Switching Marinair 31267 transformer from mic to line

[API]

I am rebuilding a few Neve 1271 lina amps to the 1272/1290 type but was planning to keep the 31267 transformer and use it in reverse mode as a mic in transformer.
But i will also need a line in on these modules and was starting to wondering if it would be possible to switch sides of the transformer via relays to use it both in "normal" and "reverse" mode.
I drew out a small sketch and i think it could be done but one thing that i was wondering about would be the caps that also would be needed to be switched in and out of each side to get correct loading of the transformers.

I know that most likely i would need a big relay and maybe it will not be economical viable but i wanted to explore the possibility and was also wondering if anyone else has done something similar?

 

[radiance]

Check the info down at the bottom of the page...
http://www.jlmaudio.com/Neve%20transformer%20info.htm

 

[API]

Yeah, i have seen that before but i dont realy like the solution.
Would like to switch the transformer from one side to the other, will try to draw something up.

 

[0dbfs]

Or the opposite. Reversing a 10468 style (ie; carnhill, hairball, etc...) with relays or switch for mic/line operation.

Also want to note that the 1063's (?) used 10468's for mic and line in with a pad network to accommodate the line-in path. Without referring to the schematic, I believe that pad was between the IPTX and the first gain-stage so the addition of the 31267 to the channel in the next-generation 1073's was perhaps an improvement to extend headroom in the input stage.

I don't see why re-configuring them with relays (or even jumpers) and changing the zoebel accordingly wouldn't work well.

Cheers,
jb

 

[Baltimore]

I believe Hairball's Lola mic pre does this with some relays.

 

[desol]

I have some 1290's with 10468's and 31267's backwards. The 10468's are the octal versions(round).
The transformers sound nearly identical...with the 10468's being slightly clearer in the upper mid's.

My backwards 31267's sound the same as the 10468's but pickup far more hum. You gotta do all the extra s*** to help them be quiet.

It's like Joe Malone said...the humbucking/noise cancellation feature that's normally in use with the line input...isn't in use when you use them backwards with one coil...
so they're very susceptible to noise. The 10468's are dead quiet. I've done all kinds of things with the 31267's to quiet them down...hooked all kinds of grounds to them,
wrapped them in mu foil, etc, etc and they're a bit better.

You have to hook them with just one coil/backwards as mic..otherwise the frequency response isn't the same. I've read the same method from a few different sources...
Anything's possible with switching...but they'll be noisy when used as mic.

 

[CJ]

i believe that the line in has different metal in the core,

if you flip a 31267/T-1452 you will have about 16 Henries of primary inductance instead of 32 Henries,

if you flip a 31267 and use only 1 coil, you will have about 4 Henries instead of 32 H,

and you lose the hum bucking,

also, the primaries on the mic and line inputs are wound as to have the Finish leads of the coil for pins 2 and 5, that is, if you wire the primaries in series, you will have finish leads for you hookup,

if you flip a 31267 you will have the starts for your hookup, this might lead to even more problems with noise, but i have not tested this theory on the bench,

now not all Neve inputs are wound the same, some use reverse wound secondaries on both Mic and Line inputs, some do not, so your mileage may vary as far as noise and flipping the line into a mic input,

FYI: the T-1454 mic input has 998 turns for series connected primaries for about 32 Henries, the T-1452 line input has 1350 turns for series connected secondaries for about 16 Henries, we are using pri and sec to more easily compare inductance,

so we have 998 T = 32 H and 1350 T = 16 H, so less turns on the mic in gives more inductance than more turns on the line in,

so the perm of the mic input core is higher than the line input, both use 26 lams ea. of LH-2023 DU,

more data here>

http://www.groupdiy.com/index.php?topic=52339.msg667198#msg667198

 

[desol]

Interesting stuff Chris....

So, is the higher inductance of the 10468 the reason i'm hearing better HF?

 

[rafafredd]

You could easily do it with a single 4PDT. I have drawn the schem for doing this sometime ago. I just can't find it. But it is not that difficult to figure it out. You could use a lorlin or other 4P3T rotary labbelled with mic/line/DI options in the same switch.

 

[API]

To busy to have been taking this any further but good to hear that you have already drawn soemthing out Rafared and that it should work.

I am slightly worried though by CJ´s findings, everything i have read about it so far the 31267 should sound identical.
I am planning to use these in a console and do not want to worry about extra noise.
Have anyone had problem with excessive noise when using these backwards in real life?

 

[CJ]

just ry it and see what happens,

or just buy another Neve input,
for the price and hassle of wiring up a relay and switch you just add a few bucks and get another xfmr,

how much does Carnhill charge for those guys anyway?

 

[desol]

Have anyone had problem with excessive noise when using these backwards in real life?

I'm pretty sure my life is real...  And yes...mine pickup more noise compared to the 10468's beside them.
I have to watch what i place in a rack above or below them.

 

[API]

Have anyone had problem with excessive noise when using these backwards in real life?

I'm pretty sure my life is real...  And yes...mine pickup more noise compared to the 10468's beside them.
I have to watch what i place in a rack above or below them.

I will indeed try them and see what happens in my setup and if the noise is worse than with my 10468 loaded preamps.

The thought of using relays was more for the fun of it and to try and solve a problem and build something.
Would want to use original Marinairs since i have them and not buy new ones for the vintage Neve amps i am rebuilding.

Desol, what i meant was in a "real life situation" or if it was measured noise only.
So was not trying to say something bad.

 

[radardoug]

You can indeed use the 31267 backwards as a mike transformer, but it is harder to get flat frequency response. However you end up with a bit of a peak at 20 kHz, sounds nice, gives it some air. You could play around with the Zobel network more than I did, but they worked fine for me.
You need to use both windings on each side, in other words the two secondaries as primary, the two primaries as secondary. If you don't you won't get good noise cancellation.

 

[Rob Flinn]

If you're getting hum from your 31267 wired backwards can I draw your attention to this post from several years back.  Look at the last few posts

http://www.groupdiy.com/index.php?topic=7745.msg92898#msg92898

 

[API]

Have been messing a bit with the 31267 transformer now hooking it up backwards but i am not getting it to work.
I hook it up backwards using only one coil on each side but i can not get it to be stable in one rebuilt 1271 lineamp.
In another i got it stable (this one had a zobel on the output coil, the other did not) but had about a 40db loss compared to a 10468 in the exact same 1271 module.

Do i need to terminate the other coils in any way to get it to work properly?

 

[CJ]

make sure you ground the shield terminal,

 

[API]

Do you mean pin 6 on the transformer?
That is tied to chassis which should go to the 0v at the powersupply but will doublecheck that.

Measured the coils on the transformers and one side they were 64 ohms and the other about 170 ohms (each coil).
Sounds about right?

 

[CJ]

that sounds about right,

try switching coils on either the pri or sec,

 

[API]

Will do. Thanks