diy solar

diy solar

DIYOSI Power Section Design

Roswell Bob

Solar Enthusiast
Joined
Dec 5, 2020
Messages
759
Location
Warner, NH
Here is a drawing of power section. Tear it apart. Please supply input as I expect to start putting tracks down next week. I will complete the schematic and get it out tomorrow or Tuesday. Keep in mind that this inverter board will be used for HF design, but is fine for a LF desoign as well. If you have any LF iron but your inverter has turned to shit then you may want to discuss what you have and we will see what we can do.
 

Attachments

  • INVERTER.pdf
    175.4 KB · Views: 25
Do you have any updates on this?

I'm no inverter expert, but I note that there's no protection, no sensing (current, voltage, temperature) or feedback, and nothing to handle reactive loads.

Hopefully you ascribe to the fail often, fail fast method of development, because there's a steep learning curve between where this circuit is and what professional inverters do.
 
Thanks of the shove Mr Stienman - this has been on my mind lately. Sorry that I have haven't been attentive to this project. SInce last post I have had a heart attack and a quad bypass. Also had a flood, a lightning strike and a car wreck. I hope to get back on it in October or November. Wife and I are getting ready for winter now. We're off the beaten path on a hilltop and winters here can be a bit difficult so today's priorities are what they are.

I have the power section 80-90% layed out. I have bus current sensing as well as bus voltage. The bus current sensor is shown on the schematic. I don't intend to do output leg current sensing. I had initially planned to use a differential amplifier for bus voltage sensing. I have gotten this past UL in the past, but not sure if they are still allowing. I also have a small arrangement of bus caps shown on the drawing to take care of reactive currents. I don't expect to see any large reactive loads. A small unloaded motor would probably be worse case.

I've designed inverters since 1977. When I was a young engineer I lost many output switches in a typical design cycle - Bipolars, Mosets, GTOs, Igbts and so forth. Towards the end of my career I was able to go through a complete development cycle without failures. As I said in previous posts this will be my last inverter design.

I will see what I have on my machine and post new drawings soon.

You seem to know a good bit about inverters. Have you any experience working with them?
 
You seem to know a good bit about inverters. Have you any experience working with them?

Very little. A friend of mine developed an inverter for motor control which we then proceeded to use on a tesla coil. The number of failures and fixes over two years taught him a lot, as you'd expect with that kind of load and environment. I expect his inverter is nearly bulletproof now:


I tend more towards the digital side of development.
 
In that schematic is only a bridge output stage, it seems to be used for dc/dc. it can used also as modulator using a cheap EGS002/ EG8010 board.
For protection the simplest way is to use a hysteresis hall sensor, for a latch protection.
From my experience autorestart protection is not good, if output is still shorted it may fail during soft start sequence.
After a protection trigger, is better to ensure the output is not still shorted before attempt to restart.
If you use autorestart it may force you to make it too sensitive, and end in troubles with some appliances.
 
I am a little bit familiar with the EGS002 board. I have one kicking around here. I seem to recall the pwm routine was a little different. I think it may be assymetrical in some manner that causes different losses in the switches. My coding skills are a little weak so I won't hammer on it. I wrote a few lines to do pwm in an 8pin PIC 12F675 a few years back. The PICs sole function is to do the PWM. I have some glue logic that takes care of hardware current limit and protection functions. If power board becomes a reality I expect there might be others interested in helping with some better code. I would do a control board with my pic or an interface board to run a EGS002. I have Altium so it is an easy task for me to crank out a control board if someone wants to write some code for it.

I've used hall sensors for most of the larger designs I did. Most LEM sensors. They were one of the first on the block and I developed a relationship with them. There is an old HP part that isolates the current sense signal using a low value resistor. The power board I have in process uses a parallel array of resistors and one of these HP/Agilent/Broadcom or whatever the name is today :)

Yes, the autostart function is probably not worth the time at this point. I have a nice hardware current limit circuit. I've found that good algorithms to start a motor or energize a transformer can go a long way in preventing tripping to begin with. I have an AIMS 2kW unit that can't start my well pump and trips most every time. If I start the inverter into the well pump it brings it up every time. So it has some kind of start function that works well. If it had an auto-restart option it would be nice. I wouldn't have to get off the couch when the old lady runs the washing machine :)

Most start up routines I have done check for ground faults and shorts as part of the power up routine before running inverter. Again, I haven't written any production code, but have been fortunate to have very capable software guys that I have been able to rely on to implement algorithms I've concocted. Many years ago there was a mentality among power guys that all protection would be done in hardware without involving microprocessors or software. With faster processors and some standup code guys that have proven they know what they are doing that mentality is gone these days.

Thanks for comments. I hope you can comment/help with design as it moves forward.
 
Back
Top