Best toroidal power transformers for audio?

Hi everyone,

I'd like to start a discussion about toroidal power transformers for use with sensitive audio projects. In my limited experience I've noticed that not all toroids are created equal. Although they are generally better than EI types when it comes to stray flux, it seems some brands do a better job than others.

What brands do you prefer using for your projects?
Are there manufacturers that specialize in power toroids for audio circuits?
Any tips or tricks for minimizing magnetic flux?

Thanks in advance for your comments!

There is a thread somewhere around here talking about the orientation of the toroid (mainly the orientation of the secondaries IIRC) having a direct impact on induced noise to the rest of the circuit.  That's all this noob has to contribute  8)

Toroidal xfmrs have to be orientated for best noise performance. That's just a fact of life. That's why PCB mount toroidals are not good for audio.
Now the main problem with off-the-shelf pwr toroidals is that they are designed for maximum power rating in a specific form factor, i.e. they are operating near maximum flux, which means a lot of stray field.
Most manufacturers use custom-made mains xfmrs, spec'd with lower flux. Operating a xfmr at 30% less flux than an off-the-shelf makes it 10 to 20dB quieter.
This is not recent practice. Leo Fender used to do that from the beginning.
Distance is also a good and cheap ingredient to quiet a xfmr.
Now you can try using floobydust to quiet a standard toroid, like those silicon-steel ribbons that you wrap the xfmr in...
And if I had to choose an off-the-shelf, I would look at the one with the lowest rating in the biggest form factor.

Abbey, thanks for the excellent info!

Toroidal xfmrs have to be orientated for best noise performance.

Click to expand...

A basic question... Do toroidal transformers tend to radiate more flux from the top and bottom (the flat sides) compared to the circumference?

...like those silicon-steel ribbons that you wrap the xfmr in.

Click to expand...

For what it's worth, I recently opened up a Vintech 1272 unit and saw one of these "ribbons" in use around the power xfmr. Looked like a custom model - taps for 48V and 24V on the secondary. (Toriod Tech was the brand)

I would look at the one with the lowest rating in the biggest form factor.

Click to expand...

Do manufacturers of off-the-shelf units list ratings for stray flux? Or, is there another rating I should be looking for?

So far, I've tried Avel Lindberg and Triad off-the-shelf models. It seems to me that the Avel Lindberg models have less stray flux than the Triads. This prompted me to ask the question - are there better off-the-shelf units to be using? Maybe someone on the forum has a favorite for low stray flux?

for all my projects I use custom made transformers. well, i have near by three manufacturers, so they can make for me transformer according to specification.

earthsled said:

Abbey, thanks for the excellent info!

Toroidal xfmrs have to be orientated for best noise performance.

Click to expand...

A basic question... Do toroidal transformers tend to radiate more flux from the top and bottom (the flat sides) compared to the circumference?

...like those silicon-steel ribbons that you wrap the xfmr in.

Click to expand...

For what it's worth, I recently opened up a Vintech 1272 unit and saw one of these "ribbons" in use around the power xfmr. Looked like a custom model - taps for 48V and 24V on the secondary. (Toriod Tech was the brand)

I would look at the one with the lowest rating in the biggest form factor.

Click to expand...

Do manufacturers of off-the-shelf units list ratings for stray flux? Or, is there another rating I should be looking for?

So far, I've tried Avel Lindberg and Triad off-the-shelf models. It seems to me that the Avel Lindberg models have less stray flux than the Traids. This prompted me to ask the question - are there better off-the-shelf units to be using? Maybe someone on the forum has a favorite for low stray flux?

Click to expand...

I'm in the middle of this myself with my micpre. I find myself needing about 24" distance from the power toroid to the (LL7901/7902) shielded mic input transformer, and orient the PT in a certain way, to make the PS excess noise completely disappear below my -125dB EIN thermal noise baseline. The toroids I'm currently playing with are Hammond and Amveco. The stray field radiation is greatest in the direction around the toroid's circumfrence, and the coupling is least when the hole of the donut points toward the mic xfmr; i.e. you can look thru the hole and see the mic transformer. I can't stress enough the importance of playing with the parts on the bench; don't take my word for what's the best orientation.

I've got an Antek on the way (inexpensive) and some "floobydust" shielding tape, sheet, etc. to see how effective that is. I think it will be necessary to completely enclose the toroid with the shield. I need about 20 dB attenuation above and beyond the best distance and orientation I can get in the unit (total excess PS noise is some 4-5 dB currently relative to EIN). Separate PS is not an option.

Stray field emission could be due to uneven winding. The Plitron units I have seen look like they take a lot of care with uniform winding, in contrast to the scatter-wound Hammond. I will probably look into a custom to get all my secondaries on one core and get the flux density down as well. Side benefit of less heat buildup...

Cheers,

Michael

From a website for Nuvotem Talema, and Irish toroid manufacturer...



I think MJK is onto something with his "uniform winding" observation.

The best toroid for audio is apparently one with a shield, preferably mu-metal. That would be good news for me. I'll be able to confirm this with some of the commercial shielding materials in the next few days. If the diagram is anything like to scale, it implies that the shielding is far more effective than the winding.

At the moment we are just in the process of creating a new Carnhill power transformer specifically for a new project.... seems like a Mu-Metal screening can is prohibatively expensive (GBP 37.00 per can in 100 off !).... so I think we will be going for a GOSS band on the transformer / mu-metal band on the interior circumference of a steel can.... we don't want it to be more than GBP 35.00 in total

The AMEK 9098 rackmount mic amps used to have a simple GOSS band on the exterior of the transformer.... so I expect that we will easily exceed that performance.

The new transformer will be approximately 27vac @ 0.5A and 50vac @ 0.023A.... so you can run a pair of 24vDC mic pres with phantom power in a 1U rack box..... that is the plan anyway.

Colin
www.audiomaintenance.com

Colin, thanks for your post. The transformer you are describing sounds like a good option for those of us who can't justify having our own custom units produced.

From what I understand, 27vac would be slightly high for the average LM317-based PSU set at 24vDC. But, the tradeoff of adding larger heat sinks may be well worth the decrease in induced hum.

-----------------------

Does anyone have a good source for silicon steel band material? Can it be purchased at a hardware store?

I use silicon steel material from broken toroidal power transformer core

Colin, when do you think Audio Maintenance will have the toroid mains transformers available?

earthsled said:

when do you think Audio Maintenance will have the toroid mains transformers available?

Click to expand...

Maybe the end of January ?

Colin
www.audiomaintenance.com

How are the triads compared to amveco and where would you find the leakage flux specs?  in the data sheet? 

Also where does one buy these "ribbons" for the transformers? and what is "floobydust"? 

I just get ones that are WAY overspec for the project. The past two times I spent 8 dollars less for just good enough I had trouble. May be the hand grenade way to do it. But everything is silent since. 50 VA for a single 1176? Yep. Don't cost hardly any more. But what I know I just stick with what has been working for me.

I did put some MU metal in my G-9 tho to get that last few DB's. So not against that kinda thing. And yeah you gotta crank the gain and spin them. They are round. Leave the lines enough room to situate them correctly.

John

migidee said:

and what is "floobydust"? 

Click to expand...

a.k.a. fairy dust. it's that magic powder fairies sprinkle and makes things work or humans heal and hearts be unbroken...

abbey road d enfer said:

Now you can try using floobydust to quiet a standard toroid, like those silicon-steel ribbons that you wrap the xfmr in...

Click to expand...

silicon-steel? Never heard of that. Do you mean grain-oriented steel? That stuff has a real and easily measurable effect, the same sheet good transformer manufacturers use themselves. I've had two projects already that had PSU toroidal transformers oriented to perfection, but still there was a 50hz magnetic hum peak above the general noise floor (shown by a very accurate FFT tool). It was some 7-10dB above the noise floor and a grain-oriented steel sheet wrap around the trafo completely eradicated it. These were both high gain tube preamps and the advantage is very real for ribbon mic usage, for example.

Dont want to hurt someones feelings but Behringer use transformers wrapped in these kind of sheets! 

I just get ones that are WAY overspec for the project. The past two times I spent 8 dollars less for just good enough I had trouble. May be the hand grenade way to do it. But everything is silent since. 50 VA for a single 1176? Yep. Don't cost hardly any more. But what I know I just stick with what has been working for me.

Click to expand...

Interesting...
How much do you usually overspec the transformer? If I have a project that uses a 25VA transformer normally, would you recommend a 50VA or something like a 100VA?

Thanks!

GOSS = Grain Oriented Silicon Steel. Apparently most transformer steel contains the element Si... something to do with increasing the electrical resistance while not affecting the magnetic properties.

I obtained some by unwinding the core of an old toroidal power transformer with open windings. Actually it was a new Hammond right out of the box with an open primary 

Magnetic shielding floobymaterial is also available at http://lessemf.com in small quantities

I spec power transformers for 2X-3X the load VA if using capacitor input rectifier filters. For 15-25VA load I would use a 50VA transformer.

Still awaiting the results of my own toroid shielding experiments. Another possibility is to reduce Bmax and potentially reduce radiated EMF through using a buck winding on the primary; drop the voltage down from 120V to ~110V.

Cheers,

Michael