The transformer is made for somebody with an Audionote Meishu.
I made the transformer for a person who wanted more transparent sound, more stage feeling, better bass and at the same time good high frequency response. So that's why it has just 0,14dB copper loss and less then -1dB at 20Hz.
btw sound and good measurements go hand in hand sometimes
(not for overallfeedback but transformers, yes)
I made the transformer for a person who wanted more transparent sound, more stage feeling, better bass and at the same time good high frequency response. So that's why it has just 0,14dB copper loss and less then -1dB at 20Hz.
btw sound and good measurements go hand in hand sometimes
(not for overallfeedback but transformers, yes)
Last edited:
This is a test, same tube circuit as before I sent, but with a C core OPT, 0,1 mm laminated, 2 watt out, no FB.
http://www.multitask.it/thd vs freq/nucleoC-4vout-noFB.jpg
The wiring is similar to Hammond; we have bought ne of this stuff to investigate the method (as I done with other brand we got).
The difference is important; the thd was reasonable low from 20 Hz with an increment a 20kHz.
At the test sound this last seem to be more detailed with a clear bass.
For the high frequency the difference we found are not so important.
Bye
Walter
http://www.multitask.it/thd vs freq/nucleoC-4vout-noFB.jpg
The wiring is similar to Hammond; we have bought ne of this stuff to investigate the method (as I done with other brand we got).
The difference is important; the thd was reasonable low from 20 Hz with an increment a 20kHz.
At the test sound this last seem to be more detailed with a clear bass.
For the high frequency the difference we found are not so important.
Bye
Walter
Walter, even if the wiring is similar you cannot tell anything conclusive. To design a proper transformer you need quite some more info. Copying the winding might be a mistake if the other things are different. If you just randomly change transformer it is like tube rolling. More likely they will perfom differently if they are different.
You cannot conclude that the Hammond is no good in general. It might be possible that samples of the same transformer are quite different and this usually happens with economical objects. I have seen some Hammonds that on paper should be the same and in reality were quite different. In particular one had a bad high frequency response above 5 KHz....
If you want to state that the C core doesn't have distortion and the E+I has a lot distiortion then you are very far from truth. Buy a Sowter and see. Most of their transformers come with specified distortion at full power within 0.5% at 20Hz and 0.05% at 1KHz.
At first sight yes.
However:
With modern materials distortions caused by the core are very small and can be neglected.
EI cores have 2 big advantages over C-cores and toroids.
1) You can make a stack of any meaningful cross-section area with different number of EI pieces, and thus, using sick stack, considerably decrease number of primary turns with all parasitic gremlins. With C-cores, you are mostly limited to selection of standard sizes.
2) By changing E and I stacking pattern, you can vary quasi-distributed air gap, making your transformer less prone to DC bias.
Thanks God! The lower distortion (as low as 0.001%) of certain types like permalloy and nickel based alloys can be useful at signal level but at power level the core distortion of a properly sized transformer will always be a lot less that of the amplifier at any output level. If the transformer has dominant distortion it is not related to the type of core.
Sorry, but with c-cores can be:
Costum made bobbins (as i do)
Costum made c-cores (i have some too)
Distributed gap c-cores are available
Multi cut c-cores are available too.
The most modern materials i have not seen for EI.
Costum made bobbins (as i do)
Costum made c-cores (i have some too)
Distributed gap c-cores are available
Multi cut c-cores are available too.
The most modern materials i have not seen for EI.
Why 0.75T? First it is not just 0.75T. It is 0.75T @ 30Hz at full power? With good C cores will be 0.85T and with the Hi-B C cores will be about 1T. It is my design choice. There is a very good reason for it. As you are are professor you should know why....
No they are manufactures and produce a range of things for a number of applications. This seems to be really difficult concept to understand for you. It is obvious!
But you almost never run an output tranformer near saturation - maybe for a couple of seconds in loud music passages. At lower inductance levels the difference is much smaller.
By the way, your measurements are extraordinally good.
If you understand physics as you tell around you can read the books, understand and put things in practice as I did. If not I really don't care!
Yes, I have not seen amorphous or nanocrystalline available as EI.
Bobbins for custom C cores had to be ordered from German manufacturer in minimum quantity of several hundreds, what was obviously overkill for 4 pcs I needed for my own projects. I'm was to lazy to make them myself from cardboard paper impregnated with glue.
After all, I built my output transformers with toroid cores. Quite surprisingly, toroid M4 GOSS were able to sustain 1.5T @ 20Hz / max power without any problem. I think Plitron pushes them even harder, taking into account size of finished transformer, 1.75T I believe.
You can push the Hi-B C core as well as it remains practically linear up to 1.8T. In fact Lundahl specifies max power for B = 1.6T (0.9T for DC + 0.7T for AC) for their custom home made cores. However that "practically" means something. Distortion will start to increase inevitably in comparison to the same core working at lower Bmax (where the latter also is the condition of highest permeability which turns into higher inductance in correspondence of the max power) and the transformer won't saturate until really subsonic. This way one doesn't need to see what's happening at 20 Hz as well.....
At the bottom of the page there is the pdf with the measurements and Vanderveen Toroidal or other C-Cores aren't any better than EI at low level (0.5W) down to 10Hz.
European Triode Festival - ETF2009 Shootout
Keeping the Bmax lower with EI makes them work as good at higher levels. Quality EI cores start to have issues around 1.3T.
So I would say that the absolute max one would want to go is 1.7T for Hi-B C-cores and GOSS toroids and 1.3T for E+I.
European Triode Festival - ETF2009 Shootout
Keeping the Bmax lower with EI makes them work as good at higher levels. Quality EI cores start to have issues around 1.3T.
So I would say that the absolute max one would want to go is 1.7T for Hi-B C-cores and GOSS toroids and 1.3T for E+I.
I tested various finished amplifiers (gNFB on and off) with different transformers / cores / tubes at 20Hz / max power at different Bmax levels. Although its not possible to separate distortions produced by the output stage and transformer core, comparing spectrum analysis results shows that most distortions are produced by the output stage.
From my own experiments I come to the conclusion that toroids GOSS cores can be pushed to 1.5T without any noticeable problem.
Moreover, music content with 20Hz @ full power is very unlikely scenario.
I wouldn't be so sure about your conclusions. Distortion is not just harmonic and music is not just a sine wave or any other simple test signal.
Agreed if we are talking about continuous power! Another reason why I consider 30Hz as lowest frequency for my designs. However if one is close to the limit at 30 Hz already he might get saturation on peaks and that causes bad bass performance that typically people attribute to lack of damping factor....
A well sorted valve amp can have quite higher peak power than its nominal RMS. Ok it is distorted but is not really an issue for very short transients. If the transformer saturates it is not the same thing. This is the "mystery" of some amps that sound louder than their RMS power says. For small power amps (10-15W or less) it matters.
Last edited:
- Status
- Not open for further replies.