Variable output voltage SMPS - 450V

[nicoverduin]

Having seen the quality of number of SMPS projects here on the forum, I have a challenge in this one. I have an electronic background int hte Tube audio field and for digital electronics. Now I need a SMPS with:
- 2 x 450V 250mA variable (digitally controlled) output voltage
- 1 x -150V 50 mA variable (digitally controlled) output voltage
- 1 x 30V 4A variable (digitally controlled output voltage

input voltage from 90Vac - 265Vac

I have been going through numerous documents on the web, order a number of recommended books (haven't arrived yet), built a booster (12V -> 220V), but I would certainly appreciate some advice in which direction to look. I have alreday had contact with Christy for first possibilities. No doubt someone will make me an offer I can't refuse, but I intend to learn and understand. But magnetcis never was my strongest point some 30 years ago.

[star882]

Two forward converters for the 450v outputs, a flyback for the -150v output, and your choice of forward or flyback for the 30v output. Run it all from an active PFC stage and adjust the feedback loops with a microcontroller.

[jackinnj]

Originally Posted by nicoverduin

Having seen the quality of number of SMPS projects here on the forum, I have a challenge in this one. I have an electronic background int hte Tube audio field and for digital electronics. Now I need a SMPS with:
- 2 x 450V 250mA variable (digitally controlled) output voltage
- 1 x -150V 50 mA variable (digitally controlled) output voltage
- 1 x 30V 4A variable (digitally controlled output voltage

input voltage from 90Vac - 265Vac

I have been going through numerous documents on the web, order a number of recommended books (haven't arrived yet), built a booster (12V -> 220V), but I would certainly appreciate some advice in which direction to look. I have alreday had contact with Christy for first possibilities. No doubt someone will make me an offer I can't refuse, but I intend to learn and understand. But magnetcis never was my strongest point some 30 years ago.

Sorenson, a division of Raytheon, has several low noise, high voltage SMPS for laboratories. The manuals can be found on the web.

I've scaled up a couple of Linear Technologies low noise push-pull switch mode controllers -- for fixed output operation -- but this has just been experimentation. Take a look at the product folders for their low-noise switchers. The low noise comes at the expense of heat production which has to be dealt with...

Magnetics and transients will bedevil this project -- Sanjaya Maniktala at National Semi has a white-paper on smps magnetics which is in the form of a power point and easy to work through -- when you see the magnitude of the harmonics that have to be dealt with you might reconsider altogether.

[Cristi]

Hello Nico, interesting project.

in my opinion, i would say that the voltages which you need should not be from zero, but from at least 10% of maximum. in this way, you may use a classic topology like flyback without too many complications. i've made something similar a while ago, based on FPS, a mcu controlled lab smps capable of delivering 5-50V at 3A. i want to add one more important aspect. the current should be adjustable too.
so, for the specs, let's say 50-450V at 10 to 250mA, -20 to -150V at 5-50 mA and 0 to 30V at 0.1 to 4A.
first stage will be a PFC, for this power level i would choose IR1150 for the simplicity and performance, the DC voltage should be 375-380V. the 50-450V supply can be made using a FSCQ1265RP or FSCQ1565RP, same for the 20-150 V supply. for the current adjust must use a small value resistor (1R at 1-3W) on the negative output for the 50-450V supply and on the positive output for the bias supply, if we connect them to the GND. for the 30V 4A can use either a FPS, but will be hard to get down to zero volts, ir can use a flyback convertor and a buck converter in the secondary.

i'll make a schematic later, just let me find some little time.

[nicoverduin]

Thank you all for the reply and support. The positive HV does not have to be from zero.
The negative however does. As this is the bias voltage for the audio tubes. The powersupply is for a digital tube tester with U/I characteristics possibilities on a computer (through USB). Bus since BIAS current is so low, I could afford to just build a converntional voltage adjus with a fixed -150V BIAS supply. In practice, I suspect the current will be a few mA.
The low voltage is for filament. The lowest voltage I have seen so far is 1.25V (pencil tubes).
The microcontroller section seems pretty easy. Just read the analog voltage, based on an input voltage (anolog and digital are possible) and switch the opto-coupler either on or off. Have to check the speed though... Don't think I can get away with a 4mHz internal clock this time.
I wil chechk the suggested documents. But now it it time for social desirable behaviour.... called work