1. Home
  2. Electronics Design
  3. Low voltage shunt regulator

Low voltage shunt regulator

Please refer to Fig 5,Pg 2, TI datasheet LM431.

What i need is a 2-terminal shunt regulator where the TOTAL current (resistor dividers + active device) is 20uA. Voltage (extended) standoff to beat least 30V, more is better, to decrease total parts cost for a 400V regulator.

The LM431 is nice, because one can add in a JFET to extend the voltage range without adding another resistive divider burden. But, its minimum operating current is 400uA - way too much.

AFAIK, the LT1004 has the lowest operating current of 10uA, but t is not programmable or extendable, an i would need 160 of the buggers for 400V.

The LT1634 also has a low 10uA operating current, goes to 5V and the "teet" pin _might_ be useable for voltage control with a divider. ZERO info n that as well as use of it with JFET/whatever to extend regulated voltage.

The LM185/285/385 spec is right at 20uA, but no control pin.

The LT1389 goes down to 0.8uA, and is available up to 4.098; the "DNC" pins _might_ be useable for voltage control with a divider; again ZERO info.

I am snubbing Maxim,so leave their vaporware out of this.

Once upon a time there was this interesting floating gate FET design; the Xicor X60008, renamed Intersil and part renamed; 0.8uA min op current. Design and specs changed over the years and i have lost touch with it. Why i mention it, is that (the initial design anyway) could be used as a 2-terminal regulator and at some point a 10V part was made (thus requiring 40 units if no added voltage extender used).

Ideas?

Reply to
Robert Baer
Loading thread data ...

Zener diode[s]?

Reply to
Frank Miles

High voltage (eg. 200V) zener diodes are pretty good. Why not use them?

Two BZD27C200P-E3-08 and you're done for less than a dollar.

I'm sure something could be cobbed together with a depletion MOSFET etc. but why bother..

Best regards, Spehro Pefhany

Reply to
Spehro Pefhany

I assume what's needed is a 400 volt shunt regulator.

Another way to do it would be to run a divider into the gate of a mosfet, and put a zener, 20 volts maybe, from source to ground. That allows for some adjustability in the divider.

But if adjustability isn't needed, a stack of plain old zeners sounds good.

--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

Actually, if I recall Bob's history, 400V is just the beginning, only a block stacked for really big shunt regulators (antique TV replacements and whatnot).

Personally I don't see what's wrong with a pair of 200V zeners -- cheap and common enough, and hard to beat the parts count.

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com
Reply to
Tim Williams

I guess if you need really high voltage, a 4500V MOSFET (eg. IXTA02N450HV) might simplify the circuit a bit but it won't likely be cheaper, and you'd still need resistor(s) that would withstand the full voltage reliably.

Reply to
Spehro Pefhany

Then the question is "why".

Tubes work awfully well in such applications.

If you can't get an exact replacement tube, just build something solid state under the cathode to force the desired behavior.

I have the manuals...

;-)

(Though specifying micro-amps does suggest a different situation than "antique TV replacements".) ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| 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.
Reply to
Jim Thompson

Tube color TVs had a CRT-high-voltage shunt regulator, to keep the colors from squirming. 6BK4 was the usual toob, a triode rated for

27KV plate voltage, nice x-ray generator. I guess they are no longer available. 
--

John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

I remember them well... growing up in "Thompson's TV Service", my dad's TV shop in the basement of our hardware store.

They're still available on eBay. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| 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.
Reply to
Jim Thompson

This has 1 us leakage at 800V

formatting link

if you were to use this along with a zener and voltage divider to drive the gate as a feed back I think you'll be just fine, although slightly high current, more so than you most likely need but it does have lowest charge.

I've never had a need for a very low current shunt leakage regulator, which I think that is what you're after? But I've used mosfets in HV circuits and I've always tried to double the handling voltage.

Just a thought../ Jamie

Reply to
Maynard A. Philbrook Jr.

Guess i should have given full requirements: 20uA to 500uA, 400V, low incremental resistance in current range.

One might think that zener diodes would do the job, but it turns out that AFAIK there are NO zener diodes that do NOT oscillate somewhere in that region of current. Some brands of (say) a 15V zener oscillate in one part of that range, while another brand (SAME size, package, rating) will oscillate in another range; a different package same band and rating would oscillate in yet a different range. So: NO OSCILLATION is another spec.

Reply to
Robert Baer

Give me a part number that you KNOW does not oscillate anywhere in the 20-500uA range. If you have a DMOSFET scheme that would do the job, please let me know.

Reply to
Robert Baer

The divider scheme you mention is something that i did a patent app a while ago (see

formatting link
The added FET at the gate is chosen to cancel the VGS of the regulating FET and thermally track that ofset. This scheme works reliably to 185C if the FETs are chosen correctly,and it works fairly well. But i want a much lower incremental resistance; temp compensation and tempco are secondary concerns.

As i mentioned elsewhere,zeners are useless due to oscillation.

Reply to
Robert Baer

Actually,i have working boards for TV use (replace/slip into the Victoreen Corotron(TM)s), 7.5KV, 12.5KV, 20KV, 25KV. They use zeners (that oscillate),those found and used oscillate the least of any.

200V zeners are the s**ts.
Reply to
Robert Baer

Actually,the lower voltage rating FETS seem to be better,especially over temperature. And voltage rating of resistors is not a problem (MiniMOX, or multiple films in series). What is the problem, is that the gain FET does not have enough gain to give a net LOW Z result. There seems to be no model for (power) FETs in the linear region,as the ASS-u-ME-ption is they are for switching, period.

Reply to
Robert Baer

That is not the app; it is to be used as a reference as well as a series constant voltage drop in a measurement system.

Reply to
Robert Baer

Those are triodes,and are not regulators per se; might as well as use a FET (smaller, similar if not more gain).

Reply to
Robert Baer

  • What is the gfs at 400V,20-500uA? In any case the gain is insufficient to give a net low-Z result.

As you might know, "i" make high voltage shunt regulators speced

20-500uA "programmed" to any voltage in 50V steps from 50V up 3750v (2000V is highest ever requested); and there is virtually no noise from nanoamps to a few mils,ZERO oscillation. BUT, for my lab requirements, the incremental resistance is too high.

What about the LT1634 and using the "test" pin for control? Anybody try that? Or the LT1389 and using the "DNC" pin(s)? Anybody try that?

With either one, the first "hurdle" is the control - does the "extra" pin give decent adjustability; the second is - will a JFET or similar work with it (and appropriate resistors) to give a composite higher voltage low Z regulator?

Looks like the Intersil part is crawling toward the unobtainium region. Sad; these were the cats meow, for truly precision voltage references that can get into the saturated reference cell region.

Reply to
Robert Baer

Can you provide any specs whatsoever on noise / ripple, voltage tolerance (initial and drift), regulation/impedance and bandwidth / response time?

More info about the application? All you've alluded to so far is "lab use", which doesn't sound like something that need be miserly with current. (I'd gladly design for you a 400V precision reference that includes zero amps (or less) in the spec, as long as you willfully ignore the wires going back to the outlet.)

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com
Reply to
Tim Williams

Please forgive me; looks like i found the answer. Dug thru more of my 10+year old regulator data sheets, and found the LM285-ADJ; 8uA min plus 10uA for divider makes it.

Toss in a 2N5116 for 30V per stage and get 14 stages for 400V.

So now, maybe one could use a depletion mode MOS FET to get about

300-400V per stage.. Question then becomes: will that work, and is there sufficient gain to get low Z?
Reply to
Robert Baer

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.