True.. the fb will make the output stage work a bit louder to compensate. So, in both cases output (vrms) is about same.😀
I do not understand why people go crazy about squeezing max watts from SE.
3 or 5 watts no big deal on loudness; but distortion profile does matter more.
And no or little FB means, high total gain is not needed. Less stages, less trouble. Higher impedance is better
Comparison is usually made on the same basis. Comparing apples with oranges doesn't make much sense.
If I can get 5W instead of 3W at the same THD, it usually means that THD will be less in the former case @ 3W if things are done right. Things are related one to the other, you know.....
Of course comparison has to be made on the same basis otherwise doesn't make sense. If not, one could actually argue that transformer quality doesn't make any difference at all. Unfortunately that is not the case.
Not criticising any posts, but I read a somewhat uncomfortable amount of adjectival description. A is "better" than B. C is "audible". Perhaps to add some perspective, if I may:
We are talking about audio (thus things audible to the general human ear), and about output transformers (thus tube applications).
A quantative example: 1 dB of transformer loss would mean e.g. that 126W output will be reduced to 100W. Audible? Sorry, according to tests not even when compared side-by-side, let alone on a day-by-day basis. (These tests were done using 'normal' somewhat educated listeners. My sympathy for those so well-endowed with super hearing that every little thing will compromise their listening pleasure.)
Then, this only applies to maximum output. Decrease the output by say that 1 dB, and the effect of the loss as far as limiting the maximum output will be moot; higher output by the tubes will cover that (other factors being equal).
We can talk about exactly what level of distortion matters, which products at which frequency etc., - the usual difference of perception will be encountered. I have read about OPTs with lower distortion that the output tubes. Then bringing primary impedance into this discussion is of secondary importance. The main effect of primary impedance will be on the performance of the stage, i.e. the tubes, not the OPT. [Obviously a very high impedance ratio will bring about higher losses than a low one (there I go myself with adjectives!), but again optimal tube performance as a result of using the correct primary impedance will override OPT effect merely regarding P/S ratio.]
Frequency response is a matter on its own, mainly regarding NFB stability as said; I would not bring that in viz-a-viz transformer losses alone. These days reasoably priced C-core transformers are available; beyond that I am sceptical regarding audibility of other small differences in OPT specs (if honest!).
We are talking about audio (thus things audible to the general human ear), and about output transformers (thus tube applications).
A quantative example: 1 dB of transformer loss would mean e.g. that 126W output will be reduced to 100W. Audible? Sorry, according to tests not even when compared side-by-side, let alone on a day-by-day basis. (These tests were done using 'normal' somewhat educated listeners. My sympathy for those so well-endowed with super hearing that every little thing will compromise their listening pleasure.)
Then, this only applies to maximum output. Decrease the output by say that 1 dB, and the effect of the loss as far as limiting the maximum output will be moot; higher output by the tubes will cover that (other factors being equal).
We can talk about exactly what level of distortion matters, which products at which frequency etc., - the usual difference of perception will be encountered. I have read about OPTs with lower distortion that the output tubes. Then bringing primary impedance into this discussion is of secondary importance. The main effect of primary impedance will be on the performance of the stage, i.e. the tubes, not the OPT. [Obviously a very high impedance ratio will bring about higher losses than a low one (there I go myself with adjectives!), but again optimal tube performance as a result of using the correct primary impedance will override OPT effect merely regarding P/S ratio.]
Frequency response is a matter on its own, mainly regarding NFB stability as said; I would not bring that in viz-a-viz transformer losses alone. These days reasoably priced C-core transformers are available; beyond that I am sceptical regarding audibility of other small differences in OPT specs (if honest!).
Which tests?? Audibility is referred to different sound. It is audible sorry.
About power loss. 1dB power loss is quite a big loss. For 100W you have to build a 126W amplifier; With 0.2 dB power loss you need to build a 105W amplifier. I can build a 105W amplifier with 4XKT88 without compromises I can NOT do it for 126W. It makes a difference for me.
Why should I waste power when I can avoid it? Why should I give up a better sound?
I have the feeling that some people cannot have a best quality transformer so it must not make a difference.....
45,
Many tests regarding what loudness increments are audible were done over years, leading to the general acceptance of how noticable e.g. 3dB and 1dB are, under which conditions. Details of these have been published over decades and accepted in textbooks. You have not noticed? Take on this research if you wish; I am only the messenger. (Look them up yourself; the internet and Proc. AES are your friends.)
My, my ..... I will give you the benefit of doubt and pardon you that sarcasm. Have a nice day, sir!
(Just for the record: I don't buy them, I design and make them. Have been doing that for the last .... never mind.)
No you have to run always 1 dB higher to get a given output unless you use 1 dB more efficient speakers. 1 dB more efficient speakers might be quite different and quite more expensive than more efficient transformers. I would say in general that more likely other factors are not going to be equal at all. Transformers with high power loss have also other problems. Higher power loss is typically the result of making things in economy. It typically means simplified design and/or smallest possible size. It's not just power loss. There are only very few exceptions of quality transformers where higher power loss is traded with universality but they are not cheap.
Many tests means nothing for me. Sorry. I trust my ears first. An article published in AES is not the bible. There are good articles and rubbish articles. Anyway the AES journal is a low impact factor journal (it means it makes little difference on how the world evolves). The average value over the last few years is not even 1 (i.e. quite low).
Me too....🙂
Last edited:
Well I took on this research when I started winding transformers quite some years ago by prototyping.
The amplifier was a 300B single ended with a 1:1 interstage transformer coupled C3g front end.
Prototype output transformers had the same impedance ratio, 3k5 : 8 ohm when I remember well, but had different DC resistances and core dimensions (both good quality silicon steel c-cores). The lower DC resistance transformer had some 60 ohm primary DC resistance and under 0.2 ohm secondary DC resistance.
The more economic transformer had some 150 ohm primary DC resistance and over 0.3 ohm secondary DC resistance.
Listening to both output stages was done over a longer period, and the obvious conclusion was that the lower DC resistance transformer was much preferred (dynamics; transparency; better loudspeaker control because of better DF). Ever since I followed that low DC resistance route (much "iron" - little "copper").
> Should that not be 3 ohm?
You are replying to a 5 year old post from a member who ask his account to be disabled.
However the DC resistance of a power (including audio power) audio transformer is normally a small fraction of the load impedance. Often 1/5th to 1/20th. So 3 ohms is kinda high for an 8 ohm load. 0.3 Ohms is awful good, though easier if deep bass is not needed.
This is implied by the <1dB losses being bandied about above. Has to be less than 1 Ohm into 8 Ohms to be <1dB.
You are replying to a 5 year old post from a member who ask his account to be disabled.
However the DC resistance of a power (including audio power) audio transformer is normally a small fraction of the load impedance. Often 1/5th to 1/20th. So 3 ohms is kinda high for an 8 ohm load. 0.3 Ohms is awful good, though easier if deep bass is not needed.
This is implied by the <1dB losses being bandied about above. Has to be less than 1 Ohm into 8 Ohms to be <1dB.
I know. Just a wanted to know if someone else would be able to answer. I ask because I have an output transformer with 136 Ohm primary and 3 Ohm secondary. The 0.3 seems very low in that light.
Aha..Thanks gorgon53. I just re-measured. 1.5ohm. Dunno what happened last night. But Is that ok? Or also very high?
Primary is 5k2 secondary 6 Ohm.
Primary is 5k2 secondary 6 Ohm.
Last edited:
Never knew that. Thanks Kay. So the meter is just not that accurate for measuring low resistance.
And that's the value you have to subtract from the readings.
Btw, »better« resistance meters provide Kelvin/four lead measurements to compensate for the probe cable resistance.
Best regards!
Btw, »better« resistance meters provide Kelvin/four lead measurements to compensate for the probe cable resistance.
Best regards!
In an even older post from 2006 Tubelab reported that he could measure big differences in OPT power losses: Why are Hammond SE OPTs dissed? Heavier than the competition...
quote:
I ordered the Hammond 1628SE's from a major supplier.... These are the most inneficient transformers that I have tested. The amp makes 5.5 watts in triode mode and 8.5 in UL with the Hammond. I get 8 watts in triode and 11 in UL with the $18 Edcor. It must take a lot of energy to magnetize all that metal...
:unquote.
quote:
I ordered the Hammond 1628SE's from a major supplier.... These are the most inneficient transformers that I have tested. The amp makes 5.5 watts in triode mode and 8.5 in UL with the Hammond. I get 8 watts in triode and 11 in UL with the $18 Edcor. It must take a lot of energy to magnetize all that metal...
:unquote.
The typical handheld DVM unable to measure correctly transformers or chokes DCR.
Instead of try old Ohm's law.
DC power supply (or battery) -> resistor-> inductor.
Measured voltage on the resistor shows current.
Measured voltage on inductor divided by current equal DCR of inductor.
Instead of try old Ohm's law.
DC power supply (or battery) -> resistor-> inductor.
Measured voltage on the resistor shows current.
Measured voltage on inductor divided by current equal DCR of inductor.