31267 trafo in 1272 clone problems..

some who have been following the Chimera PCB buildups(in the Blackmarket) by myself and a few others will recognize the problem I am having..

I am running the quad preamp with 2 Lundahl LL1538xls and 2 31267s on inputs. the channels with the lundahl trafos are clean, quiet and very nice.. the channels with the 31267s are terribly noisy(hissy) and have large amounts of 60 and 120hz hum. Hooking any mic up to it actually makes the noise worse by another 20% or so even though audio can clearly be heard through the hum and hiss.

I wired the 31267s backwards as JLMAudio proposed on the website. I tracked the issue down to the input trafos and removed the + output of the trafo to the board. the channels are now super quiet like the others so it's clear that something is not working together here.

I noticed that my Lundahls have a static resistance of ~120R inputs and ~800R output. the 31267 has ~60R input and ~150R output..

I grounded the case of the 31267 which actually made things worse for the time being. I then fooled around with different wirings and found that series wiring for both sides is the quietest, almost identical to the other channels. removing the grounding now makes things worse.. go figure..

I assume the impedence is incorrect somehow and that some kind of termination/zobel is needed to use the trafo as a direct replacement for a 10468?

any ideas?

thanks! :thumb:

(edit) I also haven't passed audio through the channels that have been rewired since I changed them..

So lets make sure were on the same page:

You are using the 31267=T-1452=Line Input in reverse as a mic input.
Green Marinair?

Who told you that you can get away with that?

First Step:

Eliminate the channel strip circuit as the culprit. Splice one of the Lundahs, (I know, pain in the rear) into one of the offending "Marinair" channels and see if it hums.

Also, your 31267 should have 60 output and 160 input. Seems to me you have the numbers backwards.

yes, yup, yeah and:

http://www.jlmaudio.com/Neve%20transformer%20info.htm

I've already moved one of the Lundahls over to the other channels and it works fine there. It's obviously the trafo unfortunately

:thumb:

OK, here is what I would do next: Splice in the Green Meanie into the circuit as a line input. That is, run Pri as input (leads 2-5 if I remember correct), and Sec into the circuit (7-10, I don't care how you wire, Series or Prallel)

Then run a line level signal into it (CD player, etc) and see if it still hums.

You were grounding Pin 6 also I assume? That is connected to three electrostaic shields, one between Pri and Sec of each of the humbucking coils, and the third which is a foil wrapper around the whole enchillada.

ok will do, however I've found that wiring in series(it's still backwards though) it doesn't hum and the whitenoise level seems identical to the Lundahls.

thanks Master CJ!

Actually, I do not even own a Neve, ...yet.
But I am familiar with the transformers! :razz:

The 31267 has the sec hooked up in paralell in the schematic that I have.

The mic in has the sec wired in series, but there's that 300/1200 impedance sw on the primary.

I think you should try grounding the windings that are not being used. Just one of the pins from each unused windings.

thanks guys, I'll be trying all of those suggestions!

:guinness: :guinness:

Here is a pic of the line in just incase you don't have it:




I do not think your problem is the difference in impedance between the Marinair line in and the Marinair mic in.

I think it has more to do that when you run the line in backwards, the windings that the mic are connected to are on the outside of the transformer, instead of being protected by the secondary and the extra E shield.

ok, wired the trafo up like normal. noise(abnormal white noise that is..) is gone, but 60 and 120 hum is still present. it's less but still very noticeable and still much more than the lundahl inputs. I think i'll be trying various wirings over the weekend but I'm worried about killing transient and frequency response..

thanks and I'll keep you updated!

ok another update:

the quietest wiring yet is exactly like the picture that CJ posted but backwards! only 60hz hum can be faintly heard.. actually quieter than the lundahls!

will have to try this when i get a chance!

Are you using that cap?

One thing I noticed is the package difference between the Black Marinair T-1454 and the Green Marinair 31267. The black model features what is definately a mu can, where as the green model has removable can halves that may or may not be mu metal. They certainly do not have that shiny mu-metal appearance. Marinair or Neve may have decided to save a few penies on the green models. I know the Lundahl's are packaged in mu cans.

You might try shielded cable on your mic xfmr wiring.

The location of your pwr supply components might be different than the original Neve boards.

When wiring the 31267 backwards, the windings that the mic is hooked to, as mentioned earlier, are on the outside. Grounding the case may be allowing your chassis hum to be electrically connected to the green transformer case, there-by inducing chassis hum into the outer mic windings.

Do you have your hot and cold polarities right? The little dots are hot.

You might try phasing your pwr transformer.
Be careful with this:

1) Unhook the Earth Ground wire from the power chord, if you are using one.
2) Connect a Hi-Z AC voltmeter between Earth Ground and your Neve chassis.
3) Connect an AC test cord to the pwr transformer. Plug in the line cord and note the voltmeter reading.
4) Reverse the polarity of the test chord and repeat step 3.
5) One of the two polarities will provide a lower voltage reading. This is the preferred polarity for that transformer. Mark the primary leads of that transformer accordingly.

The benifits obtained by this "primary polarization" are twofold:

1) Improved reliability and saftey.
2) Improved sound may also result.

You might want to play around with pin 6 of the 31267. This pin is connected to the e-shields. Try grounding and un grounding it.

thanks CJ, the cases are brass.. pin 6 has been tested grounded and ungrounded, neither of which make a difference. I have tried both reverse and normal polarity on the windings with no real difference in hum. I have 100pf caps on the output windings.. I'll up that to 1nf and see what happens. I'll be trying your phasing technique too.

thanks Trafo Master CJ!

Both the Lundahl and 31267 are wound on O type cores with one primary and secondary on each bobbin on opposite side of the core. Both bobbins mechanically speaking are sitting out of phase. So if you use both winding they will hum buck and cancel the hum.

Also CJ is right in that the wrong windings are now above the pin 6 screen windings as well. And yes the 31267 has no Mu shield. The hiss should only be tested with a 150ohm load resistor fitted to the input XLR pin 2 & 3. Because the hiss could just be that the 31267 without load reflects a higher impedance at high frequencies so the hiss is louder but only without a input connected to the mic pre.

Joe

www.jlmaudio.com

Thanks Joe, however the hum(the hiss is fine) is actually worse when terminated with a mic or 150R. I'm kind of puzzled as to the cause!

So lets say that when i have it wired like the original but turned around backwards, the hum is almost gone but when any termination is added the 60 hz hum becomes terrible and introduces 120hz hum.





:?:

ok here is an update. 31267 wired in series both sides, turned backwards. it is quiet and sounds great. I think i'll be leaving it like this.

thanks to all that helped!

:guinness: :guinness: :guinness:

Did you ended using a cap shunting the seconary, like in the original circuits?

yeah there was/is a 1000pf shunting. it did the exact same thing with or without it, changine the size up/down didn't help either.. :wink:

who knows.. it works nice now. i might pull the trafos out and put more lundahls in to make all 4 channels the same and use the 31267s in two more 1272s i'm planning.

:thumb:

I had been wondering about the disadvantages of only using one winding per side. So I take it these results mean it's not a good idea afterall? In that case, thanks all for confirming suspicions! I have also wondered, wouldn't using only half the windings cut its current handling capabilities in half? If so, then how can it posslby have the same response as a 10468 in this configuation like many people say :?. Makes no sense to me. If I am wrong someone please explain why. I hope to buy some 10368 (can acheive 600:5k using all windings, but pri and sec reversed. ~5k means same loading as intended for the 10468 with BA283) soon and try all the same permutations. 31267 are all over the place but I don't want to use them for the reasons described here, and the 10468 is just too hard to find.

It depends on the x-former. If you have a humbucker, that is, a DU core for instance, with two opposing coils instead of one coil on say. an EI core, you should really be using both coils to extract the maximum benifits of the design as far as hum and core balance.

Current is not an issue with line and mic in x-formers.