I just finnished making a regulator circuit using this http://www.tapr.org/tapr/html/pwrsup.html#LM317T mostly but also other guides. The problem is the max output is 4.5v - I was hoping to get 7.5v from this, to power my ps one lcd screen. Im not sure why its max is 4.5. I used the values from that guide, mostly. Instead of 270R I used 330R as thats the closest I had - maybe this is the reson, with 270R probably giving 5v, and something like 100R giving upto 10v? Also I used 5k pot instead of 4.7k, maybe I needed a higher value like 10k? Also in a dead 8 port hub I have a chip for converting 5vdc into 9vdc and seems to req. no components or very few if it does. Where can I buy one like this in the uk, but for 12v->24v?
Thanks for the quick reply. I would prefer to use the lm317 as I already have it. Also the lcd screen does not need a 3A supply - the power cube that comes with it is 3A but thats becuase it powers the ps aswell. The screen takes less than 1A iirc.
Ah, ok. Only thing I can think of for output voltage being so low is that your supply is a spare ATX powersupply and it doesn't have an adequate enough of a load on it to properly regulate the outputs. Each rail needs about 1A to be drawn, otherwise you get lots of noise on the rails. You're not doing something stupid like running the regulator off of +5v, are you? Not trying to insult you of course, just checking.
the circuit on the link you posted seems fine to me. things to check are as <whypick1> suggested - first and foremost that you connect to the 12V rail. 2nd is, that your psu is actually providing 12V on that rail and that it`s not an old and wore psu. 3rd is, that you have both resistors wired to the right "legs" of the lm317t - you`d be surprized how many times i`ve seen it hooke up wrong (though i doubt this is it - as you are getting Voltage out of it...) g`luck with the PSone screeny...
hmm well just checked psu with multimeter, 11.98 (incase I done the 7v wire mod and then put them back, with the 5 and 12 reversed...) so its not that. Psu is a cwt550w which is running my pc atm. Ive checked the circuit about 10 times and there is nothing obvious wrong - The only reson I can possibly think of is the resistor limits it to 4.5v. Also my soldering is not particually good - by that I mean if the lm317 overheated, would it possibly work fine upto 5v and then not go higher? And also the regulator is outputing 3.7v min and 4.5v max... should go down to 1.25v?
You've got something wrong circuitwise, with 330R and 5k the maximum output calculates as Vo=1.25(1+5000/330)=19V (though you'll only get that if your input voltage is higher). So you should get the full 1.25V-10.5V range from 12V input, with no change in maximum over half the pot's travel. I suspect the 330R resistor isn't connected to anything at one end or the other (or a dud), try a 470R if you have one, or anything up to 1k.
Is this resistor nneded? would it work without any connection from output to adjust? I will try this now
Thanks cpemma and everyone else that helped me, it now works, from 1.24-10.97v This circuit I will use for my lcd screen, which needs 7.5v. I also need two more circuits for powering one pump (liang D4 upto 24v. dunno what current it needs, but at 12v I think it was 18w) and a large 172mm fan (24v 0.55A). First I would like to use them upto 24v. This means I need to make a 12v->~30v inverter. The lcd screens psu was a switching one and had a transformer with the codes: MP-130B 1801 SMT-25H-NB238 DC6M12 0239( 01) I assume this means it has 238 windings on the primary and 25 on the secondry, which would seem about right? If so, it has a gain of almost 10x and I would need to feed it ~2.5v dc (switched quickyly to give AC effect) to get out about 25v? Do transformers work in reverse, powering secondry to get power from the primary? no diodes hidden in there or something? next questuin is how do I do the switching? using a 555 timer or something? Should I also add a fuse of some kind to prevent the output voltage from going too high? I dont want any 240v shocks
It most certainly is needed, without it the regulation doesn't exist, though you'll still get a small range of outputs. Read the datasheet. edit: our posts crossed in the wires, Congratulations!
the screen pulls 0.62A (measuerd at input to lm317). I tryed it without a heatsink assuming the screen would use half that maybe less so luckly the lm317 has thermal protection (but my fingers dont ) Im using a 7*2cm sheet of 0.5mm copper for heatsink and its not enough edit: Now using a passive nb heatsink (which was also passive on the mobo I stole it from). Its still hot enough to burn my fingers, about 70-80c. Im puzzled. A nb outputs abuot 20-30w??!!? yet this is hotter, same airflow. heatsink is only 5c lower than chip, so its not the mounting... Anyway that problems solved - cant find anything on google that will help me get 24v from a lower voltage
You pretty much always have to sink a regulator, I was running a PIC and a few other things(only ~ 75mA) from a 5v .5A regulator, and after about 5 min, it was hot. I have it bolted to a slot A heatsink now, (temp use till I can cut the HS down) it dosen't even get warm.
The regulator is wasting (12-7.5) x 0.62 = 2.8W, the bare 317T has a thermal resistance of around 50C/W so will try to go to over 170C with no sink, switch itself off at about 150C & burn your fingers. It's safe at 100C (if you don't touch it) so using the formula here you'll need a sink better (less) than 19C/W, not big but better than your bit of thin copper. I'm surprised the nb sink gets so hot, thought they'd be overkill. Don't forget the 317T tab is connected to the output pin, so if the sink touches an earthed part of the case it will spark.
