DIY inverter Mathcad file

[Roswell Bob]

I have cobbled up a short Mathcad file which calculates inverter losses and heatsink requirements for a 5kW inverter at 10kHz. The transistors will do 200% continuous given a good heatsink, This file could be used to develop both low and high frequency designs. I expect that this bridge could replace blown out guts in a commercial inverter. I will use in a 48v high frequency design. I will present that paper (DC-DC converter) design soon. I will do thermal simulations next and present the thermal contours. I have chosen Ixys fets for these calculations. If anyone would like me to run an IGBT or a SiC mosfet just give me part number. I will put together a Magnus Opus file which contains everything that a boss or client would ask for when a design is complete. Maybe we build one some day and see it it works I have Mcad, RTL, and HTML files which will not upload. I will try to convert to pdf and upload. If anybody want the other format just msg me.

 

[Roswell Bob]

pdf

 

[HaldorEE]

I have cobbled up a short Mathcad file which calculates inverter losses and heatsink requirements for a 5kW inverter at 10kHz. The transistors will do 200% continuous given a good heatsink, This file could be used to develop both low and high frequency designs. I expect that this bridge could replace blown out guts in a commercial inverter. I will use in a 48v high frequency design. I will present that paper (DC-DC converter) design soon. I will do thermal simulations next and present the thermal contours. I have chosen Ixys fets for these calculations. If anyone would like me to run an IGBT or a SiC mosfet just give me part number. I will put together a Magnus Opus file which contains everything that a boss or client would ask for when a design is complete. Maybe we build one some day and see it it works I have Mcad, RTL, and HTML files which will not upload. I will try to convert to pdf and upload. If anybody want the other format just msg me.

Looking forward to it.

Things got real at work. I am in the preproduction launch phase on one project and the research stage in the next big thing.

Lots of fun. Keep running into the "Jim Williams" guy. He must be somebody important.

- Edit -

I know who Jim Williams was. I have his collected app notes in a coffee table book. Actually useful. On paper you need the images blown up big.

Here is an on-line collection.

Best App Notes

Here's my list of Jim's best application notes (so far; this list covers App Notes 1 to 106): App Note 47, "High speed amplifier techniques...

readingjimwilliams.blogspot.com

 

[Roswell Bob]

Looking forward to it.

Things got real at work. I am in the preproduction launch phase on one project and the research stage in the next big thing.

Lots of fun. Keep running into the "Jim Williams" guy. He must be somebody important.

- Edit -

I know who Jim Williams was. I have his collected app notes in a coffee table book. Actually useful. On paper you need the images blown up big.

Here is an on-line collection.

Best App Notes

Here's my list of Jim's best application notes (so far; this list covers App Notes 1 to 106): App Note 47, "High speed amplifier techniques...

readingjimwilliams.blogspot.com

https://diysolarforum.com/attachments/diyinverter-pdf.35314/

 

[HaldorEE]

I was looking at your losses? Diode losses? I would think you would use synchronous rectification.

https://en.wikipedia.org/wiki/Active_rectification.

 

[Roswell Bob]

I was looking at your losses? Diode losses? I would think you would use synchronous rectification.

https://en.wikipedia.org/wiki/Active_rectification.

This is the output inverter - not the DC-DC converter.

With an inverter the output transistors are on while the anti-parallel diodes are conducting. It is automatic synchronous rectification. In most cases the forward voltage drop of the diode is less than the current x Rds_on product though. So the diodes take most of the current. This may not be the case for a low frequency inverter where the Rds_on can be quite low with the lower voltage fets. The last commercial inverters I did were 480v/600v mains where IGBTs were the choice. That was in 2007. I more recently did a small high frequency 2kW unit for DOD and I think I used SiC parts. In any case the Mathcad file is plug and chug so it is simple to run any switch and see what happens. But yes, your point is good and weather the diode or FET supports the reverse current will have to be reasoned through. I believe the SiC parts have a higher voltage drop, but the reverse recovery losses are quite a bit lower. They may be a better choice for 120v out. For the low frequency inverter the choice would probably favor the Gan parts. I would stay away from Gan until they show up with built-in gate drivers. They should be very soon. Good luck with work projects. I totally enjoyed that high anxiety pressure cooker environment until I didn't anymore. Never ran across Jim Williams tho I've always strayed towards LT for their parts and support. Thanks for the links. Here's one for you https://rompower.com/

 

[HaldorEE]

This is the output inverter - not the DC-DC converter.

With an inverter the output transistors are on while the anti-parallel diodes are conducting. It is automatic synchronous rectification. In most cases the forward voltage drop of the diode is less than the current x Rds_on product though. So the diodes take most of the current. This may not be the case for a low frequency inverter where the Rds_on can be quite low with the lower voltage fets. The last commercial inverters I did were 480v/600v mains where IGBTs were the choice. That was in 2007. I more recently did a small high frequency 2kW unit for DOD and I think I used SiC parts. In any case the Mathcad file is plug and chug so it is simple to run any switch and see what happens. But yes, your point is good and weather the diode or FET supports the reverse current will have to be reasoned through. I believe the SiC parts have a higher voltage drop, but the reverse recovery losses are quite a bit lower. They may be a better choice for 120v out. For the low frequency inverter the choice would probably favor the Gan parts. I would stay away from Gan until they show up with built-in gate drivers. They should be very soon. Good luck with work projects. I totally enjoyed that high anxiety pressure cooker environment until I didn't anymore. Never ran across Jim Williams tho I've always strayed towards LT for their parts and support. Thanks for the links. Here's one for you https://rompower.com/

That Rompower....

Is that for real?