If you have worked with Hitachi amorphous cores, you've probably noticed that most of them come with a big technical air gap. When both halves are pressed to each other, you can clearly see light coming from the gap at some places.
The problem is, this gap drops the permeability considerably. While it's not so bad for a SE output transformer or a filter choke, but for a small signal interstage transformer with no DC current and less windings, I need that permeability.
I've found a way to polish the surfaces, but it's a tedious process, because it is done by hand. What I'm doing is basically.
1. Taking a rough grit polishing paste and putting some on the core surfaces.
2. Rubbing the cores into each other with small, 1-2mm amplitudes.
3. Having a quick measuring jig after each rubbing and cleaning- a coil with known turns, an L meter and a core clamp. Usually L increases after each polishing attempt. So far, I'm using this method to match small cores.
I was wondering if an automated jig could be made, by using a cheap vibrating finishing sander? If the core piece A could be mounted on it, then piece B could be held against with a hand.
The problem is, this gap drops the permeability considerably. While it's not so bad for a SE output transformer or a filter choke, but for a small signal interstage transformer with no DC current and less windings, I need that permeability.
I've found a way to polish the surfaces, but it's a tedious process, because it is done by hand. What I'm doing is basically.
1. Taking a rough grit polishing paste and putting some on the core surfaces.
2. Rubbing the cores into each other with small, 1-2mm amplitudes.
3. Having a quick measuring jig after each rubbing and cleaning- a coil with known turns, an L meter and a core clamp. Usually L increases after each polishing attempt. So far, I'm using this method to match small cores.
I was wondering if an automated jig could be made, by using a cheap vibrating finishing sander? If the core piece A could be mounted on it, then piece B could be held against with a hand.
Last edited:
I am an employe of Hitachi
But for information technology
Not for metals!!! ��
I will ask for change��
Walter
But for information technology
Not for metals!!! ��
I will ask for change��
Walter
What do you mean with "small signal interstage transformer with no DC current..."?
Interstage transformers typically have primary DC current, unless used parafeed.
When you point to small signal line level transformers like input transformers and inductive volume controls, nanocrystalline toroidal cores are far superior in terms of permeability; they require special winding equipment though.
I also had the to also work with amorphous cores from other manufacturers in India. They are pretty much the same in terms of the badly cut cores air gap result. So I came to a probable conclusion that it could be a technology limitation. But why? What could be the reason a Hi-B silicon steel cut core can be almost perfect, while an amorphous core not?
What about P-P interstage transformers with little or no DC current present? Or even if we're talking about a SE interstage transformer with a low primary voltage swing and low DC current, I found out the manufacture gap can be too large, meaning the core's flux density can be far bellow saturation. The result is that I usually need more turns to achieve the needed high inductance and that results in unnecessarily high parasitic Cp and Ls. 🙁
But let's consider also small signal line level transformers.
Toroidal or C uncut core is the right choice for this, but I lack the tools for winding such. My hope is to try using cut cores.
Unacceptable.
By the way: PP interstage transformers do carry DC current; they just (are supposed to) cancel. Nevertheless a very minor airgap is good with these as DC balance is seldom perfect.
By the way: PP interstage transformers do carry DC current; they just (are supposed to) cancel. Nevertheless a very minor airgap is good with these as DC balance is seldom perfect.
A person I know who uses these cores all the time has told me what he does.
He has sheets of flat glass or stone. He puts fine grit paper on the surface. Then polishes them against it being very careful to apply even pressure. I think he has since made a jig. Then finally he uses a polish. He still does it with SE transformers.
He has sheets of flat glass or stone. He puts fine grit paper on the surface. Then polishes them against it being very careful to apply even pressure. I think he has since made a jig. Then finally he uses a polish. He still does it with SE transformers.
Yes, the problem is that this airgap is exceedingly too much just for a DC unbalance issue 🙂
That's a good one and I did that too, it works to a point. 😀 The problem lies exactly into the care of pressure not to be applied, because it would result in uneven ground spots.
One trick to apply even pressure is to push the core to the grinding surface using a low surface contact, such as a cone or ball, like this.
If the grinding surface can be made to vibrate and the core being held in a frame while low surface contact pressure is maintained, it might work?
Attachments
No matter how you do the grinding, be aware of applying the least amount of heat.
When too much heat is produced odd things happen.
When too much heat is produced odd things happen.
It's not just a big gap but the air gap is not uniform at all. I have the impression that once put together the straight side from one C to the other is not really straight.
If so that gap is likely resulting from random matching of the 2 C's or poor quality control.
Those few times I buy C cores I always demand best quality with perfect matching for minimum airgap. It is just a bit more expensive but it is worth it.
If you can polish it without overheating it should not cause any problem.
If so that gap is likely resulting from random matching of the 2 C's or poor quality control.
Those few times I buy C cores I always demand best quality with perfect matching for minimum airgap. It is just a bit more expensive but it is worth it.
If you can polish it without overheating it should not cause any problem.
Definitely true. Core splitting apart can easily occur.
Yes, so much for the high permeability of the amorphous material. 🙁
Yes, the fact the air gap is not uniform is even uglier.
These amorphous cores also seem to be highly geometrically distorted. I'm curious if it's about the fact they're amorphous and somehow hard to manufacture, or just sloppy work.
I'll definitely contact the manufacturer and insist on better quality next time, this is just unacceptable.
They are a bit more harder to manufacture as the material is very brittle and thin, so it easily breaks, but it is not an excuse for sloppy work.
Manufacture of these cores in China and India certainly needs quality control (it's bulk stuff nowadays); Indel (Metglass) and Arnold (Namglass) started manufacture some 20 years ago, and quality was better those days.
lundahl for example sells pp and ppz tx.
the ppz is zero gap, and meant for parafeed only.
pp has small gap, he says that is will not saturate if pp is sligtly unbalanced
the ppz is zero gap, and meant for parafeed only.
pp has small gap, he says that is will not saturate if pp is sligtly unbalanced
In my experience they have minimum gap and gap on request. Minimum gap is what you call zero which means 2-4 mA DC unbalance for 0.9T Bdc, depending on the transformer.
The only core with true zero gap is the un-cut core.
My attempt on an AMCC63 core - success. For 10 minutes of hand sanding towards an an aluminum plate with 240 grit sandpaper glued, the permeability doubled. There is still a visible air gap though, but my sandpaper is already worn out. I'll have to invest in better ones, probably zirconia papers.
Well respected manufacturers, at least in Germany, only sell their C cores in dedicated and marked pairs with high gloss polished interfaces. With respect to the markings, there's only one (!) way to assemble them to attain minimal gap. Even in one bobbin SM or SE cores that consist of four C's, you'll have to take care on the markings and to pair each two of the parts meticulously.
Perhaps the manufacturer of the core shown above didn't take enough care? Or both halves simply come from different pairs?
Best regars!
Perhaps the manufacturer of the core shown above didn't take enough care? Or both halves simply come from different pairs?
Best regars!
I can´t believe what I see in your picture.
Not into amorphous transformer cores at all, but I manufacture speakers and when I stack ferrite ring magnets the parallel surfaces match way better than what you show there 😱
Are you sure your cores are real ones and not fakes made in some cheesy fake factory?
They look like just out of the oven (I guess they are made by sintering) but still rough unpolished.
An unfinished product by any means.
Not into amorphous transformer cores at all, but I manufacture speakers and when I stack ferrite ring magnets the parallel surfaces match way better than what you show there 😱
Are you sure your cores are real ones and not fakes made in some cheesy fake factory?
They look like just out of the oven (I guess they are made by sintering) but still rough unpolished.
An unfinished product by any means.
Terrible!
I believe all cores in the market today are manufactured in India or China.
I still have an old Vacuumschmeltze C cores (0.1 mm lamination). I keep them like gold 😉
Not quite true.
Vakuumschmelze (do not produce silicon steel cores anymore) and Waasner offer(ed) their c-cores in standard quality and polished (Schnittqualität "C" and "A").
For normal applications (power supply transformers; gapped interstage and output transformers) their standard quality is (was) already very good.
What's coming from Asia depends heavily on quality control.
HiB silicon steel c-cores made by Nicore (of Japanese origin but manufactured in China) are very good quality; on par with European production.
Last edited by a moderator:
- Status
- Not open for further replies.