 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : High Power Solid State Inverter Relay
  Page 2 of 2 |
|||||
| Author | Message | ||||
| noneyabussiness Guru  Joined: 31/07/2017 Location: AustraliaPosts: 513 |
This would also help with the too high load dropping out problem too... I think it works !! |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
The book of words says Vcc range +10v to +39v with a 15.5v under voltage lock out. It might be a good idea to fit a suitable protective zener across CD3 to ensure that +39v can never be reached. That will give much more freedom to experiment safely. Reading between the lines (and guessing) the 39v maximum and the series resistor in the bootstrap supply suggests U1 may have an internal 39v zener diode. Reducing the series resistor excessively may cook that internal zener if its actually there. I think I would fit a decent sized external zener, maybe 24v to be certain. If its fully operational above 15.5v, it really does not need much more than that. I feel there is probably quite a bit of scope here for reliable low voltage operation with probably a fairly simple modification. Cheers, Tony. |
||||
frackers Newbie  Joined: 06/11/2009 Location: New ZealandPosts: 23 |
Entirely coincidentally, I've been playing with some 12v 400mA switching supplies like this one which uses a THX208H chip in a very minimal circuit as described here. My testing has been limited to the bench supply that can only go up to 47 volts but so far, at 40v they start up but go unstable under load - sometimes it stays alive, other times its squegging. At my full 47v, the idle current is just under 3mA and with a 95mA load (fan) its pulling 40mA from the supply so not very efficient. I've not investigated where the heat is appearing yet. I plan to change one of the feedback resistors to drop it from 12v to 5v and then use a AMS1117 linear regulator (there is space on the board for one) to give me 3.3v for my inverter controller. Robin down under - or are you up over |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
That sounds pretty much like the 12v 450mA one I tested. Starts up o/k at quite low voltage, but as soon as it has any load it tries to start up, then dies, then tries to restart again. The one I tested needed about 50v to 60v to keep running at about half full rated load. The bootstrap circuit just does not have enough puff to keep it alive after the initial kick start, with any useful load on the output as it is. If you can feed some additional current into C6 direct from the dc input that should run the chip continuously with little or no help needed from the bootstrap winding. These supplies are so cheap and so small, I feel they are well worth the trouble of persevering with. Cheers, Tony. |
||||
| frackers Newbie Joined: 06/11/2009 Location: New ZealandPosts: 23 |
Had a bit of a slip with a crock clip so now I only have 4 PSUs left  The upside of this is that I now know that the primary has 160 turns, the secondary 26 turns and the bootstrap winding is 12 turns. The isolation consists of 5 turns of the yellow tape between windings. For lower voltage operation I guess the options could be * a few extra turns could be put onto the bootstrap winding with a bit of careful pin snipping * power the chip directly from the supply side * lift the ground side of the bootstrap winding so as to put a full wave rectifier in to supply the chip. As you say Tony, worth playing with. Robin down under - or are you up over |
||||
| noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 513 |
Also just check to, some pwm IC's have a max 50% duty cycle so at lower voltages it may not be enough with the winding ratio... to go over 50% there usually is a compensation input. I think it works !! |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
I suppose the correct engineering approach would be to greatly reduce the primary turns. We don't need to be able to operate up near its 370 volt theoretical maximum dc input capability. If the primary is wound on top that should be simple. Possibly least amount of work would be to just add a series resistor and zener to supply chip power direct from the input. Only a very small amount of extra power should be required as its "almost" working. Cheers, Tony. |
||||
| frackers Newbie Joined: 06/11/2009 Location: New ZealandPosts: 23 |
Unfortunately the primary is at the bottom, then the bootstrap and then the secondary winding on top. The ferrite core was also glued which meant that the examination was somewhat destructive. According to this pdf file the chip requires between 4.8 and 9v with 16v being the absolute maximum - although it also states that the startup requires typically 8.8v I'll have to dig out a 10v zener (just as a clamp) and based on 3mA running current a suitable resistor to provide that down to 40v. I'll start at 10k and work both ways  Page 7 of the above pdf indicates that the minimum is 4.8v and the startup volts comes off the main supply via the internal output device. Robin down under - or are you up over |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Another way to skin this cat might be to power the chip from the main 12 volt regulated output. Our bootstrap winding effectively then becomes 26 turns instead of 12 turns. We then lose the input to output galvanic isolation. As we are powering from low voltage dc, and not the mains, for some applications that may not be a disadvantage. Cheers, Tony. |
||||
| Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1138 |
Wow, good discussion on the small power supplies, I havent looked at mine to see why it isnt working, intention is to take the transformer out and rewind it or perhaps design a new pcb using the components off the old board but with dual 12v and 5v isolated outputs. So for now have left the linear regulator and 12/12v isolator in place, so current draw on the linear reg will be minimal increasing proportionally with variable fan speed. Meanwhile have completed the proto design of main switch mosfet pcb and its piggy back Pic controller, try to use smd components where possible. The mosfet board has two 25 x 3mm copper bars bolted to the pcb on the top side with mosfets mounted direct on the bars, large flat strips of 0.25mm copper foil mount under the bars to give extra heat radiating surface area, the sources are clamped to the other bar using 4mm bolt and cupped washer, pcb shows mosfet solder holes... not used except for the gates., Bars will extend out sides of plastic housing with 10mm bolt holes. For this first cut both boards can be used single sided. Mosfet Board:  Controller:  Cheers Mike |
||||
| noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 513 |
Also if you are trying to control a fan from 48v have you tried the poor mans buck... using the arduino (or micromite as encouraged on here) directly pwm the 48v to 12v, using a decent inductor and capacitor in line, i have successfully have supported 3 decent fans using this method. Its not as rock solid as a dedicated chip but works non the less. I usually have a decent TVR diode and fuse inline as a just in case. Ill find the software i wrote (nothing special ) and post it here... if that ok with mods?? I think it works !! |
||||
| Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1138 |
Am using the PWM output of the Picaxe to control the fan, no havent used the other, Micromite or arduino as yet, have the picaxes lying all over the bench, seems a waste to not use them, ok for this sort of application and I dont have to try and learn something new... Cheers Mike |
||||
| Solar Mike Guru Joined: 08/02/2015 Location: New ZealandPosts: 1138 |
Minor changes to the schematic, have added a current mirror to sense the current flowing through the paralled mosfets, whatever voltage is dropped by their effective on resistance will be reflected as a voltage across the 47 ohm resistor; with 300 amps and say 6 mosfets this will be about 0.1 volts. I dont know how accurate this will be as ideally the sense fet should share the same silicon die as the ones carrying the current.  Mosfet PCB:  Cheers Mike |
||||
| Page 2 of 2 |
|||||
 Print this page |
