Author Topic: High Voltage Bench Power Supply Design  (Read 15844 times)

0 Members and 1 Guest are viewing this topic.

Offline 001

  • Super Contributor
  • ***
  • Posts: 1170
  • Country: aq
Re: High Voltage Bench Power Supply Design
« Reply #25 on: January 27, 2019, 09:42:50 pm »
It is pitfall
Common power MOSFET is actually a bunch of many small devices in parallel
Any small transistor can be overloaded
So local damage is highly expected
 

Offline BravoV

  • Super Contributor
  • ***
  • Posts: 7549
  • Country: 00
  • +++ ATH1
Re: High Voltage Bench Power Supply Design
« Reply #26 on: January 28, 2019, 09:04:32 am »
Not sure if this fits you, 0-500V at 10ma -> EDN : Regulate a 0 to 500V, 10-mA power supply in a different way

I guess you need to modify to accommodate higher current.



Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #27 on: February 06, 2019, 05:29:11 am »
Again, I'm hoping that by paralleling multiple devices I can cut down on the chances of the pass elements failing. I'm well within the DC SOA for the devices mentioned in earlier posts, and so long as I can work out a decent current limiting function the pass elements should never be subjected to the stresses that would cause them to fail. At the present time, I'm extremely busy and haven't had a lot of time to sit down and design a current limit circuit- the negative feedback loop in my current project (a push-pull 1625 power amplifier with a solid-state phase splitter) is taking longer than I anticipated to fine-tune, so this project has been pushed back a little. That is combined with the rather heavy workload that engineering majors are subjected to.

Also remember, as I mentioned, I will be cooling these devices VERY well, so it's not like they're going to be running at 60C or anything like that.

Lastly, I have no illusions that this is better than doing it with tubes. This is an application where tubes excel, but in my situation space limitations come into play.

The reason I want this to be a regulated supply, rather than just a transformer with a variac on the primary is that I sometimes use the bench supply when I want to take the power supply ripple and voltage sag out of the equation.

 
The following users thanked this post: 001

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1826
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: High Voltage Bench Power Supply Design
« Reply #28 on: February 06, 2019, 09:36:56 am »
Not sure if this fits you, 0-500V at 10ma -> EDN : Regulate a 0 to 500V, 10-mA power supply in a different way

I guess you need to modify to accommodate higher current.



Hi,

I have build this but as expected it is very slow. You could speed this up by replacing the photcell coupler by a small transformer-based supply plus a normal optocoupler. This can also drive more current then. Furthermore, the BU208A could be replaced by a linear MOSFET, now available up to 2.5kV.
 

Offline BravoV

  • Super Contributor
  • ***
  • Posts: 7549
  • Country: 00
  • +++ ATH1
Re: High Voltage Bench Power Supply Design
« Reply #29 on: February 06, 2019, 12:02:09 pm »
Hi,

I have build this but as expected it is very slow. You could speed this up by replacing the photcell coupler by a small transformer-based supply plus a normal optocoupler. This can also drive more current then. Furthermore, the BU208A could be replaced by a linear MOSFET, now available up to 2.5kV.

Mind share your circuit in schematic, as I have difficulty to visualize it.

Btw, you said very slow, do you mean the regulation ? How slow ?

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 4004
  • Country: nl
Re: High Voltage Bench Power Supply Design
« Reply #30 on: February 06, 2019, 03:02:25 pm »
With High voltages Power dissipation increases rapidly with a bit of current, but of course you already know that.

I was also thinking about building some kind of high voltage supply.
My first thoughts are around using an LM2587 Fly-back converter and a transformer of an old PC power supply in "reverse".
For a higher output voltage you can put the 230Vac windings of 2 transformers in series.

It should also be almost trivial to add a few tranistors (or optocoupler) and a Zener that throttles the LM2587 down as soon as the output voltage is 10V or so higher than the output of the lineair regulator at the end.
 

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1826
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: High Voltage Bench Power Supply Design
« Reply #31 on: February 07, 2019, 09:38:23 am »
Hi,

I have build this but as expected it is very slow. You could speed this up by replacing the photcell coupler by a small transformer-based supply plus a normal optocoupler. This can also drive more current then. Furthermore, the BU208A could be replaced by a linear MOSFET, now available up to 2.5kV.

Mind share your circuit in schematic, as I have difficulty to visualize it.

Btw, you said very slow, do you mean the regulation ? How slow ?

Give me some time, I'll post a schematics
  Wolfgang
 

Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #32 on: June 07, 2019, 01:35:17 am »
Little update:

The project is not dead, I've just been extremely busy with work, school, and many other projects. Attached is how I plan to do current limiting in some form or another. The clean and elegant way of doing this would be to connect R1 to the output of the LR8 and use the series pass transistors to do current limiting as well.

The belt-and-suspenders approach is to stick this mess in series with the rest of the circuit. The downside, of course, is that I use at least two (probably three) additional (costly) transistors. It is probably a safer approach, however.

The one other issue I'm working at is the load switch, and how to switch voltages in excess of 500V. Something tells me that you're not supposed to use a standard 250V toggle switch for this.
 

Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #33 on: June 16, 2019, 06:54:21 am »
Alright, the design has seen a revision. The LR8 is a great little device, and would be fine for most applications, but I want more than 450V. In addition to that, I think I can learn a lot more by going without it.

Here is what I came up with. Not a complicated circuit, and I'm sure by no means unique. Sorry the schematic is such a mess, it is still a work in progress.

Transformer tap switching will be accomplished using a 120:200 transformer with a voltage doubler. A relay will be used to switch between the full and 50% voltage points (convenient aspect of a doubler). With a beefy transformer, likely a .5 kVA "general purpose" control transformer (Tek transformer was commandeered for another project) and large caps, I can get decent regulation out of this.  The switching of "taps" will be done on another PCB that is the control board, which will monitor temperature of the heatsink and adjust fan speed accordingly, handle "tap" switching based on output voltage and deal with "fault" conditions (pass element short, over-temperature, insect infestation, etc).

One thing that I personally like about this is that if the pot wiper were to go open, the power supply would drop to its lowest output setting, rather than dumping the full 500V into the load. With this design, I plan to have the power supply adjustable from about 15 to 525 volts- good enough for most tubes.

Adjustable current limiting is a nice feature, but I don't feel the need for it to be very precise. As such, I'm just going to use a 5 ohm rheostat for R11.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: High Voltage Bench Power Supply Design
« Reply #34 on: June 16, 2019, 10:46:43 am »
Source, not drain resistors!

Mind that 47k pullup into huge gate capacitance will give quite low slew rate.  So low you may have problems with ripple rejection.

I don't get the 24V supply just for the opamp.  Why does it need to be regulated?  Why does it need to be so high?  It could be anywhere from 3.3V (use precision 1.24-2.5V reference; downside, Vbe of level shifter is a larger fraction of Vout) to 9V (pretty much everything as shown, resistors adjusted as needed).  Supply can be an unregulated DC wallwart, or a tiny SMPS, or maybe even something obscure since so little current is needed (e.g., coupling caps from mains, into a rectifier and shunt regulator, load side grounded -- a few mA won't trip an RCD/GFCI).

Not that 7824s are expensive or anything, but on that note, a shunt regulator with a pair of 6.3V diodes in series would get you a nice pre-reference for low noise and ripple, without the bulk of a regulator and associated caps.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: High Voltage Bench Power Supply Design
« Reply #35 on: June 16, 2019, 07:20:01 pm »
With RV1 connected that way it will have a non-linear output voltage versus position and will be quite sensitive at high output voltages/low resistance settings.  You might want to consider adding a trimmer resistor in series with RV1 to adjust the maximum output voltage.

I would not be surprised if this circuit oscillates.  I don't see any capacitors to roll off the opamp's response.  The voltage amplifier composed of Q1and R2  is adding a pole to the open-loop response as well as a nominal phase inversion.  The opamp inputs should also be swapped.

I don't see any ballast resistors in the source leads of Q2 - Q4 to help with current sharing.

I would add a resistor in series with Q5's base to limit current under transient conditions.  With high voltage supplies, the current can increase faster than the current limiter can act. eg., connecting a 1 uF cap to the output with the supply set to maximum.
 

Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #36 on: June 16, 2019, 11:59:14 pm »
With RV1 connected that way it will have a non-linear output voltage versus position and will be quite sensitive at high output voltages/low resistance settings.  You might want to consider adding a trimmer resistor in series with RV1 to adjust the maximum output voltage.

I would not be surprised if this circuit oscillates.  I don't see any capacitors to roll off the opamp's response.  The voltage amplifier composed of Q1and R2  is adding a pole to the open-loop response as well as a nominal phase inversion. The opamp inputs should also be swapped.
Op-amp inputs are an error in my schematic- the inverting input is connected to the 6.8V voltage reference. There is a .1uF capacitor from pins 8 to 5 on the 5534, which if I understand correctly serves to roll off the frequency response of said op-amp to enhance stability at low gain.

Just out of curiosity, any suggestions for how to make the voltage output track more linearly with the pot?

The 47K load resistor on Q1 was chosen to enable a less costly FET for Q1, but the slew rate issue is a fair concern and I may be better off dropping that value to more like 10K.

 

Offline peteb2

  • Regular Contributor
  • *
  • Posts: 245
  • Country: nz
Re: High Voltage Bench Power Supply Design
« Reply #37 on: June 17, 2019, 01:07:12 am »
I recently re-committed myself to revisiting an old obsession with vacuum tubes, purely from a hobby standpoint. Ridiculous it would seem because i last had anything to do with the things was back almost 40yrs when i started out with a career in electronics to become a tech. The broadcast station where i trained was fully on valves running but here and there were little islands of transistorized stuff slowly making an inroad and then suddenly the entire place was updated, all tube gear simply dumped. My training courses however covered vacuum tubes with more emphasis than semiconductors probably because the curriculum hadn't been modernized quickly enough...

So i too am now in the HV bench power supply need and have been looking at options for a while.

Certainly having something i can scratch-build, switch-mode design using 'mosfets' etc would be ideal. Sadly over the years i've actually had to make such things work again be they low voltage of high that have failed in the real world, i have seen some of the most expensive devices simply have gone BANG rendering a real repair impossible because of the cost of the actual spares.

So my approach for FWIW will be a "Valves Design", to hunt down a few of those old vaccum tubes i once used (that seem to have ebay increasingly awash with) to do this task of a High Voltage, lowish current bench supply shouldn't be too much of a job.

So far the most frightening cost will be a big mains transformer but we'll see what comes. I can't see the point of building something with very high cost semiconductors that could all be so easily be destroyed for a silly mistake at the hobby bench...

 

Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #38 on: June 17, 2019, 01:47:53 am »
I was originally going to use a big transformer out of an old 500 series Tektronix oscilloscope, but that transformer was commandeered for another project. Two suggestions for your mains transformer:

1) A 120:240 with a voltage doubler. Big control transformers that are a 120/240:240/480 are reasonably common and I see them for less than $50 on a regular basis, but they are big and awkward.

2) Have a look at Antek toroidal power transformers. They are good quality and difficult to beat for the price.

Personally, I think I can design a FET based power supply that will be reasonably robust. We shall see whether or not I am correct.

I thought very hard about doing a tube supply, but the space is a big issue, and I'm going to be employing forced air cooling in order to keep the size of my SS version down. I would probably need four or five 6L6s to deliver sufficient current. There is presently another design thread for a tube-based HV bench supply that has a lot of good information in it and which you may want to look at.
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: High Voltage Bench Power Supply Design
« Reply #39 on: June 17, 2019, 06:16:36 am »
H713, here are two ways to solve the voltage control non-linearity in the existing circuit.  The first is to go back to a fixed resistor where RV1 is and varying R12 instead.  The second is to fix the ratio of R12 to RV1 and linearly vary the reference voltage.

Reducing the value of R2 will increase the slew rate somewhat but have you calculated the power dissipation of R2 & Q1?  Also, the drain to gate capacitance of Q1 will have a significant effect due to the Miller effect.  ie., voltage changes on the drain will pump current into the gate circuit and tend to slow things down.  You may consider tying the gate of Q1 to +24 V and driving the source instead. ie., Q1 in common gate configuration.

Although I don't want to derail you from building this circuit, are you aware of something called the floating regulator?  If I were to design and build a variable HV supply, I'd consider using that topology.  If I can sum up the concept behind it, it's that it considers the +Vout terminal as the common rather than the -Vout.  The opamp can drive the gates of the pass transistors directly without using a high voltage span driver stage.  It's really just an artful re-arrangement of the existing parts.  A complication is that you'll need an additional 24 V or so power supply isolated from the main raw supply to operate the reference and op-amp.  hp, Lambda and others have been using this topology for years.  There's a description of floating regulators at the end of this document: https://booksite.elsevier.com/samplechapters/9780750686266/Sample_Chapters/02~Chapter_1.pdf
« Last Edit: June 17, 2019, 06:20:57 am by duak »
 

Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #40 on: June 17, 2019, 06:38:38 am »
I remember looking at floating regulators when I started this design process, but I'll revisit the idea. I can certainly see the advantage. Eliminated the HV driver is definitely welcome.

Yes, I did calculate the power dissipation of Q1. It will have to be fairly considerable, probably a reasonably tough TO-220 package device. The one specified in the schematic is sufficient with a 47K resistor for R2, but if I drop that much lower I'll have to choose something with a beefier DC SOA.

I'm probably going to breadboard up a sketchy prototype with some IRFP450s (I have them on hand) as the pass elements and see how it performs. If it performs as expected then I will proceed to lay out a PCB for it. LTspice simulations look pretty good, so we shall see how it behaves in the real world.

FWIW, the voltage adjust pot will probably be a 100K 10 turn. I might have some 10K helipots as well, I'll have to dig around.



 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: High Voltage Bench Power Supply Design
« Reply #41 on: June 17, 2019, 04:44:08 pm »
H713, you might be able to use a trick called bootstrapping to improve the slew rate of the driver stage.  Basically, you split the drain load resistor in two and connect the junction between them to a capacitor that connects to +Vout.  What this does is force the supply voltage for the driver stage to track the output voltage and effectively alter the time constant.  From the AC small signal standpoint, the drain load resistance has been halved, so the slew rate has been doubled.  From the DC standpoint, the total series resistance is the same so the power dissipation doesn't change.

BTW, here's a link to a floating regulator project: https://www.neurochrome.com/high-voltage-regulator/


« Last Edit: June 17, 2019, 06:14:07 pm by duak »
 

Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #42 on: June 18, 2019, 03:01:41 am »
Alright, I've redrawn the schematic so that RV1 adjusts the reference voltage. I breadboarded the circuit and the voltage adjust works quite well, however there is one significant issue, which is overcompensation. At an output of say 200V, applying a 1000 ohm load (which makes for 200 mA) results in an output voltage of about 230 volts.

Edit: Ignore the V2 attachment, it contains a significant error. The V2a attachment is correct.
« Last Edit: June 18, 2019, 03:05:05 am by H713 »
 

Offline splin

  • Frequent Contributor
  • **
  • Posts: 999
  • Country: gb
Re: High Voltage Bench Power Supply Design
« Reply #43 on: June 18, 2019, 03:00:39 pm »
For a current limit all you need is a depletion mode MOSFET with a resistor between gate and source - very easy but then you have then same SOA/power dissipation issue as with your regulator devices when the output is shorted. The IXTH6N100D2 might fit the bill - 1000V, 300W 350mA @ 500V DC SOA but costs around $6.

Presumably your regulator cct will provide a variable current limit function so a better solution (for protection, should the regulator fail short cct) might be to use a Bourns Transient Protection Unit (TPU), which looks like it would be perfect for your application:

https://www.bourns.com/docs/product-datasheets/tbu-ca.pdf?sfvrsn=9e4a8a65_38

The TBU-CA085-300-W has 850V peak withstand, (450Vrms continuous) 13 ohms max on resistance, 300mA min current trigger, 600mA max. When the current exceeds the trigger it blocks within 1us, limiting the current to around 1mA.

Not surprisingly the trigger current varies with temperature. It resets when the voltage drops below a threshold voltage, Vreset of about 16V. Best of all they're cheap at less than $2.

A selected TBU-CA085-200-WH might be better with 200mA min, 400mA max trigger.

Bidirectional versions (TBU-DT085-200-WH) are also available for the same price but shouldn't be necessary for this application, but it might help if the PS bridge rectifier were to fail short circuit.
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: High Voltage Bench Power Supply Design
« Reply #44 on: June 18, 2019, 05:37:15 pm »
H713, I'll bet it's oscillating.  Do you have an oscilloscope to see what's happening?  If not, put an AC voltmeter on the output to see if there are any spurious signals.

The NE5534 opamp is not guaranteed to work correctly if the input voltages get too close to the power supply rails.  On the data sheet this is the Common-mode Input Voltage Range.  In the latest design, the reference voltage applied to the inverting input will certainly go out of range.  At 200 V out, the ratio of R13 to R14 will put the non-inverting input right on the edge as well.  If you don't want to use a rail to rail opamp, you could make the NE5534 work by providing  a negative power supply a couple of volts below GND.

This could be done simply by using four rectifier diodes in series with the 24 V supply circuit GND connection and the general GND connection, ie., rename the two GND nets in the 24 V supply as -2.4V and connect the cathode of the rectifier stack to it.   U1 pin 4 would connect to -2.4V.  U1 pin 4 should have a 0.1 uF to GND for AC bypassing.

I think R7 to R9 are too low to help share the load current much.
 
The following users thanked this post: H713

Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #45 on: June 19, 2019, 06:17:21 am »
H713, I'll bet it's oscillating.  Do you have an oscilloscope to see what's happening?  If not, put an AC voltmeter on the output to see if there are any spurious signals.

The NE5534 opamp is not guaranteed to work correctly if the input voltages get too close to the power supply rails.  On the data sheet this is the Common-mode Input Voltage Range.  In the latest design, the reference voltage applied to the inverting input will certainly go out of range.  At 200 V out, the ratio of R13 to R14 will put the non-inverting input right on the edge as well.  If you don't want to use a rail to rail opamp, you could make the NE5534 work by providing  a negative power supply a couple of volts below GND.

This could be done simply by using four rectifier diodes in series with the 24 V supply circuit GND connection and the general GND connection, ie., rename the two GND nets in the 24 V supply as -2.4V and connect the cathode of the rectifier stack to it.   U1 pin 4 would connect to -2.4V.  U1 pin 4 should have a 0.1 uF to GND for AC bypassing.

I think R7 to R9 are too low to help share the load current much.

Yep, it was oscillating all right. Looking at the output on the scope, it probably a few hundred kilohertz. I should have guessed this by how twitchy the pot was- just touching the pot shaft caused the output to drift.

I added a 470 pF from the output of the op amp (pin 6) to the inverting input (pin 2), and that killed the oscillation- mostly.

This thing does not like capacitive loads. With a .082uF on the output, it draws almost 50 mA and the capacitor itself hums at about 3 kHz. As it turns out, poly film caps can become surprisingly effective transducers!

I replaced the 470pF cap with a 100nF cap, and it considerably improved the stability, but there is still a bit of oscillation into capacitive loads. As a side note, I can put a big electrolytic cap across the output and it doesn't care- it's the smaller film caps with very low ESR that make it go ballistic.

I tried switching the 5534 out for an OPA604 and then a TL071, and none of them were any better. I tried adding a negative rail, and it didn't seem to help much either.

If I can get it stable into capacitive loads, then I think I've got a solid design here. Load regulation is excellent, at least for a simple design like this. Connecting and disconnecting a 250 mA load (at 250V) causes the output voltage to vary no more than .1V. Quite honestly, I had a hard time really measuring much as the most precise meter I have is 4.5 digits, and even that isn't really enough to see what's going on here. I'll call that passable. It is still worlds better than the old Eico 1030. I took apart and cleaned all the pots in the Eico, and replaced the two main filter caps (most of the smaller electrolytics are original and should still be replaced). It is now a fairly usable power supply, but its output varies as much as +/- 2V with a 150mA load. 

So what about the FETs? The proof-of-concept has eased my concerns considerably. I'm testing with an IRFP450, simply because I have them in stock. I ruled out the IRFP450 from consideration early on due to its 500V limitation and the fact that it doesn't have a DC SOA in the datasheet, however even with some pretty awful things done during testing (intentionally trying to kill it), I was unable to do so. With transformer tap switching, I think the FETs will be running quite conservatively. As mentioned I will probably use the FQA8N90C-F109, though I may try a Fuji 2SK3675 as well. There is no SOA in the datasheet ( :wtf:), but if I do some worst-case-scenario testing and they hold up I would feel comfortable using them. Regardless of what FET I use, this testing will be necessary.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: High Voltage Bench Power Supply Design
« Reply #46 on: June 19, 2019, 12:53:07 pm »
A lot of the classic HEXFETs don't seem to have DC SOA today, but if you look for historic datasheets you'll usually find a curve.  Don't know why they dropped it, would guess it's just one more way to optimize cost on an old jellybean design, not having to do sampled SOA testing.

Last time I cracked open an IRF740PbF, the die size was huge, probably not much changed from the original.  I tested it to something like 140% the datasheet rating, at 300V (which is possible because they give RthJC(max), not typ).  Would expect similar from IRFPxxx's.

The capacitance of those old designs is massive though.  You may need a driver just to get enough speed to reject mains ripple.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1826
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: High Voltage Bench Power Supply Design
« Reply #47 on: June 19, 2019, 02:03:56 pm »
A lot of the classic HEXFETs don't seem to have DC SOA today, but if you look for historic datasheets you'll usually find a curve.  Don't know why they dropped it, would guess it's just one more way to optimize cost on an old jellybean design, not having to do sampled SOA testing.

Last time I cracked open an IRF740PbF, the die size was huge, probably not much changed from the original.  I tested it to something like 140% the datasheet rating, at 300V (which is possible because they give RthJC(max), not typ).  Would expect similar from IRFPxxx's.

The capacitance of those old designs is massive though.  You may need a driver just to get enough speed to reject mains ripple.

Tim

The reason for no DC SOAR shown on new devices is that is so small. The old, large dies are not prone to hotspotting, the new ones are, and it getting worse with every generation. For linear apps, you need DC-qualified *linear* MOSFETs, e.g., from IXYS. For some theory, try "Spirito Effect".
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: High Voltage Bench Power Supply Design
« Reply #48 on: June 19, 2019, 02:28:24 pm »
The reason for no DC SOAR shown on new devices is that is so small. The old, large dies are not prone to hotspotting, the new ones are, and it getting worse with every generation. For linear apps, you need DC-qualified *linear* MOSFETs, e.g., from IXYS. For some theory, try "Spirito Effect".

This is not true in general, as has been discussed many times before.  Latest generation SuperJunction transistors apparently have full DC SOA once again, despite having higher power density than anything ever before!

This only seems to be true of earlier generation MOSFETs, which were higher power density than the classic generation (taking HEXFETs as "classic", say), and which are now going obsolete.  (I think IXYS PolarHV was an example of that low-SOA generation that comes to mind?)

Clearly, there's something about the SJ design that can be changed, with little impact on switching performance or what have you, and they've taken to optimizing it, giving the wide DC SOA.  Whatever it is, it's nice that it's available!

I've also seen IGBTs with DC SOA -- I wouldn't be so inclined to take that at face value, I would definitely test it before releasing a product dependent on it -- but they are putting it out there as such!

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline H713Topic starter

  • Regular Contributor
  • *
  • Posts: 180
  • Country: us
Re: High Voltage Bench Power Supply Design
« Reply #49 on: June 19, 2019, 03:03:42 pm »
Any suggestions on improving stability into capacitive loads?

Also, is anyone here familiar with the 2SK3675? Would it be decent in this application? Allied Electronics has them on clearance for less than $2 each, so if they're tough enough,  that would be great.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf