Driver I made is doing something odd...

lasersbee said:

BTW.... the Dropout Voltage of the LM1117 is typically 1.00 Volts not 2.45V as I read it on the Data Sheet...

Click to expand...

1.2v dropout, 1.25 reference voltage. 2.45v total needed in addition to the LD's forward voltage. Same as the rkcstr driver...

bah, I think I'm just going to have to try different caps with larger capacitance... not a big fan of waiting though.

My Bad... I thought you were talking just about the Drop Out Voltage...

The best thing is to experiment (play) with is... You don't have a Laser Diode
hooked to it so you're safe to go...
Just a thought.... does your DMM have good batteries in it... :

Jerry

(900 posts/30)

Tried removing C2, didn't make any difference.

I recall phenol saying he ran into serious stability issues without C2, and that you should make sure you use low ESR, high speed tantalum caps... I also remember rkcstr saying he needed not only an extra cap for C2, but he needed a rectifier diode across that junction as well to maintain stability.

I think I may just need a bigger cap for C3... I think I'll try an electrolytic 100uF.

Well, just a thought: The datasheet says that the chip is a voltage regulator designed for very low dropout at full 1 A load. It did not say it could regulate current

And regarding tantalum caps for low ESR, I've only seen that in boost circuits like the LT1370, because of the high frequency switching to raise DC voltage. If not, stability issues will occur. This is the first time I've heard of stability issues that could have arisen from ESR in a non-boost voltage regulator.

Drat! I don't have an LM1117 in my parts bin. I can't reproduce this surge with an LM317. I've got to pick up some parts in the next day or two though. I'll pick up an LM1117 or 2 and try to reproduce the problem. (Now that I'm all curious...)

I want to see this on an oscilloscope.

cheers,
kernelpanic

kernelpanic said:

I wasn't suggesting altering the input cap. I was saying remove the sense cap and reduce the output cap. Scr*w the datasheet. Breadboard it then rip on me...

cheers,
kernelpanic

Click to expand...

Hey kernelpanic....
don't eat your socks.. man.....  ;D ;D ;D
I wasn't ripping on you... and I agree... the circuit needs to be tested on a breadboard...
The Data Sheet is only where to start....

PS. I don't have an LM1117 here either to do any testing.... need to call Digikey... :

Jerry

One thing you could try for the capacitors: put a 1 or 10 nF ceramic/multilayer capacitor in paralel with each electrolytic you are using now.

It starts off around 350mA, within a second it settles at 397mA, then just for a fraction of a second it jumps up to 510mA, then settles back down to 397mA where it stays for good.

Click to expand...

How did you measure this exactly? If you are using a multimeter or something similar, you might miss the cause. It's possible for these regulators to (briefly) oscillate at very high frequencies (tens of MHz), and you need to be looking at the output voltage or current with a scope to catch that.

Putting bigger electrolytics will not help in this case, it's likely to make things worse as their internal inductance is generally larger. Adding small ceramic caps will do a lot more.

Benm said:

One thing you could try for the capacitors: put a 1 or 10 nF ceramic/multilayer capacitor in paralel with each electrolytic you are using now.

It starts off around 350mA, within a second it settles at 397mA, then just for a fraction of a second it jumps up to 510mA, then settles back down to 397mA where it stays for good.

Click to expand...

How did you measure this exactly? If you are using a multimeter or something similar, you might miss the cause. It's possible for these regulators to (briefly) oscillate at very high frequencies (tens of MHz), and you need to be looking at the output voltage or current with a scope to catch that.

Putting bigger electrolytics will not help in this case, it's likely to make things worse as their internal inductance is generally larger. Adding small ceramic caps will do a lot more.

Click to expand...

Yeah, just a DMM... I thought maybe the ~517mA burst was some artifact of my meter and the way it measures, so I tried with another multimeter... The problem didn't seem to go away... and, unfortunately I don't have access to a scope...

I'm guessing it is oscillating though, I'm getting the impression I'm not getting the full power my DMM says either... I've hooked it up to a closed can diode that shouldn't be able to take 400mA for long and it's burning like a champ... I expected I'd get maybe a few minutes out of it, but it's been running for hours, cumulatively... The beam seems dimmer than and doesn't burn nearly as well as LOC's I've run at 400mA... I'm guessing maybe the reading on my meter is the peak current, whereas the mean current is actually lower... which I guess is at least better than the current being too high...

Any way of me testing this without a scope? Anyone here have a scope and would like to test for me?

Multimeters usually do a very lousy job at measuring intermittent signals. I doubt there is any way to 'see' what is going on without using a scope.

I would take the readings for you, but i live in holland so sending the circuit might take longer than you like. If you have more voltage regulators from the same batch, i'd be happy to hook one up here and see what happens.

But i think the best bet is to try with some ceramic caps first - if the problem is oscillation, those are more then likely to fix it.

I need to order a few parts from DigiKey tomorrow and I'll add
a couple of LM1117 to breadboard the circuit in the shop...
If I find something I'll post it here...

Jerry

Hmm, in my circuit diagram which mimics that one, there is also a rectifier diode, with opposite polarity to the laser diode, in parallel with the laser diode. I think it was meant to help avoid any potentials between the ground and power lines for the laser diode. Maybe you could try adding one of those.

All the diode does is provide polarity protection...it has no effect on the circuit.

That diode is a bit silly really. I'm not totally sure what happens if you reverse polarity on these linear regulators, but its likely to damage the regulator anyway. The laser diode on the other hand should be able to survive some reverse bias, though the capacitors may not (at least not for very long).