why bigger transformer? because to have wide frequency response, you need to have a lot of interleavings and insulation. this takes space. higher insulation distance = low capacitance P-S. more interleavings, low parasitic inductance. those are main variables to keep low.
also you need to fit a lot of wire in the window to have good inductance for good bass. if your goal is -3dB on 30 Hz by measuring frequency response you will find out that amplitude is starting to drop below 200 Hz, or maybe 300 Hz and you can hear that. you dont need filter there with S.e. amp which usually has problem with the damping factor and high output resistance. you need to move problems out from the hearing spectrum.
and there are also a losses, usually 10%, better less. you need to go with thick wire to keep them low. thick wire also takes space.. so it is clear .
all of mentioned variables you cant do with small transformer...speaking from experience...and need to see besides fr also how "nice is 1KHz " square signal under full power! you can see the ringing. etc etc
many aspects which you cant find in manufacturer label .
in attachment you can find something of my work, bigger ones 10W S.E. 5 kg each.. smaller one is cca 50mm brick, 10W PP guitar. the same output power.
Are you selling these OPTs or just making them for your own projects?
Please forgive me if I am not 100% on topic. (perhaps a warning is in place: a bit of a rant following but I might as well write it up before I completely loose it)
Some vague memory: In Push Pull the transformer needs a certain amount of current to go through it to magnetise the steel. In other words during crossover the transformer is non-linear and the bigger the core the worse it gets and it is necessary to use large amounts of NFB in a PP amplifier.
This must be why I never took a shine to PP amplifiers, especially of late. I often listen late at night at low levels next to the computer while the other members of the household are asleep.
I've got some papers on critical loudspeaker damping, spelling out that when the damping is not correct your bass response can suffer by a large amount, from memory it could be as much as four fold. The means that ideally one wants to have variable feedback which is really only possible with SE, and perhaps with triodes in PP. Clause Byrith designed such an EL34 SE amplifier.
Since most of the research is decades old and that speaker design has changed a lot this bass response impact may not be true today as much as in the past. But it does partly explain some of the heated discussions on why a certain amplifier works better with certain speakers and that certain speaker cables have a different response than other cables. Attached paper explains how it is possible to test, you'll have to search the link at the top of the paper to find the other two documents (there are three in total) as the others are too big to upload here.
And while we are on this whole OPT business: I have selected a 20% tap for my next SE UL transformer as it is close to the triode behavior and modelling suggest that it is close to what was described in David Saslaw's article (65% of pure pentode output). After I decided on the 20% tap I had a hard look at what I could come up with as a driver and I selected the 5670. (my thanks the the El Cheapo design that gave me the inspiration to look at this tube and thanks to Triode / Pentode Loadline Simulator v.1.0 (20161216 www.trioda.com) for some fatastic software to model behaviour and lay to rest some accepted design conventions ). (The often used 12AX7 has one triode slightly noisier than the other inside the same bulb, cannot remember which side, plus the two are not shielded from each other. In other words you should not use one side for one channel and the other side for the other channel and you have to know which one to use first in the signal path). It helps that the 5670 is still very cheap.
And to continue my rant: there are also differing views about a particular tube from the same manufacturer in the same equipment sounding different where the cause is only that the pins were not cleaned properly.
Pfew... a long post, I hope it is of some use to someone.
Some vague memory: In Push Pull the transformer needs a certain amount of current to go through it to magnetise the steel. In other words during crossover the transformer is non-linear and the bigger the core the worse it gets and it is necessary to use large amounts of NFB in a PP amplifier.
This must be why I never took a shine to PP amplifiers, especially of late. I often listen late at night at low levels next to the computer while the other members of the household are asleep.
I've got some papers on critical loudspeaker damping, spelling out that when the damping is not correct your bass response can suffer by a large amount, from memory it could be as much as four fold. The means that ideally one wants to have variable feedback which is really only possible with SE, and perhaps with triodes in PP. Clause Byrith designed such an EL34 SE amplifier.
Since most of the research is decades old and that speaker design has changed a lot this bass response impact may not be true today as much as in the past. But it does partly explain some of the heated discussions on why a certain amplifier works better with certain speakers and that certain speaker cables have a different response than other cables. Attached paper explains how it is possible to test, you'll have to search the link at the top of the paper to find the other two documents (there are three in total) as the others are too big to upload here.
And while we are on this whole OPT business: I have selected a 20% tap for my next SE UL transformer as it is close to the triode behavior and modelling suggest that it is close to what was described in David Saslaw's article (65% of pure pentode output). After I decided on the 20% tap I had a hard look at what I could come up with as a driver and I selected the 5670. (my thanks the the El Cheapo design that gave me the inspiration to look at this tube and thanks to Triode / Pentode Loadline Simulator v.1.0 (20161216 www.trioda.com) for some fatastic software to model behaviour and lay to rest some accepted design conventions ). (The often used 12AX7 has one triode slightly noisier than the other inside the same bulb, cannot remember which side, plus the two are not shielded from each other. In other words you should not use one side for one channel and the other side for the other channel and you have to know which one to use first in the signal path). It helps that the 5670 is still very cheap.
And to continue my rant: there are also differing views about a particular tube from the same manufacturer in the same equipment sounding different where the cause is only that the pins were not cleaned properly.
Pfew... a long post, I hope it is of some use to someone.
Attachments
With due respect: I do not know at what frequency this is but I do not like that spike / ringing at the start of the square wave. (This is assuming that the scope is properly tuned up and does not give phantom signals and that neither the source is creating it.)
how much of that is attributable to the OPT itself or/and the tube driving it?
Around the turn of the century I lived and worked in Hong Kong for a number of years. My employer warned me to be extreemly carefull when purchasing anything from a mainland Chinese as "they were not to be trusted".
Unfortunately I did buy a few amplifiers (after lots of research thinking I would have done enough due diligence) directly from China after I returned to New Zealand.
Today I had a I look on eBay and that it appears that matters have not hanged much, e.g. the labeling on this transformer states numbers that do not compute.
Unfortunately I did buy a few amplifiers (after lots of research thinking I would have done enough due diligence) directly from China after I returned to New Zealand.
Today I had a I look on eBay and that it appears that matters have not hanged much, e.g. the labeling on this transformer states numbers that do not compute.
Attachments
The amount of "easy" China(ese) bashing is becoming irritating on forums like these.
Be aware that welfare level in western countries are also possible "because of" how Chinese economy works, despite the criticism you might have.
How so does your transformer numbers do not compute?
Because adding the secondary powers does not match 150W?
It does not say a thing...
What does make sense is measuring the EI dimensions and stack, and it's lamination thickness, and check the power that a transformer with this core is able to deliver.
Checking the dimensions it seems to be an EI96.
Even with mediocre iron quality of 0.5 mm lamination thickness, a 50 mm stack could deliver 150 W.
When the transformer is wound with a lower T, and therefore "does" some 100 W secondary power, you even might have a very neat transformer.
Last edited:
i do mainly for personal experimenting and research..testing magnetic materials, differences C core, EI...one big winding company is close to my house and im visit them often.. unfortunately how it is now, shipping is acceptable only under European continent then s/h expenses can be kept low... then i can do winding-testing-shipping... abroad options are very expensive, because shipping and taxes will cost like another pair of transformers..
this is almost perfect 1 kHz square wave.. this spike is on Mhz spectrum and disappears in tube amplifier.. here it appeared because of hi amplitide low Z generator (50 Ohms).. in the amp you cant find this spike , high internal resistance of the tube rolls of this detail..
here is example of small ringing, some other OPT with C core. several kHz online
happening because of different insulation materials, not so perfect winding layers.. added parasitic inductance..
Last edited:
The discussion is interesting enough, however it is veering quite far from the topic, which is relatively affordable SE OPTs.
It is safe to assume that people understand the compromises that are made when building an OPT to a certain price point. If you need a 250w SE OPT that remains flat from 20Hz-20kHz then it is probably going to weigh 60kg and cost thousands of dollars.
Fine. That isn't really the point, however. We are mostly talking about single-ended amps that push out 5-6 watts in triode and maybe 8-10 in UL mode. Evidence suggests that many people are very happy with 1.5-2.5kg OPTs for this purpose, despite not being "perfect" down to 20Hz at 10 watts.
If someone is really into pipe organs or certain types of electronic music, I can understand them being picky about sub-35Hz response. Otherwise, a 3-4dB drop in that region basically doesn't matter, IMHO.
It is safe to assume that people understand the compromises that are made when building an OPT to a certain price point. If you need a 250w SE OPT that remains flat from 20Hz-20kHz then it is probably going to weigh 60kg and cost thousands of dollars.
Fine. That isn't really the point, however. We are mostly talking about single-ended amps that push out 5-6 watts in triode and maybe 8-10 in UL mode. Evidence suggests that many people are very happy with 1.5-2.5kg OPTs for this purpose, despite not being "perfect" down to 20Hz at 10 watts.
If someone is really into pipe organs or certain types of electronic music, I can understand them being picky about sub-35Hz response. Otherwise, a 3-4dB drop in that region basically doesn't matter, IMHO.
Last edited:
millions upon millions of cheap SE ot's were made for tv sets around the world...
the vertical output traffo is one such ot, as well as the sound output traffo,
you will be amazed at how these were made, i tore down lots of these, and they were constructed as one primary and one secondary coils built atop each other,
i never encountered any that were multi sectioned....
and yet, they sounded damned good with 6as5 and the like output tubes...
but yes, they were gapped, an onion skin type paper used as gapping material...
the vertical output traffo is one such ot, as well as the sound output traffo,
you will be amazed at how these were made, i tore down lots of these, and they were constructed as one primary and one secondary coils built atop each other,
i never encountered any that were multi sectioned....
and yet, they sounded damned good with 6as5 and the like output tubes...
but yes, they were gapped, an onion skin type paper used as gapping material...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.