Ferrite rod, side by side or in line

Hi

Trying to get good coupling between to ferrite rods with a winding on each

I tested with side by side, like this:

formatting link

Compared to aligning the 2 ferrite rods inline, end to end

I have tested it and get better coupling end to end

But I would like to do some what if experiments, so would like to simulate it, including the H field line distribution

Any recommendations?

Cheers

Klaus

Reply to
Klaus Kragelund
Loading thread data ...

The coupling is between the two wires. I guess the ferrite was designed in so the high frequencies are not coupled. You can wind both wires on one ferrite rod. That is a transformer with low pass. Test high frequency rejection spectral character.

Al

Reply to
Alan Folmsbee

But, end-to-end means it radiates over a long range (and picks up interfering signals from a similar range). Side-by-side would have less field leakage at distance. Signal is one thing, signal/(noise+ interference) ratio is the usual thing to optimize.

Reply to
whit3rd

whit3rd wrote in news: snipped-for-privacy@googlegroups.com:

You do know that a complete MAGNETIC circuit is reqired for good coupling. With two rods, you have two separated magnetic cicuits and any coupling between them will be weak, at best.

Two windings, same core gets you a transformer.

All you are currently doing is using one to "sniff" the other.

Reply to
DecadentLinuxUserNumeroUno

Yes, correct. I am transferring power accross a plastic barrier, so do not have the luxury to have the windings on the same core. I will be running in resonance, to reduce leakage effect problems

Cheers

Klaus

Reply to
Klaus Kragelund

Good point. I will do both, and test in EMC chamber to verify the difference in far field emissions

Reply to
Klaus Kragelund

why not two pieces of U shaped ferrite?

Reply to
Lasse Langwadt Christensen

U shaped would be better for sure, but I need to keep cost ultra low and U shape will probably not be cheap or easy to mount

I have looked into spiral PCB turns and a planar ferrite, but production costs is high (needs glue etc)

Cheers

Klaus

Reply to
Klaus Kragelund

Klaus Kragelund wrote in news: snipped-for-privacy@googlegroups.com:

Buy a troid ferrite and slice it in half, place the barrier between (use Nomex at 1500V / mil) I would not go any higher than a 5 mil gap as that is actually ten mils with the dual break. Coupling will occuur, but not much. power transfer capability will suffer as that gap increases. The toroid encapsulates the flux circuit and gives you the best chance of the most efficient method for this.

You could use laminated stack toroid halves if the frequency is low enough.

There is also a way to assemble laminate stacks with insulator sheets between each layer that would allow the construction of a pair of windings with high isolation between them.

As that gap increases, your "transformer" transforms into a sniffer and capacity to make power moves toward nil.

Reply to
DecadentLinuxUserNumeroUno

Ferrite pot core halves could do the same job. Some come with a hole down the centre, and you can clamp them together with a single nylon bolt.

The classic clamp for RM cores is a metal spring clip, but that's not nice if you want lots of isolation. A non-conducting nylon cable tie could apply the same kind of clamping force.

RM cores come with plastic coil formers, which make them easy to wind. The regular off-the shelf parts aren't designed for isolation, but people who make high-voltage inverters are more creative.

--
Bill Sloman, Sydney
Reply to
bill.sloman

snipped-for-privacy@ieee.org wrote in news: snipped-for-privacy@googlegroups.com:

And NOT deliver the isolation he is looking for. Learn to read entire threads, asshole, not just your cursory glance stupid shit.

Pot cores are small. Winding segregation is not easy to the level he is looking for.

Reply to
DecadentLinuxUserNumeroUno

snipped-for-privacy@ieee.org wrote in news: snipped-for-privacy@googlegroups.com:

There are plenty of hv multisegmented bobbins. And "regular off the shelf" is more like straight out of the catalog.

You have been out of the channel for far too long. My hv job was 12 years ago and I know more about it than you ever will.

Reply to
DecadentLinuxUserNumeroUno

shunning efficiency will likely increase costs.

TDK B67366A1X27 under half a buck at QTY10, under a quarter at QTY1000

needs more glue than the ferrite rods?

--
  Notsodium is mined on the banks of denial.
Reply to
Jasen Betts

The catalogs contain parts that aren't widely stocked. Broad-line distributors like element-14/Newark do keep stuff on the shelf for same day delivery.

You might, but evidence you present here isn't all that persuasive.

--
Bill Sloman, Sydney
Reply to
bill.sloman

Some pot cores are small. Others are larger

formatting link

gives quite a lot of space within which to spread out the windings.

--
Bill Sloman, Sydney
Reply to
bill.sloman

I need to bridge a 3mm gap, so no can do

By the way, using coreless spiral PCB transformer, I have about 60% efficiency. But, it's a pain to mount a PCB vertical on another PCB, so the full cost of the solution is high, since production needs special equipment to mount it

Using a standard drumcore, I could come down to 3 cents cost, and no production hassle

Reply to
Klaus Kragelund

I need to bridge a gap

Reply to
Klaus Kragelund

I am looking for about 3-5 US cents cost

Reply to
Klaus Kragelund

snipped-for-privacy@ieee.org wrote in news:9bb27510-a977-43cc-a5ff-cffa030d28a6 @googlegroups.com:

No shit.

Many times they have to fresh press both the ferrite cores and the bobbins when a REAL company with a REAL order calls up.

You really have no grasp of how business works.

Many things are "COTM", but still require minimum orders and may or may not drive a new production run, which would generate both the ordered stock and a bit for the shelves.

Today's businesses shelves are not stocked heavily as end of year audits do not like remaining stocks.

Again, you have been away too long, billy. It shows. Take your decades old 555 hater mentality and go away from here too.

Reply to
DecadentLinuxUserNumeroUno

snipped-for-privacy@ieee.org wrote in news:2a01f33e-6496-40b9-b3ee-2b65b44e8b77 @googlegroups.com:

You obviously have no grasp of what his coupling/isolation circumstance is.

You lack common sense too, because it is easily garnered by simply reading the thread. Unless you are a sloman abject idiot.

Reply to
DecadentLinuxUserNumeroUno

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.