Use two bridged to drive say a little PCB-mount 2k - 8 ohm audio transformer "backwards" to charge a HV capacitor, current feedback from each side of the "primary" inductance keeps the drive current waveform from going all sloppy, cross-coupling the G1 pins boosts each amplifier's "open loop" gain.
Seems complex. Why not just poke a sine wave into the primary through a DC block capacitor?
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Yep, but unless you want to wait all afternoon for your capacitor to charge up you need push powah to do that. Hence what you might call a "power amplifier" made out of some cheap parts.
Or you could use a HV switcher/boost converter. That's cool. Switchers are cool. Boosting 12 volts to ~250 at more than a few mA has its own fun time design probs tho
Sure, but it could be a lot simpler.
The LT3803 and one of those cheap dual-winding inductors works pretty well.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
This paper presents a new and novel technique for speed control of a brushed DC motor without employing any direct shaft transducer in its feedback mechanism. This method uses the current/voltage variation produced on the motor main input power lines as the brushes move from collector blade to the next. The frequency of variation is proportional to the motor speed and the rotor position can also be detected by monitoring the changes for each collector blade. The performance of the motor torque speed characteristic using this system as its feedback has been evaluated through laboratory testing. This configuration is very suitable for harsh environment and places with not enough space for the motor with discrete sensors.
I could probably tack on a few CW multiplier stages to the output and make an anode supply for a small CRT. Talk about complicated, just look at this thing this is nuts!
Whoops, IDK how that got pasted in there. Hey, but the paper looks pretty cool though
Those are CW multipliers... maybe you didn't look at the last pic? GH
>
Ok, I see now. JL didn't say what kind of power output he's getting from that circuit, the one I drew will charge up a 4.7uF cap to over 250 volts in a little over 250 mS in the simulation at least.
I'm gonna use it to fire a small xenon strobe repeatedly, I'd like to see what kind of flash rate I can get before the ICs melt down. I don't feel too bad about destroying a couple 20 cent LM386es...
m
It's a well-known problem, with at least one well-known solution.
Peter Baxandall apparently invented the circuit to deal the problems create d by inter-windng capacitance in the secondary of a high-turns ratio step-u p transformer. He seems to have been generating about 1mA at 1kV to drive p hotomultiplier tubes, but while I worked with a guy who had worked with him , I never met the man myself.
The circuit at the bottom of the page is my speculation about how you'd do it today.
talks about the interwinding capacitance in the transformer shown in the ci rcuit diagram and in the .asc file.
-- Bill Sloman, Sydney
om
m
John Larkin seems to have been frightened by a coil-winding machine as a ch ild.
You can design your own transformers and get them wound for you at your loc al coil-winding shop. TDK/EPCOS sell a large range of ferrite cores and for mers (quite a few of which you can buy off the shelf from Mouser).
There are quite a few parameters you can play with when you design your own transformers and inductors, so it's not an entirely trivial exercise, but you end up with lot more options than if you confine yourself to inductors that you can buy ready-wound off the shelf, as John likes to.
-- Bi9ll Sloman, Sydney
In this thing I'm sticking small current-sense resistors in line with the output of each power amp that's driving the step-up transformer, bridged, then sending negative feedback from each one to the other.
It keeps the current through the primary basically sinusoidal when the current swing is large, until the capacitor on the secondary charges up to the point that there's a high impedance looking into the primary, then amplitude of the current drops off, and some crossover distortion starts appearing in the current waveform.
I built a blocking oscillator with a BJT and the 5VSB (stand-by power) transformer from a PC power supply.
I used a 1N4148 (breakdown about 110V) as the "zener" for feedback
It seemed to work alright on the bench. I used it to to test some zeners.
--
?
I need to make a HV (10-300V) bias supply. (only a few mA) My first thought was a CW chain with some sort of FB changing the drive amplitude. But I'm not seeing how to make the feedback simple. Getting close to zero will be hard too. Maybe a fixed 300V and then a fet/resistor voltage divider?
George H.
It sounds like efficiency is not a big deal if it's a bias supply, I'd probably use a lil transformer driven by maybe a 1 transistor oscillator even, feed the primary off a linear regulator and adjust the output voltage on the secondary by manipulating the low voltage supply via the linear regulator's adjust terminal. If the current draw from the secondary is small enough (like hundreds of uA to a couple mA) and more or less constant and high precision of the bias voltage isn't required you might not even need any feedback at all, can probably get within a couple volts either side of the target just through messing with the oscillator supply voltage in that situation.
PWM on the adjust terminal could probably take you down to near zero, at some point the primary supply is just off for most of the time, volt second balance should drop the output into the dirt
this is a true cheapskate thing:
A small ferrite bar from a transistor radio should work OK as the transformer "core"
Thanks, I need pretty tight control. Some fraction of a volt out of 300. My brute farce idea looks like this.
George H.
Ya, I still think linear regulation of the oscillator supply voltage on the primary side is the way to go for small power levels, I think tighter control can still be achieved that way with feedback to the primary side with some kind of isolation.
Folks kind of pooh-poohed my idea to use two cheapo 8 pin uPs talking to each other over i2c through optocouplers to achieve that for a synchronous flyback, but doing it that way when you're just controlling a linear regulator doesn't require the same kind of tight timing and I'd think would be a lot more trivial
Maybe you don't even need strict galvanic isolation and could just get away with some large feedback resistors into an op-amp buffer.
Here's a cute HV regulator:
You can go pure class B, or have a little crossover current.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
OK just switch it on the high side. Those are a little wimpy power wise. 150 mW
George H.
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.