OEP Preamplifier transformer, noise and Mumetal grade.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

miquel

Active member
Joined
Nov 6, 2009
Messages
39
Location
Barcelona
I have not found information about this subject in the forum.

I am working with a DIY BA283 NEVE with OEP A262 A3E input transformer, of course with the OEP electromagnetic shield and connect it to ground.

The circuit works very fine, but the preamplifier captures proximity transformers electromagnetic fields (even at 30 or 40 cm….!), and inclusive noise from near AC lines.

With 70dB gain I measure the low frequency noise at preamplifier output…:

-  Far from AC lines and transformers = 70 uV or less, (is better than 90 dB signal noise at +10dB reference, and better than 100 dB signal noise at +20 dB reference). It’s possible to improve this figure varying the transformer position.

-  At 10cm near an AC line (more or less) = 400 uV output or more , (it’s worst than 75 dB signal noise).

The noise capture is evident connecting a high gain amplifier to the output, and varying the preamplifier PCB position in reference a noise sources. (In a work bench there are a lot of possible noise sources…).

I have experimented with the OEP shield and magnets and do not know if that Mumetal is genuine or is another alloy.

1.- Does anyone have experience with these transformers and electromagnetic shielding troubles?

2.- Do you have noise comparatives with Lundahl, Sowter, Cinemag, Jensen or an others?, More expensive of course…

Thanks

Miquel from Spain
 
I've been reading about magnetic shield. I have seen many Jensen guarantees 30dB of shielding, but the JT-115K-E60 has a double shielding and guarantees... 60dB...!. (Expensive of course).

If Jensen make double shielded transformers is because it is necessary in some environments, if not, I think they don’t fabricate it….!

And OEP magnetic shield…?, it’s sufficient for high gain and professional environment?

Has anyone any experience?

Miquel

 
I've used OEP microphone input transformers in the past and also found that they are VERY sensitive to external magnetic fields, of cource with the screening can. I have experience with Sennheiser microphone input transformers and I found that those are less sensitive for external fields. Most transformer manufacturers use several layers of screening material; OEP uses only one layer.
 
As well as the screening can, the design of the transformer has a big effect on how sensitive they are to hum. Some manufacturers use multiple windings in a way that helps cancel hum, which is comparable to guitar humbuckers.

Empirically, I've found that the Lundahls are better than the OEPs in this respect. The OEP screening can seems to help more with electrostatic hum than electromagnetic.
 
Used many OEP's plus can,in many projects and never had any problems with noise or electromagnetic interference.........or is that just me being lucky ?
Maybe the problem is elsewhere ?
 
I used an OEP in my first 1176 WITHOUT screens, never had an issue with noise... But I never turn the output up past half...
 
Thanks for the answers

I can be wrong, really, but I'm sure the problem is the shield, because moving the transformer (well, the PCB that contains the transformer), toward or away from possible interference sources, the hum vary (it’s like a funny game….)

In really, I am measuring output preamplifier noise at worst case, ergo at maximum preamplifier gain  with a high sensitivity amplifier, but is the common studio test, not the normal operation of course.

For the measurement I use a really low noise 100x preamplifier (without transformers jejeje) which I use only for this application (the power supply is batteries to avoid ground loops) and I can measure uV easily, but… connecting the preamplifier output to a power amplifier with a high gain, the hummmmm is clearly audible. (I have a special battery powered amplifier that it provides 1W output with only 20mV balanced input at maximum gain, very interesting tool to investigate noise).

Noise sources examples:

A big not toroidal transformer (from a bench DC power supply) has a measurable effect at about 30 cm away.

An AC power line (just under the bench wood, my God…) has a measurable effect at 5 or 10 cm.

Really, you can say…: if you haven’t mains transformers near, or no AC near, no problem, but I think the transformer electromagnetic shield should be better.

Finally I think if Jensen makes transformers for special situations (expensive of course), I think  that situation exists for all manufacturers.

I confess that maybe I'm too strict

Saludos cordials desde España,
 
I've been working on the issue of noise and electromagnetic shielding in input microphone transformers. To observe the induced magnetic field, I have provided a small transformer that I use as a source of magnetic interference at 50 Hz.

In addition to test referred to OEP, I have also tested a Lundahl LL1578 that I had in my workbench. This transformer is not as the original 1:2 (NEVE) or 1:3,2 (OEP) preamplifier, but it can connect at 1:5 and can be used to make measurements.

Hurra…!, I have solved the interference by electrostatic fields from AC lines that circulate below the workbench. It has been easier to dispose the PCB on a metal plate. One problem solved…!. Now we are going to measure the electromagnetic interference from a little noise source transformer.

I measured the noise at the preamplifier output with 70dB gain (with small error that the transformer OEP is 1 : 3,2 and Lundahl is 1 : 5, but I am not considered the difference between one and another). The reading has been done with 10 cm between the interference transformer and shielded input transformer, and also different bandwidth, namely: 0 to 300 Hz and 0 to 20.000 Hz.

I've not make readings with an rms true AC millivoltmeter, only with an analog oscilloscope and quality digital milivoltimeter, this is because my old computer and oscilloscope screen is a big source of noise that I avoid to make accurate measurements. Of course there may be rms errors…!

I've done the measurements through battery powered measure amplifier that I have described. I have been used an old analogical oscilloscope with floating ground.

Readings:

- Base noise (the input transformer is far from any interference sources), remember is the noise measure at output preamplifier:

OEP 0-300 Hz bandwidth = 100 uV noise = -78 dBu
OEP 0-20.000 bandwidth = 292 uV noise = -68 dBu

Lundahl 0-300 Hz bandwidth = 62uV noise = -82 dBu
Lundahl 0-20.000 Hz bandwidth = 265 uV noise = -69 dBu

To calculate the EIN we just have to subtract 70dB, but I think there is a problem because the EIN is too low….!. I think is probably rms error reading.

- Interference noise (from the little transformer):

Note:  I have used only the bandwidth 0-300 Hz because I need note the low frequencies (50 Hz hummm and harmonics).
70 dB gain.
10 cm separation between interference source and audio transformer.

OEP =  9100 uV noise= -38 dBu

Lundahl  =  480 uV noise = -64 dBu

Conclusions

1.- Base noise is very similar in both transformers

2.-  It is clear that the OEP electromagnetic shielding is not right, at least for high gain amplifiers. It's curious because it is apparently much thicker than the Lundahl transformers, (Lundahl has 2 or 3 very fine layers.)

3.- Llundahl is, at least, 26dB better to 50Hz electromagnetic noise influence.

(My God I can't attach an image....)

Saludos

Miquel
 
Multiple shield layers are more effective than a single thicker layer.  I recall the typical single shield gets you 30 dB, each successive shield another 30 dB.    Local orientation gets you even more. 

The best in the old days used a combination of hum bucking winding and 3 shields, for an advertised 90 dB hum reduction.  The most circulated image regarding shielding comes from Kenyon Transformer Co., with their 'nested shield' approach.  This was much copied, and can be seen in any edition of the Audio Cyclopedia.  Of course, not a single reproduction of it to be found via Google!
 
I try to attach pictures (which had only 300 or 400 Kb) and I can't, I have now downsized the jpeg to about 200 Kb, but I still do not understand what is happening. I think are interesting images. :-[
 
RuudNL said:
use OEP transformers for anything except.... microphone inputs!
Sorry to disagree there but,have eight  API type pre's and no problems with noise or hum even at high gains !
So ?????
 
Lucky you! Maybe you are using an external power supply at a 'safe' distance?
But I have experienced the same problems as Miquel; that is why I am using Sennheiser input transformers now.
I don't have serious doubts concerning the audio quality of the OEPs, but for me they are too hum-sensitive.
 
so, here's a question... could a better mu shield be fashioned using multiple layer or boxes each made of different material; aluminum, steel, copper, nickle plating? I'm sure that this was necessary at some point in history before a proper mu metal alloy was perfected and affordable.
 
I'm using OEP A262 A3E transformer in NEVE BA283, for the advice from JLM Audio. Both primaries are in series and two secondaries in parallel, thus achieving 1: 3.2 ratio. JLM Audio don’t recommend the OEP A262 A2E for overshoot and ringing tests, (also be remind it's more suitable for its original 2:1 ratio), remember with this transformer (A2E), you can also choose from two connexion possibilities: 1.- primary and secondary in series or 2.- primary and secondary in parallel, of course both 1:2 ratio.

I am not expert but I think there must be some inductance windings influence in noise susceptibilities, beyond the simple transformation ratio.

As is known, transformer manufacturers tend not to indicate the inductance in H, as they are permeability core dependent, and permeability depends to the frequency and signal intensity. (I have spoken with Sowter and told me that it’s better express the bandwidth of a given impedance in each winding and a specific signal, more than the simple inductance).

There are manufacturers that specify the inductance but do not tell at what frequency and what signal level are measured. On the contrary, OEP effectively specifies in datasheets inductance, which for A262 A3E is the next:

Frequency test = 1 kHz
Signal test = 0.27 V

Inductance:
Primary = 125 mH each
Side = 5.2 H each

My question is now:

Is it possible that sensitivity to external electromagnetic fields, (besides the quality of the shield), is related also with the windings turns number (or it’s inductance). Therefore, different transformers (for example with the same 1:2 ratio) has different behaviour to external fields.

I am sure that OEP is not bad transformer. I think is not expensive because is not hand made, but mains transformers are not made by hand….

 
For some reason I can not upload photos to the forum, for that I have loaded it in Photobucket.

I'm not saying that OEP are bad transformers, just has to be away from power transformers, especially if they are not toroidal transformers


OEP test

http://i749.photobucket.com/albums/xx138/miquel_bucket/DSC00956.jpg


Lundahl test

http://i749.photobucket.com/albums/xx138/miquel_bucket/DSC00958.jpg


Noise preamplifier for noise measurements

http://i749.photobucket.com/albums/xx138/miquel_bucket/DSC00959.jpg


Detail Lundahl detail multilayer shield

http://i749.photobucket.com/albums/xx138/miquel_bucket/DSC00960b.jpg

Miquel


 

Latest posts

Back
Top