Toroids for ESL's

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Yes, using two cores in this configuration lowers your point of saturation by a factor 2 (or one octave) compared to that of one core.

One core as a 1:40 transformation ratio and this will not be enough to drive any panel loud enough for use.
Therefore two are needed for 1:80 transformation ratio or 1:74 as measured.

Using four cores will lower the point of saturation by another factor of 2 ( for two octaves).
But this will result in a higher transformation ratio of 1:148 this could cause issues with the added transformer capacitance due to the Bifillar wound 120v windings.

In order to combat this you can use 4 cores and just one 120v winding ( or both in parallel) on each core and this will lower your saturation point by the same two octaves and still have a very usable transformation ratio of 1:74 for a large panel.
As well as combating the transformers self capacitance issue.

As Charlie had mentioned these are very affordable cores to work with.

jer :)
 
Last edited:
This is the panel crossover setup i am use now.....this come from ML thay use all the same stepup transfourmers in....
SL3, request,Ascent, Ascent i ,Odyssey, All these use the same panel....12'X48" Prodigy uses two of the same tranfourmers... pr panel 16"X48"....
ML uses a diff crossover in each...
Do you think i can setup the Antek AN-0506 this way....i know it well be diff but should work right.....this is the best sounding crossover setup type yet...i have even used....For give the Art....
Its also in my Acoustats 121 interface...with the ML tranfourmers...
:hohoho:thanks for any input
 

Attachments

  • new cross esl 002.jpg
    new cross esl 002.jpg
    468.5 KB · Views: 479
Yes, You can set them up that way.

But, it doesn't say that they will perform the same way as the ML transformer as the transformer will have a different set of characteristics and parameters than ML's transformers.

I would be a nice experiment to try though.
If you do chose to try it, let us know the outcome!!

:cheers:

jer :)
 
Thanks jer
This centerTap type crossover setup got MartinLogan the MId-range of the year in the Quest-request.
I thought it was the biger panel.... but it works well in 12" wide panels ...or any size well .....killer Mid...an topend...it just uses more of the tranfourmers windings....as input to the panels ....tone shaping...vary cool...something you dont see on other ESLs...Why ask why...
I gess it like the Bias... some think it dose not help the sound to have the best bias caps-diodes you can........i just wont it all.....
Have As much fun as you can.....MX
 
Member
Joined 2005
Paid Member
One core as a 1:40 transformation ratio and this will not be enough to drive any panel loud enough for use.
Therefore two are needed for 1:80 transformation ratio or 1:74 as measured.

Using four cores will lower the point of saturation by another factor of 2 ( for two octaves).
But this will result in a higher transformation ratio of 1:148 this could cause issues with the added transformer capacitance due to the Bifillar wound 120v windings.

In order to combat this you can use 4 cores and just one 120v winding ( or both in parallel) on each core and this will lower your saturation point by the same two octaves and still have a very usable transformation ratio of 1:74 for a large panel.
As well as combating the transformers self capacitance issue..

Instead of four Antek AN-0506 (50VA) with one 120v winding, I went with two Antek AS-1206 (100VA) and was very happy on my CLS. The setup was exactly like Jazzman has on his site - single 1 ohm series resistor and nothing on the secondaries to the ESL panels. Sounded very good.

I was lucky enough to get another four to build another interface - but now Antek has plenty (46 last I looked).

They did have issues below 150hz on my CLS, but above that they were loud and clear.

I have a set of CLS on a pair of Antek AN-0506 and it sounds nice, but I prefer the AN-1206 better...

Unfortunately, no primary side (amp side) centertap is possible with a pair of these...of course you could do the four and AS-0506 and get a primary side CT...
 
Last edited:
Member
Joined 2005
Paid Member
ML CLX uses bigger EI core for 55 to 350 Hz, and torroid for 350 to 20kz.
The EI core in the ML Request era isn't very large, but a friend's recent ML Electromotions have a tiny EI core transformer combined with electrolytics, polyester caps and iron core inductors in the X-Over. They sound like it too - compressed , grainy, boxy, compared to the upper end ML products, but still a touch of the electrostatic midrange magic.

Dayton Wright used 39 pound EI core for 30 to 7 kHz - 80kohm secondary
 
Last edited:
Theoretically the Antek 0506 and 1206 should have nearly the same core saturation points as they are both designed for the same 50Hz to 60Hz frequency range.
As the turns are adjusted as such for the same range but still have the same ratio.

However being the 1206 has approximately 50% more iron ( by weight) you may find a bit better THD performance in the lower frequency's.

I had discovered this when I was stacking my 200watt cores and the 30hz to 15hz sine waves were cleaner the more iron I used and I went as far as stacking 6 cores.
Even as far down as 10Hz the limit of my signal generator.
I could even see the Diaphragm pumping back and forth at this point.

I also adjusted the primary turns so that the saturation point was the same as well.
As the study was to see if more iron resulted in a lower THD for the bass frequency's.

I wasn't able to do THD measurement at the time but I was using my scope to analyze the waveforms.

Second with the larger core of the 1206 this also means a lesser amount of total turns.
A lesser amount of total turns should yield a lower amount of transformer capacitance as well.
This would allow a better and higher bandwidth frequency or higher transformer resonance.
This is like killing two birds with one stone making it easier on the amplifier at the higher end of the audio spectrum.

If anyone knows the procedure and would take the time to test a 1206 for the leakage inductance and capacitance parameters it would be very interesting to see the results.

I would, But, I don't have one or the funds to get some at this moment.

The 200watt cores that I use have about 1/3 to 1/2 (or so) the capacitance of Antek's 50 watt 0506 cores at around 400pf for one core.

So, I am wondering if my theory holds some truth to it, as the bigger can also lower the transformers capacitance by allowing more room for the windings and having lesser windings as well.

As I mentioned before, I still have to use four cores and one 120v winding per core in order to keep my saturation point low enough and my transformer capacitance at a reasonable level to use my cheapy 80 watt amp without over loading it.

At a normal listening level of about 90db ( even as high as 100db) and 5v peak signals I get a very good sounding low end above 80hz out of my little panel as I listen to them nearfield within .5 to 1 meter away.

My total turns count is similar to the antek's as my cores were designed for the same 50Hz to 60Hz range as well only they are rated at 210watts.

jer :)
 
Last edited:
Some people With SoundLab ESL have moved away from the Toroid used for the topend of there interfaces crossover setup.....an went with the fullrang EI type...an say thay get better sound an more output....an SL can now be driven with smaller tube amps...Dr west has a new Toroid that is to be better...the latest an greatest....but some say there just two happy with the EI type....Toroids look good on paper...an the $$ is right...
thanks for any info on ESLs
Like how the crossover in the CLX is setup...MX
 
I've been wondering about those AS-1206 transformers and what effect, if any, the shielding used between their windings might add. I don't have the capability to test and measure these transformers myself but I will buy and donate one to anyone here who could test them and post the results-- as our friend Bolserst tested the AN-506's here:
 
Last edited:
Hi,

Even power toroids can have really low distortion and HF resonance well above the audio band, and can beat EI cores in these respects.
However the problem it's rather easy to saturate them. Combined with low Rdc of the winding even 30 mV DC offset at amp's output can put significant DC current and seriously degrade the performance. Therefore a series resistor(0.3 - 0.6 Ohm or so) is a must for toroids IMO.

Regards,
Lukas.
 
I would be happy to test a 1206 Charlie.
Now that I have fixed my frequency counter I have been able to get some very accurate and repeatable measurements with my cores.
And they are pretty consistent from core to core.

I have not had any major issues yet with the DC offset, although I have not tested for them.
Typically my added primary windings that I use on my cores have a self resistance of about .3 ohms.
I have in the past added a .47 ohms resistor in series as per spice simulation.
But, I notice no difference in the sound at all or very little if any.
So, I stopped adding it.
It could be that I just don't hear as well at high frequency's any more.

I used to add as much as 2 or 3 ohms but this degraded the sound greatly and attenuated the sound.
By using spice (circuitmaker2000) I found that no more than .33 ohms was needed by using my measured data.
This is when I found that the primary winding already had this value of resistance.

Good point however, I will look more into this when I get back into some more testing of my cores.

I don't have any experience with any I-E cores except on one that I have that came out of a PushPull 6CA7/EL34 Knight kit amp.
I didn't use it for fear of arcing at high voltages and then I would be out a good tube amp.
I had ruined two good 6V6 transformers because of this before.
It did sound good with my Micro ESL running a 500v bias though.
Now that I have measurement technique down pat, I my revisit this transformer for a comparison sometime.

Jer :)
 
Last edited:
Hi,

2-3 ohms is IMO too much. As far as I know most of trafo's non-linear components are in parallel with primary, so the output impedance of the amp should be low enough for low distortion.
My problems were caused because I have used trafos with very low Rdc(<0.1 ohm for all of them).
If you have Rdc ~0.3 ohms there should be no problems as long as the amp has low dc offset.
You can for sure do some measurements if core saturation point is changed depending on series resistance if you have an interest ; ).

Regards,
Lukas.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.