Again!
Until you don't show a practical test that explain how is wrong the reverse mode yours are only word.
I think you are not able to do this, that's all.
And, as asked by me in other thread, please let show us your project with your amazing OT that are beautiful, I suppose.
Walter
Until you don't show a practical test that explain how is wrong the reverse mode yours are only word.
I think you are not able to do this, that's all.
And, as asked by me in other thread, please let show us your project with your amazing OT that are beautiful, I suppose.
Walter
Imho, until a reverse test is compared to a forward (standard) test under the same power/loading conditions, and shown to have the same results, then the reverse method remains an unverified test method, no matter how much credence is attributed to the reverse method.
I have no doubt the reverse method is easier to set up and make high power measurements, and comparisons between OPT's using that method may provide better insight in to those OPT's.
The difficulty of doing a forward test using a very low distortion high voltage generator, in a quality distortion measurement assessment loop, is no easy feat indeed, and is an indicator of why the two test methods haven't yet been able to be cross-compared. Last year I was doing a bout of OPT measurements and finished off making some distortion tests - even with a 300W ss amp in bridge mode, the test power level could only attain about 1W level, so I started to prepare a parafeed valve amp (after seeing a comment by the late Patrick Turner on the topic) to act as the high-voltage generator but got bogged down.
HI
when I ask to try to test in reverse mode is because someone can have a better configuration to do this way better.
I think this solution is the best way.
The share of information is helpful to reach the solutio and it is very easy to say: it is wrong.
I have published also two curves for both systems and are perfectly the same, the level is different for the well known limit but if most of people say that the level of test is not important for OT this means the my test is perfect.
All the step was published on AUdioreview magazine from Fabrizio Montanucci that is the technical director of magazine and this a great fact.
He is also a engineer that time ago developped and done the Unison Unico 150 project went then in production. Plus other project sometimens left as proto
Last, I have done hundreds of test on different OT that in second time are involved in a amp and every time I know very well which is the weight of OT alone
waltube, you continue to purport equivalence without providing 1:1 test results for verification. Your comments portray the same assumptions as you proffered in the other thread.
Can I suggest you just present results and links, and refrain from snake oil like comments like "the level is different for the well known limit but if most of people say that the level of test is not important for OT this means the my test is perfect."
Can I suggest you just present results and links, and refrain from snake oil like comments like "the level is different for the well known limit but if most of people say that the level of test is not important for OT this means the my test is perfect."
My last phrase is perfectly fine.
And, please read your posts on OPT char. thread around post 180, or you forgot?
https://www.diyaudio.com/community/threads/opt-characterization.313957/page-10
And, please read your posts on OPT char. thread around post 180, or you forgot?
https://www.diyaudio.com/community/threads/opt-characterization.313957/page-10
On diyaudio already was a thread where Timpert proved that the method Walter used was not very accured and that the results where not in line with the forward method. He also showed how to improve the backward test from Walter.
Concering distortion test, if you understand how a transformer works you are not so much interested in this kind of tests. Transformers have very less distortion, a lot less then any tube.
not needed Walt is done before by others, i not have to repeat that.
Read this link again https://www.diyaudio.com/community/threads/opt-characterization.313957/page-3
Post 22, 30, 35, 44, 46 and many more, specially 91 and 105
Last edited:
Again!
If there is the way to do the reverse test better, fine!
One the reason is to share the info
What do you think, I put test not words or poor diagram at +/80 mV and from this you can know everything of OT!
And you can't understand in which way the OT works and the fact that the Zs in not zero!!
In addition with reverse mode you can test accurately the OT changing the Zs of source to simualate the different Rp of tubes even they are non so linear
Bye
If there is the way to do the reverse test better, fine!
One the reason is to share the info
What do you think, I put test not words or poor diagram at +/80 mV and from this you can know everything of OT!
And you can't understand in which way the OT works and the fact that the Zs in not zero!!
In addition with reverse mode you can test accurately the OT changing the Zs of source to simualate the different Rp of tubes even they are non so linear
Bye
Ho ho Walt, your test was NOT to be ment changing Z source, that was not what your wrote here befor. You dit something else, you put a Load resistor and not measured with the impedance of the tube. Or do you want to change your mind now and do exactly what Timpert told you to do?
btw my diagram was a lot better and more acurate then yours, i got a far more real frequency response as in real world and you did not. I tested it also with a tube and got those nice results.
Just a few observations:
1 Watt rms into 8 Ohms requires 2 x (root of 2) Volts rms. That is 2.828 Volts rms in round numbers.
2.7 Volts rms into 8 Ohms is less than 1 Watt.
A transformer can be driven by a negative resistance. That negative resistance can cancel out the DCR.
Most designers do not / will not do that (and make it work over the 20 -20kHz audio range).
If you will not design your amplifier that way, why do that test?
The output transformer itself, as it is actually used in your amplifier will always have higher distortion than the results of that kind of test.
There are designers on the Tubes / Valves threads that do design negative resistance into the amplifier, for the purpose of woofer control and response.
For a 1kHz sine wave (not a low frequency, and not a high frequency) . . .
Please list 5 models of single ended output transformers that are rated to have only 0.2dB insertion loss, and there price (Cost).
Please list 5 models push pull output transformers that are rated to have only 0.2dB insertion loss, and their price (Cost).
And, be sure you only list currently produced transformers, do not list rare or ancient transformers, and do not list transformers that are made of Un-Obtanium.
Thank You!
1 Watt rms into 8 Ohms requires 2 x (root of 2) Volts rms. That is 2.828 Volts rms in round numbers.
2.7 Volts rms into 8 Ohms is less than 1 Watt.
A transformer can be driven by a negative resistance. That negative resistance can cancel out the DCR.
Most designers do not / will not do that (and make it work over the 20 -20kHz audio range).
If you will not design your amplifier that way, why do that test?
The output transformer itself, as it is actually used in your amplifier will always have higher distortion than the results of that kind of test.
There are designers on the Tubes / Valves threads that do design negative resistance into the amplifier, for the purpose of woofer control and response.
For a 1kHz sine wave (not a low frequency, and not a high frequency) . . .
Please list 5 models of single ended output transformers that are rated to have only 0.2dB insertion loss, and there price (Cost).
Please list 5 models push pull output transformers that are rated to have only 0.2dB insertion loss, and their price (Cost).
And, be sure you only list currently produced transformers, do not list rare or ancient transformers, and do not list transformers that are made of Un-Obtanium.
Thank You!
Last edited:
PP is much easier, but SE only a very small number.
Tango has some models and Hasimoto, also the Dutch designer Menno van der Veen (vdv3035) and then there will be a long silence I think.
Fortunately my transformer, a current design, also with only 0.15dB AND designed for full output at 20HZ (not 25Hz or even higher which makes the 0.2dB a lot easier)
N. Crowhurst was very critical but he was right.
btw what about price, is 50 euro much, 100, 200, 400 each transformer? Less then 250 each TOP transformer is for me the max. (with a good c-core)
Tango has some models and Hasimoto, also the Dutch designer Menno van der Veen (vdv3035) and then there will be a long silence I think.
Fortunately my transformer, a current design, also with only 0.15dB AND designed for full output at 20HZ (not 25Hz or even higher which makes the 0.2dB a lot easier)
N. Crowhurst was very critical but he was right.
btw what about price, is 50 euro much, 100, 200, 400 each transformer? Less then 250 each TOP transformer is for me the max. (with a good c-core)
Last edited:
to 6A3
It is evident that 2,7 is type error; but if was intentionally the difference is 0,3 dB, do you think it is dramatic?
Why you speak about negative resistance? nobody speak about it. I use a standard ss amp from different brand and/or kit with a low Zout (verified on test set ( as usual).
And I can't understand why you are speaking around this spec; it is not necessary
"
The output transformer itself, as it is actually used in your amplifier will always have higher distortion than the results of that kind of test.
"
excuse me but I can't understand this concept; maybe is a limit of my english. Please help me to understand
"There are designers on the Tubes / Valves threads that do design negative resistance into the amplifier, for the purpose of woofer control and response."""""
Please give me some better details.
The last part in another aspect that I can' understand.
I have done lot o f test from commercial brand ( as listed partially, some are missing) and my personal protos done from more than one builder but mainly from FIAT in Roma and aren' in market. ; all of them are paid from me, this must to be clear.
Aren't gift.
And aren't words or diagram with +/-80 mV.
The insertion loss is a parameter that is important but with the possibility to have a wide range of results it became secondary just because it is possible to understand the job done by OT itself. The copper has a impact and also in this aspect we are playing around including the search of a good iron.
And in my case I will ask to change some aspect of the proto I am testing and if the results are not fine as expected I will ask to change some parameters; about commercial, it is evident that I do only the check.
Two notes
It is evident that 2,7 is type error; but if was intentionally the difference is 0,3 dB, do you think it is dramatic?
Why you speak about negative resistance? nobody speak about it. I use a standard ss amp from different brand and/or kit with a low Zout (verified on test set ( as usual).
And I can't understand why you are speaking around this spec; it is not necessary
"
The output transformer itself, as it is actually used in your amplifier will always have higher distortion than the results of that kind of test.
"
excuse me but I can't understand this concept; maybe is a limit of my english. Please help me to understand
"There are designers on the Tubes / Valves threads that do design negative resistance into the amplifier, for the purpose of woofer control and response."""""
Please give me some better details.
The last part in another aspect that I can' understand.
I have done lot o f test from commercial brand ( as listed partially, some are missing) and my personal protos done from more than one builder but mainly from FIAT in Roma and aren' in market. ; all of them are paid from me, this must to be clear.
Aren't gift.
And aren't words or diagram with +/-80 mV.
The insertion loss is a parameter that is important but with the possibility to have a wide range of results it became secondary just because it is possible to understand the job done by OT itself. The copper has a impact and also in this aspect we are playing around including the search of a good iron.
And in my case I will ask to change some aspect of the proto I am testing and if the results are not fine as expected I will ask to change some parameters; about commercial, it is evident that I do only the check.
Two notes
- FIAT ha one of the best wound machine in the world; it came from famous brand from Switzerland this allow us to have a good quality of the coils
- Sometimes I buy iron around the world even the results are good but not as expected ( but we will play around) ; like Metglas or Finemet Hitachi plus some minor brand; but with standard C cores 0,1 mm laminated the results are interesting; this will be good for a commercial aspect, in case
Another this transformer against that transformer:
_I am looking for a 6C33 transformer, 600 ohms/8 ohms.
1)There is a Toroidy toroidal (31 Henry, 54 ohm primary) 600ma max ( I need only 300 mA though) - really 31H, will rock the boat except with a very tight drive I’m afraid.
2) Aliexpress amorphous core one, (5.4 Henry, 37 ohms prim). Inductance: 670R (4.5h) static DC maximum 300MA unsaturated inductance parameter
3) I see no small E I core one. I’ld be happy with a lower Primary Impedance - even 300 - 400 ohms is fine.
is there any experience?
_I am looking for a 6C33 transformer, 600 ohms/8 ohms.
1)There is a Toroidy toroidal (31 Henry, 54 ohm primary) 600ma max ( I need only 300 mA though) - really 31H, will rock the boat except with a very tight drive I’m afraid.
2) Aliexpress amorphous core one, (5.4 Henry, 37 ohms prim). Inductance: 670R (4.5h) static DC maximum 300MA unsaturated inductance parameter
3) I see no small E I core one. I’ld be happy with a lower Primary Impedance - even 300 - 400 ohms is fine.
is there any experience?
I think that the Toroidy is preferable seen the few details; don't worry about the specification of current
Walter what is your problem with +/- 80mV? It was just an OUTPUT level on a measurement I made. I also made a measurement at +/- 2.5V OUTPUT and it was the same picture just a different voltage. Are you trolling here? Don't you know that the transformer is linear in this range?
Why are you trolling all the time?
If you want to prove the quality of a transformer, the best/easiest way for frequency response is the forward test method, simply because almost anyone with standard equipment can do such a test. The backward test is limited for frequency response and probably stability because many standard power amplifiers are not good enough for such a test, even you have not shown the real high frequencies.
You need to add a series resistor that represents the actual Ri of the tube, NOT the load representing the transformer impedance (see all of Timpert's comments in earlier comments)
As N. Crowhurst explains, any higher (output) impedance than zero will affect the transformer distortion. Although your backward method is more interesting for harmonic distortion measurements (but in a limited but standard audio range) any series resistance will have an effect on the distortion so what you want to show with your method is not a real life deal.
For a frequency test, it is possible to do a test with a very wide voltage range as long as you use the truly representative output impedance resistance of the tube OR use a tube output. Distortion is not a problem here.
I have access to Swiss-made Meteor and Micafill machines as well as Blume & Redecker and Bobifil machines.
That won't make any difference... Only the design counts and so far I have not seen anything special at Fiat.
Ohh i forgot a lot of other transformer test what can be very very interesting but it doesn't make any sense to tell about that here at the moment.
Why are you trolling all the time?
If you want to prove the quality of a transformer, the best/easiest way for frequency response is the forward test method, simply because almost anyone with standard equipment can do such a test. The backward test is limited for frequency response and probably stability because many standard power amplifiers are not good enough for such a test, even you have not shown the real high frequencies.
You need to add a series resistor that represents the actual Ri of the tube, NOT the load representing the transformer impedance (see all of Timpert's comments in earlier comments)
As N. Crowhurst explains, any higher (output) impedance than zero will affect the transformer distortion. Although your backward method is more interesting for harmonic distortion measurements (but in a limited but standard audio range) any series resistance will have an effect on the distortion so what you want to show with your method is not a real life deal.
For a frequency test, it is possible to do a test with a very wide voltage range as long as you use the truly representative output impedance resistance of the tube OR use a tube output. Distortion is not a problem here.
I have access to Swiss-made Meteor and Micafill machines as well as Blume & Redecker and Bobifil machines.
That won't make any difference... Only the design counts and so far I have not seen anything special at Fiat.
Ohh i forgot a lot of other transformer test what can be very very interesting but it doesn't make any sense to tell about that here at the moment.
Last edited:
Fiat isn't better i guess. The details that matter don't show Fiat either
Last edited:
600 Ohm for a 6c33 is not the best choice. It's not the most linear tube so a bit higher would be beter , 800-1000 for instance.
What is the price level you are looking for, min-max?
tubes4all,
Is your transformer insertion loss: -0.15dB at 1kHz, and -0.2dB at 20Hz?
0.05dB variance from 1kHz to 20Hz, without negative feedback?
I probably misunderstood what you said.
Thanks.
The fact that zero Ohm impedance driving a transformer makes its distortion lower has been known, probably long before Norman Crowhurst (but with credit to him for so much that he did). But to do that with vacuum tubes, they would have to use infinite negative feedback from the primary.
Well, I would rather have a little bit of distortion from the transformer, instead of having to deal with an infinite negative feedback zero Ohm impedance tube amplifier driving the primary.
waltube,
I can not remember the posters name, he posted many single ended amplifiers, using Russian tubes. The amplifiers use current feedback to control the woofer (according to its impedance).
I should not have mentioned negative resistance, but one poster said to drive the secondary with a zero Ohm impedance, and another poster said to consider the winding DCR or the drive would not be zero impedance. At least as I can remember the posts.
We used a negative resistance amp to drive a variable bandwidth series mode crystal filter.
The terminating resistance was adjustable, lower resistance gave narrower bandwidth, but with 6 filters in cascade, the losses were very large, and the signal to noise was bad.
A negative resistance amplifier in front of each filter kept the signal level constant at the termination, no matter the filter bandwidth; so signal to noise was restored.
This is probably the best application of a negative resistance amplifier I can think of.
Is your transformer insertion loss: -0.15dB at 1kHz, and -0.2dB at 20Hz?
0.05dB variance from 1kHz to 20Hz, without negative feedback?
I probably misunderstood what you said.
Thanks.
The fact that zero Ohm impedance driving a transformer makes its distortion lower has been known, probably long before Norman Crowhurst (but with credit to him for so much that he did). But to do that with vacuum tubes, they would have to use infinite negative feedback from the primary.
Well, I would rather have a little bit of distortion from the transformer, instead of having to deal with an infinite negative feedback zero Ohm impedance tube amplifier driving the primary.
waltube,
I can not remember the posters name, he posted many single ended amplifiers, using Russian tubes. The amplifiers use current feedback to control the woofer (according to its impedance).
I should not have mentioned negative resistance, but one poster said to drive the secondary with a zero Ohm impedance, and another poster said to consider the winding DCR or the drive would not be zero impedance. At least as I can remember the posts.
We used a negative resistance amp to drive a variable bandwidth series mode crystal filter.
The terminating resistance was adjustable, lower resistance gave narrower bandwidth, but with 6 filters in cascade, the losses were very large, and the signal to noise was bad.
A negative resistance amplifier in front of each filter kept the signal level constant at the termination, no matter the filter bandwidth; so signal to noise was restored.
This is probably the best application of a negative resistance amplifier I can think of.
Again!
Probably you haven't understand well the type of test.
Mayb.in the future.
Thats'all
What N. Crowhurst ment is that if you put 10W input in a transformer a good transformer gives 9,5W back. That is 95% or 0,2 dB.
For the transformer i use the loss is 0,15dB, 96,6% (about 45 Ohm dcr for the primairy winding)
tubes4all,
I should have put it another way, information for the newbees.
The transformer insertion loss is 0.15dB, 0.2dB, or 0.5dB, etc. at 1kHz.
If the transformer primary inductance and transformer primary distributed capacitance are significant at 20Hz and 20kHz respectively, the insertion loss is greater at those frequency extremes.
The frequency response rolls off at the low frequency end and at the high frequency end.
Negative feedback can flatten out the frequency response, but the tubes have to put more power into the transformer to get the same power out, versus the power needed at 1kHz.
Just my opinion.
I should have put it another way, information for the newbees.
The transformer insertion loss is 0.15dB, 0.2dB, or 0.5dB, etc. at 1kHz.
If the transformer primary inductance and transformer primary distributed capacitance are significant at 20Hz and 20kHz respectively, the insertion loss is greater at those frequency extremes.
The frequency response rolls off at the low frequency end and at the high frequency end.
Negative feedback can flatten out the frequency response, but the tubes have to put more power into the transformer to get the same power out, versus the power needed at 1kHz.
Just my opinion.