Does a C-core transformer have inherently better performance vs. an IE core? If true, one suspects C-core would be more common.
ANK Audiokits - EL34 Triple C-Core Pentode Monoblocks Tube Amplifier
What about dual tube rectifiers?
TX!
ANK Audiokits - EL34 Triple C-Core Pentode Monoblocks Tube Amplifier
What about dual tube rectifiers?
TX!
Dual rectifier power supplies existed in the 1930,s so yes hype "wins the day " neither is "C" core an "innovation " .
What is really the point is stray flux especially in high impedance tube circuits reduced by circling the transformer with a tight band soldered at the ends to hypothetically present an external "shorted turn " or better still pot it in a ferrous metal enclosure .
Toroidal transformers are better than E core at this --not perfect ,they do leak but nothing like the E core but we are talking $$$$ here and on a production line pennies count.
C core are "better " than E core but still not as good as a toroid ---then we have "R" core yes better than E core but expensive .
In any case moving your transformer some inches away from sensitive tracks on a PCB also makes a big difference but them --again--we have cost to the manufacturer.
Way back in the late 80,s I made a decision to completely isolate the PS by using monoblock construction not only for the power amps but for the power supplies .
Did it make a difference---???-- you bet it did seen on my scope big improvement BUT I was building it for my own pleasure not for several 1000 shareholders or large share holders like CEO,s .
You pays your money and takes your choice only two things are free in life ( as you don't pay at their inception )--life-Birth & Death but I noticed in the USA corpse,s get sued or rather their relatives and yes it applies to the UK next of kin.
What is really the point is stray flux especially in high impedance tube circuits reduced by circling the transformer with a tight band soldered at the ends to hypothetically present an external "shorted turn " or better still pot it in a ferrous metal enclosure .
Toroidal transformers are better than E core at this --not perfect ,they do leak but nothing like the E core but we are talking $$$$ here and on a production line pennies count.
C core are "better " than E core but still not as good as a toroid ---then we have "R" core yes better than E core but expensive .
In any case moving your transformer some inches away from sensitive tracks on a PCB also makes a big difference but them --again--we have cost to the manufacturer.
Way back in the late 80,s I made a decision to completely isolate the PS by using monoblock construction not only for the power amps but for the power supplies .
Did it make a difference---???-- you bet it did seen on my scope big improvement BUT I was building it for my own pleasure not for several 1000 shareholders or large share holders like CEO,s .
You pays your money and takes your choice only two things are free in life ( as you don't pay at their inception )--life-Birth & Death but I noticed in the USA corpse,s get sued or rather their relatives and yes it applies to the UK next of kin.
Some advantages and disadvantages of C-cores. Keep in mind some of them apply to the toroid as well.
-Grain oriented. The same way as the toroid, the C-core is a wound core from a strip of soft-magnetic material. This permits a high working flux density over the whole core, whereas grain oriented E-I laminations will have the flux perpendicular at the corners of the material
-Freedom of core winding material. C-cores can be made from HiB, permalloy, 25um amorphous and nanocrystalline material.
-Cutting. This is an advantage as well as disadvantage. It gives the freedom of disassembling the core, flexibility in air gapping, using a square coil which can be wound on a classic winding machine.
But cutting inevitably leads to loss of permeability, lamination shorting (some losses, although the shorts can be acid etched on some materials) and potential source of mechanical noise if the gap is not glued and has cutting imperfections.
-No laminations assembly. This saves manufacturing time.
As a transformer winder, they are my favourites due to the flexibility they offer.
R-core is like an uncut C-core with a round surface area with existing bobbins on top of it. The bobbins have a mechanical gear for coupling and are rotated by a special machine for winding.
-Grain oriented. The same way as the toroid, the C-core is a wound core from a strip of soft-magnetic material. This permits a high working flux density over the whole core, whereas grain oriented E-I laminations will have the flux perpendicular at the corners of the material
-Freedom of core winding material. C-cores can be made from HiB, permalloy, 25um amorphous and nanocrystalline material.
-Cutting. This is an advantage as well as disadvantage. It gives the freedom of disassembling the core, flexibility in air gapping, using a square coil which can be wound on a classic winding machine.
But cutting inevitably leads to loss of permeability, lamination shorting (some losses, although the shorts can be acid etched on some materials) and potential source of mechanical noise if the gap is not glued and has cutting imperfections.
-No laminations assembly. This saves manufacturing time.
As a transformer winder, they are my favourites due to the flexibility they offer.
R-core is like an uncut C-core with a round surface area with existing bobbins on top of it. The bobbins have a mechanical gear for coupling and are rotated by a special machine for winding.
Both have their pro's and con's as it was stated in detail in the previous post.
I do like my C-Core output transformers but I have classic lamination style inputs and outputs that even sounds great. So I think there is no objective winner in all disciplines.
What I often see in transformers that they are designed for cheap when it comes to EI cores but C-cores are mostly designed for performance because they are higher priced models.
I'm sure with a heavy EI core it can be the same performance in the end, but then they have to be designed for performance optimum, not low price.
Be aware , in the early days of audio there were no C-cores, just massive dimension of conventional transformers and their performance was optimised to the max.
P.S. Build my mono power amps with dual tube rectification. Sounds good to me.
I do like my C-Core output transformers but I have classic lamination style inputs and outputs that even sounds great. So I think there is no objective winner in all disciplines.
What I often see in transformers that they are designed for cheap when it comes to EI cores but C-cores are mostly designed for performance because they are higher priced models.
I'm sure with a heavy EI core it can be the same performance in the end, but then they have to be designed for performance optimum, not low price.
Be aware , in the early days of audio there were no C-cores, just massive dimension of conventional transformers and their performance was optimised to the max.
P.S. Build my mono power amps with dual tube rectification. Sounds good to me.
Last edited:
@ Schmitz77:
You're mistaken that "in the early days, there were no C-cores".
Just one example (1956!):
http://www.tubebooks.org/Books/lockhart.pdf
In the beginning, when McIntosh was still a US company, they used c-cores exclusively.
Quality was higher on the priority list those days.
Search this site and see what you find on todays MC275 output transformers....
You're mistaken that "in the early days, there were no C-cores".
Just one example (1956!):
http://www.tubebooks.org/Books/lockhart.pdf
In the beginning, when McIntosh was still a US company, they used c-cores exclusively.
Quality was higher on the priority list those days.
Search this site and see what you find on todays MC275 output transformers....
Last edited:
Early is relative. In audio, it started with the cinema talkies amps of the 1940s. WE used conventional types only.
Earlier were the tube radios, they date back to the 1920s. In those times, there were no C-cores available. But the first toroidal was very early.
What you show is a book from 1956 (thank you!). This was called the golden age of tube audio. Not the early age of tube audio. Tubes were in production for more than 40 years at that time.
I think Fairchild used them, too. But can't find the Ad anymore.
Earlier were the tube radios, they date back to the 1920s. In those times, there were no C-cores available. But the first toroidal was very early.
What you show is a book from 1956 (thank you!). This was called the golden age of tube audio. Not the early age of tube audio. Tubes were in production for more than 40 years at that time.
I think Fairchild used them, too. But can't find the Ad anymore.
Last edited:
In addition to the things mentioned above, the C core transformer (in particular double C core) fit well as audio output transformer, the core has made with grain-oriented steel or with other special materials (amorphous, etc). Some important companies already use them, e.g. Audionote, Ancient Audio, McIntosh. For sake of information a report an interesting link:
The Audio Note Transformer Design Philosophy Article By Andy Grove And Peter Qvortrup
The Audio Note Transformer Design Philosophy Article By Andy Grove And Peter Qvortrup
A couple of points that get overlooked.
There are two basic classes of cores found in audio. Tape wound cores (toroids, c-cores and R-cores) and those with stamped laminations. People far too often assume a typical characteristic of one class cannot exist in the other. This causes gross over-generalizations to be created and perpetuated.
From a core topology the EI lamination forms a shell configuration for a magnetic path that is identical to a "double C-core". A singe C-core forms a core configuration that has the same magnetic path as a UI stamped lamination. If one wanted they could make a core from stamped doughnut shaped laminations and get a truly gapless toroid.
When using a grain oriented material, tape wound cores keeps the grain direction 100% in line with the magnetic path but this only really applies to M-6 and the square loop nickels. Amorphous, nanocrystalline and round loop nickels are not oriented materials so the grain orientation benefit falls out of the mix.
If very thin material thickness is required then tape wound cores are the best choice. Amorphous and nanocrystalline are 0.1mm or thinner tapes and quite brittle so tape wound cores are thee only economically viable option.
Cutting a tape wound core introduces a substantial airgap over a comparable toroid and even precise polishing does little to minimize it. There are cut cores like the uni-core that go to great lengths to increase the surface area of the gap to reduce effective gap size but I have never seen them used in practice.
Alternately stacking stamped laminations greatly reduces the effective airgap and historically has been the goto approach for small 80% nickel signal transformers. There are EE Lams that take advantage of this.
From a mechanical noise issue, particularly in power transformers it is very difficult to keep the gap faces of a C-core transformer from causing noise. Alternately stacked EI's spread the gap forces around the entire core and when vacuum impregnated with varnish gives a much more "solid structure". It is important to remember that the use of goss in stamped laminations has its own issues. From a core perspective the toroid is king but the lack of a straight winding structure makes for possible noisy windings which again a vacuum varnish impregnation goes a long way to fix. Here is where the R-core pops its head up with the uncut core of a toroid and a straight winding layer of a core type transformer for power transformers it seems to be a winner. An added plus of the R-core is the round cross sectional core allows for the wire to go down like butter. They do make round cross sectional toroids in which the first layer goes down beautifully but after that all bets are off. The one down side of the R-core is the tape the cores are wound with is on the thick side but for 50-60hz operation this may be an acceptable compromise.
I guess the point I am trying to make is the simple proclamation of one core type being better than another is complex situation specific to an individual design and framing it as a universal fact just leads to confusion and mis-information.
dave
There are two basic classes of cores found in audio. Tape wound cores (toroids, c-cores and R-cores) and those with stamped laminations. People far too often assume a typical characteristic of one class cannot exist in the other. This causes gross over-generalizations to be created and perpetuated.
From a core topology the EI lamination forms a shell configuration for a magnetic path that is identical to a "double C-core". A singe C-core forms a core configuration that has the same magnetic path as a UI stamped lamination. If one wanted they could make a core from stamped doughnut shaped laminations and get a truly gapless toroid.
When using a grain oriented material, tape wound cores keeps the grain direction 100% in line with the magnetic path but this only really applies to M-6 and the square loop nickels. Amorphous, nanocrystalline and round loop nickels are not oriented materials so the grain orientation benefit falls out of the mix.
If very thin material thickness is required then tape wound cores are the best choice. Amorphous and nanocrystalline are 0.1mm or thinner tapes and quite brittle so tape wound cores are thee only economically viable option.
Cutting a tape wound core introduces a substantial airgap over a comparable toroid and even precise polishing does little to minimize it. There are cut cores like the uni-core that go to great lengths to increase the surface area of the gap to reduce effective gap size but I have never seen them used in practice.
Alternately stacking stamped laminations greatly reduces the effective airgap and historically has been the goto approach for small 80% nickel signal transformers. There are EE Lams that take advantage of this.
From a mechanical noise issue, particularly in power transformers it is very difficult to keep the gap faces of a C-core transformer from causing noise. Alternately stacked EI's spread the gap forces around the entire core and when vacuum impregnated with varnish gives a much more "solid structure". It is important to remember that the use of goss in stamped laminations has its own issues. From a core perspective the toroid is king but the lack of a straight winding structure makes for possible noisy windings which again a vacuum varnish impregnation goes a long way to fix. Here is where the R-core pops its head up with the uncut core of a toroid and a straight winding layer of a core type transformer for power transformers it seems to be a winner. An added plus of the R-core is the round cross sectional core allows for the wire to go down like butter. They do make round cross sectional toroids in which the first layer goes down beautifully but after that all bets are off. The one down side of the R-core is the tape the cores are wound with is on the thick side but for 50-60hz operation this may be an acceptable compromise.
I guess the point I am trying to make is the simple proclamation of one core type being better than another is complex situation specific to an individual design and framing it as a universal fact just leads to confusion and mis-information.
dave
Ditto! 🙂
Dave, have you had experience with 80% nickel C-cores? Do you suspect their cut surface will be a major limitation to their permeability?
P.S. From my overall experience, the worst cut surfaces have been the ones of amorphous and nanocrystalline cores. Now with time, it seems some manufacturers get better at it. Amorphous iron surfaces are very hard and probably wear the instruments fast. Nanocrystalline core structure is must easier to machine, but it tends to chip into grains, making a pitted surface that results into an air gap. A fine abrasive tool is probably needed to avoid chipping.
Dave, have you had experience with 80% nickel C-cores? Do you suspect their cut surface will be a major limitation to their permeability?
P.S. From my overall experience, the worst cut surfaces have been the ones of amorphous and nanocrystalline cores. Now with time, it seems some manufacturers get better at it. Amorphous iron surfaces are very hard and probably wear the instruments fast. Nanocrystalline core structure is must easier to machine, but it tends to chip into grains, making a pitted surface that results into an air gap. A fine abrasive tool is probably needed to avoid chipping.
Dave-- if your analysis is correct and being a person who has been in constant touch for many years with American consumer /rights groups /class actions and many more I would direct you to the FTC--Truth in Advertising -
Truth In Advertising | Federal Trade Commission
If a case could be made would you let me know as I could forward it to the appropriate group ?
Truth In Advertising | Federal Trade Commission
If a case could be made would you let me know as I could forward it to the appropriate group ?
Totally pointless because all the claims made by the company are subjective. Only descriptions of sound. Now had they posted falsified data or something that you could independently verify then you might have something. I really don’t see anything inflammatory about Dave’s post.
No. All of my C-core experiences are with the amorphous and nano materials. It is interesting to note that the permeability ranges for nanocrystalline and 80% nickel cores for the most part overlap yet for a given turns per core area I see a greater than 4X increase in inductance from 1X1 alternately stacked 80% nickel EI's compared to the "ungapped" nanocrystalline samples I have used. The means for a comparable inductance you need either double the turns or double the core area for a similar design. I am referring to low flux situations where meeting a minimum inductance value is a primary design criteria.
In his book Ruben Lee gives a list of effective air gaps for various stacking arrangements of EI lams and a 1X1 configuration gives 1/10th the effective gap of a butt gap. (0.0005" vs .005") Granted the rough face of but gap leaves a lot to be desired over a lapped surface of a c-core but the sheer increase of the gap area of 1X1 stacked laminations puts you in the middle between the minimally gapped toroid with its gap over a large surface area and the small gap area of a cut and polished core.
The thing most people miss here is that the effective gap is determined by both the physical space between them and the surface area of that gap. Some even go as far to make the silly argument that toroids do not have an air gap ;-)
dave
I am not talking of Dave,s post but the implied subjective advertising masquerading as "professional advice " or "specialist advice ".
My question arised from the searching of a high permeability material for step-up input transformers. The C-core nanocrystalline just doesn't have enough and on top of that, even 0.18mm HiB of the same size has better permeability. So I'm aware I'll have to invest into 80% nickel, but I fear about it being into a C-core due to the cutting potentially screwing all the permeability needed.
@daanve, technically common nanocrystalline material is specified at 1.26T saturation density, although in practice I've found the knee to begin at around 1T. Same goes to common amorphous despite the claimed 1.56T. So I'm designing my transformers and chokes up to 1.0T maximum working flux density.
@daanve, technically common nanocrystalline material is specified at 1.26T saturation density, although in practice I've found the knee to begin at around 1T. Same goes to common amorphous despite the claimed 1.56T. So I'm designing my transformers and chokes up to 1.0T maximum working flux density.
Last edited:
If your application fits (DC), nanocrystalline toroids are superior for small signal transformers.