How to get a negative 1000V at 0.1mA?

[Artlav]

I got a photomultiplier tube that takes a negative 1000V to drive and takes about 80 uA of current.
I need to provide this from a 12V supply.

So far i've been doing that with a push-pull converter out of another project, fed over an isolated 12V->12V DCDC brick, which is an extreme overkill, bulky and pricey.
I've looked around for pre-made DCDC power modules that can produce -1KV, and while there are some from XP Power  and Traco, dear god are they expensive.

The question now is - how to make one?
It has to be regulated, with low ripple, and i have a strong preference for something that won't require winding inductors/transformers.

I'm not sure even what topology to go for, and google was surprisingly unhelpful. There are some variants of a boost converter that can do negative voltages (buck-boost?), and it appears to be possible to attach a voltage multiplier to one.
Would that work in practice?
Or is there a more straightforward solution?

[Seekonk]

There are plenty of 300V type inverters on ebay. Most are not isolated and neg is referenced to ground. a multiplier could still be used to multiply to neg HV. Look at LCD TV inverter circuits and use the transformer used to supply HV to fluorescent tubes and multiply that.

[jeroen79]

Why a negative kilovolt? Negative to what?

Voltage is relative.
If you have something battery powered then   it will be negative or positive depending on what you call zero.

If you drop the 'negative' would you then be able to find a suitable design?

[Siwastaja]

Such high voltages are often produced using isolated topologies (which then may or may not internally connect one of the terminals to the input ground, getting rid of the isolation!), meaning using these topologies it doesn't matter if it's negative or positive, as long as you have a way to choose which end of the transformer you ground.

I'd do it with a flyback.

[NiHaoMike]

Is efficiency a big concern? The easiest solution would be to hack a CCFL inverter. If high frequency HV diodes are hard to find in your area, a 240V transformer used backwards in an inverter circuit with a voltage multiplier connected to it can work.

[JS]

Flyback is the most reasonable topology, you could get the inductor from am old tv and be close enough but from 12V to 1kV without a transformer or dual inductor with the correct turns ratio isn't easy to get. Wounding your own is probably the best so you can get a reasonable voltage at the output to regulate it later from a fixed 1000V+ or with feedback around your converter.

The best option there would depend on
-load changes (the consumption is steady or the load changes all the time)
-input voltage range (are the 12V fixed or it's a battery that changes by a few volts during it's discharge?)
-how precise you need the output to be?
-Do you need to be able to change that voltage while working?
-How much ripple and noise you can afford to have?

All those variables would make to the best choice, as you could use a fixed duty cycle and then use a linear regulator or regulate with the duty cycle as SMPS do, or a combination of both as clean efficent supply do.

JS

[ebclr]

This product may be is what you are looking for 

https://www.ebay.com/itm/DC-6V-12V-1000V-1500V-High-voltage-Power-Boost-Step-up-Module-Generator-Board/172793939362?hash=item283b5245a2%3Ag%3AWcIAAOSwwvZZUGOj&_sacat=0&_nkw=1000V+coverter&_from=R40&rt=nc&LH_TitleDesc=0

[Alex Nikitin]

Have a look at this appnote:

http://www.analog.com/media/en/technical-documentation/application-notes/AN118fb.pdf

Flyback converters are not usually recommended for PMTs, a resonant Royer converter is much better and can be easily controlled. PMTs require a low noise and good stability supply, accurately adjustable in a good range (usually 300-1500V).

Cheers

Alex

[Wolfgang]

If you want to homebrew this, a simple way is to use a Royer converter on a 6V small PCB line transformer (1.5VA, e.g.), then a voltage multiplier to obtain ca. 1.2-1.3kV.
An RC ripple filter with 1-2uF caps and 100kOhm resistors should do (takes a while to charge but ripple is very low).
Regulation can be done by sampling the output voltage by a resistive divider, then regulating the supply voltage of the Royer converter.
There are some CCFL circuits by Jim Williams showing this principle.

I tried such a circuit for an isolation tester (1kV, 100uA max) and a Geiger counter tube supply, both worked well.

[Artlav]

Why a negative kilovolt? Negative to what?

Because the tube requires electrons to flow from the photocathode towards the anode, which requires it to be at a 1KV below the anode.

Sure, i can use a positive 1KV supply, but then i'll either have to have two sets of batteries or power it over an isolated brick, since the tube is connected to the rest of the circuitry (preamp, ADC, etc).

Is efficiency a big concern? The easiest solution would be to hack a CCFL inverter.

Not really care about efficiency, unless it's completely dreadful.

I've considered CCFL inverters, but they are unregulated and designed for constant current operation, so i don't see an easy way of making one produce a stable voltage.

Flyback is the most reasonable topology
...
-load changes (the consumption is steady or the load changes all the time)
-input voltage range (are the 12V fixed or it's a battery that changes by a few volts during it's discharge?)
-how precise you need the output to be?
-Do you need to be able to change that voltage while working?
-How much ripple and noise you can afford to have?

Where can i find examples of a flyback design producing a regulated negative voltage?

On the list,
-Load is almost perfectly stable per tube, but not between tubes and between days.
-Input is a battery, so 11-14V or so.
-Very precise. Amplification depends exponentially on the bias voltage, so being a 100V off would make the signal be off by a factor of 10.
-No need to change the voltage past the initial setup.
-The less the better

This product may be is what you are looking for 

It's unregulated.

Have a look at this appnote:

So many convoluted self-wound transformers...
I would strongly prefer to avoid this unless there is no other way.

[In Vacuo Veritas]

http://www.ceratecinc.com/transformer.php

[mikeselectricstuff]

Why a negative kilovolt? Negative to what?

Relative to the incoming supply ground, obviously

[JS]

Flyback is the most reasonable topology
...
-load changes (the consumption is steady or the load changes all the time)
-input voltage range (are the 12V fixed or it's a battery that changes by a few volts during it's discharge?)
-how precise you need the output to be?
-Do you need to be able to change that voltage while working?
-How much ripple and noise you can afford to have?

Where can i find examples of a flyback design producing a regulated negative voltage?

On the list,
-Load is almost perfectly stable per tube, but not between tubes and between days.
-Input is a battery, so 11-14V or so.
-Very precise. Amplification depends exponentially on the bias voltage, so being a 100V off would make the signal be off by a factor of 10.
-No need to change the voltage past the initial setup.
-The less the better

Then 100µA isn't too much for a linear regulator, even regulating 200V, pass transistor must live up to the task but that's not hard to find.
The problem you have is the line regulation, as your input changes quite a lot, you could regulate it down, then use a flyback with a fixed duty cycle, use a linear regulator to keep the voltage at the output stable and clean.
If you want efficiency you could use some feedback across the switching stage, that makes it deal with the line regulation and get a better high voltage source to start with, then the final linear regulator will take care of the edge, but you might be fine regulating 50V or less at that stage.

This product may be is what you are looking for 

It's unregulated.

You could use something like this for the topology I described but I don't know how well or close it will remain, 1500V output is a lot and you probably can't adjust it, even if not regulated, or maybe it just gets X times the input, so you adjust the input so the output is 1100V or something and then add the linear regulator after. With this you will avoid having to deal with any transformers, but the one in the module.

JS

[HighVoltage]

Ultravolt makes a perfect product for you

Model: 1V15-N0.8
15V input
- 1000V output, 0,8mA

[Siwastaja]

Where can i find examples of a flyback design producing a regulated negative voltage?

Any standard isolated flyback design. Being isolated, you can freely connect the output+ to your ground reference.

Consider outputting, say, 1050-1100V and using a linear regulator to go further. Both for transient response and noise.

[Artlav]

Ultravolt makes a perfect product for you

At $270 a pop it's anything but perfect.

[BrianHG]

A10P-12 from XP Power.
https://uk.rs-online.com/web/p/products/1238299/?utm_medium=buyNow&cm_mmc=aff-_-uk-_-findchip-_-1238299&utm_source=supplyFrame

[HighVoltage]

Ultravolt makes a perfect product for you

At $270 a pop it's anything but perfect.

I bought a few used once over the years for 10 to 20 Euro each
Is this for a production run or a one off project

[BrianHG]

You can design your own with one of these transformers:
https://www.digikey.com/product-detail/en/wurth-electronics-midcom/7508110151/1297-1122-ND/4800054

You will need to make your own oscillator driving the secondary side, and on the primary, you will need to double the voltage with 2 diodes and 2 caps.  This will get you in the 1kv range.  Your method of regulation will need some work.  It can be cheap as Zeners shunting the output, or, using a real switchmode controller for the input side.

[chris_leyson]

Bourns make some nice CCFL transformers, http://uk.farnell.com/bourns-jw-miller/pm61300-4-rc/transformer-smd-ccfl-6w-13v-to/dp/1795421 Although the Bourns PM61300 is not recommended for new design. You can even get CCFL inverters from Mouser and maybe Digikey. If you add a diode and cap to the output and change the current sense feedback to voltage feedback then job done. Somewhere around here I've got a modified CCFL inverter board that generates +1.2kV for a small CRT.

[reboots]

Here's a power supply intended to drive a Hamamatsu R1548 photomultiplier tube at +1250V, from a 2011 effort to create a portable nuclear analyzer. If you are familiar with this application, you will understand that it requires very good voltage regulation (and you may recall the significance of the year 2011). The effort never progressed past this point, but the power supply prototype appeared to perform well. Perhaps you can use this circuit topology as the basis for a negative supply.

The power supply uses a LT1054 switched capacitor voltage converter to drive a voltage multiplier through an off-the-shelf Eaton/Coiltronics CCFL backlight transformer. The CTX210659-R transformer is sadly obsolete, but its characteristics were not especially exotic. Recovering a signal from the anodes is an exercise for the astute observer, and probably involves a large film cap.

All design credit is due to AERCO Electronics. I was responsible for the physical implementation, and can't offer any additional information or application advice. Have fun!

[Seekonk]

I haven't used a PMT since '86 when I greatly increased the dynamic range of the tube. I like that upside down approach.

[Ian.M]

Microchip's LR8 linear regulator can handle an input to output differential of up to 450V.  Its usage is similar to a LM317, except at a much lower current and much higher voltage.

Design a floating output inverter + voltage multiplier to deliver between about 100V and 300V above the output voltage you want, then regulate it down with the LR8, with a 400V unidirectional TVS diode across it to protect it against overvoltage if the output is ever shorted, (+ reverse voltage during shutdown) and ground the positive output terminal to get a negative output on the other terminal.

It will waste a bit of power as the LR8 has a minimum load current of 0.5mA, but assuming you are boosting to 1200V, and size the divider chain to draw 0.5mA,  you can keep its dissipation under 3/4W, well within the capabilities of the SOT-89 or D-PAK packages.

[Mazo]

Mains transformer(220-30-40.... whatever volts) to something like 6V.Low voltage windings in parallel driven by class AB amp(don't really see a point in switching design at this power level) high voltage windings in series rectified,filtered and feedback from the output changing the amplitude of the sine applied.(all in all a handful of transistors and op-amps if I am not missing something).Don't forget to calculate the frequency of the sinewave needed to not saturate the cores.Also be careful with the isolation stress as it is over a kV.
Probably the hardest part is the voltage(current) control AM sinewave oscillator(attenuator).Look at jfets 
Or you can just do all that with a MCU if you don't like the analog way.
All numbers cited are just rough mental calculations.

[voltsandjolts]