Powering phone from LM317 regulator. Voltage not constant?

[nharrer]

Hi.

I have an old cell phone whose battery is dead. So I thought I just power it straight from a power supply so I can use it for some stationary tasks like listening to music on my desk.

I removed the dead battery and hooked up a lab power supply and set the voltage to 4.2V. Which worked without a problem. Since the lab power supply is bulky and I need it for other things I thought I would use a LM317 to build a little circuit to power the phone. On a breadboard I tested the standard LM317 circuit with C_in 100nF ceramic cab, C_out a 10uf tantalum, R1 330 Ohm and R2 a pot adjusted, so that V_out is 4.2V. For the test I used again the lab supply for V_in with 9V.

Weirdly the phone starts to boot but goes black after a few seconds. I measured the regulated voltage and it jumps around 4.2 - 3.8V as soon as the phone starts booting. I also measured the current, which jumps around from 100mA to 800mA. I also used a 0.1 Ohm resistor in series to measure the current with a scope. In infinite persistence, I can see that the current sometimes jumps up to above 1A. But only very shortly. You can not see that on the multi meter.

The output voltage of the LM317 does not seem to be very stable. Even at around a 600mA load, the voltage is already down to around 3.8V. I guess the high current jumps are causing the voltage to drop so low that the phone shuts down? Am I asking to much of the LM317? Or it's something else I don't see?
I mounted a little head sink to it. Because first I thought it would go into thermal shutdown. But the funny thing is if I increase V_in to 12V it seems to be more stable and I usually get the phone to fully boot up. Which causes to LM317 to get even hotter.

Any suggestions? Note, that I am a real novice. That is the first time that I try to use a voltage regulator. Maybe I am doing it all wrong...

Regards,
Norbert.

[Benta]

Reduce R1 to 120 ohms. This has been thrashed to death on this (and other) forum.

[Cliff Matthews]

Yeah but the last post on that was ages ago.. 11 days at least 

[nharrer]

Yes. I read some of those. And I tried that. But no change. In fact it's even worse. Now the phone rarely starts with 12V V_in, which it did with R1 at 330 Ohm.
I thought the 120 Ohm was for problems with minimal load? But I guess I have too much load. But still under 1.5A.

[bktemp]

Did you try a resistive load to verify your circuit?
Breadboards can have a high contact resistance up to 1 ohms for cheap boards. That gives a significant voltage drop at 1A.

[nharrer]

Thanks bktemp. I think that was the problem. And not only the breadboard, but also the flimsy jumper wires connected to pretty long alligator clip leads that I used.
I used a 5 Ohm load. The voltage at the leg of the LM317 was constantly 4.2 and at the load resistor only 3.6V.

In the meantime I also found this thread where the OP had the same symptoms:
https://www.eevblog.com/forum/beginners/lm317-regulation-droop/msg686401/#msg686401
Here the problem was also the breadboard. Building the circuit on a proto board solved his problem.

Unfortunately I seem to have done something bad to my phone. It now draws 700mA even if it is turned off which used to be nearly zero. Turned on it's over 2A. Surprisingly it still turns on and it seems to work. But I guess something is (partially) shorted out now.

Anyways. Thanks for all the input.

[Benta]

If your phone was turned off at some point with R1 at 330 ohms, you can be certain that it was fed overvoltage from your LM317.
How on earth did you choose 330 ohms? Even the datasheet shows 240 (which only applies to LM117/217).

[nharrer]

I was experimenting a lot. I used 120, 240 amongst others. 330 was just the last try before I posted here. However I measured the voltage when the regulator output was open circuit and it was not above 4.2V. But you are right. I did leave it plugged in for a while when the phone was turned off. Maybe it went instable when I was not looking.
Not much harm done though. The phone was old, had a cracked screen and a dead battery. I was about to throw it out. But at least it had a final mission as a guinea big .

[Zero999]

Does the LM317 get hot?

The 330R vs 120R probably isn't the cause here, as when the phone is is booting, it's almost certainly drawing enough current to satisfy the LM317's minimum load requirement.

[nharrer]

Right, it was not the resistor. With the help of bktemps post we figured out that the resistance of the breadboard and crappy leads were causing a pretty significant voltage drop at higher currents. I went right into this trap for young players. Not that I am that young anymore, but EE-experience wise I am an infant.