Is this a practical HV supply?

[keantoken]

Hi all.

I had this idea for a 300V supply, for say something using tubes or even in the KV range.

This is very simple, but I haven't seen much of it around (maybe I just haven't looked enough?). Assuming that the transistor is within its limits, is this practical?

It could also be used for say, an avalanche pulser, though you would probably use a switching supply for this in order to use a cheaper transformer.

As an afterthought, if the base current for the cascode affects accuracy too much, we need only drive it with an FET off of a 9V supply.

Also, DC accuracy is not guaranteed unless you adjust with a potentiometer and even then any drift will be amplified by about 120. As better solution would be perhaps to use an adjustable Zener+BJT, but we might be trading accuracy for lower accuracy.

- keantoken

[Minion]

I found this for use with Nixie tubes but I hear works well with Other tubes ....

http://www.ledsales.com.au/kits/nixie_supply.pdf


Puts out 170v DC from 12v AC (so you can use the same Xformer for Plate and heater) and I believe you can get more voltage by useing a bigger inductor and Caps....


CHeers

[keantoken]

This is a strange switching supply I designed (simulated, haven't built yet). I had this crazy idea and it worked... Not sure how it compares to the others, but it doesn't need no 555, just throw it together out of spare parts.

Q3 improves efficiency, but it can be troublesome and isn't needed for the functionality of the circuit.

- keantoken

[lineup]

Good try, keantoken.

I recommend a fairly high voltage precision Voltage reference
and using VbE multiplying technique
you can get a very good high voltage regulatort.

TL431 is one 30 Volt volt precision adjustable zener-diode.
Very low price/performance! is another benefit.
The TL431 output voltage can be filtered RC, to improve parameters
if you need.

30V x 5 = 150 Volt, for example

Assisted with TO126 or TO220 transistor to cope with power=heat



Hi-volt-reg_120v
_lineup2006
Preferably Like my article for some years ago:
http://www.diyaudio.com/forums/showt...68#post1567768

Why re-invent everything when somebody already did some good work
to re.-construct The Wheel



Regards
your
Lineup of sweden

[sawreyrw]

keantoken,

Here are a few comments on the circuits you posted.

Regarding the first one:
1. The circuit should work, but the 115k resistor needs to be increased to 295k to get an output voltage of 300 volts.
2. The base current of the transistor may cause significant error in the output voltage (depending on the current gain), therefore, I would use a MOSFET as you suggested. I would bias the gate of the MOSFET at around 10 volts.
3. The raw DC supply voltage should be a little higher, say 315 volts, to allow for component tolerances.
4. Using 315 volts and assuming the maximum output current is 1 mA, I would increase R4 to 5K. (This is not a big deal.)
5. What is the cap for? Any noise on the raw DC supply will be coupled directly to the TL431 sense pin and upset the output. You might want to connect the cap from the sense pin to ground.
6. You may have stability problems with this circuit.

The second circuit you posted is called a boost mode DC to DC converter. Here are some comments.
1. The output voltage will be highly dependent on the parameters (Is and Bf) of Q4, and will not be very accurate with different devices of the same type. This is because Is and Bf vary significantly from device to device. You need to use a better error amplifier.
2. The 180 uA current source will charge the gate capacitance of M1 very slowly. This means that is will switch slowly and there will be significant switching losses. You need a better gate drive circuit.
3. Many IC vendors (including LTC) make boost controllers that will far outperform a discrete circuit. If you really want to make such a circuit, consider them.

Rick

[keantoken]

Thank you for your comments.

I wasn't thinking about accuracy when I designed the second circuit. If I made it in real life I would use an FET for the voltage sensing transistor and if not at least use a pot to trim the voltage to the desired value.

One of these days I will get around to trying to make a boost circuit that is good enough for tubes so that I can experiment.

My reasons for using C1 in the first circuit were to decrease the impedance of R3 and R1, so that any parasitic properties of the TL431 wouldn't interfere with the voltage sensing. I suppose we don't need this unless we are using a transistor as the voltage sensor.

Thanks,
- keantoken

[keantoken]

Hey, can someone tell me a good method for discharging 200V caps?

I'm experimenting with a circuit similar the second one I posted with a photoflash cap and have been discharging the "fun" way, but that's getting a little old.

I don't have any High-voltage, high-watt resistors.

My ears appreciate your suggestions.

Thanks,
- keantoken

P.S. I have realized that making a good discrete DC booster is harder than it looks, so I'm going to try and make my own 555 circuit. I might still try a discrete one after this one is done.

[sawreyrw]

You probably don't need "high voltage high wattage" resistors to do this. The energy stored in the cap is 1/2 C V^2, so the resistor has to have an energy rating greater than that.

[star882]

Just use some common light bulbs (in series for higher voltage if needed) to discharge the capacitor. Use some smaller ones (line voltage Christmas tree bulbs or nightlight bulbs are ideal) and you'll get visual feedback.

[keantoken]

Okay, thanks.

- keantoken