My apologies, I discovered I had measured the transformer with the wrong primary ends (inversed). During the previous measurements, the clumped capacitance was higher. Now with the correct wiring it is lower and it results of a -3dB point at 65kHz instead of 58kHz with the secondary grounded. Of course, the square waves show a slightly higher Q factor of the Cp and Ls resonance, this is quite normal.
I will redo these measurements together with the THD measurements and post them when the times comes.
Best regards!
I will redo these measurements together with the THD measurements and post them when the times comes.
Best regards!
Last edited:
The primary has a natural completely symmetrical CT. It can be used in a P-P configuration. The problem is that you'll have 1.5k of primary impedance per side. It might work for a parallel P-P or a very low Rp tube.
it should work ok end to end is 3k?
46 tube data
https://frank.pocnet.net/sheets/021/4/46.pdf
My bad. I was on a couple of beers when I wrote this post and I still do.
Half of the primary turns results in 1/4 of the primary impedance - 750R.
Using the transformer in a P-P configuration will also lower the Cp, because of the fact the primaries will be referenced to ground from the center tap.
Half of the primary turns results in 1/4 of the primary impedance - 750R.
Using the transformer in a P-P configuration will also lower the Cp, because of the fact the primaries will be referenced to ground from the center tap.
In the case of the (PP) amplifier working in class A, both tubes are conducting all the time and they swing continuously across the whole secondary. The impedance of the whole secondary is "seen" by both tubes.
But in the case with class B, one tube conducts during a half cycle, while the other does not. So basically one half of the primary works per half of a cycle too. Half the primary will have 4 times less the impedance than the whole primary.
But in the case with class B, one tube conducts during a half cycle, while the other does not. So basically one half of the primary works per half of a cycle too. Half the primary will have 4 times less the impedance than the whole primary.
Yes I already told ya class B PP 46 amplifier and there is a link to the data sheet Plate to plate Imp is suggesting 3500ohms or there abouts
I have just started. As I said I am using a EI-120 core.
I have the perfect wire size to wind 2720 primary turns with the selected geometry but don't have the right wire size for 140 secondary turns that will minimize copper loss. So I will use a smaller wire and accepet a bit higher Rdc. Power loss should still be 0.5 dB which is acceptable, I think.
I have the perfect wire size to wind 2720 primary turns with the selected geometry but don't have the right wire size for 140 secondary turns that will minimize copper loss. So I will use a smaller wire and accepet a bit higher Rdc. Power loss should still be 0.5 dB which is acceptable, I think.
That's great! You'll be probably finished before me. I have to urgently wind the second transformer, because the mandrels will be used by a colleague soon.
if you two gentleman would like an independent listening review on which would sound better I would be obliged
but then the winner would need to wind the other one so i could use them the pair
A chance I won't be able to see my transformer again.
The EI-120 core I am using has a 40x50 sq mm area which is indeed 19 sq cm when considering a typical stacking factor of 0.95.
For 200 Vrms at 20Hz (about 13W into 3K) with low distortion 1920 turns are insufficient because the induction B is way too high for EI cores. They might be ok with Hi-B C-cores.
Winding 2720 primary turns total induction Bdc + Bac (with 130mA DC and 200Vrms signal) is 0.92T at 20Hz. At 30Hz it's even better.
Having more turns means smaller wire size has to be used resulting in a bit higher copper loss however this is still ok with efficiency around 90%. In my experience this is a better choice if high frequency response is good enough.
If one wants better efficiency without compromising other specs better to go with C cores....
For 200 Vrms at 20Hz (about 13W into 3K) with low distortion 1920 turns are insufficient because the induction B is way too high for EI cores. They might be ok with Hi-B C-cores.
Winding 2720 primary turns total induction Bdc + Bac (with 130mA DC and 200Vrms signal) is 0.92T at 20Hz. At 30Hz it's even better.
Having more turns means smaller wire size has to be used resulting in a bit higher copper loss however this is still ok with efficiency around 90%. In my experience this is a better choice if high frequency response is good enough.
If one wants better efficiency without compromising other specs better to go with C cores....
I am not sure. I have just finished winding the transformer but then need to put the core, impregnate laminations and test it. tomorrow I am going back to UK. I will return back in one month. I don't have any equipment in the UK. I will likely send the OT to my friend Waltube who has certainly better gear to test it thoroughly....
to 45
Ok I am waiting you stuff while I am testing the "strange tubes"
to 50AE
please use the test set as described on my old post about LITZ trafo ( where I am still running, Hardly!!!).
It is a smart way to test fine the trafos.
You need a good ss amp, 30-50 watt max
You can omit the Ia bias, for now
Walter
Ok I am waiting you stuff while I am testing the "strange tubes"
to 50AE
please use the test set as described on my old post about LITZ trafo ( where I am still running, Hardly!!!).
It is a smart way to test fine the trafos.
You need a good ss amp, 30-50 watt max
You can omit the Ia bias, for now
Walter
Attachments
Thank you for suggesting your method. But there is no way I want to test my transformer without Ia. We're testing a transformer for a classic SE application, so there must be current through the primary.
There's no way I can have a 4kV power supply either. A CCS sounds like a much better idea.
There's no way I can have a 4kV power supply either. A CCS sounds like a much better idea.
In my opinion you can test the stuff without Ia first to understand the high frequencies that are impacted not so so heavy by L but parasitic.
For bass without current you can understand the face.
We are check a new method to get the bias current with a better method.
The test on primary winding is more important with a large swing .
Walter
For bass without current you can understand the face.
We are check a new method to get the bias current with a better method.
The test on primary winding is more important with a large swing .
Walter
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.