Something that dates from the late 60's, originated for power transformers by Ordean Kilte and simultaneously by me for audio. In power transformers it allows the core to always set to zero when voltage is removed. This disallows turn on over currents in high voltage transformers and allows fusing for use rather than eventual catastrophic current draw. In OPT's the same construction techniques allow for a near complete self demagnetization of the core, just as in powers, and this defeats the polar saturation on the back half of the BH curve, before H crosses zero.
On an oscilloscope, looking at a square wave, there is always an initial S curve to the rise in B, as it is always coming from rest. Looks like a squiggle at zero crossing. If you look carefully you will see no discontinuity in this or a sinusoidal wave form when crossing zero H, well into core saturation. This means that low frequency impedance is very stable and all oscillation is self canceling. The leading and trailing edge of the square wave is always exactly square, with a tiny amount of high frequency oscillation, if the impedance (inductance) match is not ideal. As in using the same transformer for 300 B or 2A3 tubes. Not audible and not destructive, just a ripple in the top and bottom plane.
The sonic benefits are a removal of the typical hooded character found in PP OPT's and the elimination of non linear bass response in SE OPT's.
In power transformers, instantaneous joule throughput is remarkable. In a guitar amplifier, an iron set from us allows a tube amp to produce the power tubes maximum instantaneous power output, repeatedly. This without damage to the tube or transformers, Means a 50 watt amp is a stadium level beast, but the playability and sonic character are the same from whispers to waterfalls.
Just brings the E/I core into rough parity with amorphous core. You give up absolute clarity and cleanliness for a huge dollop of micro detail information and a slight softening to that absolute clarity and cleanliness.
Bud
On an oscilloscope, looking at a square wave, there is always an initial S curve to the rise in B, as it is always coming from rest. Looks like a squiggle at zero crossing. If you look carefully you will see no discontinuity in this or a sinusoidal wave form when crossing zero H, well into core saturation. This means that low frequency impedance is very stable and all oscillation is self canceling. The leading and trailing edge of the square wave is always exactly square, with a tiny amount of high frequency oscillation, if the impedance (inductance) match is not ideal. As in using the same transformer for 300 B or 2A3 tubes. Not audible and not destructive, just a ripple in the top and bottom plane.
The sonic benefits are a removal of the typical hooded character found in PP OPT's and the elimination of non linear bass response in SE OPT's.
In power transformers, instantaneous joule throughput is remarkable. In a guitar amplifier, an iron set from us allows a tube amp to produce the power tubes maximum instantaneous power output, repeatedly. This without damage to the tube or transformers, Means a 50 watt amp is a stadium level beast, but the playability and sonic character are the same from whispers to waterfalls.
Just brings the E/I core into rough parity with amorphous core. You give up absolute clarity and cleanliness for a huge dollop of micro detail information and a slight softening to that absolute clarity and cleanliness.
Bud
John,
I have no magic bullet for looking into the out of phase core behavior to point you to. Thom Thomas of Temple and I have been scratching our heads over this for 20 years.
Ordean eventually removed the references to this from the second edition of his design book, due to not being able to find a smoking gun. This, even though he had run several tests, with in excess of a million turn on moments for each, without ever generating an over current pulse at turn on.
The difference is immediately obvious in listening. However, except for a more sinuous B/H curve, when viewing the exciting current at no load, and a lack of crossover offset at H zero, there is really no way I know of to view what is really going on in the core, and only a dim view of what is happening in the window.
I did build a multi level coil many years ago, alternating multiples of one and two turn windings per layer, with single layer primaries, to allow me to look at the phase relationships within the core window. This was instructive in visualizing how the phase coherence is corrupted by magnetostrictive flux at the gaps, across the saturation scale.
These experiments eventually led me to designing a dielectric circuit in the coil, utilizing different dielectric constants to help strip out the effects of distributed capacitance and allow a stronger coupling capacitance at the antenna barricades. Those experiments led me to "voicing" OPTs' for electric guitars. The bulk of our signal production is used up by the guitar amplifier industry due to this. And, there are no financial trials to overcome.
Bud
I have no magic bullet for looking into the out of phase core behavior to point you to. Thom Thomas of Temple and I have been scratching our heads over this for 20 years.
Ordean eventually removed the references to this from the second edition of his design book, due to not being able to find a smoking gun. This, even though he had run several tests, with in excess of a million turn on moments for each, without ever generating an over current pulse at turn on.
The difference is immediately obvious in listening. However, except for a more sinuous B/H curve, when viewing the exciting current at no load, and a lack of crossover offset at H zero, there is really no way I know of to view what is really going on in the core, and only a dim view of what is happening in the window.
I did build a multi level coil many years ago, alternating multiples of one and two turn windings per layer, with single layer primaries, to allow me to look at the phase relationships within the core window. This was instructive in visualizing how the phase coherence is corrupted by magnetostrictive flux at the gaps, across the saturation scale.
These experiments eventually led me to designing a dielectric circuit in the coil, utilizing different dielectric constants to help strip out the effects of distributed capacitance and allow a stronger coupling capacitance at the antenna barricades. Those experiments led me to "voicing" OPTs' for electric guitars. The bulk of our signal production is used up by the guitar amplifier industry due to this. And, there are no financial trials to overcome.
Bud
@BudP -Posting #17
Hello Bud
thanks for the kind and wellspoken answer. Yes I agree and transformer design is really an "art"...
I think I've to try out one of your designs, but first I have to do other things first. Is it ok that I'll drop you a PM then?
And can you give me some more background of O-Netics, website etc.
Joao
Hello Bud
thanks for the kind and wellspoken answer. Yes I agree and transformer design is really an "art"...
I think I've to try out one of your designs, but first I have to do other things first. Is it ok that I'll drop you a PM then?
And can you give me some more background of O-Netics, website etc.
Joao
Joao,
O/Netics has no web site. It would be up to me to set it up and run it and I just do not have time. A friend has gotten a design to the point of mounting it, but I still have to provide a huge amount of data and it is far down the list of things to spend time on. You, or anyone, may PM me. We do a fair amount of custom design work and our price vs performance seems to please everyone.
LinuksGuru,
I must make some assumptions, as the nomenclature you provided is not an industry standard one. "orient Hi-B" almost certainly refers to grain oriented, M6 grade, silicon steel. There is an M3 grade, we use it on our level two and three OPT's, but it is only available through Thomas and Skinner here in the USA. In the Metric system, Z11 is equivalent to M6. I have attached a metric core data sheet just for grins and the relationship data is on the bottom. There is some question about the availability of M6 grade core in China. Seems you must either import it from Japan or be part of the military supply chain to obtain it, or even M19, which is a non grain oriented material that in the old days was referred to as "Audio Grade" core.
Bud
O/Netics has no web site. It would be up to me to set it up and run it and I just do not have time. A friend has gotten a design to the point of mounting it, but I still have to provide a huge amount of data and it is far down the list of things to spend time on. You, or anyone, may PM me. We do a fair amount of custom design work and our price vs performance seems to please everyone.
LinuksGuru,
I must make some assumptions, as the nomenclature you provided is not an industry standard one. "orient Hi-B" almost certainly refers to grain oriented, M6 grade, silicon steel. There is an M3 grade, we use it on our level two and three OPT's, but it is only available through Thomas and Skinner here in the USA. In the Metric system, Z11 is equivalent to M6. I have attached a metric core data sheet just for grins and the relationship data is on the bottom. There is some question about the availability of M6 grade core in China. Seems you must either import it from Japan or be part of the military supply chain to obtain it, or even M19, which is a non grain oriented material that in the old days was referred to as "Audio Grade" core.
Bud
I think this argument is like asking a civil engineer if steel truss or concrete bridges are best. Or like asking if Ceramic or Alnico is the best speaker magnet material.
I think that in all cases the details of the design matter more than the material used. You could make a good or poor design using either.
Then there is the issue of budget. Very, very few of use ever buy the "best" of anything. What most people look for is the best at a given price. So if you are looking for the best $100 part maybe using more of the cheaper steel sounds better.
I think that in all cases the details of the design matter more than the material used. You could make a good or poor design using either.
Then there is the issue of budget. Very, very few of use ever buy the "best" of anything. What most people look for is the best at a given price. So if you are looking for the best $100 part maybe using more of the cheaper steel sounds better.
ChrisA,
Absolutely correct. It will be very difficult to find amorphous core transformers at that $100 price. You can get our level one, up to 40 watts, for about 40% more than that. You can get many OPT's from many other mfg's, ones that have absolutely nothing to apologize for, and they will cost less than $100.
And really, until you EnABL your drivers, spending for more than our level one is not worth the money. I know, I know, everyone will disagree and point to their own system as more than worthy, but so long as you are missing that last 40 to 50 dB of downward dynamic range, as a clear and coherent signal source, out of your speakers, all you are going to get are hints of what the best OPT's can provide.
I am being quite serious here. And I will accept that correctly done Maggies, electrostatics, full ribbon systems and well done horns will help to explore that realm, and will show more than hints.
Bud
Absolutely correct. It will be very difficult to find amorphous core transformers at that $100 price. You can get our level one, up to 40 watts, for about 40% more than that. You can get many OPT's from many other mfg's, ones that have absolutely nothing to apologize for, and they will cost less than $100.
And really, until you EnABL your drivers, spending for more than our level one is not worth the money. I know, I know, everyone will disagree and point to their own system as more than worthy, but so long as you are missing that last 40 to 50 dB of downward dynamic range, as a clear and coherent signal source, out of your speakers, all you are going to get are hints of what the best OPT's can provide.
I am being quite serious here. And I will accept that correctly done Maggies, electrostatics, full ribbon systems and well done horns will help to explore that realm, and will show more than hints.
Bud
Hello Bud
in your attachment of your post #25 the shown cores are EI-cores. Are your transformers only EI-cores or can you provide "Double-C-cores" as well?
Next questions: do you manufacture only OPTs, what about Interstages?
Onto OPTs, and let's take one for 300B in single ended - that's the interest of most audiophiles...
Can you explain some details of your design like dc-resistance and inductance of the primary, frequency response etc...
Joao
in your attachment of your post #25 the shown cores are EI-cores. Are your transformers only EI-cores or can you provide "Double-C-cores" as well?
Next questions: do you manufacture only OPTs, what about Interstages?
Onto OPTs, and let's take one for 300B in single ended - that's the interest of most audiophiles...
Can you explain some details of your design like dc-resistance and inductance of the primary, frequency response etc...
Joao
Joao,
We currently only build E/I core. Economics dictate that you do one or the other and the price of C Cores means you are strictly limited in range, fewer core sizes to cover the power / inductance range. We have built C core transformers in the past and probably still have valid links with suppliers. However, buying a pair of core sets is VERY expensive and so you end up having to buy a fair number of them and so must then continue down that road.
I manufacture outputs, interstage, input splitter, plate chokes, power chokes and power transformers for tube amplifiers. The only "fixed" designs are what we provide for out Guitar amplifier DIY business. We also have a number of Guitar amplifier boutique manufacturers, and each of them has custom requirements.
I have only designed two outputs for the 300 B tube. One of them is push pull, for Lynn Olson's Karna and the other for Wright Sound's model 7, a single ended output that paid more attention to cost and physical size restraints than to ultimate performance.
The PP OPT in circuit is flat from 18 Hz to 40kHz and the phase is flat from 20 Hz to 35 kHz under load at full power (30 watts, though 40 was available), but I have no curves to hand. I may be able to obtain some from Gary Pimm, depending upon what he kept from that build.
The Wright sound OPT is not particularly great. Flat from 20 to 22 kHz with phase flat from 20 to 18 kHz. The major problem being not enough physical size ,so I had to settle for less than optimum performance. In a better optimized SE OPT we should see the top end extend out to 40 kHz as with the PP OPT. George never provided me with any loaded test data and I do not bother with unloaded data, as it is misleading at best. I will never know how this OPT really performs objectively, as George passed away recently and there are no amplifiers available to investigate. Subjectively it was very clear, detailed and with considerable internal gradient information. It also allowed that 300 B characteristic of displaying musical beauty to unfold.
I am planning a SE 300 B amplifier at this moment. It will use a solid state, push pull, two stage driver stage feeding a push pull to single ended nickle core interstage and then directly to the 300 B. The SS section will come from Gary Pimm's second generation SS Tabor amp and will have his CCS as the input splitter. Hopefully this will side step the problem of getting enough raw power to drive the 300 B without the noise and distortion inherent in every driver stage using tubes I have encountered. Gary's first generation SS Tabor can be found here.http://www.pimmlabs.com/
LinuksGuru
I will quote myself from earlier in the thread
The outputs and interstage devices I have had experience with were generally smaller in amorphous C core construction and a bit bigger in grain oriented C core construction than equivalent E/ cores. For power transformers, E/I core has the best performance for the $, toroid's the best space utilization for the performance and I know nothing about C Core power transformers.
Bud
We currently only build E/I core. Economics dictate that you do one or the other and the price of C Cores means you are strictly limited in range, fewer core sizes to cover the power / inductance range. We have built C core transformers in the past and probably still have valid links with suppliers. However, buying a pair of core sets is VERY expensive and so you end up having to buy a fair number of them and so must then continue down that road.
I manufacture outputs, interstage, input splitter, plate chokes, power chokes and power transformers for tube amplifiers. The only "fixed" designs are what we provide for out Guitar amplifier DIY business. We also have a number of Guitar amplifier boutique manufacturers, and each of them has custom requirements.
I have only designed two outputs for the 300 B tube. One of them is push pull, for Lynn Olson's Karna and the other for Wright Sound's model 7, a single ended output that paid more attention to cost and physical size restraints than to ultimate performance.
The PP OPT in circuit is flat from 18 Hz to 40kHz and the phase is flat from 20 Hz to 35 kHz under load at full power (30 watts, though 40 was available), but I have no curves to hand. I may be able to obtain some from Gary Pimm, depending upon what he kept from that build.
The Wright sound OPT is not particularly great. Flat from 20 to 22 kHz with phase flat from 20 to 18 kHz. The major problem being not enough physical size ,so I had to settle for less than optimum performance. In a better optimized SE OPT we should see the top end extend out to 40 kHz as with the PP OPT. George never provided me with any loaded test data and I do not bother with unloaded data, as it is misleading at best. I will never know how this OPT really performs objectively, as George passed away recently and there are no amplifiers available to investigate. Subjectively it was very clear, detailed and with considerable internal gradient information. It also allowed that 300 B characteristic of displaying musical beauty to unfold.
I am planning a SE 300 B amplifier at this moment. It will use a solid state, push pull, two stage driver stage feeding a push pull to single ended nickle core interstage and then directly to the 300 B. The SS section will come from Gary Pimm's second generation SS Tabor amp and will have his CCS as the input splitter. Hopefully this will side step the problem of getting enough raw power to drive the 300 B without the noise and distortion inherent in every driver stage using tubes I have encountered. Gary's first generation SS Tabor can be found here.http://www.pimmlabs.com/
LinuksGuru
I will quote myself from earlier in the thread
The outputs and interstage devices I have had experience with were generally smaller in amorphous C core construction and a bit bigger in grain oriented C core construction than equivalent E/ cores. For power transformers, E/I core has the best performance for the $, toroid's the best space utilization for the performance and I know nothing about C Core power transformers.
Bud
Dave - you're certainly welcome to test them, and Bud, some of the modifications you suggested (along with a few minor others) have been performed on these.
Yes, it's a great shame that George never embraced the use of higher sonic value components, or at least allowed space in his chassis for those who do.
Yes, it's a great shame that George never embraced the use of higher sonic value components, or at least allowed space in his chassis for those who do.
Looking For Body & Weight Of Tone
Really interesting thread.
I'm curious which of the material types mentioned will lend more weight or body to an instrument's sound/tone? I'd like a bit more weight in the sound of my system and am willing to try getting it and possibly other gains by upgrading output transformers. Any ideas?
Thanks,
Really interesting thread.
I'm curious which of the material types mentioned will lend more weight or body to an instrument's sound/tone? I'd like a bit more weight in the sound of my system and am willing to try getting it and possibly other gains by upgrading output transformers. Any ideas?
Thanks,
TV Man,
I would need to know a bit more about your system and what you mean by weight and tone, before offering my opinion.
In a general sense, a properly designed and built commercial core OPT, using M3 and treating differential dielectrics within the coil structure as an E Field lens, as opposed to amorphous core construction, which will have slightly better distortion performance but be somewhat lacking in gradient information, for the internal structure of notes and transients, might be a better choice for you. However, a lot depends upon your speakers and your definition of weight. The two types will be roughly comparable in cost, at about $1000 per pair.
At the next level down, they will again be roughly comparable in cost but the above differences will be more noticeable and the cost around $500 per pair.
The last, but not least level will have very few amorphous core contenders and the commercial core has a noticeable advantage in information retention, with a wider gap relation in distortion. Cost here will be around $300 per pair.
Below this point the commercial core transformers reign, but their construction details begin to erase the level of gradient detail the above transformers attain.and though smooth and melodious, they are not going to allow you to hear what it sounds like you are after. Costs here run from $120 per pair up to $200 per pair.
Bud
I would need to know a bit more about your system and what you mean by weight and tone, before offering my opinion.
In a general sense, a properly designed and built commercial core OPT, using M3 and treating differential dielectrics within the coil structure as an E Field lens, as opposed to amorphous core construction, which will have slightly better distortion performance but be somewhat lacking in gradient information, for the internal structure of notes and transients, might be a better choice for you. However, a lot depends upon your speakers and your definition of weight. The two types will be roughly comparable in cost, at about $1000 per pair.
At the next level down, they will again be roughly comparable in cost but the above differences will be more noticeable and the cost around $500 per pair.
The last, but not least level will have very few amorphous core contenders and the commercial core has a noticeable advantage in information retention, with a wider gap relation in distortion. Cost here will be around $300 per pair.
Below this point the commercial core transformers reign, but their construction details begin to erase the level of gradient detail the above transformers attain.and though smooth and melodious, they are not going to allow you to hear what it sounds like you are after. Costs here run from $120 per pair up to $200 per pair.
Bud
Agent 5,
Nickle core is limited in the sizes available in E/I core construction. Tape / cut core nickle can be very good, but there are still some niggles with what level of coupling can be allowed in the coil structure, before Q sets in and you have both rising phase and frequency response, to a peak, at about 30 Khz. E/I M3, used in interstages, is also under this threat.
Bud
Nickle core is limited in the sizes available in E/I core construction. Tape / cut core nickle can be very good, but there are still some niggles with what level of coupling can be allowed in the coil structure, before Q sets in and you have both rising phase and frequency response, to a peak, at about 30 Khz. E/I M3, used in interstages, is also under this threat.
Bud
Links Guru,
Sorry, I missed your post. I no longer know what the various nomenclature refers to in performance with amorphous core. You might ask Dave Slagle at Intact Audio on hid forum. He is a current advocate of all types of C Core construction and may be current with what is being offered in the market place. Alloys do seem to come and go in that material as it's usage changes. It is also possible Per Lundahl at Lundahl would respond. They have their cores wound for them and I am sure have only a couple of materials they have settled on.
Bud
Sorry, I missed your post. I no longer know what the various nomenclature refers to in performance with amorphous core. You might ask Dave Slagle at Intact Audio on hid forum. He is a current advocate of all types of C Core construction and may be current with what is being offered in the market place. Alloys do seem to come and go in that material as it's usage changes. It is also possible Per Lundahl at Lundahl would respond. They have their cores wound for them and I am sure have only a couple of materials they have settled on.
Bud