Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Rebuilding My Power Supply

Status
Not open for further replies.
I got lost ages ago with what your upto, But I do enjoy reading it :D 217A:nailbiting: all a little extreme for flashing LEDS:happy:, I have visions of slightly twitching components writhing around the floor at your place, legs in the air with small wisps of smoke still escaping. and the sounds of manic laughter from behind a closed door:hilarious:. You kill fets without fear,remorse or regret. I wish I was as callous, I might learn a bit quicker :woot:
Thank you for that, it caused me to browse back through this thread, I hope it'll be of help to other, it was educational to me. Since I need to go back in and make a few more changes, I'll take and add some more photos of it completed, hopefully for the "LAST" time. Then, on to my next ???????
OH yea, It's a Power Suppy;)
 
If the next is another psu, I am going to buy in some fets before they all become extinct :D
 
I added the scr and associate parts and it didn't do what I had hoped for, actually, it didn't do anything, at high current, the LED would light dimly, and when output was shorted, the light would blink, but the SCR did not appear to fire. The FET was dead though and it's tab was loose, not sure how that happened. After I replaced the FET, I tried to get the SCR to trip by shorting out put but it never did. I put a Fluke 87 ammeter in series and another Fluke across the output as a short and got a reading of 2.84 V and 20.96 amps - no screen shot - . I then put a load on the out put of 4 4 ohm resistors, each on a switch and took screen shots as I added more load and the last test was to put a .5 ohm across the output and the reading were 11.76 V, down from 18 and the current was 24.76 amps. (WOW), surprised me!!! so I tried it again, and again, the 50 watt resistor was starting to smoke. The screen shots are labelled.
https://www.dropbox.com/sh/wfpkr5w0hcrz7a2/RoElm-fUVF
 
You have a very, ermm 'unique' building style! Love the pics tho.
I am trying not to think about the cruelty aspect
 
Well I've been very happy with my power supply, the transistor version didn't seem to every fail, just get hot, the new fet model doesn't get hot and I thought it wasn't going to fry FETs, WRONG !!! I left the leads laying on the table, the negative side was driving lights and then they started dimming and flickering and I realize the leads laying the table were shorting out, I tried to get them real quick, but I didn't and by the time I did, the damage was done. When I put it back together, I paralleled 2 FETs, I'm considering putting a high current shut down in it if I kill these FETs. After it was back running, I put the FET in the vise and squoze it and the plastic popped off nicely, I think I see why it is DEAD, Is this MELTED silicone?
 

Attachments

  • 20140219_182903.jpg
    20140219_182903.jpg
    255.3 KB · Views: 578
  • 20140219_182914.jpg
    20140219_182914.jpg
    232 KB · Views: 573
Hi,

Wow that looks nasty, and sorry to hear that it blew up again.

We can add current limit if you like. Sometimes this isnt too hard to do.

You know what, i was looking at some "pre assembled" buck switchers too using the Simple Switcher IC's and i might get one just to try it out. They are listed on Amazon and come in a variety of forms, some with seven segment displays built RIGHT ON the board itself, and it displays voltage i would presume.

But i really like the idea of adding external switch transistors to the basic IC chip, although i would definitely add current limit to mine because these things will always blow up the parts if we dont have that feature.

Another idea would be to build one with an isolated output. That would be a bit harder though as we would have to design a transformer and find a way to power it with the chip. Maybe using a bridge circuit.
 
Last edited:
I think I know why I was frying the 4905s so badly, no over current at all, I had put a 30 amp fuse in some time back and forgot to change it back. One really nice thing about the external switch is that it take all the punishment and it dies to save the LM2576 and when you buy them in quantity, they are a lot cheaper.
Go to http://www.datasheetcatalog.com/ and type in simple switcher. There's a ton of them.
 
Hi,

You mean there are a lot of Simple Switcher IC chip types or a lot of different switches?

I have seen that there are a lot of Simple Switcher IC chips, and in particular i have seen one like the LM2576 but with faster clock rate of 150kHz, which is three times faster than the LM2576 so it could get by with a much lower inductance. It's the LM2596 chip. It works the same (almost) as the LM2576 except for the clock rate and the small feedback capacitor recommended for some input or output voltages. Board layout might be a little more critical too but might not be too bad. I am currently investigating these chips and have an order placed for a couple boards with the chips already mounted and caps and pot and everything, but it's going to take about two weeks to get them unfortunately probably due to shipping location.

What i noticed about these chips though is that the efficiency isnt all that great. Some input/output voltage combinations will only yield a 75 percent efficiency for example. But with a MOSFET replacement for the output switch transistor (as you have been experimenting with) we should see a marked increase in efficiency say 85 percent or better depending on output current. The drawback is that when we add a MOSFET we loose current limit, so we have to come up with a way to get that back without sacrificing efficiency.
The way i did it on my Li-ion charger works very well (i wanted lower current limit) but it eats up some efficiency due to the sense circuit. It's unfortunate that with a 5v output a 0.7v sense resistor drop is a lot of loss. With a charger it doesnt matter too much, but with battery operated equipment we want high efficiency.
 
UPDATE, I have lost a few sets of IRF4905 FETs that were used as my switchers for the positive side of my power supply, but I think I let the leads short, they are a lot stiffer than they were originally, any way after lots of help on another subject, Heat Transfer, I changed from a 200 ohm resistor feeding paralleled FETS to a totem that has 2K ohm resistors feeding the base of the 2N2222 and 2N4403 transistors and I think these will last a bit longer. I took some screen shots of before and after of the gate voltage. Here's the schematic and .asc file also
I was pulling 16 amps at ~ 18 volts for the scope shots.
 

Attachments

  • LM2576-LT1270A DUAL POWER SUPPLY.asc
    10.3 KB · Views: 365
  • LM2576-LT1270A DUAL POWER SUPPLY.png
    LM2576-LT1270A DUAL POWER SUPPLY.png
    58.5 KB · Views: 3,200
  • 16AMPS2.png
    16AMPS2.png
    158.8 KB · Views: 603
Hi,

The new drivers look like they work much better than a single resistor, as expected. But why the 2k series base resistors?
 
what would you suggest, please, I used 2K because I couldn't find my 3K at the moment. My thinking is that 2 or 3 k is low enough to drive it full on and in another place, I've been loosing the the 2N4403 with 25v on it.
 
I haven't figured this out yet, may not even try to, but I using my power supply to on an H bridge and the bridge caused a short circuit and the power supply died, again. I replace the failed parts and I couldn't get it to work again until I went back to the pre totem design, I made some fet totems out of an 8 pin P & N FET chip and tried that too. I could get the voltage to come up but as soon as I applied a load, the voltage dropped to about 2 volts. This was the case or no voltage or 24 volts for every thing I tried until I went back to the NO totem design. Also, after going back to the no totem design while still in bread board mode, I had it working properly and switched in the transistor totem while lightly loaded and the voltage dropped and wouldn't come back after switching out the totem until I turned it off. I was watching the wave forms on the scoped and it seemed mess up the phase relationship between the switcher and transistor collector and couldn't recover. Anyway, here's the schematic, again, I tried using 100 ohms from collector to fet and it became unstable as load changed, so I stayed with 2oo ohms and the voltage stays pretty steady. I also tried a TIP125 darling in place of the TIP42c and that didn't seem to help either.
 

Attachments

  • LM2576-LT1270A DUAL POWER SUPPLY.png
    LM2576-LT1270A DUAL POWER SUPPLY.png
    56.7 KB · Views: 3,233
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top