Diode pump voltage doubler to drive FET

Go To Last Post
45 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm trying to drive an N channel FET from a Mega328 running on 3V (8MHz). I'm having trouble finding any through hole mounting device that will turn on properly at 3V. Since I don't need fast switching, I wondered about using a square wave output from a counter/timer to drive a simple diode pump. Has anyone tried this? Do you think it would fly?

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, a charge pump will work BUT - you can only deliver as much power out of the pump as the source will provide. AVR GPIO pins cannot deliver much power. Charging and discharging the FET gate capacitance IS power and depending on the switching rate and the FET, it can be substantial.

 

There are quite a few charge pumps out there on the market. Doublers, Inverters, Doubling Inverters, and more. There are also a number of FET drivers that have charge pumps built in. I would choose any of those over a scratch-built AVR powered device.

 

If you don't need a really big FET, there are quite a few that can be driven from 3V logic and in an SOT-23-3 package. BSS-138 is one N-channel part. Those are about the easiest SMT devices there are. I can solder them even without a magnifier (though, maybe not for much longer). 

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

Last Edited: Fri. Oct 13, 2017 - 05:40 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for the reply. I need a fairly hefty FET, it needs to be able to handle around 250mA, which is not a lot, but it's switching a motor that runs from 1.5V, so I need a pretty low on resistance.

It's a one-off, so I might try my diode pump from a square wave output. I can remember doing something similar many years ago on a PIC to produce a negative voltage for RS232.

Unless I'm completely wrong, the gate capacitance shouldn't be a factor, since the on/off cycles will be seconds rather than microseconds.

 

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks. Have ordered some NSS12201L parts. Looks hopeful.

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Yes a simple diode/cap voltage doubler (look up term charge pump) would work fine...at high motor currents (say 20 ..30 amps), you'd have to worry about the ramp on of the fet as the doubler "slowly" built up voltage---since the ramp on dissipates a lot of heat at those currents.

At 250ma, that is not likely a concern (unless you were ramping up 100000's of times a second)   ...30 amps vs 0.25A...the heating is 14400 times greater (or less depending on your view)!

 

Anyhow, take a look  the PSMN022 FET for your generic "junkbox" needs.  Works good at 3.3V gate drive

 

Better yet, look at the ZXMS6004...it is cheap & offers many good things like short circuit protection & overtemperature protection---not that you'd ever short something out!!

https://www.diodes.com/assets/Datasheets/ZXMS6004N8.pdf   

 

You can also get it in sot-23 package , and adapter boards are easy to get for soldering/prototyping either one.  with your low currents, you could easily solder on 3 protoboard wires direct to the pins.

 

 

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Sat. Oct 14, 2017 - 03:49 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks,

I have also ordered some of the ZXMS6004 parts. They look perfect.

 

John

 

Four legs good, two legs bad, three legs stable.

Last Edited: Sat. Oct 14, 2017 - 11:12 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The voltage doubler is the front end of a Cockroft-Walton circuit.  If you need more voltage try a Google on it.  I have one that is generating in excess of 10kV from a 5V input.  But be careful, I accidentally fried a USB port on my laptop, and apparently fried my AVRISP mII at the same time, as it doesn't seem to work any more.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have one that is generating in excess of 10kV from a 5V input.

You've made quite an alarm clock!

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:
You've made quite an alarm clock!

It's true.  It woke me up a couple of times when I put my fingers in the wrong place

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

is generating in excess of 10kV from a 5V input

 

 apparently fried my AVRISP mII at the same time 

 

So that's covered under warranty, right?  cheeky

 

JC 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:
So that's covered under warranty, right?
Quote:

I never thought of that.  What a good idea.  Maybe I can send it in and get a new one.  It's only about 7 years old.  AS7 sees in but can't connect.  AS6.2 doesn't even see it.  I'm glad I was using it instead of my ICE when I picked up the board in a daze and put my fingers on the bare wires on the back.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

MarkThomas wrote:

The voltage doubler is the front end of a Cockroft-Walton circuit.  If you need more voltage try a Google on it.  I have one that is generating in excess of 10kV from a 5V input.  But be careful, I accidentally fried a USB port on my laptop, and apparently fried my AVRISP mII at the same time, as it doesn't seem to work any more.

That must use a lot of diodes!

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If one of the (FE)transistors works OK, I'll abandon the charge pump, as I'd like to keep the AVR in a low power mode most of the time.

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

John_A_Brown wrote:
That must use a lot of diodes!

Two diodes and two capacitors for each stage and I have 19 stages.  It will spark a 7mm gap.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

How do you get 10kV from 5V with "just" 19 stages? There must be also an inductor or transformer somewhere, right?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

El Tangas wrote:
How do you get 10kV from 5V with "just" 19 stages? There must be also an inductor or transformer somewhere, right?

 

Yes, there is a 10:1 transformer.  The primary side is fed 5V through a 2n2222 transistor, and when it is switched off the back EMF through the primary is limited to the breakdown Vceo of the transistor of about 40V, so theoretically the secondary produces 400V.  I don't think it is really that big, but big enough.  I switch the transistor on and off with an ATmega328P, so I can use the UART-USB-PuTTY to play with the on time and the pulse rate to maximize things, which is how some of the 10kV got into my USB port and AVRISP mkII. 

 

I am surprised continually exceeding the Vceo of the transistor doesn't destroy it, but a knowledgeable friend says if the current isn't too big it wont fry the transistor.  I have run it for hours and hours watching the spark and it is still working, but I noticed recently the ATmega is getting really hot, so it might be time to replace it

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Maybe the question should be why isn't the out put 2.6 MV with 19 stages. Theoretically, with ideal components, each stage should double the voltage.

Letting the smoke out since 1978

 

 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

digitalDan wrote:
Theoretically, with ideal components, each stage should double the voltage.

No, each stage does not double the voltage.  Each stage adds the amount of the first stage, which is the doubler, so it would go 400, 800, 1200... not 400, 800, 1600...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What's that smell? ... hot fiberglass resin

There's a photo of the front panel of a PC off into the interior at a PCB fire (short between ground and power planes and the power supply didn't current limit)

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

MarkThomas wrote:
I am surprised continually exceeding the Vceo of the transistor doesn't destroy it, but a knowledgeable friend says if the current isn't too big it wont fry the transistor.
Likewise for Vbeo which is used to generate sub-nanosecond rise-time pulses (test high bandwidth probes, EMI generator)

MarkThomas wrote:
I have run it for hours and hours watching the spark and it is still working, but I noticed recently the ATmega is getting really hot, so it might be time to replace it
Current from collector into base thru mega328P's ESD diode into the power rail.

That diode will become leaky then eventually short.

May have a Vcc over-volt that the mega328P's Vcc shunt is dissipating.

Most voltage regulators are one quadrant (current source only)

A few regulators are two quadrant (current source and current sink)

Recommend adding a shunt regulator to Vcc to limit Vcc to the AVR max of 6V; IIRC, a jellybean 431 can sink up to 100mA.

Or, build a two quadrant voltage regulator; there are amazing high current and high bandwidth op amps.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

MarkThomas wrote:
Maybe I can send it in and get a new one.
Atmel AVRISP mkII is long past its EOL.

There's one clone (Waveshare) and several or many function-like (LUFA AVRISP2)

Some on this forum operate the clone with likely many more operating a LUFA AVRISP2.

 

https://www.waveshare.com/wiki/USB_AVRISP_XPII

eBay Stores

Waveshare

AVR Tools

Programmers | Debuggers

http://stores.ebay.com/WaveShare/Programmers-Debuggers-/_i.html?_fsub=2822958010&_sid=826039540&_trksid=p4634.c0.m322

http://www.fourwalledcubicle.com/AVRISP.php

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
Current from collector into base thru mega328P's ESD diode into the power rail.

I am running the device off a 5V battery.  I set it on top of an old transistor tester that has a metal case and when I turned it on I noticed instead of the spark jumping across the spark gap it was jumping from the power plug to the metal tester case, and then probably to the underside of the perfboard.  I suspect that got high voltage into all kinds of places it shouldn't go.  Like you suggest it probably fried the ESD protection diode.  I programmed another ATmega328 and put it in and now it is working again.  I am going to just let it run like this for a while and see if the same thing happens.  If it does I will think about your suggestion.

 

gchcapman wrote:
Recommend adding a shunt regulator to Vcc to limit Vcc to the AVR max of 6V; IIRC, a jellybean 431 can sink up to 100mA.

What about just using a 5.1V Zener.  I have a bunch of those.

 

gchapman wrote:
Atmel AVRISP mkII is long past its EOL.

Mine is certainly past it's EOL.  I have been using it to save on my ICE, so now I am using the ICE and everything is good.  I am being especially careful around high voltage.  I have a Geiger counter project going too, which is how I got started with all this high voltage stuff.  It runs at 390V, but it is running when I program the ATmega that controls it and compute counts/min, so I am being really careful I don't fry my ICE too.  I tend to be a little cavalier with my projects and plug in power backwards more times than I like to admit, and sometimes put 12V into a 5V place.

Last Edited: Mon. Oct 16, 2017 - 02:48 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

MarkThomas wrote:
I set it on top of an old transistor tester that has a metal case and when I turned it on I noticed instead of the spark jumping across the spark gap it was jumping from the power plug to the metal tester case, and then probably to the underside of the perfboard.
Have you considered an enclosure for it?

Reasons :

  • Safety
  • You're running out of spare parts wink

The stored energy inside CRTs was enough to kill some TV technicians.

IIRC, it's 10microA across one's heart to begin fib.

If the enclosure is a no go then consider PCB feet; there may be enough of an air gap.

MarkThomas wrote:
What about just using a 5.1V Zener.
That'll work though Zeners are leaky.

The 5V battery is likely 4 series NiMH cells so simply recharge it.

Another good Zener to stock is 6.2V as that's stiff and would be well away from a charged NiMH 5V battery at room temperature.

MarkThomas wrote:
Mine is certainly past it's EOL.
laugh

MarkThomas wrote:
It runs at 390V, but it is running when I program the ATmega that controls it and compute counts/min, so I am being really careful I don't fry my ICE too.
An assumption is it's shared ground (390V, Atmel-ICE, USB hub, laptop)

The laptop is in some risk if/when the 390V reverses polarity during a fault.

Olimex's isolated AVRISP is STK500v2.

 

https://www.olimex.com/Products/AVR/Programmers/AVR-ISP500-ISO/

 


https://www.altex.com/3M-Bumpon-Protective-Bumper-Pad-Clear-SJ-5312-P146150.aspx

https://www.altex.com/3M-Bumpon-Protective-Bumper-Pad-Clear-SJ-5303-P144138.aspx

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
Have you considered an enclosure for it?

I have, but I don't know how much longer I am going to play with it to make it worth while.

 

gchapman wrote:
consider PCB feet

My perfboard has feet, however one of them is pretty close to one of the capacitor wires, and I get shocked if I touch it.  I have some nylon feet, and now that you mention it I think I will replace the aluminum ones with nylon.  That should help.

 

gchapman wrote:
Another good Zener to stock is 6.2V as that's stiff and would be well away from a charged NiMH 5V battery at room temperature.

I have a big bag of 5.6V Zeners someone gave me.  That is a little better.  Maybe I will just solder one in.  The 5V battery I have is one of those rectangular things my wife got at a conference that is supposed to be used to charge a cell phone in a pinch.  I'm not sure what is inside, but I recharge it off the USB port on my laptop.  It has a regular USB A port for output and USB micro for charging.

 

gchapman wrote:
An assumption is it's shared ground (390V, Atmel-ICE, USB hub, laptop)

Yes, all shared ground.  I run it off a 9V battery and it has an LCD screen to show counts/min and battery voltage.  I'm using a Chinese Nano to run the thing, so the only danger is when programming the Nano or setting parameters.  I pretty much have all the parameters dialed in so I don't need the USB-PuTTY connection any more to change voltages and such.  Maybe I will take the Nano out of its socket if I need to reprogram it at some point in the future so at least my ICE will be safe.  There is still the USB port worry if I need to talk to it.  O well, that little laptop has been through the wars and still has 2 USB ports working.  What's life without a little risk?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Thanks. I think 10kV shoulod be enough to turn on my original IRF520.

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, you could use the first stage of the 10kV generator to charge up the MOSFET gate. Power up the inductor, then close the switch transistor, limiting the flyback with a proper zener instead of the transistor breakdown.

 

It either works or punctures the gate, killing the MOSFET cheeky

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Not sure what the best way to do it is.  I got the idea out of a Nuts&Volts article.  My knowledgeable friend calls it brute force.  I am working on a more sophisticated circuit to power my Geiger counter.  The 10kV was just an experiment, and I like watching the sparks.  I wonder if using a Zener isn't pretty much the same thing as using the transistor breakdown, although the Zener is designed to break down.  The transistor I am using has been in there a long time and still seems to be doing it's job.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, it is "brute force", but how else would you do it? I think it's a good way. But in this case I was replying to the OP. Maybe he could use the flyback (limited by a zener) to power the MOSFET gate, or would that be insane even to try? Who knows...

 

Anyway, I'm curious if you observed a loss in the gain of the 2N2222 over time? I heard that subjecting BJTs to many breakdown cycles causes them to lose gain, that sometimes can be partially recovered by annealing at maximum rated temperature (that is 200 ºC or so) over a few hours.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

El Tangas wrote:
I'm curious if you observed a loss in the gain of the 2N2222 over time

I have an old Tektronix Type 575 Transistor Curve Tracer, so I can actually measure the gain.  I could measure the gain of a new transistor, put it in the sparker and run it for a few weeks, and then measure the gain again to see any change.  I will do that for you. 

 

Actually, I am using obsolete TN2219 transistors in a TO-92 like package, which have pretty much the same specs as a 2n2222, but with higher maximum current, probably due to the metal heat dissipater sticking up like on a TO-220.  I bought several hundred of them at a liquidator some years ago and use them for everything.  I just said 2n2222 because everyone knows what that is, and it is very similar.  I will make that measurement and let you know in a couple of weeks after I get a bunch of breakdowns.  It is running at about 2 kHz.

 

mark

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I was attempting humo(u)r.

 

I remember building a Cockroft-Walton circuit many years ago as an ioniser, back when having a stream of negative ions was regarded as a good thing. It was a project that end up gathering dust, both literally and figuratively.

I built in in a red plastic box about the size of a small lunchbox, which had housed the line isolation section of a 1200/75 baud modem made by GE for the British Post Office "Viewdata" system, which soon changed its name to "Prestel" before disappearing without trace. It could be considered the forerunner of the web in some ways. The problem was that nobody was interested in being a subscriber, as there was hardly any information on line, and no body wanted to provide information, since there were no subscribers to consume it. The French version lasted a lot longer, as the French Post Office stopped issuing paper telephone number directories, and supplied a terminal to each household instead, thereby breaking the chicken-and-egg-log-jam, which is, incidentally, excellent spread on hot toast.

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Actually the TN2219a data sheet has TN2222a in parenthesis, and that data sheet shows the heat dissipater and calls the package a TO-237, which looks to me like a TO-92 with the heat dissipater.  I bought 3 or 400 of them at a penny a piece, figuring it would be a lifetime supply.  The 2n2222a are a little more costly, and pretty much the same.  I like the higher current capability of the 2219.  The TN22219a data sheet says 1A continuous current.

 

John A Brown wrote:
The French version lasted a lot longer[/ quote]

I remember being in Paris right after everybody had terminals in their houses.  We used it to make train reservations, and to look up stuff.  I think that was the beginning of the web, as there was actual content available.  I remember thinking it was a pretty cool idea.  I never did anything about it.  Much like the time I was offered a chance to invest in a car phone company in the late '70s.  I thought that was a crazy idea.  Who would want a phone in their car.  Now I am not a rich man.  O well, I am happy enough, and I have hundreds of transistors and diodes and capacitors.  That said, I got a big box of older capacitors from my friend that I had to sort and bag.  There are some green ones that I thought I saw as being the type used in certain guitar amps for that "good" sound, so if anyone needs an odd cap let me know and I might just have some.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

OK, I put a fresh transistor in my Tektronix 575 Transistor Curve Tracer and measured a current gain of about 260 using .02mA per step on the base current.  I will leave the tester set up as it is and put the transistor back in after running it on the Cockroft-Walton circuit for a couple of weeks and see if the gain has changed.  I will let you know

 

BTW, the "5V" battery I got from Wired magazine measures 4V after a full charge.  I wonder if the abuse from the C-W circuit messed with it, or if it has always been that way and is the reason it wont charge a cell phone.  I have another one that measures 5V that I might try to see if I can get a bigger spark, but I dont want to mess it up too.

 

I put nylon feet on my board.

 

That's all for now.  laugh

 

mark

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

MarkThomas wrote:
BTW, the "5V" battery I got from Wired magazine measures 4V after a full charge.
Well, that has the (IIRC) ineffable "infinite value" - theusch

Equivalent that's more expensive, has variable voltage output (cell phone, notebook PC), and recharge while driving the car or truck :

PowerStream Power supplies, battery chargers, batteries and packs, dc/dc converters, injection molding

PowerStream

Universal lithium ion battery pack, industrial version with outputs 5V to 19V

https://www.powerstream.com/PST-MP3500.htm

Useful energy storage 45 watt-hours

...

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

In searching for a low power consumption high voltage GM tube supply I found this Maxim Integrated application note.

 

https://www.maximintegrated.com/en/app-notes/index.mvp/id/3757

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Having that gizmo sparking all the time doesn't interfere with any of your (or your neighbors) electronic equipment?

Letting the smoke out since 1978

 

 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Current's too low

LucidScience

Gallery Project : ROCK DISAGGREGATOR

http://www.lucidscience.com/gal-rock%20disaggregator-1.aspx

...

... it did manage to destroy a laptop and a camera that were in the same building. A 200 Kilovolt pulse at 10 nanoseconds through 10 large inductors is certainly a good source of radiated energy!

...

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"infinite value"  I like that.

 

sbennett wrote:
In searching for a low power consumption high voltage GM tube supply I found this Maxim Integrated application note.

That looks interesting.  The circuit has 7 stages of Cockroft-Walton.  I will look at that harder.  Thanks!

 

DigitalDan wrote:
Having that gizmo sparking all the time doesn't interfere with any of your (or your neighbors) electronic equipment?

Doesnt seem to.  My neighbor runs enterprise routers and nodes, so if it messes him up he deserves it.  For a while he had a node set to actively interfere with unknown wifi signals, and our wifi didnt work half the time.  I dont think there are any hams in the area to pick up the signals.  I cant hear it on my AM/FM radio.  It's running off a battery so it is not getting into the mains.

Last Edited: Mon. Oct 16, 2017 - 10:47 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
A 200 Kilovolt pulse at 10 nanoseconds through 10 large inductors is certainly a good source of radiated energy!

Thats a real EMP!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Cool, thanks yes

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:
Better yet, look at the ZXMS6004...
TI's TIOS101 are also protected NFETs for IO-Link though it can drive a 1.5H inductor.

Texas Instruments

TIOS101 (ACTIVE)

Digital Sensor Output Drivers With Integrated Surge Protection

http://www.ti.com/product/TIOS101/description

via http://www.mouser.com/new/Texas-Instruments/ti-tios101-drivers/

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

TI's TIOS101 are also protected NFETs for IO-Link though it can drive a 1.5H inductor.

Those are fine parts, but pretty lousy drivers for any sort of current (for efficiency)...looks like a 0.15Amp load drops 1.1Volt !!  That's like a 7.3ohm mosfet...These days even 0.1 ohm is somewhat lousy.  For low current, some of the other feature might make it a winner.

 

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

El Tangas, I ran the Cockroft-Walton spark generator for about 8 hours on and off.  I don't usually run it if I am not in the room because I am afraid something might start on fire, but last night I turned it on and went out for about 6 hours and noticed it was on when I got back.  The transistor breaks down 2000 times per second, so 8 hours is almost 58 million breakdowns.  The transistor gain seems pretty much unchanged..  I will run it some more and get back to you with the results

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for the experiment. Meanwhile, I remembered in which context I found those claims about the gain degradation. It was when researching this type of oscillator:

 

As you can see, the transistor is in reverse, so the emitter breaks down instead of the collector. Eventually I found someone who had observed the gain degradation, now I can't find it again... But I suppose this abuse of the transistor might be more damaging than having the collector break down.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

avrcandies wrote:

Yes a simple diode/cap voltage doubler (look up term charge pump) would work fine...at high motor currents (say 20 ..30 amps), you'd have to worry about the ramp on of the fet as the doubler "slowly" built up voltage---since the ramp on dissipates a lot of heat at those currents.

At 250ma, that is not likely a concern (unless you were ramping up 100000's of times a second)   ...30 amps vs 0.25A...the heating is 14400 times greater (or less depending on your view)!

 

Anyhow, take a look  the PSMN022 FET for your generic "junkbox" needs.  Works good at 3.3V gate drive

 

Better yet, look at the ZXMS6004...it is cheap & offers many good things like short circuit protection & overtemperature protection---not that you'd ever short something out!!

https://www.diodes.com/assets/Datasheets/ZXMS6004N8.pdf   

 

You can also get it in sot-23 package , and adapter boards are easy to get for soldering/prototyping either one.  with your low currents, you could easily solder on 3 protoboard wires direct to the pins.

 

 

 

Just wanted to say that the ZXMS6004 devices eventually found their way here from China(must learn to read Amazon listings more thoroughly...).

They are perfect. Work at 3V, big enough for sausage-fingered me to use easily. Take virtually no current, low on-resistance.

 

Thanks to avrcandies for the suggestion.

 

Four legs good, two legs bad, three legs stable.