I have built some self-oscillating DC-DC converters using those DSL/ISDN modem isolation transformers. They usually have isolation voltages up to 1 or 2 KV RMS and are cheap as chips. Works great!
----Android NewsGroup Reader----
Great minds do sometimes come up with similar solutions. Or kluges.
That last one is actually in production.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
I have one product that uses a 1:1:2:2 ISDN transformer to make 120 volts at 1 mA, for a UV photodiode. No failures with a couple thousand in the field.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Hey, we're talking about a 750V auxiliary supply, you've got NO problem if it leaks to the low voltage current limited input. Your big problem is on leakage on its OUTPUT, which has nothing to do with the xformer. Whatever you connect to that output, THAT has to hold the HV.
And "how much" depends on the material. Most plastics I think degrade over exp(volts)*time. Ceramics/glasses tend to have a threshold, a trapping capability or something, below which the longevity is either more steeply exponential, or even hyperbolic (unlimited below threshold). Compare ratings of monolithic isolators that use SiO2 (conventional die process, capacitive coupling) versus polyimide (built-up, copper windings, magnetic coupling).
The same will be true of conventional windings (almost all polymer insulated, of course), with the added gotcha that very high voltages will generate partial discharge and corona problems.
So, figure whatever the breakdown is (1-10 sec), and take maybe 1/20th of that for operation in the 50 year range.
Tim
-- Seven Transistor Labs Electrical Engineering Consultation Website: http://seventransistorlabs.com
Back then a sales engineer wanted to hard-sell me on Tiny Logic, saying that the CD4000 series would go obsolete. I told him that I don't believe that and that CD4000 is much cheaper. Needless to say the sales guy was wrong. This design is so cost-sensitive that we later even backed down from 5% resistors to 10%, 20% and untrimmed.
-- Regards, Joerg http://www.analogconsultants.com/
The top of the secondary will exhibit 750V against the primary which sits at or near zero (logic level). That's what can cause the degradation if not rated properly. And yes, you do have a problem because in many environments an uncontrolled breakdown to whatever, even chassis, is considered a hazard. As it should be.
-- Regards, Joerg http://www.analogconsultants.com/
Aside from cost, the 4000 series will operate at much higher voltages, which often comes in handy, and it is not guaranteed to work at modern (low) supply voltages. It's not like there is a one-for-one equivalency.
That said, I expect the availability of anything but simple gates and a few popular functions such as PLL, counters and such like will continue to shrink. CD4086, for example, are now non-stock, but not yet EOL. You can still buy 1960s-design 7400 gates at Digikey, albeit at rather high prices.
--Spehro Pefhany
The good (or sometimes bad) thing about some of the Tiny parts is that they have brutal drive capability at insane speeds... like 5 volts swing in 600 ps, from a 15 cent part. There is a 1 ns flipflop, too, as fast as some ECL families.
I haven't designed in a 4000-series part in ages, but then my stuff isn't very parts-cost sensitive.
-- John Larkin Highland Technology, Inc picosecond timing laser drivers and controllers jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
That was my other argument. With CD4000 I did not need any voltage regulator.
That's a "vintage panacea chip", those go out of style easily. I stick to the jelly bean variety. Schmitt inverters, two-input NAND, stuff like that. They are sold in such high qties that you and I will probably not live long enough to see that go obsolete. For example, even Digikey has tens of thousands CD40106 in stock and they only do that for high-demand items. Rochester has 150,000 of them.
-- Regards, Joerg http://www.analogconsultants.com/
For high speed purposes I use them as well. Occasionally for glue logic when I really only need one section.
Compared to modern lower voltage parts they feel like driving an old
34-horse VW Beetle. But often that's all you need. The beauty is that you don't need a logic rail for CD4000. One just has to keep in mind that they become like molasses at 5V and lower.-- Regards, Joerg http://www.analogconsultants.com/
Most of my stuff has a uP and/or an FPGA, so I generally run 3.3 or
2.5 volt logic levels. So I need HCT-type analog multiplexers and such.-- John Larkin Highland Technology, Inc picosecond timing laser drivers and controllers jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
I was recently at the APEC conference and met the Renco guys there. Seems like they know what they are doing. Will have to let them bid on a design
Cheers
Klaus
They have come through where others tucked their tails and said it can't be done. It's an attitude I really like.
-- Regards, Joerg http://www.analogconsultants.com/
Does a ferrite bead work as well as a resistor in the gate lead of that type of circuit, or do you really need an actual low value resistor to effectively damp parasitic oscillations?
-- ----Android NewsGroup Reader---- http://usenet.sinaapp.com/
A ferrite usually works. A resistor is cheaper and allows you to tune the rise/fall, to minimize EMI.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
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.