only limitation I see is the amount of control voltage but it should be plenty for many designs.
No reason why it wouldn't work, as long as the CV stays within CM limits.
You could probably power this with a single negative rail, that would increase the admissible CV range.
 

Latest posts

Top