LM317 adjustable voltage regulator problem

[boylesg]

I have solderd up this circuit.

Vin = 24V
R1 = 1k
R2 = 10k or 12k or 8.2k or 8.2k + 4.7k in parallel.

It doesn't matter what I do with R2 I cannot get Vout below 20V

Are there some details they don't bother mentioning in the data sheets. Like R1 must be under ......R?

How do you get these bloody things to work properly?

I don't understand how the fixed voltage regulators can be so easy to use and these things impossible to use.

 

[Jestin_cubetech]

select R1 value between 12 - 240 E

try this configuration

R1=220E
R2=3.3K

out put voltage =20V

 

[boylesg]

Never mind, I just realised what I am doing wrong.

I had assumed the pinout was the same arrangement as the fixed voltage regulators which I have been using up until now, i.e. input on the left, output on the right and common, or in this case adjust, in the middle.

Just looked at the datasheet and realised the pinout is totally different.

The circuit diagams in the datasheets are arranged in the same manor as for the fixed voltage regulators which re-inforced my incorrect assumption.

One again that's what happens when technical people write user manuals - they make assumptions about the reader's experiance level. If it were up to me I would arrange the LM317 symbol with a pinout that reflects the pinout of the actual device.

The whole point of writing user manuals is to make them idiot/Greg proof!

- - - Updated - - -

select R1 value between 12 - 240 E

try this configuration

R1=220E
R2=3.3K

out put voltage =20V

Bloody nora. Shouldn't an important piece of info like that be explicitly stated some where in the datasheet?

None of the online calculators for LM317 seem to mention such a restriction either.

Assuming I have the damn thing wired up correctly, what would be the effect of using a greater than 240R value for R1?

 

[godfreyl]

Just looked at the datasheet and realised the pinout is totally different.

It works better if you look at the datasheet before you build the circuit.

One again that's what happens when technical people write user manuals -

Seriously? You're blaming the guy who wrote the datasheet because you didn't bother to read it?

The whole point of writing user manuals is to make them idiot/Greg proof!

This isn't idiot-proof enough? It's at the top of page one:

 

[IanP]

Assuming I have the damn thing wired up correctly, what would be the effect of using a greater than 240R value for R1?

You should rather decrease the value to something below 240ohs ..
120ohm would be a better choice as this thing needs some minimum load, and values >240ohm set the sum R1+R2 to values which would not satisfy the "minimum load", or "no load" condition ...
:wink:
IanP

 

[alexan_e]

Bloody nora. Shouldn't an important piece of info like that be explicitly stated some where in the datasheet?

Assuming I have the damn thing wired up correctly, what would be the effect of using a greater than 240R value for R1?

The datasheet does mention a minimum output current in order for the regulator to operate properly (check min load current) and that is what this low value resistor is all about.
If your load can guarantee that min current then you can use a higher value resistor but if the output current falls below the rated current then the output voltage will increase.

Alex

 

[FvM]

It has been emphasized in a recent LM317 thread (this year, please search edaboard if you are interested) that the guaranteed minimum current of plain LM317 (in contrast to better specified LM117) is as high as 10 mA. In so far, R1 of 120 ohm is required if the voltage divider is the only load.

There has been also a controverse discussion, if the datasheet voltage divider suggestions should take care of LM317 specifications and no-load case. Perhaps there should be a "mind your step" warning sign :grin:

In my view, practising art of electronics requires - among other skills - some reading. People who don't like it can still consider more intuitive arts as painting or woodwork.

 

[boylesg]

Seriously? You're blaming the guy who wrote the datasheet because you didn't bother to read it?

Well not really. At the end of the day I was a dic%head for not reading the datasheet thoroughly. :smile:

Given that, during my C++ days, I have been involved in manual writing and fell into a similar trap in assuming knowledge that the reader did not necessarily have, it was more an observation of a problem than outright blaming of the manufacturer for my mistake.

But also at the end of the day a top notch user manual should explicitly cover such constigencies.

By making the pin arrangement on the symbol not correspond to the pin arrangement on the actual device in contrast to the all the fixed voltage regulators I have used, the manual writer is making just that much more diffcult for inexperienced users like me to use the device.

And user manuals are ideally designed to make it EASY for inexeperienced users to successfully use the device.

All the necessary information was there if I had have read it ALL but never the less the above inconsistency could be remedied and thus make the user manual a bit more 'robust'.

- - - Updated - - -
If you want a shocker of a user manual, you should see the manual for my DMM. They guy whoe wrote it clearly does not have a good command of the english language. It is usable but it is a bad reflection on the compan in my opinion.

The datasheet does mention a minimum output current in order for the regulator to operate properly (check min load current) and that is what this low value resistor is all about.
If your load can guarantee that min current then you can use a higher value resistor but if the output current falls below the rated current then the output voltage will increase.

Alex

What will the effect be of inadequate load current? Increased rippling?

- - - Updated - - -

You should rather decrease the value to something below 240ohs ..
120ohm would be a better choice as this thing needs some minimum load, and values >240ohm set the sum R1+R2 to values which would not satisfy the "minimum load", or "no load" condition ...
:wink:
IanP

Well I was going to make it 220R simply due to the fact that I don't have any 240R resistors. I already have to series to resistors for R2 and I dodn't want to have to series two resistors for R1 as well.

What is the recommended wattage for the resistors anyway? Don't think I have seeen any mention of this in the datasheet.

 

[godfreyl]

What is the recommended wattage for the resistors anyway? Don't think I have seeen any mention of this in the datasheet.

OK, you're right - they do assume some knowledge.

Current = Voltage / Resistance
Power = Voltage * Current

Or more simply:
Power = Voltage squared / Resistance

 

[boylesg]

OK, you're right - they do assume some knowledge.

Current = Voltage / Resistance
Power = Voltage * Current

Or more simply:
Power = Voltage squared / Resistance

I finally got the prick of thing to work with a steady output of 17.65V after I put in all the protection diodes in. It has taken me a week of extremely late nights to get this figured out, among other household tasks of course.

For some I just couldn't get that circuit diagram in the datasheet sorted in my head while transposing it to a circuit board due to that inconsistent LM317 symbol.

But now that I have done it once, perhaps I will find it easier next time. :smile:

No doubt I am having the same problem with the DC-DC boost converter I have been trying to solder. Ahhh well I will get it sorted sooner or later - I am not ready to quit on it yet.

 

[godfreyl]

You can never assume there's a connection between the symbol and the pinout. For example there's only one symbol for an NPN transistor, but it is common for different transistors to have different pinouts even when they are in the same package.