LM317T Voltage Regulator Input Resistor Cremation.

[Darky8]

Hi Folks!

I have a device with about 4 LM317T's which all seem to have tiny 10ohm SMD Resistors on their voltage in.

One particular one is doing about +33v down to 12.5v. However as soon as power is applied, the input resistor is toasted.

I tested across the resistor on the working ones, and it lowers the input voltage by only about .2 volts!

I've replaced this particular IC, but still no joy. I've not seen these devices used with such a resistor. Is it needed? And what could be causing their destruction?

Many Thanks

D +

[budm]

Re: LM317T Voltage Regulator Input Resistor Cremation.

The resistor is a cheap way of using it as a protection, it cost less than a fuse. The one with 0.2VDC drops only has about 20mA flowing through the 10 Ohms resistor so power dissipation is in very small mW.
Resistor will burn up when too much current flowing through, I would check the resistance on the output pin
'+33v down to 12.5v.' That is not a good way of dropping the Voltage. that is about 20 Vdrops on the 317, 20V x current draw will be the power that the 317 will have to dissipate.
BTW, what kind of device is this? Is the 33 Vdc regulated? 317 max DC in is around 37V.
https://cdn.badcaps-static.com/pdfs/...4077910403.pdf

[Darky8]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Thank you.

It's from a hifi DAC.

The 317 works along side a 337 to drop +/-33v to +/-22.5 (sorry did I put 12.5!) to power the two audio out channels.

As far as I can tell they are fed straight from a pcb mounted transformer.

[budm]

Re: LM317T Voltage Regulator Input Resistor Cremation.

https://www.badcaps.net/forum/attach...8&d=1475960052
I would check the resistance between the input pin and GND, the output pin and GND.
Those black Tantalum caps 6.8uF 35VDC, the failing mode are shorted circuit.
So the 10 Ohms resistor for LM337 (Negative reg IC) is the one that failed, right?
BTW, any pictures of the whole board?

[kc8adu]

Re: LM317T Voltage Regulator Input Resistor Cremation.

if those tants are on 33v line there is no margin.current limiting kept them from becoming fireballs.
very high on the suspect list.

[budm]

Re: LM317T Voltage Regulator Input Resistor Cremation.

if those tants are on 33v line there is no margin.current limiting kept them from becoming fireballs.
very high on the suspect list.

+1, so far if it is fed from the bridge using linear transformer, so if the AC line is high, the 33VDC will go even higher.

[Darky8]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Firstly sorry I can't get any more pictures ATM, not at home.

It's the LM317 that's frying the resistor, the LM337 works fine as do 3 other LM317's (But they are around 12v feed, with 5v output).

So there are two transformers and 3 1000uf 35v caps, nothing else in the power supply at all.

however, these solid caps don't sit with me at all well. The unit has had questionable repairs/soldering, plus it's an 1990's DAC. So I de-soldered all the caps in order to clean up the board, and solder them back properly. The PCB was encrusted with dark brown crusty stuff (?), along with just general dirt.

As always, I took photo's of everything and put the right capacitors back, the right way around. They might not be the exact ones that we're in the same place, but they are the correct capacitance/voltage/polarity.

All other regulators work, so I have disconnected pretty much all the circuitry fed by this 317, and it still does the same.

I just managed to delete a more in depth post (Grrrr!), so will post more details once home.

Thanks so much for your help. This is a critical and highly valued piece of 1990's classic HiFi.

D ~

[6BW6]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Keep in mind the max input on a 317 is 40V..... wouldn't happen to have any spikes by any chance, if so try a 39V zener clamp.

Tantalum's darn things, I switched a function generator on.... 'nothing', then some other time I switched a Wavetek signal generator on.... 'nothing', in both cases tantalum's went s/c on me so I replaced with low esr electrolytic's.

If blowing resistors a case try a small 6V lamp while you investigate, at least it's limiting the current.

Dave

[budm]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Well, you need to do the resistance test as I suggested.
https://www.badcaps.net/forum/attach...8&d=1475960052
So the picture is before the work was done? It did not look like it was worked on.
The reason I ask about the resistor feeding the 337 because it has dark color which I thought it was due to heat.

[Darky8]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Hi, I did check the resistance on the output pin. And it's 3.8k to GND.

[Darky8]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Hi, folks!

Sorry this has been a bit of a long time coming. But I enclose some much better pics.

The LM317/LM337 combo supply +/-22~v to the two ceramic output PCB's. Pins 1 & 2 -22v, GND, then 4 & 5 +22v.

The +33 and -33v are fed to only these two regulators, all the rest are running lower voltage and outputting ~5v.

The lone 337 works fine (as do all the rest of the other 317's) with stable negative output.

I have removed both ceramic line boards, and found that a trace goes round to two SMD transistors via another resistor, both of which look toasted. I removed both transistors, so I'm pretty certain that there is no load on the output to speak of.

The positive side of the 33v line is shared through the coils of at least two of the 3 24v rated relays. So that when this regulator is powered, so are two relays.

It's all powered from the two transformers, on the board from 240v AC with the only other components being 3 1000uf 35v caps.

I wan't to recap this board, but I wan't to find out what's going on before I do so at least I learn something whilst improving it.

Kind Regards

R +

[stj]

Re: LM317T Voltage Regulator Input Resistor Cremation.

if your output resistance is 3.8k then chances are, the part it's driving is shorted and will be getting pretty damn hot.

[Darky8]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Yeah, I'm going to try and get everything disconnected from the output to see if it works then.

If it does (which I doubt it will), we'll pop them back in one by one and see.

I might also knock up a quick off-board LM317 circuit, and try and power things with that.

[stj]

Re: LM317T Voltage Regulator Input Resistor Cremation.

replace the limiting resistor with a jump wire if your brave,
then feel each chip quickly when you power it up.
something will get hot other than the regulator!

[Darky8]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Actually I know where that resistance is coming from.

It's the two resistors for the adjust! DOH.

I would have to remove them, then measure the output pin.

[stj]

Re: LM317T Voltage Regulator Input Resistor Cremation.

you cant, they set the output.

[Darky8]

Re: LM317T Voltage Regulator Input Resistor Cremation.

If I remove them, I don't need to power the circuit. I can simply measure the resistance between the Ouput and GND, then put them back afterwards.

I can't see it being a problem on the output side. I've pretty much isolated it and it still melts resistors.

We'll see what it does with some new caps.

[budm]

Re: LM317T Voltage Regulator Input Resistor Cremation.

As requested in post 4. How about the resistance of the input pin? If the TANT cap at input pin is shorted then you will blow that 10 Ohms resistor.

[Darky8]

Re: LM317T Voltage Regulator Input Resistor Cremation.

Hi, Sorry I missed the input pin test.

I tested the LM377 (the working one) and got 1.39-1.44M/ohms stable at input pin.

I tested the LM317 input pin and started at a low reading, but grew rapidly until I removed my meter at around 22M.

I also found that the TAN across the input was charged to 0.38v, the others had no charge.

The 3 6.8pf 35v caps measured

6.66pf 0.76 ESR
6.48pf 0.60 ESR
6.52pf 0.22 ESR

Could I swap these for 10uf 50v just to test...??

[budm]

Re: LM317T Voltage Regulator Input Resistor Cremation.

So you test the input resistances of the working one in the circuit, but how about on the one that the resistor keep burning up?