hey guys,
i'm interested in using R-core transformers for power supplies (at least moderately-sized ones). i have heard some people say toroids have some drawbacks as power transformers (not so good isolation, more HF noise transmission?). EI cores have a lot of flux leakage and low efficiency. i know the flux leakage of R-cores is low (on the order of toroids?), but how are their other properties such as HF isolation and load regulation vs. toroids? is one type of transformer inherently better at rejecting RF noise, or is it all in the actual implemtation? i wonder how a good R-core "sounds" compared to a typical toroid like the Plitron units...
cheers,
dorkus
i'm interested in using R-core transformers for power supplies (at least moderately-sized ones). i have heard some people say toroids have some drawbacks as power transformers (not so good isolation, more HF noise transmission?). EI cores have a lot of flux leakage and low efficiency. i know the flux leakage of R-cores is low (on the order of toroids?), but how are their other properties such as HF isolation and load regulation vs. toroids? is one type of transformer inherently better at rejecting RF noise, or is it all in the actual implemtation? i wonder how a good R-core "sounds" compared to a typical toroid like the Plitron units...
cheers,
dorkus
R-core transformers for power supplies
They look very promising to me. Split bobbins have good high frequency isolation due to lack of interwinding capacitance. One
of biggest limitations of any type of transformer is high flux desity as result of the desire to get the most power from the smallest physical size transformer. This is particularly a problem for many torroid transformers.
http://www.custommag.com/CMIHOME.HTG/rcore2.htm
http://www.custommag.com/CMIHOME.HTG/rcore3.htm
http://www.custommag.com/CMIHOME.HTG/rcore4.htm
http://www.electroassemblies.com/r-core.htm
They look very promising to me. Split bobbins have good high frequency isolation due to lack of interwinding capacitance. One
of biggest limitations of any type of transformer is high flux desity as result of the desire to get the most power from the smallest physical size transformer. This is particularly a problem for many torroid transformers.
http://www.custommag.com/CMIHOME.HTG/rcore2.htm
http://www.custommag.com/CMIHOME.HTG/rcore3.htm
http://www.custommag.com/CMIHOME.HTG/rcore4.htm
http://www.electroassemblies.com/r-core.htm
A lot of it depends on who makes them. Most commercial winders use as little copper as possible. They will tell you that more will increase your IR losses, but what it really means that they are cutting costs anyway they can; and if it hums, then that is your problem.
So.....there is no easy answer, although the general principles you have stated are correct.
Jocko
So.....there is no easy answer, although the general principles you have stated are correct.
Jocko
hi fred,
yah, i was looking at that site (custommag.com) too. they espouse a lot of the benefits but didn't mention load regulation or HF isolation. i guess load regulation is dependent more on core size than anything else though? in any case i emailed their sales department to see if they stocked any standard values. i'm very interested in using them for a DVD/SACD player power supply that i'm rebuilding (i fried the stock one) and a little amp i'm building for a "personal" stereo.
hmm, what is the difference between a regular split-bobbin and "true" R-core? digikey carrys some low-profile transformers that are said to be split-bobbin... i wonder if they are better sounding power xformers than, say, the Aveco or whatever encapsulated toroidals (the popular ones in the blue molded case). they only come in low VA sizes but they should be adequate for digital or low-current line level circuits.
p.s. yeah jocko i agree, it all depends on the individual implementation... a bad xformer is a bad xformer no matter how you cut it. i wonder how the Custom Magnetics ones are. they're one of the few US suppliers i've found on the web so far. if correctly done though, i'm wondering if R's are preferrable to a typical toroid... maybe a listening test is in order.
yah, i was looking at that site (custommag.com) too. they espouse a lot of the benefits but didn't mention load regulation or HF isolation. i guess load regulation is dependent more on core size than anything else though? in any case i emailed their sales department to see if they stocked any standard values. i'm very interested in using them for a DVD/SACD player power supply that i'm rebuilding (i fried the stock one) and a little amp i'm building for a "personal" stereo.
hmm, what is the difference between a regular split-bobbin and "true" R-core? digikey carrys some low-profile transformers that are said to be split-bobbin... i wonder if they are better sounding power xformers than, say, the Aveco or whatever encapsulated toroidals (the popular ones in the blue molded case). they only come in low VA sizes but they should be adequate for digital or low-current line level circuits.
p.s. yeah jocko i agree, it all depends on the individual implementation... a bad xformer is a bad xformer no matter how you cut it. i wonder how the Custom Magnetics ones are. they're one of the few US suppliers i've found on the web so far. if correctly done though, i'm wondering if R's are preferrable to a typical toroid... maybe a listening test is in order.
May not answer your question, but.....
I've built lots of supplies using big toroids. I like them. I have also built supplies with those cheap split-bobbin types.
I have gotten good results with them by putting a big resistor in series with the filter cap. Yeah, it wastes power, but limiting the time the transfomer sees as something close to a dead short is a good thing. The transformer acts in a very non-linear element at that time.
Back in the days when we had choke input filters, we didn't have some of the problems we encounter today.
We also didn't listen as closely, either.
Jocko
I've built lots of supplies using big toroids. I like them. I have also built supplies with those cheap split-bobbin types.
I have gotten good results with them by putting a big resistor in series with the filter cap. Yeah, it wastes power, but limiting the time the transfomer sees as something close to a dead short is a good thing. The transformer acts in a very non-linear element at that time.
Back in the days when we had choke input filters, we didn't have some of the problems we encounter today.
We also didn't listen as closely, either.
Jocko
hi jocko,
yeah i've heard rumblings about how those good old-fashioned RC and LC filters had some very desirable properties. a few months ago i was all for choke regulation, was about to use it every place i could, but then when i started figuring out the resonance freqs, critical inductance, blah blah blah i got discouraged. (sonnya didn't help in that regard 😛) i basically accepted the fact that if the resonant frequency of the supply wasn't out of the audible range it would impart some coloration on the sound... getting a 10Hz or lower resonance from a reasonably-sized supply turned out to be non-trivial.
however i'm game for using a small-value series resistor to reduce the punishment on the transformer and rectifiers, at least for lower-power supplies where the power loss would not be so bad. doesn't it reduce the noise fed back into the AC line through the transformer by damping the supply lines as well?
yeah i've heard rumblings about how those good old-fashioned RC and LC filters had some very desirable properties. a few months ago i was all for choke regulation, was about to use it every place i could, but then when i started figuring out the resonance freqs, critical inductance, blah blah blah i got discouraged. (sonnya didn't help in that regard 😛) i basically accepted the fact that if the resonant frequency of the supply wasn't out of the audible range it would impart some coloration on the sound... getting a 10Hz or lower resonance from a reasonably-sized supply turned out to be non-trivial.
however i'm game for using a small-value series resistor to reduce the punishment on the transformer and rectifiers, at least for lower-power supplies where the power loss would not be so bad. doesn't it reduce the noise fed back into the AC line through the transformer by damping the supply lines as well?
you know...
people are always talking about minimizing impedance of the supply lines, so resistors are bad, very bad... but if you use enough supply capacitance won't the impedance of your supply essentially be the ESR of your filter caps? so who cares if the supply feeding it is not zero impedance, ideally it's just acting as a "charger" for the filter caps anyway. assumming you have enough capacitance to satisfy your current requirements, wouldn't a bit of resistance after the rectifiers be ok? i guess for a power amp that's a lot of capacitance but for a preamp it should be fine. ok maybe the regulation would suffer a tad too (good argument for chokes) but i think it would be acceptable, particularly if you're feeding regulators downstream anyway.
people are always talking about minimizing impedance of the supply lines, so resistors are bad, very bad... but if you use enough supply capacitance won't the impedance of your supply essentially be the ESR of your filter caps? so who cares if the supply feeding it is not zero impedance, ideally it's just acting as a "charger" for the filter caps anyway. assumming you have enough capacitance to satisfy your current requirements, wouldn't a bit of resistance after the rectifiers be ok? i guess for a power amp that's a lot of capacitance but for a preamp it should be fine. ok maybe the regulation would suffer a tad too (good argument for chokes) but i think it would be acceptable, particularly if you're feeding regulators downstream anyway.
Dorkus,
I have found that R-Cores can work very well, I used to pull them out of old Sony and Luxman equipment.
By the way what ever happend to Son of Dork? Did the design by committee idea kill it?
Jam
I have found that R-Cores can work very well, I used to pull them out of old Sony and Luxman equipment.
By the way what ever happend to Son of Dork? Did the design by committee idea kill it?
Jam
hi jam, long time no see. 🙂
Son of Dork is still alive but hibernating for the moment. but all the random stuff i've been posting about as of late (r-core, resistor comparisons, etc.) are building blocks. i'm also starting to re-think the design requirements of it based on how my system is evolving. for instance, i realized that i need something to switch my subwoofers between mains feed (for music) and dedicated LF input (for SACD or movies). i also need a pretty flexible triggering system so that it can work well with my surround sound processor with a minimum of fuss (otherwise my roommate will never be able to use it). i also still don't know what active stage to use. i was going to stick to Borbely JFET designs and custom shunt attenuators, but i'm starting to wonder if i should work on two different grades of modules to keep costs reasonable (the requirements for subwoofer and rear channels is different from mains/center). etc. etc. so, i am still thinking about stuff, but working out the simple building blocks first. once i pick what kind of resistors i like, i may start work on the attenuator, as well as audition the PGA2310 chip as a possibility for the rear/subwoofer boards. i'm also going to move on from my JFET buffer to the SE balanced line amp. i'll post something in the SoD threads soon. 🙂
Son of Dork is still alive but hibernating for the moment. but all the random stuff i've been posting about as of late (r-core, resistor comparisons, etc.) are building blocks. i'm also starting to re-think the design requirements of it based on how my system is evolving. for instance, i realized that i need something to switch my subwoofers between mains feed (for music) and dedicated LF input (for SACD or movies). i also need a pretty flexible triggering system so that it can work well with my surround sound processor with a minimum of fuss (otherwise my roommate will never be able to use it). i also still don't know what active stage to use. i was going to stick to Borbely JFET designs and custom shunt attenuators, but i'm starting to wonder if i should work on two different grades of modules to keep costs reasonable (the requirements for subwoofer and rear channels is different from mains/center). etc. etc. so, i am still thinking about stuff, but working out the simple building blocks first. once i pick what kind of resistors i like, i may start work on the attenuator, as well as audition the PGA2310 chip as a possibility for the rear/subwoofer boards. i'm also going to move on from my JFET buffer to the SE balanced line amp. i'll post something in the SoD threads soon. 🙂
I believe you have had an Audio Epiphany
thanx fred, actually i got inspiration for the "charger" idea from a DIY preamp article in Positive Feedback many years ago, written by one of the designers at Audio Note. not a fan of their tube gear they make but he had a lot of interesting ideas that challenged some of the conventional (misconceived?) notions in audio design with some pretty good common sense. so i can't really take credit for an epiphany, only for separating the signal from the noise. 😉
thanx fred, actually i got inspiration for the "charger" idea from a DIY preamp article in Positive Feedback many years ago, written by one of the designers at Audio Note. not a fan of their tube gear they make but he had a lot of interesting ideas that challenged some of the conventional (misconceived?) notions in audio design with some pretty good common sense. so i can't really take credit for an epiphany, only for separating the signal from the noise. 😉
PHRED AND PHOCKO.
XFORMERS,
Sorry to say so,to me it is not who makes them but how they're made and what they're made off.
A decent book or study on magnetix is very handy here.
I always liked the good old c-cores and ei with a good electrostatic screenie and if possible a center tap as well.
Toroids?Too much bandpass and I hate it when they buzzzzzzzzzz.
As with everything in all technologies there is good stuff and not so good stuff around.
Always try to design in a two stepped way:
Isolate the transformers from the rest of the world and isolate the psu from the transformers.
Expensivo but pays of in the long run.
Et voila Jocko,another two Eurocents dans ze pottieee.
Buona notte signori,😎
XFORMERS,
Sorry to say so,to me it is not who makes them but how they're made and what they're made off.
A decent book or study on magnetix is very handy here.
I always liked the good old c-cores and ei with a good electrostatic screenie and if possible a center tap as well.
Toroids?Too much bandpass and I hate it when they buzzzzzzzzzz.
As with everything in all technologies there is good stuff and not so good stuff around.
Always try to design in a two stepped way:
Isolate the transformers from the rest of the world and isolate the psu from the transformers.
Expensivo but pays of in the long run.
Et voila Jocko,another two Eurocents dans ze pottieee.
Buona notte signori,😎
Re: PHRED AND PHOCKO.
i think that's what jocko meant by who makes them... who makes them affects how they're made.
fdegrove said:
i think that's what jocko meant by who makes them... who makes them affects how they're made.
QUITE RIGHTLY.
Dorkus,
As with all good company you did get some added value in the process and hey this time it was free to boot!
Could be right,it probably was what the good chap meant.😉
Dorkus,
As with all good company you did get some added value in the process and hey this time it was free to boot!
Could be right,it probably was what the good chap meant.😉
I've used the R-cores in production, and also the Plitron toroidals.
First about the Plitrons, I didn't like them very much - they tended to hum and buzz too much, and I had a few units short-circuit internally and melt. I quickly switched to Ulveco toroidals, and these were better; quieter, and no self-destruction.
The R-cores were invented by a company called Kitamura Kiden, and they combine the benefits of split-bobbin designs - excellent isolation between primary and secondary windings, with lower levels of distortion (the cross section of an R-core is always circular and without edges, while split-bobbins have rectangular cross-sections). And after having used several hundreds of these R-core transformers, I haven't had a single failure.
The only downside with the R-cores is that their volumetric efficiency is clearly lower than it would be for a toroidal. For example, in my smaller Connoisseur 4.0 preamp, I am using one R-core transformer of approximately 200VA per channel. If I were to change to a toroidal, I could likely fit a 400VA unit in the same space.
And speaking of toroidals, as a special type of R-core, Kitamura Kiden also makes round cross-section toroidals. Kitamura originally didn't have toroidal winding equipment in-house, and so the winding work used to be farmed out to an outside subcontractor. I didn't use any of these versions, but they were used by a couple of other Japanese audio manufacturers, with seemingly acceptable results.
Subsequently, Kitamura acquired their own toroidal winding equipment and started offering toroidals that were built completely in-house. These I tried, but so far I haven't had acceptable results with them - in general buzz and hum has been higher than I would prefer, and the winding quality has not been so clean. But Kitamura has only had a few years experience in winding toroidals, and they are improving. I'll probably soon have them make some more prototypes for me, and if I like what I see, I'll probably use them in production.
If not, I'll try again next year.
regards, jonathan carr
First about the Plitrons, I didn't like them very much - they tended to hum and buzz too much, and I had a few units short-circuit internally and melt. I quickly switched to Ulveco toroidals, and these were better; quieter, and no self-destruction.
The R-cores were invented by a company called Kitamura Kiden, and they combine the benefits of split-bobbin designs - excellent isolation between primary and secondary windings, with lower levels of distortion (the cross section of an R-core is always circular and without edges, while split-bobbins have rectangular cross-sections). And after having used several hundreds of these R-core transformers, I haven't had a single failure.
The only downside with the R-cores is that their volumetric efficiency is clearly lower than it would be for a toroidal. For example, in my smaller Connoisseur 4.0 preamp, I am using one R-core transformer of approximately 200VA per channel. If I were to change to a toroidal, I could likely fit a 400VA unit in the same space.
And speaking of toroidals, as a special type of R-core, Kitamura Kiden also makes round cross-section toroidals. Kitamura originally didn't have toroidal winding equipment in-house, and so the winding work used to be farmed out to an outside subcontractor. I didn't use any of these versions, but they were used by a couple of other Japanese audio manufacturers, with seemingly acceptable results.
Subsequently, Kitamura acquired their own toroidal winding equipment and started offering toroidals that were built completely in-house. These I tried, but so far I haven't had acceptable results with them - in general buzz and hum has been higher than I would prefer, and the winding quality has not been so clean. But Kitamura has only had a few years experience in winding toroidals, and they are improving. I'll probably soon have them make some more prototypes for me, and if I like what I see, I'll probably use them in production.
If not, I'll try again next year.
regards, jonathan carr
haven't tried plitron units yet but i had a pair of custom toroidals wound for me by Toroid Corporation of Maryland and they seem pretty good... they're not totally silent, a bit of a buzz when charging up the supply or current is high but in normal operation they are very quiet. i forget the exact rating but i think it's around 300VA each, pretty beefy for the rating. wasn't cheap ($140 for the pair) but i think reasonable considering they were custom done and spec'ed out for me. the engineers are very helpful and gave me free design consultation.
R-CORES
Hi fellows,
We already had a few threads going regarding these xformers:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=5777&highlight=RCORE
http://www.diyaudio.com/forums/showthread.php?s=&threadid=3019&highlight=RCORE
And there is more on it if you care to do a search on the forum.
Cheers,😉
Hi fellows,
We already had a few threads going regarding these xformers:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=5777&highlight=RCORE
http://www.diyaudio.com/forums/showthread.php?s=&threadid=3019&highlight=RCORE
And there is more on it if you care to do a search on the forum.
Cheers,😉
hi frank,
yah i saw those threads but i still had some lingering questions. my main question was with the degree of isolation they provide, and it seems it is good. i was peeking in my Sony SACD changer today and i saw a Kitmura R-core used for the analog power supply. interestingly, i've found that this Sony doesn't benefit as much from power line conditioning as other devices i've tried. my editor also found that the top model SCD-XA777ES, which also has R-cores, did not like power line conditioning or fancy power cords - it sounded just fine plugged into the wall. i'm not sure if it has anything to do with the R-core but who knows. the engineering that goes into these players is quite good in some ways though... yes of course none of us would build a cd player like this but given the constraints of mass production and profit margins there are some nice design touches.
yah i saw those threads but i still had some lingering questions. my main question was with the degree of isolation they provide, and it seems it is good. i was peeking in my Sony SACD changer today and i saw a Kitmura R-core used for the analog power supply. interestingly, i've found that this Sony doesn't benefit as much from power line conditioning as other devices i've tried. my editor also found that the top model SCD-XA777ES, which also has R-cores, did not like power line conditioning or fancy power cords - it sounded just fine plugged into the wall. i'm not sure if it has anything to do with the R-core but who knows. the engineering that goes into these players is quite good in some ways though... yes of course none of us would build a cd player like this but given the constraints of mass production and profit margins there are some nice design touches.
R-CORES.
Hi Dorkus,
To me it does not necessarily have to be R-cores but it seems that that particular feature was singled out by Sony,and others to obtain exactly what you experience,i.e.good isolation from the vagaries of the powergrid.
Moreover I also think that those x-formers will not throw as much digital supersonic hash back into the mains.
That alone won't cut it of course but it helps.
Good xformers and psu isolation definitely help sonics.
I don't know if these R-cores are commercially available in the US or Europe so in the meantime I try to source wel made double C xformers with electrostatic shield and center tap.
To my ears a good transformer will almost be selfregulating and will filter out a lot of HF hash that rides the powerlines.
It therefore comes as no surprise that filtering the powerline in front of the player does not improve things.
Actually I wouldn't be surprised if it would degrade the sound,especially the micro,makro dynamic range.
Jonathan Carrs' findings seem to confirm these impressions and I'm sure he can also tell you that good powertransformers are a major ingedient to improved sonic performance.
These don't come cheap unfortunately.
Cheers,🙂
Hi Dorkus,
To me it does not necessarily have to be R-cores but it seems that that particular feature was singled out by Sony,and others to obtain exactly what you experience,i.e.good isolation from the vagaries of the powergrid.
Moreover I also think that those x-formers will not throw as much digital supersonic hash back into the mains.
That alone won't cut it of course but it helps.
Good xformers and psu isolation definitely help sonics.
I don't know if these R-cores are commercially available in the US or Europe so in the meantime I try to source wel made double C xformers with electrostatic shield and center tap.
To my ears a good transformer will almost be selfregulating and will filter out a lot of HF hash that rides the powerlines.
It therefore comes as no surprise that filtering the powerline in front of the player does not improve things.
Actually I wouldn't be surprised if it would degrade the sound,especially the micro,makro dynamic range.
Jonathan Carrs' findings seem to confirm these impressions and I'm sure he can also tell you that good powertransformers are a major ingedient to improved sonic performance.
These don't come cheap unfortunately.
Cheers,🙂
yeah, i can't stand having to use power line conditioning. it really shouldn't have as much effect as it does if the powers supplies were better designed. in the case of the Sony, yes some things about the supply are not what we would call optimal but i think they've made sound overall design choices which make upstream conditioning less necessary and maybe even detrimental.
in particular, i don't like isolation transformers. you are right, they mess up the dynamics. there's already a xformer in every one of my components, i dont' need another one to get in the way. that said, there is one isolation transformer that works extremely well - the Ensemble Isolink. it is for use with digital transport/dac only but the xformers look big enough for a power amp. i believe they custom wind their own EI cores, with heavy shielding and vibration isolation. this improved even the Sony CD player, it is an amazing product! all the benefits of power line filtering with NONE of the drawbacks, i could hear no coloration whatsoever. clearly a concept well-executed. unfortunately, it also costs $2800 for the dual version, but worth it if you can afford it in my opinion. this is no surplus transformer thrown into a metal case - it's all beautifully crafted custom work.
i think i can get R-cores here in the states from the company i mentioned early on. i emailed their sales department, we'll see if they write back next week.
in particular, i don't like isolation transformers. you are right, they mess up the dynamics. there's already a xformer in every one of my components, i dont' need another one to get in the way. that said, there is one isolation transformer that works extremely well - the Ensemble Isolink. it is for use with digital transport/dac only but the xformers look big enough for a power amp. i believe they custom wind their own EI cores, with heavy shielding and vibration isolation. this improved even the Sony CD player, it is an amazing product! all the benefits of power line filtering with NONE of the drawbacks, i could hear no coloration whatsoever. clearly a concept well-executed. unfortunately, it also costs $2800 for the dual version, but worth it if you can afford it in my opinion. this is no surplus transformer thrown into a metal case - it's all beautifully crafted custom work.
i think i can get R-cores here in the states from the company i mentioned early on. i emailed their sales department, we'll see if they write back next week.
- Status
- Not open for further replies.