he will see this when he wakes up and logs on...
he can reply or not, up to him....
posting here is never obligatory...
he can reply or not, up to him....
posting here is never obligatory...
actually the amp design always starts with the speaker, what sensitivity?
and therefore how much power?
so knowing these, you can decide what amp to use...
my advice is to start with low voltage amps at first, as you gain more
experience, you can then move up to the bigger voltage tube amps...
and therefore how much power?
so knowing these, you can decide what amp to use...
my advice is to start with low voltage amps at first, as you gain more
experience, you can then move up to the bigger voltage tube amps...
I´m not lundahls advocate, but for me here is easier way to get things done. (maybe toroids from poland also, sometimes)
That sim photos shows everything. For 20....20k -3dB :
.........20k5.....9k....
MIN > 31H > 18,6H
MAX <178mH & <90mH
parasitic H for ll1688 is 7mH
It seems both opt will exceeds hifi limits🙂
No it doesn't show what happens when both are at 11.5W. That was my question. With 9K distortion is about the same and so overall is more convenient.
With 20.5K 1250V would a better idea for me but still efficiency would be on the lowish side in class A1. The bias points don't have to be the same. It doesn't make sense.
20kHz -3 dB is poor surely for a 9K transformers. It means more likely response is -1 dB or more down already at 10 kHz which is clearly audible!
It should be no more than -1 dB @ 20kHz to be ok.
I heard from one of the transformer designer , he does not think the working voltage above 800 should be used.this is what he said about 1000V OP transformer
"The transformer must be dried in a controlled environment and then inserted into a pre-evacuated oil bath and then have a vacuum pulled on the immersion tank, which is also immersed in a pre-evacuated oil reservoir in the vacuum tank. This is the infamous PCB oil. Even with that precaution layer insulation for every layer of wire would be a needed precaution, thus increasing distributed capacitance and thereby limiting high frequency response to - 3db at 18kHz by inducing phase shift.
Without question you can obtain output transformers that will last at least 100k hours without this oil bath. Their high frequency roll off will begin at about 9 kHz and by 20kHz likely exceed -30 db. In addition the core will need to be sized to obtain enough winding width and height to allow for the large amount of layer insulation required and the dead air space required from layer to layer around the edges of the layer insulation, to provide adequate creepage distance. This large core will induce phase shift in the low frequencies just from an increase in magnetic path length. Hence the -15 db at 20 Hz often found in large SE OPT's and the phase shifted low frequencies that "sound" as though they are weak and flabby. "
"The transformer must be dried in a controlled environment and then inserted into a pre-evacuated oil bath and then have a vacuum pulled on the immersion tank, which is also immersed in a pre-evacuated oil reservoir in the vacuum tank. This is the infamous PCB oil. Even with that precaution layer insulation for every layer of wire would be a needed precaution, thus increasing distributed capacitance and thereby limiting high frequency response to - 3db at 18kHz by inducing phase shift.
Without question you can obtain output transformers that will last at least 100k hours without this oil bath. Their high frequency roll off will begin at about 9 kHz and by 20kHz likely exceed -30 db. In addition the core will need to be sized to obtain enough winding width and height to allow for the large amount of layer insulation required and the dead air space required from layer to layer around the edges of the layer insulation, to provide adequate creepage distance. This large core will induce phase shift in the low frequencies just from an increase in magnetic path length. Hence the -15 db at 20 Hz often found in large SE OPT's and the phase shifted low frequencies that "sound" as though they are weak and flabby. "
Last edited:
pieter t got the feeling that he is wasting his time in this thread, so he will be silent, until maybe (but I bet never):
popilin comes up with real world measuring results (yes please less algebra and simulation but measurements of the real performance).
Then a good evaluation of how output transformers who are wound equally but have different core material influence sound in the same circuit using the same speakers.
But it will not happen 🙄
I did show real world measurements about magnetic permeability, which curiously is the point here, just read again.
http://www.diyaudio.com/forums/tubes-valves/305872-opt-845-tube-4.html#post5036389
Although this is nothing compared to the amazing amount of measurements provided by you, isn't it? 😀
Equations tell more about a transformer than that some people think, mostly without knowing the equations.
I know, the usual chitchat: Core A open the highs, core B open the midrange, core C have a better soundstage, core D has a better expression of tonal accuracy...
I can provide even more poetic "evaluations"... 😛😀
BTW, if two transformers having different core are wound equally, I just would not trust on the winder. 🙄
Last edited:
Read better popilin: I did not say "different core", but I said "different core material".
I have 4 c-cores here with exactly the same dimensions (based on SU75b geometry) but they differ in grade and/or material:
1. HiB GOSS 0.27mm;
2. HiB GOSS 0.35mm:
3. Amorphous alloy;
4. Finemet (nanocrystalline).
So I can exchange these cores using exactly the same wound coil.
This way I can do meaningfull measurements and sonic evaluation.
(My final contribution here 🙂)
Pieter, what was your conclusion ,my head spinning looking at your arguments, lost track what each one is saying.
I don't agree on that. Transformers precisely made for the 845/211 to work up to 1250V a least don't have to use PCB oil.
With high voltage trasformers things need to be looked at and solved more carefully. For transformers that have to work above 500V I usually get a safety test that do not do for lower voltages. I have a friend that makes transformers for industry that can do that for me.
A Megahertz transformer TU110SE is a 10K transformer that goes surely flat up to 25-30 kHz AT LEAST and can work with 1250V plate voltage without problems. I have made this example because this is available for diyers but it is expensive.
Unless my logic is wrong, if two cores has exactly the same dimensions but are made with different material they are indeed...er...different, isn't it? 😀
When I do a transformer design, I start from the core, and its limitations.
As there are many variables involved, transformer design is full of compromises, and you should already know it.
You think that you are evaluating different core material influence in sound, but I am afraid that you are just evaluating different transformers as a whole.
If you have a standardized design for all cores (read it as core material) good for you, it saves a lot of time, but do not count on me.
Last edited:
basics
If an output valve can deliver 10 Watts into a load of 3000 and the speaker system is 8 ohm,
we use transformer to match the load from using the formula P= e2/r
where P is the power in Watts
e is the voltage across the load
r is the resistance of the load
We can determine the signal voltage that is found on the primary of the transformer
10= e2/3000 or e2= 3000 and e=sqrt 3000= 173.2V
Assuming zero loss for secondary of 8ohm impedence
10=e2/8 or e2=80 e=sqrt 80= 8.9
Turns ratio of transformer= 173.2/8.9= 19.35:1
If we assume transformer is only 90% efficient
Power corrected for loss of 1Watt
9= e2/8 0r e2=72 and e =sqrt 72=8.49
Turns ratio of transformer= 173.2 /8.5 = 20.376.
I am making some progress
If an output valve can deliver 10 Watts into a load of 3000 and the speaker system is 8 ohm,
we use transformer to match the load from using the formula P= e2/r
where P is the power in Watts
e is the voltage across the load
r is the resistance of the load
We can determine the signal voltage that is found on the primary of the transformer
10= e2/3000 or e2= 3000 and e=sqrt 3000= 173.2V
Assuming zero loss for secondary of 8ohm impedence
10=e2/8 or e2=80 e=sqrt 80= 8.9
Turns ratio of transformer= 173.2/8.9= 19.35:1
If we assume transformer is only 90% efficient
Power corrected for loss of 1Watt
9= e2/8 0r e2=72 and e =sqrt 72=8.49
Turns ratio of transformer= 173.2 /8.5 = 20.376.
I am making some progress
Last edited by a moderator:
@kinku, in case you have not seen this yet, Patrick Turner has a very informative page
about transformers....output-trans-theory
about transformers....output-trans-theory
in the choke testing that i did, i used the same coil
and same sized cores., 3/4, but two different materials,
Z11/M6 and H50.....indeed there was a difference...
That's the point regarding the OT. Even 120Hz might not be a ideal for assessing the OT behavoiur at low frequency. The first part of the design to decide how many tunrs, air gap etc. is based on low frequency performance (below 50-60Hz). Then the geometry is chosen to get things right at the other end. For chokes 120Hz is ideal.
For the OT, as you go up in frequency the higher inductance from a lower loss material becomes more and more irrelevant in the sense that ususally the applied load in much less than the associated reactance. So, generally speaking, in SE OT's from about 150-300 Hz (depends more on the design compromises) the charateristic impedance you see reflected when looking into the primary is truly the secondary load multiplied by the squared turn ratio. The lower loss is not needed to make it work perfectly. The sound might be different but lower loss doesn't mean necessarily better sound. Maybe the H50 is a bit too lossy for SE OT application (I don't know as I have never used it) but for the other typical materials the lower the frequency the less losses count and the effective permeability in SE transformers will be ruled by the air gap (at least in sensibly made transformers). That's why it is a design choice.
H50 is the common material available to us,
although recently i observed that magnetizing currents
seem to be lower now, maybe it has improved
but the sellers do not have the information...
most of them are from Taiwan of China...
i have the Z11, an M6 equivalent and H18,
as well as the RM18 core materials
Bud Purvine posted this in another thread linked by Kinku,