ESL transformer.

In my very first ESL built in 2008 (a hybrid with 12 x 48 perf metal stators) I used the Just Real Music M299 transformers. I'm a bit treble deaf so it could be my hearing-- but to me those M299's just sounded anemic on the treble end.

After reading one of Calvin's posts about cheap toroids, I replaced the M299's with a tandem of inexpensive Farnell 50VA 230V/2x6V power toroids and I was amazed at the improvement... the highs were pristine and I drove them with ridiculous power and couldn't hurt them. I've evolved through several ESL panels since then but I'm still using those very same Farnell toroids. Highly recommended for hybrid ESLs!!

I also did a quicky listening test (a few hours) with a tandem of 50VA Antek 2x115V/2x6V power transformers. These sounded just fine too (way better than the M299s) but I believe their dual 115V secondaries have higher capacitance and I don't trust them to be as robust (arc resistant) as the single 230V secondary toroids.


Wow thank you very much!!

I am sold on using toroidals for the audio transformers.
 
Wow thank you very much!!

I am sold on using toroidals for the audio transformers.

I recommend two of these per speaker with 6V input windings in parallel and 230V output windings in series and the ground from the bias supply connected between the 230V windings.

Understand that the tandem pair is suitable for a hybrid speaker only, with the woofer crossing in not lower than about 220Hz with a 24db/octave filter slope.
 
Last edited:
I should be good then, the crossover is ~600Hz :up:

Most new builders cross the woofer in too low in an effort to retain the ESL magic down into the mid bass region, and their speaker suffers from mid bass sickout as a result. But you are wanting to go in the other direction and crossover at 600Hz.... just wondering why.

What are the dimensions of your panel?
 
What about these

TSTA 0040/002 - Mains transformer for tubes - Shop Toroidy.pl

used for tubes so ignore the high voltage secundairy.

but when you chose 2x115 you got a centertap, and since it has 3.15 secund you got a ratio of 230/3.15 = 73 only one would be enough for mid high panels ? since you got a centertap.

not sure how good they perform but they are rated up to 4kv test... not sure if you would paralel the 3.15-0-3.15 i think you chose one >? between 0-3.15 or 3.15-3.15 but then your stepup is halved

price is less then 2 trannies from farnhell

i am gone buy one it hink if they do work ok that would be my ticket to make affordable magnetic planar esl hybrides. i wanted to make for so long
 
Last edited:
Most new builders cross the woofer in too low in an effort to retain the ESL magic down into the mid bass region, and their speaker suffers from mid bass sickout as a result. But you are wanting to go in the other direction and crossover at 600Hz.... just wondering why.

What are the dimensions of your panel?

what also works rather well is take a verry slow slope and cross at 500-600 or so, it is sort of a compromise, and fitting the woofer in with such slow 6dB oct slope is more easy to. and phase is also rather good with these slopes. and you can fill in the mid bass suckout by shifting the slope of the woofer a bit. just something i used on my planars. i thought it sounded rather integrated. correct me if im saying something stupid :) it just worked for me
 
Most new builders cross the woofer in too low in an effort to retain the ESL magic down into the mid bass region, and their speaker suffers from mid bass sickout as a result. But you are wanting to go in the other direction and crossover at 600Hz.... just wondering why.

What are the dimensions of your panel?


There are four panels that are 5"x 5". I am not building from scratch just trying to get some Janszen 1-30's running to be used with KLH woofers. The crossover is a high pass filter which consists of a 20uF cap and 1.5mH inductor. My math may be off but it looks like the cut off is actually higher than I thought at 918Hz.

One unit works the other doesn't. The unit that works uses a voltage multiplier with a 1:4 power transformer for the bias. The unit that keeps blowing audio transformers uses a half wave rectifier for the bias, the power transformer is a 1:8 ratio. I have been thinking of changing the bias supply in the unit that keeps blowing audio transformers to voltage multiplier because it's the only thing I can think of that is destroying the audio transformer. Maybe during power off the higher voltage transformer is throwing much larger transient spike?
 
Hi Guys,
Off late I've been mucking around with my ML Odyssey, the step up trannie
on the primary side has a CT connection, why is this so & what is it's function ?
Yes I have the schematics but I'm still at lost. Would appreciate if someone can
enlighten me on this.

Thanks
 
There are four panels that are 5"x 5". I am not building from scratch just trying to get some Janszen 1-30's running to be used with KLH woofers. The crossover is a high pass filter which consists of a 20uF cap and 1.5mH inductor. My math may be off but it looks like the cut off is actually higher than I thought at 918Hz.

One unit works the other doesn't. The unit that works uses a voltage multiplier with a 1:4 power transformer for the bias. The unit that keeps blowing audio transformers uses a half wave rectifier for the bias, the power transformer is a 1:8 ratio. I have been thinking of changing the bias supply in the unit that keeps blowing audio transformers to voltage multiplier because it's the only thing I can think of that is destroying the audio transformer. Maybe during power off the higher voltage transformer is throwing much larger transient spike?

For panels that small, the high crossover makes sense.
(I imagined you were building a single large panel that could play down to 200Hz before the roll off kicked in significantly)
 
For panels that small, the high crossover makes sense.
(I imagined you were building a single large panel that could play down to 200Hz before the roll off kicked in significantly)

No sir they are tiny panels, basically tweeters that go down lower.




Do you think I can get away with using two 10VA transformers? I roughly calculated the panel capacitance and load and it looks like I don't need much current, ~400mA max.

I would like to try and use these. AN-0105 - 10VA 5V Transformer - AnTek Products Corp


I was having an issue of one unit being louder than the other. I think it's because the half wave bias supply is ~1.3kV while the multiplier unit has 1kV bias. The original audio transformer are 1:50, if I use two 5v transformers it will be 1:46 which might actually help balance out the output.
 
…Do you think I can get away with using two 10VA transformers? …I would like to try and use these. AN-0105 - 10VA 5V Transformer - AnTek Products Corp
Two of those would be a good match for driving the JansZen tweeter panels.

… One unit works the other doesn't. The unit that works uses a voltage multiplier with a 1:4 power transformer for the bias. The unit that keeps blowing audio transformers uses a half wave rectifier for the bias, the power transformer is a 1:8 ratio. I have been thinking of changing the bias supply in the unit that keeps blowing audio transformers to voltage multiplier because it's the only thing I can think of that is destroying the audio transformer. Maybe during power off the higher voltage transformer is throwing much larger transient spike?
Can you provide any more details on this? What fails in the transformer? Primary or Secondary? Shorted or open? Also, did it fail while you were playing music? Or between turning off/on the supply? A pic of the electronics may be helpful, perhaps there is a mistake in the wiring…have not heard of this problem before.
 
… if someone does we might be of help to each other and finally create some affordable ones ? i think i stated this sort of colab earlier in a thread. but it never came to something
If I recall correctly, you were still trying to figure out what your panel capacitance would be and how low you wanted to go.
ER Audio Mini Panels #27 - #30

From golfnut’s post you can see that the transformer design needs to be tailored to the type of ESL you are building. If it is a segmented design the winding configuration for a C-core or EI-core transformer is rather simple. You just need to decide on how low you need it to play. This is the approach Acoustat used in the Spectra 11, and what I have been using. Oh, Capaciti and Audio Exklusiv also use it.
About to take the ESL plunge #102, #105
Driving ESL transformer more efficient on SS-amp
Driving ESL transformer more efficient on SS-amp


What about these
TSTA 0040/002 - Mains transformer for tubes - Shop Toroidy.pl…not sure how good they perform but they are rated up to 4kv test... not sure if you would paralel the 3.15-0-3.15 i think you chose one >? between 0-3.15 or 3.15-3.15 but then your stepup is halved price is less then 2 trannies from farnhell i am gone buy one it hink if they do work ok that would be my ticket to make affordable magnetic planar esl hybrides. i wanted to make for so long
I think you would be better off with two small toroids, a proven solution. Remember the 4kV rating is for isolation between windings. It does not mean that the individual windings can handle that voltage…I doubt they can reliably. Also, the 3.15 Vrm input @ 50Hz would limit you to a hybrid with crossover above 500Hz if you are using an amplifier rated more than 100W into 8 ohm. Usually hybrids are crossed lower than that and/or use more power.
 
Old thread, new question ... about the suitability of an array of 100V line transformers for ESL stepup use. It's strange that these kind of dedicated audio transformers are not mentionned anywhere as an option to buld a stepup transformers array for ESL's. Why that?

My motivation for rising this question is as follows: After the revision of my Quad63's, I have a single panel left. Oh, no, no ... it's not the same story as the famous extra screw/washer left after a terminated amplifier revision ... It's an extra, 9th panel I separatly bought. And now, actually having read some recent posts about experiments with trinaural stereo using a bandwidth-restricted center, I suddenly have a use for this single panel: This by now solitary Quad panel would make a perfect center for frequencies from 400Hz upwards. So now I need, amongst other gear, a decent stepup transformer for this use.

At the same time, an el-cheapo outlet nearby offers a whole surplus box full of tiny, high quality 100V line transformers rated 6W at 4Ohms. These specs makes for a stepup ratio of approx. 1:20. Therefore an array of 6 such transformers would give a decent 1:>100 ratio. As the secondary winding is rated at approx. 5Vac for standard fullband use, a 2*3 serial/parallel array would specify at least for 40Vac input in this passband application above 400Hz, promising an AC HT of some 4kV. Perfect for this application, it seems. But, as always ... wait a moment ... is there a drawback I do not see? I am slightly reluctant about the fact that there will be worst-case up to 4kV between a primary and a seconary winding? So, does such an array seem reasonably safe for the driving amplifier?

Your advice is welcomed!
 
Last edited:
I think your concern about very high voltages between primary and secondary is valid. The insulation materials were probably not designed for so much voltage. In the transformer world they often refer to this feature as “hi-pot”. It might work for a while on short bursts of voltage but over time you could experience arcing inside.
 
Thank you. This caveat/aspect was new to me, and indeed, inter-windings-isolation equals not inter-windings-isolation. I better know now what really happened some time ago when I heard strange noises and perceived a decent smell emanating from a tansformer during my first and innocent DIY-steps along with ESL's.

Meanwhile and thanks to your input I also have learned that for us europeans, "hi-pot" may translate to EN 61 558-2-4 ("Trenntransformator") and even better EN 61 558-2-6 ("Sicherheitstransformator"), both classes with a specified isolation withstanding 5kV. Beware: there are two other specs, EN 61 558-2-1 ("Netztranformator") specified at 1.5kV and EN 61 558-2-2 ("Steuertransformator") specified at 2.5kV. Interestingly, transformers sold for driving mobile indoor 12V-halogen lights are specified as the 5kV-type of "Sicherheitstransformatoren", and also transformers to drive kid's toys, these named "Spielzeugtransformatoren" (EN 61 558-2-7), are specified as a subgroup of the "Sicherheitstransformatoren". I am quite sure that there must be similar regulations for everyday-use-transformers worldwide. And so there are plenty enough of transformers around which fit into the 5kV class and therefore also seem fit enough for ESL array use. Make your choice wisely ...

And meanwhile, I also contacted the distributor of these aforementionned, high-quality swiss-branded, but surplus and el-cheapo line transformators. They were not able to state any specs about the isolation between primary and secondary of theirs own product ... Since it is out of interest an well suited as is for a regular 100V line use. This reminded me of the good old days at Rolls-Royce, when asking about the motorization's horsepowers: Anecdotically, the anwer would have been a decent and elegant ... "enough". Ok, then. My conclusion is to better resort to some surplus 230V/12V toroidals indended for indoor mobile halogen lighting to build up my ESL array, if I look for a cheap solution. It seems much safers than to build an array out of this wonderbox' specifically unspecified 100V line transformers.
 
Last edited:
Hi,

if You build efficient panels -meaning hybrid panels- simple power toroids work very well.
Their rather low primary inductance won't suffice for low-freq useage.
A pair of 230/9V gives around 1:50 for large panels with large capacitance around 2nF, a pair of 230/6V gives around 1:70 transformation factor for smaller capacitance panels.
50-120VA cores are a good choice, but even small 20VA types will allow for enough power for medium and smaller sized panels.
I used to order standard power toroids from BSAB in Germany, which are single-primary/single-secondary types with an insulation specced to 4kVrms.
Connected in parallel-prim/serial-sec the insulation value of the combi doubles.
In all those years I had not one single transformer fail neither due to internal flashovers nor any other problem.
I regard them as best buys at their comparably low price level.

jauu
Calvin