I am getting the idea that if you build your own transformers, then you can make better amps because you can give the tubes, etc the exact required voltages.
But where to get less that 10 tons of laminates at a time?
(plus, and ideas about ferrite cores having the same magnetism, or permeability, as fe/si, please post because ferrite seems to be more available)
But where to get less that 10 tons of laminates at a time?
(plus, and ideas about ferrite cores having the same magnetism, or permeability, as fe/si, please post because ferrite seems to be more available)
Last edited:
it isn't that hard to find smaller regional manufacturers that do custom - I'd much rather pay them as little as 2-3x the standard part cost in many cases for exactly what you want
the biggest issue in power supply transformers is that “standard” parts push Bmax – the least iron, copper they can ship means the most profit
the biggest issue in power supply transformers is that “standard” parts push Bmax – the least iron, copper they can ship means the most profit
Agree with Nigelwright, but add that winding your own components is a complex undertaking, and requires a lot of skill & patience. Even so, you may do a lot of work and still not end up with a better wound component as Nigel has said. You may however end up with an outstandingly good transformer, but study, many repetitions of building components and experience will be the best teacher. I would never discourage someone from trying.
I am not sure where to locate larger ferrites, but you can try:
Thomas and Skinner
for laminations. T&S have sold me a few pounds at a time in the past.
Good luck
Terry
I am not sure where to locate larger ferrites, but you can try:
Thomas and Skinner
for laminations. T&S have sold me a few pounds at a time in the past.
Good luck
Terry
these look good -- but not E-cores, they are C-cores -- how do you wind a C-core, will it work? I have never seen an amp w/ C-cores.
Looks like a (socialist) government will have to mandate E-cores...
An externally hosted image should be here but it was not working when we last tested it.
Looks like a (socialist) government will have to mandate E-cores...
First you wind your turns on the bobbins. These can be mounted to a motor for easier winding just be sure the wire feed spool is 30 times the bobbin length away for smooth winding. Then after the bobbins are wound you glue the cores to each other. (If i have to tell you to place the bobbins on before you glue the cores, you might want to reconsider the project!)
Johan Potgieter wrote in 2005:
I have been using C-core output transformers extensively for tube amps over the decades. A rough comparison to E-cores is that they can handle about 1.5 times the flux and at lower distortion. In between there are of course grain oriented E-I cores, where the grain lies in the direction of the major legs. This gives somewhat better results than ordinary E-I, but it depends on the type of steel.
In RSA we can get C-cores of M5 type material, with permeability graphs available which are a great help. I design my own transformers according to procedures outlined in "Radio Designer's Handbook" by F Langford-Smith, using data from Crowhurst. I have found these predictable. The Bartolucci products look fine, although I would have expected at least the leakage reactance specs also (hope I did not overlook them!). The equivalent capacitance is also important - this is sometimes rather neglected. (In practice the equivalent capacitance can easily be the significant limitation to high frequency response rather than the more generally quoted leakage reactance.)
Yes, it is advisable to fill the whole winding window to get maximum windings with lowest copper loss. It is difficult to give more specific advice here; I would recommend using them though - locally they are not much more expensive than ordinary cores. I have unfortunately never been able to measure specific distortion or find data on that.
I have been using C-core output transformers extensively for tube amps over the decades. A rough comparison to E-cores is that they can handle about 1.5 times the flux and at lower distortion. In between there are of course grain oriented E-I cores, where the grain lies in the direction of the major legs. This gives somewhat better results than ordinary E-I, but it depends on the type of steel.
In RSA we can get C-cores of M5 type material, with permeability graphs available which are a great help. I design my own transformers according to procedures outlined in "Radio Designer's Handbook" by F Langford-Smith, using data from Crowhurst. I have found these predictable. The Bartolucci products look fine, although I would have expected at least the leakage reactance specs also (hope I did not overlook them!). The equivalent capacitance is also important - this is sometimes rather neglected. (In practice the equivalent capacitance can easily be the significant limitation to high frequency response rather than the more generally quoted leakage reactance.)
Yes, it is advisable to fill the whole winding window to get maximum windings with lowest copper loss. It is difficult to give more specific advice here; I would recommend using them though - locally they are not much more expensive than ordinary cores. I have unfortunately never been able to measure specific distortion or find data on that.
I wish I had these Columbians living in my apartment complex!
Click for youtube, great vid -- fun pronunciations
Oops sorry, the have laminates, but it would still be nice to learn if ferrite has improved to match fe/si
An externally hosted image should be here but it was not working when we last tested it.
[was]
Thanks, but those are ferrite cores, and I cannot find supporting material saying that they have the permeability (magnetism) of si/fe. If ferrite does, then it is some new ferrite, and not something that we would find in the surplus category.
Looks to me from the alibaba site that China is the place -- my FB friends are 90% in China (but mostly euro) and my best friends own a buffet in Connecticut (mmmm), but I am stuck here in Canada for the moment.
Last edited:
From lenardaudio.com:
The C core is approx 20% more efficient compared to the best quality grain orientated E I laminations. The square shape of the E I laminations is not an efficient conductor of magnetism. A % of the induced magnetism is spewed outside of the core and lost. A further % is lost as eddy currents within the laminations which causes the laminations to get hot. E I mains power trannys are noticeably hot when touched. C cores are an efficient conductor of magnetism, but C cores do not have an aesthetic square shape and are more difficult to mount onto a chassis. C cores are mostly used in industrial electronics and were approx twice the cost of E I laminations. Today, C cores are only slightly more expensive than E I laminations and it makes no sense not to use the superior quality C core in all output trannys, including power trannys. [not to be confused with gender-benders]
I guess I would use hose clamps -- I always do, when tape is inappropriate
The C core is approx 20% more efficient compared to the best quality grain orientated E I laminations. The square shape of the E I laminations is not an efficient conductor of magnetism. A % of the induced magnetism is spewed outside of the core and lost. A further % is lost as eddy currents within the laminations which causes the laminations to get hot. E I mains power trannys are noticeably hot when touched. C cores are an efficient conductor of magnetism, but C cores do not have an aesthetic square shape and are more difficult to mount onto a chassis. C cores are mostly used in industrial electronics and were approx twice the cost of E I laminations. Today, C cores are only slightly more expensive than E I laminations and it makes no sense not to use the superior quality C core in all output trannys, including power trannys. [not to be confused with gender-benders]
I guess I would use hose clamps -- I always do, when tape is inappropriate
all technology is mature -- think of the difference windows made, or the cooking pot, think about it
back to hifi, I put a lot of effort into a top-down understanding of amps because amp knowledge seems to be empirical, meaning experimental. everybody experimented with independent regions and then experimented to glue them together, and this before component technology
my thought is to develop a top-down view, to actually abstract the functions, which are both AC and DC, and involved what I am conceive of as the path of least resistance of frequency developed at one end (needle) resulting in frequency at the other end (speakers) with frequency all through out (waves) -- it is a signal however (particle), that appears to follow paths of least resistance. resistor resistance is tuned the power from the transformer and the needs of the tubes, etc. so, if the voltage can be tuned at the transformer there is less need of resistor tuning. the big companies may have done this (in the 50s- 60s), but that cannot be done now as voltages are off the self. even so, since it is all empirical, then the circuits may have been designed for off the self transformers and the requirements fed to the engineers who then ordered transformers to match the experimental circuits with the same off the shelf voltages -- the art was progressing right to the very end with the introduction of the transistor
back to hifi, I put a lot of effort into a top-down understanding of amps because amp knowledge seems to be empirical, meaning experimental. everybody experimented with independent regions and then experimented to glue them together, and this before component technology
my thought is to develop a top-down view, to actually abstract the functions, which are both AC and DC, and involved what I am conceive of as the path of least resistance of frequency developed at one end (needle) resulting in frequency at the other end (speakers) with frequency all through out (waves) -- it is a signal however (particle), that appears to follow paths of least resistance. resistor resistance is tuned the power from the transformer and the needs of the tubes, etc. so, if the voltage can be tuned at the transformer there is less need of resistor tuning. the big companies may have done this (in the 50s- 60s), but that cannot be done now as voltages are off the self. even so, since it is all empirical, then the circuits may have been designed for off the self transformers and the requirements fed to the engineers who then ordered transformers to match the experimental circuits with the same off the shelf voltages -- the art was progressing right to the very end with the introduction of the transistor
lenardaudio.com gives a pretty good case for building the transformers, and notes that they are the most expensive part of the whole tube amp.
The other page I looked at there shows a "best practice:" push pull pre-amp, with two layers of triodes, and an ultra-linear driver with a bias circuit in front of it.
To make the driver transformer, you have to tune the windings carefully, which means that you will not get the transformer off the self. If you want/need to substitute other parts, then further customization is necessary.
He makes many of the points that I did in my writing on wikiversity (that I had to halt because of a "maladaptive" sock-puppet problem there -- that you would not believe). One is that the "academic" stuff is just experimental (what I call empirical) and that experiments can't be glued together -- you have to have a top-down conception of what is going on (what constructivists call constructed knowledge and I think of as wiki structured). He quotes Albert Einstein
Here is the page I referenced (click)
The other page I looked at there shows a "best practice:" push pull pre-amp, with two layers of triodes, and an ultra-linear driver with a bias circuit in front of it.
To make the driver transformer, you have to tune the windings carefully, which means that you will not get the transformer off the self. If you want/need to substitute other parts, then further customization is necessary.
He makes many of the points that I did in my writing on wikiversity (that I had to halt because of a "maladaptive" sock-puppet problem there -- that you would not believe). One is that the "academic" stuff is just experimental (what I call empirical) and that experiments can't be glued together -- you have to have a top-down conception of what is going on (what constructivists call constructed knowledge and I think of as wiki structured). He quotes Albert Einstein
Here is the page I referenced (click)
Last edited:
I just want to add that the above constructive ideas apply to psychology as well -- I am finishing my masters. Experimental psychology (which is empirical and academic) it the most bogus thing imaginable, and doesn't just not work -- it hurts a lot of people.
If people correctly abstracted ideas like God/evolution meant them to do, then this would be a far better world!
If people correctly abstracted ideas like God/evolution meant them to do, then this would be a far better world!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.