I would agree. This is about output transformers, not simply where to buy the cheapest.
I am obliged for much information given, some of it refreshing the old grey cells, but much an update from those more in the practice than I am.
Regarding l.f. vs h.f. aspects, I am depressed to learn that losses start clouding the picture as low as you say, Bud. Not to go off at a tangent too far, but in principle, somewhere the 'own' inductance of a coil system must start coming in (meaning Zl is then rather from the inductance of the core-less coil, although intuitively I will not be surprised if it is at super-audio frequencies!). I will have to think out that Zl will drop so fast that in effect it will drop below the load resistance again at a sufficiently high frequency.
So out again with a transformer in the morning, putting it as a load on my amplifier to look at the response and phase angle over an audio signal sweep. I cannot do too accurate measurements, but with some care and good judgement a picture should emerge. (High time I get myself good spectrum analyzer stuff.)
I have never pictured having to see this as an antenna-to-antenna problem; usually inter-sectional capacitance is kept low to avoid h.f. attenuation.
Whatever, thanks again for insight from members.
LinuksGuru,
I agree; for serious work no scrap-yard stuff! Over here though, as said before, we are fortunate to have transformer iron stockists, who will even punch laminations in non-standard patterns. Just for the record, C-cores with M5 steel are normally stocked, but can be had with M4, M2H and M0H (at a price, naturally!)
I am obliged for much information given, some of it refreshing the old grey cells, but much an update from those more in the practice than I am.
Regarding l.f. vs h.f. aspects, I am depressed to learn that losses start clouding the picture as low as you say, Bud. Not to go off at a tangent too far, but in principle, somewhere the 'own' inductance of a coil system must start coming in (meaning Zl is then rather from the inductance of the core-less coil, although intuitively I will not be surprised if it is at super-audio frequencies!). I will have to think out that Zl will drop so fast that in effect it will drop below the load resistance again at a sufficiently high frequency.
So out again with a transformer in the morning, putting it as a load on my amplifier to look at the response and phase angle over an audio signal sweep. I cannot do too accurate measurements, but with some care and good judgement a picture should emerge. (High time I get myself good spectrum analyzer stuff.)
I have never pictured having to see this as an antenna-to-antenna problem; usually inter-sectional capacitance is kept low to avoid h.f. attenuation.
Whatever, thanks again for insight from members.
LinuksGuru,
I agree; for serious work no scrap-yard stuff! Over here though, as said before, we are fortunate to have transformer iron stockists, who will even punch laminations in non-standard patterns. Just for the record, C-cores with M5 steel are normally stocked, but can be had with M4, M2H and M0H (at a price, naturally!)
jrenkin,
Yes these are O-Netics transformers, but the ones found at C3 Amps and Rhodes are only for guitar amps. Anyone can PM me, I will shift you over to my private email when you do, but a PM is always welcome.
As for differences..... hmmm.... Our level one OPT's are likely to be 3 to 4 times the amount of retained signal information, over anything you have heard. Their FR and phase are flat from 20 to 40kHz for both PP and SE at any power level, including up to hard saturation at 20Hz. They will not oscillate and their impedance curve down in the sub sonic realm is very flat, while phase and frequency response roll off very smoothly. Compared to say, a Lundahl 1620, they have slightly but noticeably more distortion, are not as dynamic in transient response, are roughly as clear in tonality and timbre but will have more internal information to note and transient structures. Of the two, I like the 1620 slightly more.
When O-Netics level 3 is in the picture, the low level coherent signals are completely retained, so hall resonances, deeply buried musical signals etc are made clearly apparent. They have another 2 times the information content of level one and provide a complete ease to all sounds, regardless of their complexity or amplitude. Compared again to a 1620 they are slightly softer in presentation, just as clear tonally and transiently, are as vivid, but do leave the Lundahls sounding faintly cartoon like when it comes to deeply complex music, such as Cantate Domino on XRCD or a 24/192 rendering of Carmina Burana, from a live stage production. I wont go into tape or LP's here, but you can expect the same differences with those sources.
These comparisons come from Gary Pimm's system, using his second generation Tabor SS amp, with plug and play interstage and output transformer connections. None of this should be taken as faulting the Lundahl 1620, these are fabulous output transformers, Lundahl and I have just taken two different paths and thus offer different strengths.
Please note that these are very small differences and are really only available from speakers that can exceed the typical downward dynamic limit pointed to by Meyer Sound, of about 40 db below average signal levels, typical with modern dynamic speakers. Ribbon based systems and CD / LeCleach horn systems will provide greater downward dynamic range, as will any EnABL'd dynamic speaker system.
My previous statement about purchasing less revealing transformers still stands. Unless you really do have a very revealing speaker system (and of course we all are certain we do!) there is no real reason to spend from $150 to $600 per output transformer. Even if an exquisite level of signal retention and expression into your listening room is your eventual intention, I would still suggest you utilize under $100 apiece OPT's for all of your development and system proofing work, including hum and noise reduction. Then, spend your money on high resolution speakers, output transformers, speaker cables and signal sources, in that order.
Bud
Yes these are O-Netics transformers, but the ones found at C3 Amps and Rhodes are only for guitar amps. Anyone can PM me, I will shift you over to my private email when you do, but a PM is always welcome.
As for differences..... hmmm.... Our level one OPT's are likely to be 3 to 4 times the amount of retained signal information, over anything you have heard. Their FR and phase are flat from 20 to 40kHz for both PP and SE at any power level, including up to hard saturation at 20Hz. They will not oscillate and their impedance curve down in the sub sonic realm is very flat, while phase and frequency response roll off very smoothly. Compared to say, a Lundahl 1620, they have slightly but noticeably more distortion, are not as dynamic in transient response, are roughly as clear in tonality and timbre but will have more internal information to note and transient structures. Of the two, I like the 1620 slightly more.
When O-Netics level 3 is in the picture, the low level coherent signals are completely retained, so hall resonances, deeply buried musical signals etc are made clearly apparent. They have another 2 times the information content of level one and provide a complete ease to all sounds, regardless of their complexity or amplitude. Compared again to a 1620 they are slightly softer in presentation, just as clear tonally and transiently, are as vivid, but do leave the Lundahls sounding faintly cartoon like when it comes to deeply complex music, such as Cantate Domino on XRCD or a 24/192 rendering of Carmina Burana, from a live stage production. I wont go into tape or LP's here, but you can expect the same differences with those sources.
These comparisons come from Gary Pimm's system, using his second generation Tabor SS amp, with plug and play interstage and output transformer connections. None of this should be taken as faulting the Lundahl 1620, these are fabulous output transformers, Lundahl and I have just taken two different paths and thus offer different strengths.
Please note that these are very small differences and are really only available from speakers that can exceed the typical downward dynamic limit pointed to by Meyer Sound, of about 40 db below average signal levels, typical with modern dynamic speakers. Ribbon based systems and CD / LeCleach horn systems will provide greater downward dynamic range, as will any EnABL'd dynamic speaker system.
My previous statement about purchasing less revealing transformers still stands. Unless you really do have a very revealing speaker system (and of course we all are certain we do!) there is no real reason to spend from $150 to $600 per output transformer. Even if an exquisite level of signal retention and expression into your listening room is your eventual intention, I would still suggest you utilize under $100 apiece OPT's for all of your development and system proofing work, including hum and noise reduction. Then, spend your money on high resolution speakers, output transformers, speaker cables and signal sources, in that order.
Bud
There has been some discussion on using OPTs at other than designed to primary impedances by switching the load to different taps in this thread:
http://www.diyaudio.com/forums/tubes-valves/161015-parallel-tubes-2.html
For example I just got a set of Edcor 30W 5K PP transformers which I plan to use with a parallel PP 6P1P amp. The transformers are rated at 30W, however the tubes in PPP will probably deliver less than 20W so there should be no issues in this configuration.
However, I plan to remove half the tubes which will run 10K p-p instead of 5K p-p, and to compensate I will run the 8 ohm load on the 4 ohm tap to maintain 10K matching.
What will be the ramification of doing this?
http://www.diyaudio.com/forums/tubes-valves/161015-parallel-tubes-2.html
For example I just got a set of Edcor 30W 5K PP transformers which I plan to use with a parallel PP 6P1P amp. The transformers are rated at 30W, however the tubes in PPP will probably deliver less than 20W so there should be no issues in this configuration.
However, I plan to remove half the tubes which will run 10K p-p instead of 5K p-p, and to compensate I will run the 8 ohm load on the 4 ohm tap to maintain 10K matching.
What will be the ramification of doing this?
The only thing that could suffer is your half power point in the bass. It might rise from 20 to 40 Hz and you would have a bit more distortion at 20 Hz, but not likely twice as much. So, just barely noticeable unless you are an organ fan.
However Edcore may have provided more inductance, out in the actual operating realm than was actually needed to provide a half of a db down at 20 Hz. If so, these cautions are meaningless. If you know the inductance at 1 vac & 120 Hz you can typically multiply it by 10 to obtain the working inductance, out at minimum volume and above. Take that number and multiply it by 2 times Pi and 20 to obtain the load impedance you have enough inductance to match for -3 db at 20 Hz. For -0.5 db divide that number by 2.76.
Bud
However Edcore may have provided more inductance, out in the actual operating realm than was actually needed to provide a half of a db down at 20 Hz. If so, these cautions are meaningless. If you know the inductance at 1 vac & 120 Hz you can typically multiply it by 10 to obtain the working inductance, out at minimum volume and above. Take that number and multiply it by 2 times Pi and 20 to obtain the load impedance you have enough inductance to match for -3 db at 20 Hz. For -0.5 db divide that number by 2.76.
Bud
Silver wire ???
Does it makes any sense to build windings of PP output trafo with silver-coated wire? Secondary DC resistance is typically around 0.15 Ohms, primary - 100 - 150 Ohm, so losses are about 3 - 7% only.
So far I have not seen any serious engineering book outlining sonic benefits of using silver wires in output transformers of audio amplifiers.
It is clear why silver wire is being used in step-up transformers for example, but output ones is a different story.
Any feedback is greatly appreciated. Please mention reliable scientific/engineering source of information posted.
Does it makes any sense to build windings of PP output trafo with silver-coated wire? Secondary DC resistance is typically around 0.15 Ohms, primary - 100 - 150 Ohm, so losses are about 3 - 7% only.
So far I have not seen any serious engineering book outlining sonic benefits of using silver wires in output transformers of audio amplifiers.
It is clear why silver wire is being used in step-up transformers for example, but output ones is a different story.
Any feedback is greatly appreciated. Please mention reliable scientific/engineering source of information posted.
munroc,
For winding forms, known as bobbins, you might try Ram Sales in ca.. they have a web presence. They may also be your only hope of getting small quantities of core material, as the major mills have increased their minimum orders to 20 to 30 pounds of one size or grade. This equates to nine 18" long rows of a 3/8 " center tongue width core or two 9 inch rows of a 1 1/2" center tongue width core. Read this as expensive. EIS inc, also on the web, handles the various hold down tapes rated for transformer use, dielectric materials in sheet form, varnishes and magnet wire in 10 pound spools.
End product of all of this is that you will end up with more material than you can use, of wire or core, and so you might as well buy too many bobbins also.
LinksGuru,
To my almost certain knowledge there is no objective reason to use insulated silver winding wire in place of 4 nines commercial copper wire. Both are protected from surface contamination and both will provide equivalent performance in any form of objective test you may care to utilize.
However, if you are looking to judge by matters of sonic taste, then silver wire has a slight advantage in the realization of coherent micro dynamic information. To utilize this you must have superior grades of everything down stream from the transformer and the rest of the transformer should also have high perm material in the core and high permittivity dielectrics and a useful dielectric circuit, in the coil. For all of that, you will gain about 10% better resolution of those micro dynamic signals. This % of change is strictly an anecdotal experience upon my part, with absolutely no validity beyond that point.
Bud
For winding forms, known as bobbins, you might try Ram Sales in ca.. they have a web presence. They may also be your only hope of getting small quantities of core material, as the major mills have increased their minimum orders to 20 to 30 pounds of one size or grade. This equates to nine 18" long rows of a 3/8 " center tongue width core or two 9 inch rows of a 1 1/2" center tongue width core. Read this as expensive. EIS inc, also on the web, handles the various hold down tapes rated for transformer use, dielectric materials in sheet form, varnishes and magnet wire in 10 pound spools.
End product of all of this is that you will end up with more material than you can use, of wire or core, and so you might as well buy too many bobbins also.
LinksGuru,
To my almost certain knowledge there is no objective reason to use insulated silver winding wire in place of 4 nines commercial copper wire. Both are protected from surface contamination and both will provide equivalent performance in any form of objective test you may care to utilize.
However, if you are looking to judge by matters of sonic taste, then silver wire has a slight advantage in the realization of coherent micro dynamic information. To utilize this you must have superior grades of everything down stream from the transformer and the rest of the transformer should also have high perm material in the core and high permittivity dielectrics and a useful dielectric circuit, in the coil. For all of that, you will gain about 10% better resolution of those micro dynamic signals. This % of change is strictly an anecdotal experience upon my part, with absolutely no validity beyond that point.
Bud
Last edited:
Thanks, Bud, I suspected this from the beginning - silver wire transformer is a marketing tool. But I'm was keen to get an evidence from industry professional.
May I ask why do you want to do this? Good winding equipment cost a plenty of money, high-quality DIY transformer design required digging arcane material in several academic-level engineering books 400 - 700 pages each, + ability to be a decent-level programmer.
If this is for fun only its OK, but if you are interested in good result IMHO better to order from factory or small specialized firms like Electra-Print, Lundahl, or Bud if he has something in the price range you would like to pay for.
LinuksGuru,
cut the ** will you!
It's the second time I see you standing on a pedestal frowning at people asking 'noob' questions by being counter-product and highlighting just the negative aspects as you did in my thread (http://www.diyaudio.com/forums/tubes-valves/157658-bare-r-cores-bobbins-where-2.html#post2037043), and oh.. programming, now step down from your high horses and don't be such a snob, I tell you what I have worked as a SMPS designer involving transformer designing for 2 world MAJOR Telecom companies with an arsenal of measuring equipment you can just dream of but I don't go and pull of peoples dreams for that!
Look at your self asking all kinds of delicate questions about transformer designing, why would you have the right to ask and get all the answers but at the same time frowning at others who are behind you in knowledge, nobody tells you it's impossible, the word 'leecher' comes to ones mind.
And winding machines costing $$$, there do exist small machines driven by hand perfectly useful for couple of transformers and people have even built their own winding machines of a scraped hand drilling machine attached to a wood box all in total for maybe $30 or so, one guy even did a winding machine for toroid transformers from child bicycle rim and all this exists in video form uploaded on YouTube and plenty of, and for some it's not a question of money but because its called DIY and giving a bit of challenge.
Shame on you!
Cheers Michael
Ultima Thule,
I agree with what you point to, though I have no reason to think Links was being negative in quite the manner you portray.
I am also quite negative about diy folks winding there own transformers. Not those used in telecom, but those used as power and output transformers, ones that see the power grid or high DC plate voltage. My concern is that it is flatly impossible to pre warn a new designer / winder about the very extreme safety issues involved. My tutelage lasted 7 years, under the careful eye of Nathan Grossener, the inventor if Metglass, Albert Fisher, a well know Chicago based transformer designer and Leroy Carson the chief designer at Better Coil and one of the most respected designers in the world of transformers. Also the inventor of black box thermal modeling that is used by ALL reputable transformer designers. This sort of understudy, with people of this caliber was possible only because almost no one was interested in learning about what looked to be a dieing field and they were very willing to spend the time and provide their expertise. I was not confident of my skills for about 4 years of 8 to 10 hour days, working as a junior designer in a Chicago transformer company.
That ordinary folks, EE's and other newly interested parties want to learn is splendid and I do try to provide some guidance, but just the thought of them making transformers in these classes scares me badly. I suspect that Links was just trying to provide some notion that this might not be an easy and straight forward thing to learn and I will also caution anyone involved in trying this out that it is not a sport, but a potentially deadly business and the modes of death are not obvious.
Bud
I agree with what you point to, though I have no reason to think Links was being negative in quite the manner you portray.
I am also quite negative about diy folks winding there own transformers. Not those used in telecom, but those used as power and output transformers, ones that see the power grid or high DC plate voltage. My concern is that it is flatly impossible to pre warn a new designer / winder about the very extreme safety issues involved. My tutelage lasted 7 years, under the careful eye of Nathan Grossener, the inventor if Metglass, Albert Fisher, a well know Chicago based transformer designer and Leroy Carson the chief designer at Better Coil and one of the most respected designers in the world of transformers. Also the inventor of black box thermal modeling that is used by ALL reputable transformer designers. This sort of understudy, with people of this caliber was possible only because almost no one was interested in learning about what looked to be a dieing field and they were very willing to spend the time and provide their expertise. I was not confident of my skills for about 4 years of 8 to 10 hour days, working as a junior designer in a Chicago transformer company.
That ordinary folks, EE's and other newly interested parties want to learn is splendid and I do try to provide some guidance, but just the thought of them making transformers in these classes scares me badly. I suspect that Links was just trying to provide some notion that this might not be an easy and straight forward thing to learn and I will also caution anyone involved in trying this out that it is not a sport, but a potentially deadly business and the modes of death are not obvious.
Bud
Wow, so much emotions for so simple questions "why DIY and not ..." !
BTW, when I'm was in need for custom audio transformers, winding it myself was the first option. I dropped it almost instantly since no material was available at meaningful price, 100 kg was min for even quite expensive amorphous alloys. I had (20 years ago) manual winding machine borrowed for some time. It was built smart way by a former submarine officer, yet not capable if layering thin wires in nice orderly fashion as factory equipment does (a drawback of any DIY machinery).
The second option was to order it from specialized company. One which would make it for sure asked $3000 for the design and samples. Others wouldn't do it fine for sure since they used primitive calculus like the one from MRB.
Third option is to pay 2 x 300 - 500 EURO for a pair from Tamura, Hashimoto or Plitron. After all, even the price is crazy, I will obtain working products.
I dig quite deep into this issue just to find how complex it is when it comes to REAL professional design & quality. I wrote my own CAD program and discovered later sad fact - there is no ferromagnetic's material data publicly available to feed into this software. Academic material I have (huge 400 - 700 pages books) is of Russian/Soviet-era origin, and all data sheets are for the alloys manufactured back in these days.
Just a basic problem. In real life push-pull output stage *never* fully balanced, there is *always* DC current flows over primary coil. Under certain circumstances, 10% discrepancy of idle current of output tubes may lead to the 2x decrease of primary coil inductance and 5x increase of nasty 3rd order harmonic distortions. Now the question Ultima Thule (aka Michael), from (as you say) some snob on the pedestal: how would you *precisely* re-calculate transformer properties if there are no manufacturer's data - how permeability is changing under flux caused by *both* DC and AC??? The required "mu/aw0/flux" chart is nowhere to be found from today's manufacturers! Air gap is not used in PP trafos like the ones found in SETs. I assume there is another way apart from the one described in Russian books (BudP certainly knows and uses it), yet it will require digging formulas from another 500-page book!
Bud, you got it 100% right.
Last edited:
I am going to go slightly off topic here and post an insulation system design we did for Mackie Designs a number of years ago, to replace a noisy power toroid in one of their medium sized mixers. This is the document submitted to CE for international approval of the safety insulation system.
You will have to use your imagination to see the insulation materials as they would be applied to a coil. This does not utilize a pre-molded CE rated bobbin, as they loose too much winding volume. Instead the creepage and clearance requirements (6mm across any surface and 0.040" across any gap) are made solely with dielectric tapes. This design also had very tight thermal, regulation and emitted field characteristics. This will give you a general idea of just how complex a robust, totally safe linear power transfromer is to make.
All of our high voltage power transformers are made just like this and are then vacuum impregnated with a polyester resin that seals the windings from oxygen, to keep corona at bay and also turns the transformer into an obsidian hard lump.
Bud
You will have to use your imagination to see the insulation materials as they would be applied to a coil. This does not utilize a pre-molded CE rated bobbin, as they loose too much winding volume. Instead the creepage and clearance requirements (6mm across any surface and 0.040" across any gap) are made solely with dielectric tapes. This design also had very tight thermal, regulation and emitted field characteristics. This will give you a general idea of just how complex a robust, totally safe linear power transfromer is to make.
All of our high voltage power transformers are made just like this and are then vacuum impregnated with a polyester resin that seals the windings from oxygen, to keep corona at bay and also turns the transformer into an obsidian hard lump.
Bud
My interest in transformer materials was one more of curiosity. I have seen references to people winding their own transformers. I have looked around and it seems like obtaining the core material, in small quantities, was not really all that possible. The direct interest relates to the desire to build something that is transformer related (a motor growler). For that application, it looks like my best source of materials will be scrap industrial transformers. I posted the question here since it was an active discussion and there was some mention of transformer materials. I did not want to resurrect an old thread where similar things were discussed in the past. I apologize for the hijack.
For my current needs, transformers that are commercially available will satisfy my needs, for the point that I am at in the world of tube amplifiers. I have been following this thread since I found it quite interesting. I have been enjoying BudP's input to these discussions.
For my current needs, transformers that are commercially available will satisfy my needs, for the point that I am at in the world of tube amplifiers. I have been following this thread since I found it quite interesting. I have been enjoying BudP's input to these discussions.
As an addendum to the previously posted insulation system layout. The basic relevant CE 6500 criteria are as follows'
1.) A minimum of 6 mm of creepage and clearance (8mm for units shipped to Finland). This means that for all adjacent or coil crossing wires from windings of opposite polarity insulation must be provided on all sides of the wire in such a way that there will be 6mm of distance from any part of the wires to any part of a winding of opposite polarity. This also requires that any coil wire that is bound within the windings of opposite polarity has a 6mm creepage and clearance from any other bound wire of a winding of opposite polarity. Insulation coating on magnet wires is specifically ignored here, just as if there was none at all.
2.) All dielectric barriers must be constructed of three separate layers, any two of which will be able to provide the full dielectric withstand of 3750 vac for one minute without breakdown. Breakdown limit does allow a small amount of current to flow across the surfaces of the insulation, anywhere from 5 micro amps to 125 micro amps, depending upon the standards for the particular type of equipment the transformer will be used in. Any dielectric barrier that is not three layers of material, as in bobbin flanges, must be a contiguous material with a minimum thickness of 0.040" and be able to withstand the above voltage stress tests. All of these barriers must INSURE a physical separation and a 6mm creepage distance. This means they must be physically fixed and sized in such a way that a poorly trained manufacturing technician cannot corrupt their purpose. All dielectric stress tests will be performed after the transformer has been run for eight hours, at 10% above it's rated load in all secondary windings.
3.) All secondary windings must withstand a direct short circuit, with all other windings loaded to plus 10% of rated load, without emitting molten material (much less catching fire) or in any way dislodging or distorting the dielectric withstand materials by more than 5% in any direction. These short circuit tests are to be applied consecutively, with no more than 10 seconds between shorting of one and un-shorting of the previous and the previous winding is to be taken back to it's 10% overload and must provide 90% of it's original voltage for the duration of the tests on other windings.
4.) At no time during any of these stress tests is the transformer to exceed it's maximum rated insulation temperature classification in any winding, which depends upon the lowest rated material in the dielectric structure, or adjacent to those structures that might cause deformation of that, or other dielectric materials. This means using a thermal fuse in the primary, and also that a low VA winding of opposite polarity must be wound out at the last winding position and a thermal fuse placed on it too, if the primary VA under short circuit of that low VA winding, is not high enough to cause the primary to heat up and trip the thermal fuse for the primary. Resettable fuses with proven life times are acceptable, but they are bulky and expensive. There are some exceptions to this requirement, for units used in "professional" circumstances. Oddly enough Guitar amps fall under this classification, but they must be current fused so that they will not fail the abuse tests and those are not any less stringent. We build all of our Guitar amp power transformers in such a fashion that they cannot ever have a turn on, over current inrush, due to core or coil offset voltages, for just this reason.
There are other tests for softening of lead wire under compression and a very rigorous test of all of the materials found in the coil to prove that they will not deteriorate each other or the insulation on the magnet wire. And then there are subsequent tests of these same abused units to ensure that they meet specific requirements for their eventual usage. There are no exceptions to any of this, if you fail one part, the test halts, you are charged the full fee and get to start all over again the next time.
In my world, the above is everyday stuff, minimum basic requirements for being in business. Then there are the FDA regulations and after that the medical component industry and beyond that are space rated applications. This is why having folks just decide to make their own transformers, for high voltage or even mains connected use, scares me so badly. Without the above proof of suitability the usual 300 k hours of mean time before failure that is industry standard, cannot be guaranteed. To not build to that lifetime standard almost certainly guarantees failure and subsequent fire and deadly shock hazard.
So, you have been warned, this is not a playground.
Bud
1.) A minimum of 6 mm of creepage and clearance (8mm for units shipped to Finland). This means that for all adjacent or coil crossing wires from windings of opposite polarity insulation must be provided on all sides of the wire in such a way that there will be 6mm of distance from any part of the wires to any part of a winding of opposite polarity. This also requires that any coil wire that is bound within the windings of opposite polarity has a 6mm creepage and clearance from any other bound wire of a winding of opposite polarity. Insulation coating on magnet wires is specifically ignored here, just as if there was none at all.
2.) All dielectric barriers must be constructed of three separate layers, any two of which will be able to provide the full dielectric withstand of 3750 vac for one minute without breakdown. Breakdown limit does allow a small amount of current to flow across the surfaces of the insulation, anywhere from 5 micro amps to 125 micro amps, depending upon the standards for the particular type of equipment the transformer will be used in. Any dielectric barrier that is not three layers of material, as in bobbin flanges, must be a contiguous material with a minimum thickness of 0.040" and be able to withstand the above voltage stress tests. All of these barriers must INSURE a physical separation and a 6mm creepage distance. This means they must be physically fixed and sized in such a way that a poorly trained manufacturing technician cannot corrupt their purpose. All dielectric stress tests will be performed after the transformer has been run for eight hours, at 10% above it's rated load in all secondary windings.
3.) All secondary windings must withstand a direct short circuit, with all other windings loaded to plus 10% of rated load, without emitting molten material (much less catching fire) or in any way dislodging or distorting the dielectric withstand materials by more than 5% in any direction. These short circuit tests are to be applied consecutively, with no more than 10 seconds between shorting of one and un-shorting of the previous and the previous winding is to be taken back to it's 10% overload and must provide 90% of it's original voltage for the duration of the tests on other windings.
4.) At no time during any of these stress tests is the transformer to exceed it's maximum rated insulation temperature classification in any winding, which depends upon the lowest rated material in the dielectric structure, or adjacent to those structures that might cause deformation of that, or other dielectric materials. This means using a thermal fuse in the primary, and also that a low VA winding of opposite polarity must be wound out at the last winding position and a thermal fuse placed on it too, if the primary VA under short circuit of that low VA winding, is not high enough to cause the primary to heat up and trip the thermal fuse for the primary. Resettable fuses with proven life times are acceptable, but they are bulky and expensive. There are some exceptions to this requirement, for units used in "professional" circumstances. Oddly enough Guitar amps fall under this classification, but they must be current fused so that they will not fail the abuse tests and those are not any less stringent. We build all of our Guitar amp power transformers in such a fashion that they cannot ever have a turn on, over current inrush, due to core or coil offset voltages, for just this reason.
There are other tests for softening of lead wire under compression and a very rigorous test of all of the materials found in the coil to prove that they will not deteriorate each other or the insulation on the magnet wire. And then there are subsequent tests of these same abused units to ensure that they meet specific requirements for their eventual usage. There are no exceptions to any of this, if you fail one part, the test halts, you are charged the full fee and get to start all over again the next time.
In my world, the above is everyday stuff, minimum basic requirements for being in business. Then there are the FDA regulations and after that the medical component industry and beyond that are space rated applications. This is why having folks just decide to make their own transformers, for high voltage or even mains connected use, scares me so badly. Without the above proof of suitability the usual 300 k hours of mean time before failure that is industry standard, cannot be guaranteed. To not build to that lifetime standard almost certainly guarantees failure and subsequent fire and deadly shock hazard.
So, you have been warned, this is not a playground.
Bud
Last edited:
Thanks Bud for your reply, informative and well appreciated!
The transformer designing I worked with was no sc. "Telecom transformers", SMPS stands for Switch Mode Power Supply and they were connected directly to the mains voltage and have to fulfil all the standards as any generic transformer you are making (creepage, clearance distances, flame class, pollution degree) and yes I have done the CE/UL certification and that did incorporate not only the transformer design and parts used in it but the whole product so I am very familiar with all that you describe here just that it's couple of years ago since I worked with it as I am doing other things these days thus have become a bit rusty.
That said I would like to say Bud you are really providing golden information for many here without going of with a notion or assuming whether somebody is incompetent or not (and that's how it should be in the first place), keep on doing that, I do appreciate at lot what you post here though I haven't taken part actively in this thread before.
I am still waiting for an answer from LinuksGuru on the question I had in the R-core thread where he could order just a couple of custom made transformers from China (or did you have to buy 100 Kg/$3000 from China??), and as Bud thinks it's a laborious task to make own transformers I'm sure many appreciate to know where LinuksGuru can buy them.
BTW, if anyone knows where I can buy C-cores (I have skipped the idea of R-cores as they seem to be even harder to get hands on) in small amounts I would be very happy to know as I need transformers for my upcoming PP KT88 amp, I am taking part in one group design going on over at this link: http://www.diyaudio.com/forums/tubes-valves/156699-mullard-5-20-kt88-pp-blocks-22.html You are all welcome with constructive ideas and support!
Cheers Michael
The transformer designing I worked with was no sc. "Telecom transformers", SMPS stands for Switch Mode Power Supply and they were connected directly to the mains voltage and have to fulfil all the standards as any generic transformer you are making (creepage, clearance distances, flame class, pollution degree) and yes I have done the CE/UL certification and that did incorporate not only the transformer design and parts used in it but the whole product so I am very familiar with all that you describe here just that it's couple of years ago since I worked with it as I am doing other things these days thus have become a bit rusty.
That said I would like to say Bud you are really providing golden information for many here without going of with a notion or assuming whether somebody is incompetent or not (and that's how it should be in the first place), keep on doing that, I do appreciate at lot what you post here though I haven't taken part actively in this thread before.
I am still waiting for an answer from LinuksGuru on the question I had in the R-core thread where he could order just a couple of custom made transformers from China (or did you have to buy 100 Kg/$3000 from China??), and as Bud thinks it's a laborious task to make own transformers I'm sure many appreciate to know where LinuksGuru can buy them.
BTW, if anyone knows where I can buy C-cores (I have skipped the idea of R-cores as they seem to be even harder to get hands on) in small amounts I would be very happy to know as I need transformers for my upcoming PP KT88 amp, I am taking part in one group design going on over at this link: http://www.diyaudio.com/forums/tubes-valves/156699-mullard-5-20-kt88-pp-blocks-22.html You are all welcome with constructive ideas and support!
Cheers Michael
Here is another insulation system submission document that is a bit more on topic for those who must build their own. This is a guitar output transformer that is in the products of a major manufacturer.
The CE6500 standard for output transformers differs from the power transformer only in that the creepage and clearance is defined by the AC RMS voltages of windings of opposite phase added together. The range is 1 mm for up to 30 vac rms combined, and 3 mm for 60 to 900 vac rms. Creepage and clearance does climb to 6 mm and more as combined voltages rise above this point. This means that most sensible outputs need to have 3 mm of creepage and clearance, fixed just as it is in the power transformers.
These output transformers are viewed as secondaries to the power transformer and so, hazardous voltage and safety low voltage requirements are applied. This means that there must be two layers of insulating materials, each of which which must withstand the dielectric stress voltage of 2750 vac for one minute.
As you go through this document, notice that the secondaries are physically offset from the bobbin walls by a tape that has NO effective dielectric strength. It still must meet the temperature class and pollution class of the rest of the materials. This 3 mm wide tape is the key to a number of benefits. The space left unused does not generate leakage inductance. The secondary windings are subject to linear antenna field events and the core phase discontinuity that occurs close to the window walls at low flux levels does not seriously disturb the proper phase relation ship of primary to secondary at high levels of flux density as it moves farther out into the window area and into the windings. And finally it meets the CE requirement for fixed creepage and clearance.
The secondaries are not short circuited for this class of approval but they are loaded to 10% over current. The coil materials must also pass a very rigorous test of all of the materials found in the coil, to prove that they will not deteriorate each other or the insulation on the magnet wire.
Transformers, both power and output, that are built to these standards are very tough to destroy and when failure does occur they will still protect a user from shock hazard and, because all of the materials are self extinguishing, from a fire hazard once the power is removed from the transformer. Naturally, with a set of thermal fuses in place even this limited hazard is removed.
Bud
The CE6500 standard for output transformers differs from the power transformer only in that the creepage and clearance is defined by the AC RMS voltages of windings of opposite phase added together. The range is 1 mm for up to 30 vac rms combined, and 3 mm for 60 to 900 vac rms. Creepage and clearance does climb to 6 mm and more as combined voltages rise above this point. This means that most sensible outputs need to have 3 mm of creepage and clearance, fixed just as it is in the power transformers.
These output transformers are viewed as secondaries to the power transformer and so, hazardous voltage and safety low voltage requirements are applied. This means that there must be two layers of insulating materials, each of which which must withstand the dielectric stress voltage of 2750 vac for one minute.
As you go through this document, notice that the secondaries are physically offset from the bobbin walls by a tape that has NO effective dielectric strength. It still must meet the temperature class and pollution class of the rest of the materials. This 3 mm wide tape is the key to a number of benefits. The space left unused does not generate leakage inductance. The secondary windings are subject to linear antenna field events and the core phase discontinuity that occurs close to the window walls at low flux levels does not seriously disturb the proper phase relation ship of primary to secondary at high levels of flux density as it moves farther out into the window area and into the windings. And finally it meets the CE requirement for fixed creepage and clearance.
The secondaries are not short circuited for this class of approval but they are loaded to 10% over current. The coil materials must also pass a very rigorous test of all of the materials found in the coil, to prove that they will not deteriorate each other or the insulation on the magnet wire.
Transformers, both power and output, that are built to these standards are very tough to destroy and when failure does occur they will still protect a user from shock hazard and, because all of the materials are self extinguishing, from a fire hazard once the power is removed from the transformer. Naturally, with a set of thermal fuses in place even this limited hazard is removed.
Bud
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.