low power, HV-in regulator with depletion mosfet

also

Or sober, either. However, if you've been posting drunk this evening, I'll cut you some slack. It doesn't make you look very good, though.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 USA 
+1 845 480 2058 

hobbs at electrooptical dot net 
http://electrooptical.net

That is not challenging requirement, make the 1K's safety resistors: Please view in a fixed-width font such as Courier.

. . . 48V . | . | . | . [1.0K,1W] . | . | . ----------+ . | | Hobbs Diode . | | . | | .--------| . | 63V | | | | | | . | ___/ | ------- | _|_/ . | // \ 15V --- 0.47u | --- 0.47u // \ 15V . | --- 500mW --- | --- --- 500mW . | | | | | | Hobbs . | | | | | | OVP . -+----------+------------+---------+---------+----------+---- . | . --- . /// . . .

Thanks for the plug, but I didn't invent any of that--I read it in app notes when I was a kid. You probably don't need the inverse parallel diode if the total output capacitance is only 470 nF, and a lot of the OVP issues are a lot less worrying if the input supply is only 48V. (If the TL783 came in smaller packages, it would be a pretty good one-chip solution, but it doesn't.)

The app note just says 'HV', and since most of the parts they list are

500V or 1 kV, that is liable to be rectified line voltage, which is quite another kettle of fish.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 USA 
+1 845 480 2058 

hobbs at electrooptical dot net 
http://electrooptical.net

nknown,

nt

I never trust HV parts like that, and it costs almost as much as the whole 78LXX circuit. I/O ripple attenuation is ridiculous enough to be buried in the output noise, unless the HV ripple is over 10V.

The Hobbs ref is a joke. Widlar authored the original app note with all tho se protection features. The input shorting problem at the time was mainly d ue to "sputtering" aluminum electrolytics from the 60-70s era, lots of ripp le current and running hot to boot.

unknown,

78LXX circuit. I/O ripple attenuation is ridiculous enough to be buried in the output noise, unless the HV ripple is over 10V.

A buck for a one-chip solution isn't too bad, and it has guaranteed specs. Something has to hold back the input voltage, and surely it would be comforting for that something to have overcurrent, thermal, and second-breakdown protection like the TL783.

You're kidding. ;)

Widlar authored the original app note with all those protection features. The input shorting problem at the time was mainly due to "sputtering" aluminum electrolytics from the 60-70s era, lots of ripple current and running hot to boot.

IIRC the LM78xx datasheet says to use the inverse diode if you have more than about 10 uF of output capacitance. Modern devices have lower ESR than the old aluminums, so that problem has got worse rather than better, I think. But 470 nF wouldn't be a big problem, I don't think.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 USA 
+1 845 480 2058 

hobbs at electrooptical dot net 
http://electrooptical.net

--
For someone who claims to be non-competitive, you belie your own 
claim.

He's mentioned them here, on and off, for at least 5 years, probably much longer. Quite interesting devices that I would never have thought of looking for otherwise.

"Normally closed", automatically current limiting transistors that can stand off high voltages, several uses for that.

There was a particularly nice application of protecting a buffer input from accidental HV, the FET current limits to below the ESD diode destruction level.

--

John Devereux

Just the facts, m'aam.

I've been using them for decades. I've posted circuits with them in the past. They are good to know about, especially for high-voltage, low-power situations.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

situations.

I stole the idea of using back-to-back LND150Ks for input protection in a design I'm just sending off today. At zero bias, they look like 1k resistors.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

Right. A pair of back-to-back LND250s makes a great input protector for a signal conditioner. It limits input current to about 1.5 mA up to hundreds of volts, but behaves like a < 2K resistor normally.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

situations.

We need a dual-chip version in some tiny package.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

First time I ran into them was in loop-powered instrumentation. They were not widely available at the time. Gradually, they've become more mainstream.

Yes, nice idiotproofing. An extremely challenging spec to otherwise meet without degrading the performance on a low-level input.

past.

situations.

4 pins, 2 depletion fets, provision for the shared crossover resistor or a short. What would be nice is maybe 10 mA, 100 ohm fets.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

also

You bloviated. You gave no specifics. Your Pecksniffian echo listed a bunch of BS.

How about some facts? Please delineate your "concerns". BS doesn't count.

Of course you know you can only be rebutted if you present a tangible statement of a problem, so you won't, or, more exactly, you can't. What a fraud. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85140   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

past.

situations.

A 6-pin SC70 or SOT23 would be great, ideally two dice rather than one. I don't know that you can put two chips in such a small package, so some thick SOI or something like that might work. The sources need to be in the middle, of course, if we want to be able to put in a resistor, which I expect makes junction isolation more difficult.

I agree about the lower resistance, for sure. Sometimes I can't use them because of their Johnson noise. and putting a big bypass across them sort of defeats the purpose. If they made them with gate protection zeners, you could cross couple them like this

D S S D

0----*-------* *-*---------*-* *-------*---0 | | V | | V | | | ------- ------- | | ----- ----- | | | | | *-----------}--RRRRR--* | | | *--RRRRR--------------*

which would reduce the dissipation a bit in fault conditions. (If it weren't for the body diodes, it would do a lot more.)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

also

People, including me, post impersonal technical stuff, up for discussion, and a few crabby farts like you respond with content-free insults. You are incapable of *discussing* electronics. You contribute almost nothing.

Amateurs might not understand my listed concerns; they are welcome to ask questions, to which I will respond in what I intend to be a helpful manner. For you to not understand them is incredible.

Fred's suggestion about the depletion-bootstrapped 78L would have been appreciated, if he handn't coupled it with a lame insult. As it turns out, his reference was worth discussing, in that it has lethal hazards.

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

past.

situations.

I guess semiconductor physics keeps us from having a 10-ohm, 1 mA, 500 volt bidirectional limiter.

I've used a normally-off SSR with some downstream logic (in software) to turn it off when things are overloaded, before the internal mosfets blow up.

There's also a single-resistor trick:

--

John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom timing and laser controllers 
Photonics and fiberoptic TTL data links 
VME  analog, thermocouple, LVDT, synchro, tachometer 
Multichannel arbitrary waveform generators

as

past.

uations.

t -

Interesting, so you tie the gate to the source and for input voltages less than 2 volts or so, it looks like a resistor, above that the current is limited to ~1-2mA. (?)

By back to back you mean two in series, but with D-S polarity reversed? (Then conduction through the body diode??) Or in parallel... which also seems to make one body diode forward biased? (with a resistor in parallel.)

(I've never used a depletion fet.)

George H.

ETS as

fets

d

the past.

situations.

in

1k

or a

  |     |

lt

turn it

OK, I guess in parallel would just let out the magic smoke (As too much current went through the body diode during over voltage.)

So I find FET symbols mostly confusing. Your arrows in the above are on source side, and pointing in the direction of current flow... like an npn bjt?

George H.

.highlandtechnology.com  jlarkin at highlandtechnology dot com