diyAudio

diyAudio (https://www.diyaudio.com/forums/index.php)
-   Planars & Exotics (https://www.diyaudio.com/forums/planars-and-exotics/)
-   -   Quad 63 (and later) Delay Line Inductors (https://www.diyaudio.com/forums/planars-and-exotics/338927-quad-63-delay-line-inductors.html)

Hans Polak 13th March 2020 05:36 PM

Bolsert,

While I was working on my posting, your version preceded mine.
Many thanks for all the info, I will have a close look at it and where necessary integrate this in my LTSpice model.

Hans

Hans Polak 13th March 2020 06:47 PM

Bolsart,

A few questions on your excellent work if you allow me:

1)Do the values for the audio transformer that you supplied refer to one single transformer, or did you measure the set of two as one big transformer.

When regarding this as the data for just one transformer, I get 2 times 1.48H in the secondary, which seems too much, leading to a sharp rise of the impedance above 20Khz.
When regarding your data to be valid for the combination of the two audio transformers, I just get one single 1.48H or two 0.74H inducters which seems O.K.

2) The 35pF secondary cap that you specify is a bit of a mystery to me.
Should this not be the capacity between the primary and the seconday winding ?

Hans

martin clark 13th March 2020 11:16 PM

Quote:

Originally Posted by bolserst (https://www.diyaudio.com/forums/planars-and-exotics/338927-quad-63-delay-line-inductors-post6119878.html#post6119878)
Finally located the measurements I took back in 2016....

That's a huge service & of interest to the rest of us - many thanks!

Hans Polak 14th March 2020 01:07 PM

2 Attachment(s)
Steve,

I would stil like to get some of your support in knowing how you defined your transformer data.
Was the data provided for just one transformer, so the Main Amp sees 2x par. 250mH=125mH with a total gain of 244, or did you measure both as being one big transformer making the main amp see 250mH with a total gain of 122.

In the image below, I have shown for both possibilities the FR and the Impedance.
However I had to increase the secondary 3K65 resistor to 65K for FR and Impedance versions to get the same unloaded FR with 7.5dB peak around 20 Khz as you showed, and also the impedance needed this higher value to keep it from getting too low around 20Khz.

You also see a rapid rise in impedance above 20Khz that is steeper than the real ESL63 is showing, that could mean that the 1.48H leakage may be a bit too large ??

But even with the unloaded FR curve being the same as in your picture, I still see rather large differences with the loaded FR.

Do you still have some additional ideas ?

Hans

Hans Polak 14th March 2020 02:38 PM

In the above image, the red line is with complete load and the green line with LC line but without elements
Could it be that the smoothing that Arta applies to your curves, let them look less zig zag ?

Hans

Hans Polak 14th March 2020 09:40 PM

Instead of increasing the secondary 3K65 to 65K, I can also place a series resistor of 60k in the 35pF secondary cap and leave the 3K65 intact.
This gives better results in the FR and in the Impedance curve.
Will show te results tomorrow,

Hans

stokessd 14th March 2020 10:14 PM

Quote:

Originally Posted by bolserst (https://www.diyaudio.com/forums/planars-and-exotics/338927-quad-63-delay-line-inductors-post6119878.html#post6119878)
Finally located the measurements I took back in 2016. Thank again to stokessd for letting me borrow one of the interface units he was purchasing.


Man, I have no recollection of that, but it sounds like something a nice guy would do; and I'm clearly a nice guy so...

:D

I'm glad to help.

Sheldon

bolserst 15th March 2020 04:07 AM

3 Attachment(s)
Quote:

Originally Posted by Hans Polak (https://www.diyaudio.com/forums/planars-and-exotics/338927-quad-63-delay-line-inductors-post6120002.html#post6120002)
…Do the values for the audio transformer that you supplied refer to one single transformer, or did you measure the set of two as one big transformer.

Sorry for the confusion. The parameters are for a single transformer. You can modify the primary inductance of each transformer as long as you modify the secondary inductance at the same time such that: LS = LP x N^2, where N=step up ratio = 122. Depending on the signal level, it could have a range of values, I just chose a nominal value when documenting the parameters. It will really only affect the LF response, so probably not of much importance for your effort. Attachment #1 is screen grab of a spice model I made back in 2016 when I was checking the measurements. It looks to match pretty well. One thing I notice is that I added in a source resistance of 0.3 ohm, which sounds like a reasonable value for all the connections and wiring between amplifier and my test setup. Looks like I also added in about 10pF of stray capacitance to account for other capacitive loads seen by HV windings not related to the winding capacitance(see response to next question below). Usually winding capacitance is so much larger than stray capacitance that it can be ignored, but the ESL63 transformer is a unique design where such details can become important.

Interestingly(at least to me) if stray capacitance can be ignored, the frequency of the HF resonance is at the same frequency for two transformers (wired up per ESL63) as for one. The reason being that the both LL and CS are in series. So the net leakage inductance doubles but net winding capacitance halves, so the product is still the same...thus same resonance frequency.

Quote:

… The 35pF secondary cap that you specify is a bit of a mystery to me.
Should this not be the capacity between the primary and the seconday winding ?
This is the self capacitance of the secondary winding which is made up of many layers. Basically each pair of winding layers acts like the plates of a parallel capacitor. With many layers, you have a group of capacitors in series. The net effect is a capacitive load on the secondary that resonates with the leakage inductance. Attachment #2 provides some details on winding capacitance. You can download the full book from here: http://www.tubebooks.org/Books/RDH4.pdf
The capacitance between primary to secondary winding is very small with the ESL63 transformer because the end of the secondary closest to the primary is grounded.

Quote:

Originally Posted by Hans Polak (https://www.diyaudio.com/forums/planars-and-exotics/338927-quad-63-delay-line-inductors-post6119926.html#post6119926)
… I have changed the coupling factor of the audio transformer from 1 to 0.9998.

You probably already know this, but the coupling factor(K) is just another way to define leakage inductance. If you are including the leakage inductance(LL) as an inductive component in series with the secondary winding, then you should set K=1. Alternatively, you can remove the inductor representing LL and set K to a value that appropriately represents the amount of leakage inductance.
The relationship between LL and K is:
K= sqrt(1-LL/LS)

Quote:

Originally Posted by Hans Polak (https://www.diyaudio.com/forums/planars-and-exotics/338927-quad-63-delay-line-inductors-post6120666.html#post6120666)
…even with the unloaded FR curve being the same as in your picture, I still see rather large differences with the loaded FR. Do you still have some additional ideas ?

As I mentioned previously, I haven’t done any modeling work on the LC transmission line. My work was geared toward understanding the transformer design. Based on review of the Baxandall treatise, I would guess that the differences you are seeing will likely come down to proper modeling of the shorted winding coupled to each of the inductors. This is mentioned to dampen and gently roll off the top end. See also text related to Figure 3. in Attachment #3. Really makes one appreciate Walker’s accomplishment from 40 years ago!

bolserst 15th March 2020 04:14 AM

Quote:

Originally Posted by martin clark (https://www.diyaudio.com/forums/planars-and-exotics/338927-quad-63-delay-line-inductors-post6120240.html#post6120240)
That's a huge service & of interest to the rest of us - many thanks!

You are very welcome.
Iím glad to hear there are people besides myself that find this interesting.

Quote:

Originally Posted by stokessd (https://www.diyaudio.com/forums/planars-and-exotics/338927-quad-63-delay-line-inductors-post6121189.html#post6121189)
Man, I have no recollection of thatÖ

Ha! And I was surprised by how many measurements I found that I had done that I donít remember taking. :D

Hans Polak 15th March 2020 02:55 PM

2 Attachment(s)
Steve,

Thanks for all your info's and the confirmation that your data concerned just one Audio transformer.

I could replicate your LTSpice sim, but also looked at your original measurement, showing a peak around 19kHz of +7.5dB with your HV probe connected.
To get the same +7.5dB I had to add a small serial resistance of 13K to the secondary 35pF Cap.

To show what a huge difference this 2.7Meg//5pF probe causes, I have made sims with and without this probe.

But I'm still faced with a large zig zag ripple when connecting the delay line.
So I will have to play around with the Delay Line Inductor.
When I had one to measure, It would be of great help.

One last question:
In your posting #19, the first attachment, You have shown the inductance of the secondary open transformer for a certain frequency while increasing the input voltage.
Why isn't the inductance flat up to the point where saturation starts ?

Hans


All times are GMT. The time now is 12:16 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 16.67%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Copyright ©1999-2021 diyAudio

Wiki