The LM317 is quite the hot head ...

[eccerr0r]

Re: The LM317 is quite the hot head ...

Now, the trick is how would one make a digitally adjustable LM317 that will regulate down to 0V?

I suspect that the same trick of using some sort of regulator to make sure the negative rail doesn't go too far down else you'd have a lot of dead region.

In any case some offset trickery is needed as now a 0 to 3.3V signal needs to be -1.25 to +11 or whatever...

Plus the 3.3V signal needs to be regulated too else an RC low pass filter is not going to give you an accurate output...

Now question is: will this work: (note that I cheated with a LM337 to regulate negative voltage. LM358 is not RRIO so it will have trouble getting to the high voltages but should get close to zero output in this configuration. I also did not account for input leakage which will be significant with the values of the resistors used here.)

[EasyGoing1]

Re: The LM317 is quite the hot head ...

http://www.edn.com/design/analog/436...h-a-PWM-signal

See this is why I seek out friends like you guys ... thats good stuff! I could potentially do this project without the digital pot that I spent a whole $5 for three.... lol

And I just happen to have a couple of 741s laying around ... hmmmm....

[EasyGoing1]

Re: The LM317 is quite the hot head ...

Now, the trick is how would one make a digitally adjustable LM317 that will regulate down to 0V?

I don't have a requirement for 0V in my project ... cool circuit though ... send me a 337 and a 358 and ill test it

[eccerr0r]

Re: The LM317 is quite the hot head ...

I don't have a requirement for 0V in my project ... cool circuit though ... send me a 337 and a 358 and ill test it

At the very least you still need the LM358 for using that EDN article design as if you used the LM741 it won't even go below 2.5V most likely.

The LM337 is not absolutely necessary, do whatever you want to get a somewhat low impedance, regulated, negative 1.25V to 2.5V rail.

[EasyGoing1]

Re: The LM317 is quite the hot head ...

do whatever you want to get a somewhat low impedance, regulated, negative 1.25V to 2.5V rail.

k, not quite sure what all that meant or what a “rail” is ... but I only need to be able to make somewhat small adjustments in a voltage range of roughly 3 to 5 volts... the more resolution I have available to me in my adjustments, the better.

For example being able to go from 3.032 volts to 3.039 would be much better than if I could only go to 3.131 or something like that...

[EasyGoing1]

Re: The LM317 is quite the hot head ...

At the very least you still need the LM358 for using that EDN article design as if you used the LM741 it won't even go below 2.5V most likely.

Check out this design for 0 to 3V using a 317.

... when I tested the link I just provided in preview mode, it seemed to have inserted %20's in the link ... dont know why ... so here it is again just in case (this one actually worked for me when the other didn't)

http://www.edn.com/design/analog/432...able-regulator

[eccerr0r]

Re: The LM317 is quite the hot head ...

Yeah once again it requires a negative supply of some sort, there's lots of ways to skin the cat, but the cat needs to be skinned.

[EasyGoing1]

Re: The LM317 is quite the hot head ...

Yeah once again it requires a negative supply of some sort, there's lots of ways to skin the cat, but the cat needs to be skinned.

Why is this mandatory?

[eccerr0r]

Re: The LM317 is quite the hot head ...

It's the only way to get the LM317 to get to 0V. Can't get around it any other way other than perhaps something to regulate down externally, duplicating the function of the LM317 except handling the 0 to 1.25V region.

[EasyGoing1]

Re: The LM317 is quite the hot head ...

It's the only way to get the LM317 to get to 0V. Can't get around it any other way other than perhaps something to regulate down externally, duplicating the function of the LM317 except handling the 0 to 1.25V region.

Well actually my question was WHY is it necessary that the circuit be capable of 0V ... i didnt doubt the reasons for actually making a 317 produce 0V ... I just wanted to know why 0 was required is all...

[eccerr0r]

Re: The LM317 is quite the hot head ...

Oh not for your problem of a battery charger, it was a question I posed as a brain teaser.

[EasyGoing1]

Re: The LM317 is quite the hot head ...

http://www.edn.com/design/analog/436...h-a-PWM-signal

From that link

You can change the op amp's gain by choosing proper values for R4 and R2

I wouldn't even know where to start choosing values for these resistors...

[EasyGoing1]

Re: The LM317 is quite the hot head ...

http://www.edn.com/design/analog/436...h-a-PWM-signal

So I actually built this circuit using my Arduino Due to control it ... which only does PWM up to 3.3 volts. HOWEVER, it does offer 12 bit PWM resolution as opposed to the regular 10 bit, so I have 4095 different levels of PWM to play with vs 1024.

So I arbitrarily picked a 1k resistor for R4 and R2 ... and with my PWM at 0, the output voltage was like 3.2 volts ... I was not able to get it to go higher until I kicked the PWM up to 3000 or roughly 2.something volts ... and then it hit the ceiling of around 10 volts on the LM317 before I reached the end of my PWM abilities .... so in other words, I don't have a very wide "sweep" of adjustment with this design...

How do I go about figuring out which resistors to use? I swapped in and out dozens of different resistors with different results, none of which gave me what I needed...

Also, the potentiometer had little effect on things, except once or twice with different values for the other two resistors, at which point adjusting the 10k pot would move the output on the 317 about 2/10ths of a volt but then would swing back where it was as I reached the opposite end of the wiper on the pot ... kinda weird if you ask me.

Ideas?

[eccerr0r]

Re: The LM317 is quite the hot head ...

Check the voltage at the input of the 741 (keep in mind the impedance of your multimeter will load down the circuit, even if you have a 10Mohm DMM). See if it is doing the range you expect. Check this for the top end.

If you're insisting on using a 741, you will have some nonlinear operation as you see at bottom end due to the inherent issues with the 741. Using the 741 you will need a negative supply to rid of this bottom end issue.

[EasyGoing1]

Re: The LM317 is quite the hot head ...

Using the 741 you will need a negative supply to rid of this bottom end issue.

See now you're starting to make sense ... where does one conjure up ... negative voltage ... exactly??? Some kind of voodoo?

[EasyGoing1]

Re: The LM317 is quite the hot head ...

You know the tiny motors that operate the vibration mechanism in a gaming controller ??? ... if I were driving that with a 317, do those motors pose the same risk to a 317 as a coil does to a transistor in terms of blow back once power is disconnected? Should the motor have a protection diode on it?

[eccerr0r]

Re: The LM317 is quite the hot head ...

See now you're starting to make sense ... where does one conjure up ... negative voltage ... exactly??? Some kind of voodoo?

Easier to just use a LM358 instead of the LM741.

Else you'll either
- build another power supply and connect it backwards (or another winding on the transformer). You can use a 9V or even a few batteries to generate it in the mean time.
- create a charge pump of some sort to synthesize the negative voltage, perhaps using a 555 or some switching regulator.

You know the tiny motors that operate the vibration mechanism in a gaming controller ??? ... if I were driving that with a 317, do those motors pose the same risk to a 317 as a coil does to a transistor in terms of blow back once power is disconnected? Should the motor have a protection diode on it?

Yes, it also generates some inductive back EMF too. A small vibrator motor probably doesn't generate that much however.

[EasyGoing1]

Re: The LM317 is quite the hot head ...

Easier to just use a LM358 instead of the LM741.

Would you happen to know if the 358 is common in any specific electronics that I might have laying around ??? I feel like I'm doing my part when I salvage stuff rather just buying more ... not to mention a bird in the hand ... three days for shipping is murder for a US citizen ... lol

- create a charge pump of some sort to synthesize the negative voltage, perhaps using a 555 or some switching regulator.

No 555's laying around and everything else you said you may as well have been speaking chinese.... charge pump?? this is a real thing in electronics?

Yes, it also generates some inductive back EMF too. A small vibrator motor probably doesn't generate that much however.

The reason I asked about that, is because I built a small little 317 variable buck converter kinda thing ... in fact this is the circuit I used:



I forget the values for the voltage divider, but the second resistor is a pot ... 5k I think ... anyways ... whats happening is that Ill use the thing ... three or four times over maybe a weeks time then allofasudden, I plug it in and its passing ALL of the source voltage and the resistor on he adjustment pin does nothing... so if im feeding it 12 volts then thats whats at the output of the 317 ... so if I then replace the 317 with another one ... it works beautifully till suddenly ill plug it in and it only passes the source voltage ... till I replace the 317... I am now on my 5th 317 the last three happening in the last month... this could bel

No letting the smoke out and no stressing of the 317 has happened at all ... it never once got hot ... so i was thinking maybe when I run those motors ... then shut it down ... they might be hitting the 317 from the wrong side which would explain me no seeing the problem util the next time I plug it in. Other than that im stumped

[eccerr0r]

Re: The LM317 is quite the hot head ...

You'll have to look around for a LM358 or LM324, they are fairly common (they've been around almost as long as LM741).

MC3043, LM2904, TS271/TS272/TS274 are alternatives. A RRIO LM6142 or TLV2371/TLV2372/TLV2374 will work but are overkill. The LM741 is an ancient chip that has so many drawbacks, but the rail problem is not one of the most serious ones (there are many other low noise op amps that also don't go to rail that are commonly used due to noise and input characteristics where the LM358 is not so great at). You may find a LM358 in old crappy ATX power supplies like Deer used to generate PG. Good power supplies likely use a supervisory chip instead and not have that op amp.

The NE555/LM555 is a very common chip, it's an all purpose timer chip. Due to its output characteristics, it can be used to make charge pumps like in
http://www.electronicecircuits.com/e...supply-circuit .

And I don't know if you have a large capacitor on the input else you may be backpowering the LM317 which is harmful. A protection diode across input and output (anode on output, cathode on input) can protect against back powering.

Unfortunately likely you will have to go buy parts unless you have a lot of junk to scavenge from... I suppose I have way too much scrap to scavenge from now, though I did spend quite a bit from Jameco to fill the shelves...

[kc8adu]

Re: The LM317 is quite the hot head ...

how are you terminating the charge cycle?
when in cv mode you need to watch current and cut it off at about 50ma for a typical 18650.
use a mosfet switch and disconnect the cell with it.
then it does not matter if your 317 can hit 0.
this also prevents issues with the 317 getting voltage to its output when off without the drop of a diode.
do not float charge li-ion.