SMPS 230 VAC to 115 VAC 800W

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
subwo1 said:
If it still did not work to your liking after the filter addition, you may be able to add feedback to regulate the output.

Thanks for your constructive help subwo1, the problem with using the 340VDC as the primary feed to drive a short width pulsed output is that the end use devices to be driven are 115VAC (160Peak) and they have a mob of spike suppressors (180V) that BLOWUP when hit with 340V peak, so I am doing the hard yards and building a sinewave converter as a second stage feed by the first stage which is working very nicely now at 100Khz that feeds a 2:1 transformer core, that also now gives me the advantage of having the correct peak voltage and DOUBLE THE CURRENT to start with, .... which now feeds the sinwave converter that is almost completed, I will keep you posted as to the results, I am sure a bit of feedback will be needed for regulation so I have allowed 10% more turns on the secondary of the 100Khz transformer which I think will allow me to vary the PWM that generates the sinewave giving me some form of regulation, I am making this unit now capable of 20Amps out of the secondary so it will do the job easily.... thats my theory at this stage ...... lets see how I go. I do enjoy intelligent constructive responses.
Kind regards.
 
For simplicity and cost reduction, you can use a half bridge driving a 1:1 transformer.

Also, I think that you can make the circuit much simpler if you tie one of the ends of the load to a virtual ground created at the mid point between ground and 340V, and switch the other end between 340V and ground with a simple half bridge.

The virtual ground may be kept perfectly centered with a circuit like this (replace the voltage supplies by capacitors):

An externally hosted image should be here but it was not working when we last tested it.


(It's the same circuit that I proposed in another thread to prevent rail pumping in class-D amplifiers.)
 
TeeJay said:


...lets see how I go. I do enjoy intelligent constructive responses.
Kind regards.
Thanks TeeJay. I hope it continues to have good progress.


Eva said:
For simplicity and cost reduction, you can use a half bridge driving a 1:1 transformer.
I played with it a little on the simulator and it seemed to do well.:) I'd like to investigate it more in the future.
 
Eva said:
For simplicity and cost reduction, you can use a half bridge driving a 1:1 transformer.

Also, I think that you can make the circuit much simpler if you tie one of the ends of the load to a virtual ground created at the mid point between ground and 340V, and switch the other end between 340V and ground with a simple half bridge.

The virtual ground may be kept perfectly centered with a circuit like this (replace the voltage supplies by capacitors):

An externally hosted image should be here but it was not working when we last tested it.


(It's the same circuit that I proposed in another thread to prevent rail pumping in class-D amplifiers.)

Thanks EVA, in this proposed idea of artificially centre tapping the mains do I loose the 2:1 current advantage ? in other words if the output is drawing 5 amps is the primary drawing 5 amps from the mains supply?

Also can you tell me what the signs are of the core saturating, does it just appear resistive and get hot or not?
 
OK guys, here is the challenge!
I am having trouble getting my design past 9 amps SO...

I am prepared to pay seriously for the learning exercise!!!
you must be able to also explain the various decisions and technical reasons for using certain components.


I will pay US$5000 for a working design/proto that meets these conditions.

1. INPUT 210VAC-250VAC 50-60Hz

2. OUTPUT regulated 115VAC **60Hz Sinewave @ 15Amps continous into a resistive or inductive load.

3. The order of importance is the lowest cost, size, parts count, design wins.

4. Must have the 2 to 1 current advantage e.g.
For an OUTPUT load of 10 amps the input should be approx 5 amps (2:1 Tran effect) plus or minus efficency losses.

5. Normal EMC Spec's for Australia applies. ( Available on the Internet.) or I can supply.

6. Parts used must be readily available.

7. Time limit 30 days from this posting.


Regards TJ
 
TeeJay said:
OK guys, here is the challenge!
I am having trouble getting my design past 9 amps SO...

I am prepared to pay seriously for the learning exercise!!!
you must be able to also explain the various decisions and technical reasons for using certain components.


I will pay US$5000 for a working design/proto that meets these conditions.

5. Normal EMC Spec's for Australia applies. ( Available on the Internet.) or I can supply.

Regards TJ
drop #5 or expect to pay more :)

If Australia and Finland would be a bit closer to each other I would be intrested...
 
If you are considering paying $5000 for a working proto, then why won't you contact the power supply manufacturers in your neighborhood? Many of them are willing to do a custom design, although the $5000 might be a problem. Your advantage is that those guys know their stuff, and you are dealing with a company with which you can make hard agreements.

And don't drop that EMC requirement. You need it anyway, so it's part of the total set of requirements, just like the input/output voltages and currents.

30 days is not much time to go from scratch to a completed product. Having a PCB made already takes at least a week (or cost will be unacceptable). Although the $5000 sounds tempting, I think I won't be the right person for this due to lack of practical experience with these power levels. I would get it working, but probably not within 30 days.
 
DaBit said:


Thanks for your reply DaBit..
30 days is not much time to go from scratch to a completed product. Having a PCB made already takes at least a week (or cost will be unacceptable). Although the $5000 sounds tempting, I think I won't be the right person for this due to lack of practical experience with these power levels. I would get it working, but probably not within 30 days.

You do not have to build a board, if you are confident your design works, I can make boards same day at material costs so if that was an issue, it is not anymore.

The 30 days is not a hard and fast item , but I had to start somewhere, I am open to offers and will listen to any offer and cost.

I could go to commercial people but this is a learning exercise as well and I figure that if some of the people on this forum give there time willingly, they should bennefit as much as anyone else, when this kind of situation comes along, I expect a few professionals are lurking around here somewhere!!!
TJ
 
TeeJay said:

You do not have to build a board, if you are confident your design works, I can make boards same day at material costs so if that was an issue, it is not anymore.


With power supplies, the board layout is a very important part. It can make or break the design. And without testing the board, one can never be confident that the design works.

If no more-capable person pops up, I might give it a shot. But let's see what happens first.
 
TeeJay said:
OK guys, here is the challenge!
I am having trouble getting my design past 9 amps SO...

I am prepared to pay seriously for the learning exercise!!!
you must be able to also explain the various decisions and technical reasons for using certain components.


I will pay US$5000 for a working design/proto that meets these conditions.

1. INPUT 210VAC-250VAC 50-60Hz

2. OUTPUT regulated 115VAC **60Hz Sinewave @ 15Amps continous into a resistive or inductive load.

3. The order of importance is the lowest cost, size, parts count, design wins.

4. Must have the 2 to 1 current advantage e.g.
For an OUTPUT load of 10 amps the input should be approx 5 amps (2:1 Tran effect) plus or minus efficency losses.

5. Normal EMC Spec's for Australia applies. ( Available on the Internet.) or I can supply.

6. Parts used must be readily available.

7. Time limit 30 days from this posting.


Regards TJ

Dear TeeJay,

I've got a working prototype
It costs 150USD.
Regulation is between +/-10%.
The weight is 13kg.
It only uses 1 part (+ a few for soft-start)
Size is 220mm*88mm.
It's called: toroid transformer.

Best regards,
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.