Current Flowing Indicator.

[Mark Hennessy]

The resistor between OUT and ADJ is not critical. The datasheet recommends 240 ohms because it gives a convenient current of ~5mA that makes the numbers easier; all the LM317 does is maintain a fixed 1.25V between OUT and ADJ, while endeavouring to draw as little current as possible via the ADJ pin so that the output can be determined by resistor values alone. If you choose 1.2k, the current is ~1mA, and that's about as low as one should go if the ADJ current is to remain insignificant compared to the current in the two resistors.

In other words, deviating from the "default" 240 ohms so that the potentiometer you happen to have covers the range the require is absolutely fine, providing the operating current is more than ~1mA.

Or, worry ye not

Joe, if you add a capacitor from ADJ to ground, then the diode from ADJ to OUT becomes mandatory. I tend to use them as a matter of course.

[Joe]

I saw that on the data-sheet, Mark, but thanks for the reminder.

I've always been a bit puzzled by the 240Ω resistor as when reading about them the words "usually" or "typically" tend to go with it. It's also odd that it's an E24 value which gives the idea that it's important.

There's a calculator and a table of values here and, as can be seen, they are all over the place.

I know I'm an arch "Amateur Twiddler" but have only blown one LM317 up. That was when testing a magic eye and I accidentally touched the HT and found that "They don't like it up 'em." The odd thing was that I'd instinctively been reaching for a blocking capacitor before realising that I wanted DC to get through; it was only after the event that using a diode occurred to me.

Those panel meters do look worthy of investigation, David, though if you could blow one I'm not very confident that it would survive my tender ministrations. I can see myself building a bigger box.

I never expected fitting an LED to be so involved but I'm not knocking it as it's all good fun and this thread has gone down some very interesting side-roads on the way.

- Joe

[Skywave]

I never expected fitting an LED to be so involved but I'm not knocking it as it's all good fun and this thread has gone down some very interesting side-roads on the way.
- Joe

And we have all added to our knowledge of various things as well.

Al.

[Nick]

Hi Joe - as I have mentioned before, I've been engaged in electronics and radio work since I was a youngster and over that time period, I have looked at all manner of cct. diags., but I have never seen the Greek symbol [lower case omega] ever used as a symbol for resistance - until now! Now if that symbol was in use in the 1920's - 1930's, that would explain it: my experience with kit from that era is very thin

Oddly, today I was at Spitalfields Antiques Market (its on every Thursday and I work round the corner) - sometimes (not often) there are things of interest.

One stall had obviously cleaned out the workshop of an old radio guy, and there were cards of resistors etc... guess how they were marked?

Oh, he wanted 5 quid a card. I passed...

EDIT: Just thought! I could buy one, measure the component value and then we'd know what the lower-case omega meant...

[AlanBeckett]

Nick,

Do you think they are really lower case Omegas or just badly written Upper case?

Alan

[pwdrive]

Hawkeye! I wouldn't like to be a s**t house rat in your garden..

Lawrence.

[Nick]

Do you think they are really lower case Omegas or just badly written Upper case?

Fair question. Don't know really. Kind of odd way of spending your life, measuring all your resistors and sticking them individually on cards...

[Joe]

OK, Dr Mark and David, I surrender and will put a digital ammeter on this.

I noticed earlier that set to low output voltages the LED comes on with no load connected. I think I know why - I've even done some sums:-

   

When R2 is set to zero the only thing across the output is R1. This means that a minimum of 3mA is flowing through Rs and the resulting 3mV across it is enough to set the 741 off. I found a 0.2Ω resistor but this was the same, in fact shorting it didn't effect a cure as the resistance in the wires seemed enough.

I've studied the LM311 data-sheet which Al supplied, and the SGS-Thomson version, but am still a bit unsure as to how to rig one up. What I need to be able to say it it is "If there's only 3mV difference (in fact we'll say 5mV and play safe) between your inputs stay off."

I want to fully re-build this anyway, and fit a meter, but a comparator would be something new to play with and learn about but I've no idea where to start.

Experience and knowledge is never wasted.

- Joe

[Mark Hennessy]

The op-amp doesn't behave predictably with both inputs near to the negative supply rail. Either supply rail, indeed. I've mentioned this more than once - pay attention at the back

To prove the point, you could temporarily use a battery to make pin 4 of the op-amp negative with respect to 0V, and you'll probably find that it behaves itself. If you try this, do make sure the maximum supply range of the op-amp isn't exceed.

Op-amps vary enormously in this regard - some do work close to one of the rails, some work at both, but most don't - as always, study the datasheet. Some op-amps (e.g. the TL06x/07x/08x series) go one stage worse - as the two inputs get close to the positive (IIRC) rail, they swap around, causing all sorts of weird problems!

A comparator is essentially an op-amp with very poor properties as an amplifier, although you could choose to make one work as an amplifier if you really wanted to. Often, they have an open-collector output stage, thus making it easy to interface with logic circuits. Check the datasheet carefully, as an open-collector output won't light the LED in your circuit - in which case you need to swap the two inputs around, and wire the LED (and series resistor) between output and pin 7.

Good idea to add the current meter. And while you're at it, add a volt meter if you don't already have one. Or, if you're feeling adventurous, make a single meter switchable.

Perhaps needless to say, but the current meter will probably need an external shunt resistance, and this shunt resistance can be shared with the LED circuit. Even if the meter has an internal shunt, this can be shared. No need to put the two in series - that just wastes volts...

All the best,

Mark

[Joe]

Mis-understood again. I thought that by 'Near the rail' you meant that the voltages on the inputs couldn't be too near to the supply voltage. Presumably this is why +/- supplies are best for them?

I have studied the data-sheets but the required information isn't on them in a way I can understand.

I had a play with another 741 on a breadboard and that was plain weird: I'd only connected the power and a 1kΩ resistor and LED to the output and it wouldn't go out with nothing on the inputs or even with them shorted.

I'm starting to think and EM80 would be easier!

- Joe