Need some clues

[xtrmi]

Ok so I put the 5k pot back in there between adj and gnd and a 120ohm between out and adj. I still got the same behavior. Maybe the 317 is dead?

 

[marcbarker]

If the resistor from the output to the adjust pin for an LM317 is higher than 120 ohms then the 10mA in it ...[cut!]

AG, I'm confused now!... How can this resistor being higher than 120 ohm, still have "the 10 mA" in it ? please explain

And xtrmi, you'll need to post a schematic (again!) , or reference something posted earlier, with annotated voltages.

 

[audioguru]

If your LM317 has an idle current of 10mA (some will have less idle current) and a 300 ohm resistor from the output to the ADJ pin then the 10mA will create a voltage drop of 300 ohms x 10mA= 3.0V. But the LM317 regulates the voltage by setting and sensing 1.25V from the output to the ADJ pin and therefore with 3V it will not regulate and the output voltage will rise without a load.

Since your LM317 dropped its voltage even if the resistor is 120 ohms and the input voltage is high enough then the LM317 is broken.

 

[xtrmi]

I will try to get to there store and buy one today. I would be willing to try other regulators if you think they would be better.

 

[marcbarker]

I think I see now.

My understanding is now that the internal design of the 317 is different to fixed voltage regulators such as 7805, and that difference is that the internal ground returns have been re-arranged such that only the ground return(s) associated with the voltage regulation are connected to the ADJ pin, all the others return via the OUT pin.

The reason for doing is, is to reduce as much as possible the current flowing out of the ADJ pin, so that it becomes more like a voltage-sense pin. These ground returns exiting via the OUT pin are what constitutes the [up to] "10 mA" that you mentioned, which you are describing as "Idle Current". What threw me was I found no mention of 'idle pin current' in the datasheet, so I presumed (incorrectly) that you meant Adjust Pin current.

This current is also the reason why these variable regualtors have a minimum output [load] current. The choice of Current Programming resistor needs to take into account the maximum amount of this ground return current exiting the OUT pin, so officially for a 317 it has to be a maximum of 120 ohm.

 

[marcbarker]

I would be willing to try other regulators if you think they would be better.

For this project a 317 is as good as any other regulator, I think you should stick with it!

 

[RCinFLA]

On the LM317, the adjust pin draws a small amount of current that must be insignificant compared to the current run through the adjust resistor to ground. If I remember correctly I think it is like 50 uA. This is why the applications circuit lists a 240 ohm resistor from output to adjust pin. The output to adjust pin is always 1.25 vdc so a 240 ohm resistor draws 5.2 mA which swamps out any variation in the adjust pin current around 50 uA.

The output voltage is going to be 1.25v + 5.2mA * Radjust.

Also, the LM317 is spec'd at 40 vdc max. so with 44 v input you are exceeding capability of part.

Hope you have a very large heat sink when you get it working. If you draw 0.5 amp with output set to 1.5 vdc you are going to dissipate over 20 watts on the LM317.

 

[audioguru]

..... This is why the applications circuit lists a 240 ohm resistor from output to adjust pin.

No.
You are wrong.
The 100uA max current of the ADJ pin has nothing to do with the max value for the resistor from the output to the ADJ pin.

The applications circuit with a 240 ohm resistor uses the more expensive LM117 that has a minimum load current of 5mA (1.25V/240 ohms= 5.2mA). But the LM317 has a minimum load current of 10mA so it needs a 120 ohm resistor (or less).

Maybe you want to gamble. Maybe there are some LM317s (made on a good day) that have a minimum load current of 5mA, especially if the input voltage is lower than 40V.

 

[Hero999]

This is why the applications circuit lists a 240 ohm resistor from output to adjust pin. The output to adjust pin is always 1.25 vdc so a 240 ohm resistor draws 5.2 mA which swamps out any variation in the adjust pin current around 50 uA.

The output voltage is going to be 1.25v + 5.2mA * Radjust.

The minimum current from the output terminal is 10mA, therefore the minimum resistance for R1 is 120R.

Also, the LM317 is spec'd at 40 vdc max. so with 44 v input you are exceeding capability of part.

That's not true. It's specified at 40V maximum differential from the input to output, therefore you could have an input of 44V and an output of >4V and not exceed the maximum rating of the part.

 

[audioguru]

The input can be 1040V and the output can be 1000V and the voltage ratings of the LM317 will not be exceeded until the output is shorted to 0V.

 

[marcbarker]

The input can be 1040V and the output can be 1000V and the voltage ratings of the LM317 will not be exceeded until the output is shorted to 0V.

Do think it'd work regulating this? high voltage

 

[audioguru]

I saw that little spark before.

 

[RCinFLA]

Statement about 1040v input and 1000v output being okay.

There is usually a capacitor place on output to improve transient response of the regulator. If not there there might be a cap to ground on the circuit the power supply is supplying. There will also be stray capacitance to ground on Vadj pin due to resistor termination capacitance and PCB layout.

Now - power supply is off, what is voltage across this output to ground cap. (hope you say zero volts). What is this cap's voltage the instant the Vin 1040 vdc shows up. (hope you say zero because caps don't change voltage instantaneously). So what is the startup instantaneous voltage across the regulator?

 

[Ghosty_Ghoul]

That shouldn't be a problem as long as the capacitor charges quickly enough when the supply is applied.

If there's a concern that the maximum Vin-Vout will be exceeded as a result of a turn-on transient then add a 39V zener between the input and output terminals which will protect it against transients.

 

[marcbarker]

I thought I'd repost this just in case we didn't notice it.

Is XTRMI going to make the LM317 circuit with a 120 ohm resistor replacing his 300 ohm resistor so that it regulates properly?

 

[xtrmi]

Yeah I did that, but the lm317 is broken so tomorrow I'll head to frys to buy another one. Then I'll post here about it. This has been a big help. Hopefully it will work and can understand the regulator better. Obviously I didn't know enough I was just copying achematic. By the way my friend did the same with a 240 and he claims his is ok. I even see some 240 being used in the application notes but they may be loaded more or something i'm not an expert yet. <-- lol

 

[audioguru]

The datasheet for the LM317 shows a more expensive LM117 on the first page using the 240 ohm resistor that it works well with.
Many people copy that circuit and use it wrongly for their LM317.

The ICs are not all made exactly the same. There are good ones and not-so-good ones. Some (not all) need 120 ohms to keep their output voltage from rising without a load.

 

[marcbarker]

The ICs are not all made exactly the same. There are good ones and not-so-good ones. Some (not all) need 120 ohms to keep their output voltage from rising without a load.

To explain why 240 ohms works, I offer the theory:

I'll take a guess that traditionally, 117's that didn't quite meet spec get re-badged as 317's, those that didn't make 317 spec get melted down again. Now in the 21st century, I'll guess that the "117 yeild" is so much higher from a 117/317 wafer, that few are 317 spec, and chances are is the "317" you buy is a 117 spec chip, badged "317" for marketing reasons.

 

[audioguru]

You would think that a manufacturer of an LM317 that has a better spec due to a modern better yield would sell it as an LM117 maybe with a reduced price.
Or improve the spec's on their datasheet.

 

[marcbarker]

You would think that a manufacturer of an LM317 that has a better spec due to a modern better yield would sell it as an LM117 maybe with a reduced price.
Or improve the spec's on their datasheet.

Trouble is there is huge demand for "317"s in plastic. Meanwhile, people cling on religiously to standard datasheets, quoting them like a bible or rule-book.