A coworker is having issues with a 3-terminal voltage regulator. I don't know which part, but for the sake of discussion let's use one like the LM317T. Anyway, in circuit its output looks like a sawtooth wave instead of being well regulated at around 1.5 volts. Is there a likely simple explanation, or somewhere I can direct him to investigate (such as input bypass cap failure)? Or do I need to get more details from him (amplitude, DC bias, frequency range) before anyone can take a quick crack at this?

 

Anyway, in circuit its output looks like a sawtooth wave instead of being well regulated at around 1.5 volts. Is there a likely simple explanation, or somewhere I can direct him to investigate (such as input bypass cap failure)? Or do I need to get more details from him (amplitude, DC bias, frequency range) before anyone can take a quick crack at this?

IS he using LM317L, the TO-92 version?

Is the LM317T mounted on heatsink? What is the input voltage and load current draw? Nature of the loading?

If the frequency is high, say over couple of hundred Hertz, it points to oscillation from instability(input, output caps problem).

Low frequency oscillation of a few Hertz or slower usually points to thermal overload limiting engaging and resetting(heat sink, mounting etc..).

 

All diagnoses are possible with a schematic. Is there a filter cap on the rectifier?

 

I seriously doubt I'm going to see any schematics; I'm not on that project. I can ask tomorrow for more details about the regulator output they're seeing. Apparently this circuit worked properly until within the past week or so, so it seems like SOMEthing failed. I was wondering if perhaps an input or output bypass cap failure (assuming they're designed into the circuit) would cause this behavior, or perhaps the load open-circuited. The guys trying to diagnose this don't seem to be really into working at the component level; I just seem to remember at sometime in the past seeing or hearing about conditions where voltage regulators... well... don't.

When I find out a little more tomorrow, I'll let you know.

 

When used as voltage regulators, they really need at least an 0.1uF cap right at the input to ground and another at the output to ground.

They also need at least a 10mA load on them at the output to provide guaranteed regulation, which is usually provided by a 120 Ohm R1 from OUT to ADJ. Since Vref (the voltage on OUT referenced to the ADJ terminal) is nominally 1.25v, there will be nominally 10.416mA flowing through R1; in order to get 1.5v out, R2 from ADJ to GND would need to be ~24 Ohms.