little pcb-mount CTs

I don't see the connection to Digikey listing non-stock items. You have

it's an indication that it might not be there for long

go lookup 5V FPGAs on digikey, there is quite a few available, but 10 times more non-stock

-Lasse

I don't get it. In-circuit, the input impedance is ohmic (winding factor and output load resistor) plus some small DC resistance; there's no dependence on energy storage in an inductor, no L/R time constant involvement.

How does any problem occur 'at 400 Hz', in a transformer?

Not enough inductance. The old rule of thumb among RF guys is Z=4x the impedance of the circuit it's connecting to, at the desired frequency. When the Z of a winding is only 2-3ohms and you want a 400ohms burden resistor that's a wee problem :-)

Exactamente. It could be the "writing on the wall". Maybe not but many times it was.

I don't mind that they keep obsolete stuff on the search engine. (There is an "in stock" button.)

George H.

If there were history that showed a trend of fewer available parts over time that might mean something. But there are many reasons why parts are not stocked, mostly having to do with stocking agreements. Digikey handles 100's of thousands of parts and they get a little something from the makers to do their stocking.

I'd be surprised there were *any* available. They stopped introducing them years and years ago. What's your point?

IncandescentLinuxUserNumeroUno may not be AlwaysWrong, but he does make some dumb choices. Dissing me may not be particularly stupid, but dissing Joerg is exceedingly foolish.

IncandescentLinuxUserNumeroUno shares this much with krw - he can't imagine that his own ideas on a subject could possibly be incorrect.

Who said the non-stock items are obsolete???

With a 100 ohm burden resistor, the 100:1 version of the Coilcraft LPR6235 is down 3 dB at about 900 Hz, and -8 dB at 400 Hz. There's not much to be gained from reducing the burden resistor, because the secondary copper is over 300 ohms.

If I was running at exactly 400 Hz, I might almost use it and cal around the error. Or resonate the secondary or something. But my range is maybe 300 to 1KHz.

A 0.1 ohm resistor and my little ISDN transformer looks like it would work. I'll need an opamp to add some gain ahead of the ADC. But that's all parts that we have.

On Tue, 5 May 2015 18:44:44 -0700 (PDT), Bill Sloman Gave us:

He is an idiot, and so are you. I said nothing about "a DSL transformer". What I said was that ISDN most certainly IS still in WIDE USE.

You also need reading comprehension course work.

AND you are still a Goddamned, retarded line length dumbfuck.

AlwaysWrong, in vintage form. I'm well aware that Joerg isn't an idiot, as is everybody who can understand his occasional posts about the more demanding circuits than you can comprehend.

But decreasing, as Joerg was pointing out. It's an obsolescent technology, just like the antique news-reading software to which you are sentimentally attached.

We don't all share your problems

And from my point of view, you are addicted to antiquated news-reading software.

Except that it can avoid another opampif the amplitude isn't sufficient.

You could cal that out if you can send a 50-100Hz pulse stream of known amplitude through the primary side and do an FFT. Or modulate it on. That gives you the attenuation for each bin.

A pilot signal would also work but they often razz me at clients for that because I want too many of the timers in their uC.

That's ok as long as the transformer looks like it'll be around in the distributions channnels for a long time.

We order directly from Talema, shipped from India. They seem to make them to order.

The packaging seems to be common, so other people are willing to make these if we need them to.

AH, now I see. I checked input winding resistance (against burning the thing up), but not output winding resistance. Other problems (inductance) one can deal with by reducing the burden resistance, but it'd take a negative-impedance amplifier to null DCR of the secondary.

There can be a saturation problem, too, at low frequencies.

The Talema ISDN has 4 windings, 1:1:2:2. If I put the pairs in series, I can get a 1:2 step-up that I can connect to a 0.1 ohm shunt. At 1 amp, I'll get 100 mV RMS in and 200 mV RMS out. That starts to look ratty at about 180 Hz. A 1:4 (or 1:5!) stepup would be nice but won't work at 300 Hz.

CTs are a nuisance. The shunt+coupler seems to work a lot better.

The ISDN pretty much quits being a transformer at about 112C.

Could you give a few more words on that? What's special about a dual secondary?

Cheers

Phil Hobbs

If you can happen to come across a dual secondary type... you can use the flux of one (no current load = no drop on DCR) to drive current through the other. Resulting current (not voltage) is exactly proportional to input, to much lower frequencies and higher fluxes.

Tim

It's the "zero flux" CT idea. One secondary is the sense winding, and you drive the other secondary to get zero sense signal.

It's cool because it's super linear, and no core will saturate at zero flux!

The DCCTs use this, plus a second-harmonic trick to work down to DC.