I have some power transformers with an obvious lamination stacking characteristic. Instead of all E-I laminations alternating orientation for the entire stack, there are 'stripes' visible in the stack due to the alternation of smaller stacks. Instead of E:I:E:I:E:I:E:I:E, there are alternating 'bunches'of laminations.
I don't have one in front of me, but I recall it looking something like EEEEE:IIIII:EEEEE:IIIII:EEEEE, if that makes sense. The point is that there is a pattern instead of alternating every other one.
I'm wondering if anyone knows the effect of this stacking method.
These transformers were for an old Heathkit BE-2 battery eliminator with switchable rectifier wiring: HWR or a doubler I assume was also HW.
IIRC from school, a HWR induces a DC flux in the core due to the asymmetrical rectified waveform. I imagine the alternating lamination 'clusters' produces a different eddy current and/or flux leakage property. I'm trying to convince myself it nets a 'distributed gap' different from and perhaps better for some purpose than the alternating pairs method.
If it does result in a distributed gap, that might 'humor' the alleged DC flux from HW rectification.
If that is true, it makes me wonder if such a magnetic structure might tolerate service as a low-fi SE OPT for a guitar amp or similar. (I have dozens of them, unused, they're not much use for anything else, and they are so ugly they have an inherently primitive look...I forgot to mention they have open windings for a Variac-like rotating pair of contacts to access a ground-bare secondary winding. I'm imagining...a variable tap OPT (it is isolated, not an auto-transformer like a typical Variac/Powerstat).
Thanks
Murray
I don't have one in front of me, but I recall it looking something like EEEEE:IIIII:EEEEE:IIIII:EEEEE, if that makes sense. The point is that there is a pattern instead of alternating every other one.
I'm wondering if anyone knows the effect of this stacking method.
These transformers were for an old Heathkit BE-2 battery eliminator with switchable rectifier wiring: HWR or a doubler I assume was also HW.
IIRC from school, a HWR induces a DC flux in the core due to the asymmetrical rectified waveform. I imagine the alternating lamination 'clusters' produces a different eddy current and/or flux leakage property. I'm trying to convince myself it nets a 'distributed gap' different from and perhaps better for some purpose than the alternating pairs method.
If it does result in a distributed gap, that might 'humor' the alleged DC flux from HW rectification.
If that is true, it makes me wonder if such a magnetic structure might tolerate service as a low-fi SE OPT for a guitar amp or similar. (I have dozens of them, unused, they're not much use for anything else, and they are so ugly they have an inherently primitive look...I forgot to mention they have open windings for a Variac-like rotating pair of contacts to access a ground-bare secondary winding. I'm imagining...a variable tap OPT (it is isolated, not an auto-transformer like a typical Variac/Powerstat).
Thanks
Murray
Murray..
You are correct....
The minumun gap is when stacking EI laminations in a 1x1 manner....
The max gap is when it is butt stacked....All E's in one side and all I's in the other...
Then if you need more gap then you add a spacer between the E and I...
Then any groupings between these extremes are to make gaps to size...
Yes it makes the transformer more tolerant to DC inbalance...
I have stacked 2x2, 5x5 and 10x10 in some of my designs...
It also linearizes the BH loop...
The trade-off is that the inductance falls off and in power transformers this will increase the excitation cuurent, which increase transformer losses...
A trade-off is typically made.. from all my analysis 2x2 seems to be most optimal in power transformers...
This is common in Hi-End output transformers from the 50's.... It allows some DC imbalance from Push-Pull pairs... you just need to make sure you still have enough inductance to meet the low frequency requirement....
You are correct....
The minumun gap is when stacking EI laminations in a 1x1 manner....
The max gap is when it is butt stacked....All E's in one side and all I's in the other...
Then if you need more gap then you add a spacer between the E and I...
Then any groupings between these extremes are to make gaps to size...
Yes it makes the transformer more tolerant to DC inbalance...
I have stacked 2x2, 5x5 and 10x10 in some of my designs...
It also linearizes the BH loop...
The trade-off is that the inductance falls off and in power transformers this will increase the excitation cuurent, which increase transformer losses...
A trade-off is typically made.. from all my analysis 2x2 seems to be most optimal in power transformers...
This is common in Hi-End output transformers from the 50's.... It allows some DC imbalance from Push-Pull pairs... you just need to make sure you still have enough inductance to meet the low frequency requirement....
This is good to know as I had the exact type question.....I had just recently dis-assembled a core & noticed the Three by Three arrangement.
__________________________________________________________Rick.......
__________________________________________________________Rick.......
I believe sometimes the builder do it in this way to finish the work quickly, not for any technical justification. Perhaps I´m wrong.
"What is taper / gamma ? "
You will have to ask Jacques on that one to be certain. But I am aware of a "stacking factor" (gamma?) that has to do with how densely the lams pack together in the bobbin compared to the ideal geometrical calculation. Typically around 95% to 99%. This has to do with shearing effects on the edges of the lams leaving a little residual kerf on the side edge and also to any systematic bend in the lams and to surface roughness and any enamel coating. By packing more lams together in matched bunches, one should be able to squeeze them together a little better (giving a slightly higher inductance and volt/turns for the same bobbin size). Of course that comes with a cost in leakage inductance, due to the poorer interleave around the gaps, as mentioned earlier above.
You will have to ask Jacques on that one to be certain. But I am aware of a "stacking factor" (gamma?) that has to do with how densely the lams pack together in the bobbin compared to the ideal geometrical calculation. Typically around 95% to 99%. This has to do with shearing effects on the edges of the lams leaving a little residual kerf on the side edge and also to any systematic bend in the lams and to surface roughness and any enamel coating. By packing more lams together in matched bunches, one should be able to squeeze them together a little better (giving a slightly higher inductance and volt/turns for the same bobbin size). Of course that comes with a cost in leakage inductance, due to the poorer interleave around the gaps, as mentioned earlier above.
Last edited:
Ah, good to hear from you Cerrem...been a while.
I found some good info in the last week that intersects with what you said, and had some other interesting details. Thanks for the in-depth explanation.
Magnetic Metals has an article on their web site http://www.magmet.com/pdf/TransformDesignConsiderat.pdf that gives one view of this interleaved stacking method, comparing butt-stacking through 3x3 interleaving (it says 4x4 but the graph only shows through 3x3).
I also found the Reuben Lee book Electronic Transformers and Circuits on vias.org with more information about this. He summarizes I think up to 1x1 through 16x16 compared to EI butt stacking in terms of an approximate air gap equivalent, and stacking factor effects, for a particular lamination size. I took this to be a general trend with varying results for different lamination sizes.
I had begun to suspect how broad the effects of altered stacking patterns might reach...the topic broadened to a whole range of things (as do most things one digs into), like asymmetrical waveforms including varied rectifier circuits, and altering the harmonic levels when needed.
I guess it's just one small snapshot of the whole of the art of magnetics design.
I had been curious if there was actually a way to estimate the equivalent gap and what kind of 'off-label' repurposing opportunities this might present some for orphan transformers with little practical use in their original application...I realize there are many compromises involved...there is not a huge equivalent gap from this, but knowing the number of turns on the secondary (open frame allows to to see it) lets me estimate the number of primary turns which would let me estimate Bdc if I wanted to take the thought a bit further into the weeds. Keeps me off the street, you know?
For Osvaldo, it took me a while to find a text discussion of this, and found Cerrem's reply today (it may have been posted for quite a while), but there appear to be a broad range of reasons a particular stacking pattern is chosen...in magnetics for other than OPT applications as well. This was in context with discussion of transformers physically larger than a given lamination material's availability. A special lamination structure might be made to accommodate a larger winding assembly, by overlapping multiple lamination pieces to give the required dimensions. Similar to the distributed gaps that result from alternating orientation of EI (or other lamination shape) laminations, there could be multiple gaps within the overall magnetic path length. I don't have an example handy (I didn't print the page), but imagine multiple 'I' laminations interleaved (with a brick wall as a very primitive visual analogy) to give a longer side to a magnetic core assembly ...I don't know, maybe a three-phase power transformer is a place that might be a solution.
I still never wound those 6C33C-B OPT's you (Cerrem) designed! ...could it have been as long ago as 1998?... I've moved, pursued other interests, briefly owned an Adams Maxwell 1260 traverse I couldn't get wiring info for and had to get rid of for space reasons...I found in boxes in the garage a brand new soldering iron station and box of EH/Sovtek tubes, sockets & power resistors for a project I fully intended to build, but moved in 1998 and forgot about. All that stuff ended up in the garage and I just found it! No idea what schematic I was intent on...I think it was on the Svetlana site, which has disappeared. I think I found what I was looking at previously connected to Eric Barbour and a German valve site (better print it now)...TBD
I still have two rolls of skived electrical (not plumbing) PTFE in two thicknesses, and two boxes of M6 lams in sizes I can't even remember, that I got way back then.
The bright side is all this stuff would cost me a heck of a lot more in 2012 dollars! Shipping is not friendly for laminations! I got a quote on another size and was shocked - pretty unattractive when I looked at the minimum order and the shipping! They don't even want to both with the short box they offered many years ago.
Makes recycling more and more attractive...
I haven't found an easier way to get laminations apart from a varnished transformer out than poking a chisel-tip Xacto knife at the corner of each layer and prying. Nuisance is relative...it's a lot less work than unwinding wire by hand (if you're counting turns).
Then there are the 'Perma-Fil' potted transformers...seems like epoxy...I do not have the determination to negotiate with those assemblies! Hence the desire to 'repurpose' found transformers. (I scrapped 1800# of essentially useless (too small, too costly to ship, weird low voltages, rodent-infested boxes, paper-wasp-eaten insulation (I hope the varnish killed them), etc.) transformers last year at an average of $0.30/pound :O(. )
Is this a hobby or a disease?
Say, does anyone know what ever happened to Midwest Transformer in Gray's Lake Illinois (maybe they merged with another company?)
I found some good info in the last week that intersects with what you said, and had some other interesting details. Thanks for the in-depth explanation.
Magnetic Metals has an article on their web site http://www.magmet.com/pdf/TransformDesignConsiderat.pdf that gives one view of this interleaved stacking method, comparing butt-stacking through 3x3 interleaving (it says 4x4 but the graph only shows through 3x3).
I also found the Reuben Lee book Electronic Transformers and Circuits on vias.org with more information about this. He summarizes I think up to 1x1 through 16x16 compared to EI butt stacking in terms of an approximate air gap equivalent, and stacking factor effects, for a particular lamination size. I took this to be a general trend with varying results for different lamination sizes.
I had begun to suspect how broad the effects of altered stacking patterns might reach...the topic broadened to a whole range of things (as do most things one digs into), like asymmetrical waveforms including varied rectifier circuits, and altering the harmonic levels when needed.
I guess it's just one small snapshot of the whole of the art of magnetics design.
I had been curious if there was actually a way to estimate the equivalent gap and what kind of 'off-label' repurposing opportunities this might present some for orphan transformers with little practical use in their original application...I realize there are many compromises involved...there is not a huge equivalent gap from this, but knowing the number of turns on the secondary (open frame allows to to see it) lets me estimate the number of primary turns which would let me estimate Bdc if I wanted to take the thought a bit further into the weeds. Keeps me off the street, you know?
For Osvaldo, it took me a while to find a text discussion of this, and found Cerrem's reply today (it may have been posted for quite a while), but there appear to be a broad range of reasons a particular stacking pattern is chosen...in magnetics for other than OPT applications as well. This was in context with discussion of transformers physically larger than a given lamination material's availability. A special lamination structure might be made to accommodate a larger winding assembly, by overlapping multiple lamination pieces to give the required dimensions. Similar to the distributed gaps that result from alternating orientation of EI (or other lamination shape) laminations, there could be multiple gaps within the overall magnetic path length. I don't have an example handy (I didn't print the page), but imagine multiple 'I' laminations interleaved (with a brick wall as a very primitive visual analogy) to give a longer side to a magnetic core assembly ...I don't know, maybe a three-phase power transformer is a place that might be a solution.
I still never wound those 6C33C-B OPT's you (Cerrem) designed! ...could it have been as long ago as 1998?... I've moved, pursued other interests, briefly owned an Adams Maxwell 1260 traverse I couldn't get wiring info for and had to get rid of for space reasons...I found in boxes in the garage a brand new soldering iron station and box of EH/Sovtek tubes, sockets & power resistors for a project I fully intended to build, but moved in 1998 and forgot about. All that stuff ended up in the garage and I just found it! No idea what schematic I was intent on...I think it was on the Svetlana site, which has disappeared. I think I found what I was looking at previously connected to Eric Barbour and a German valve site (better print it now)...TBD
I still have two rolls of skived electrical (not plumbing) PTFE in two thicknesses, and two boxes of M6 lams in sizes I can't even remember, that I got way back then.
The bright side is all this stuff would cost me a heck of a lot more in 2012 dollars! Shipping is not friendly for laminations! I got a quote on another size and was shocked - pretty unattractive when I looked at the minimum order and the shipping! They don't even want to both with the short box they offered many years ago.
Makes recycling more and more attractive...
I haven't found an easier way to get laminations apart from a varnished transformer out than poking a chisel-tip Xacto knife at the corner of each layer and prying. Nuisance is relative...it's a lot less work than unwinding wire by hand (if you're counting turns).
Then there are the 'Perma-Fil' potted transformers...seems like epoxy...I do not have the determination to negotiate with those assemblies! Hence the desire to 'repurpose' found transformers. (I scrapped 1800# of essentially useless (too small, too costly to ship, weird low voltages, rodent-infested boxes, paper-wasp-eaten insulation (I hope the varnish killed them), etc.) transformers last year at an average of $0.30/pound :O(. )
Is this a hobby or a disease?
Say, does anyone know what ever happened to Midwest Transformer in Gray's Lake Illinois (maybe they merged with another company?)
- Status
- Not open for further replies.