Thanks all of you. I do have a multimeter. I will test it an supply the results. I can definitely understand that the demand spike is a root cause of the issue but forgive my ignorance as I thought the draw or amperage would always be the same since the lm317s simply control the voltage given the circuit I'm using? As a good guess is that the motors probably draw 1amp at most. I don't believe more is possible? And the motors are only separated from the mains by the heating coils resistance and 4 bridge diodes to rectify to dc. I'm only speculating. I'll attach a picture of the supply if it helps

 

I was a little off on specifications

 

Adding another LM317 in parallel, or a boost transistor added to one LM317 would likely solve the problem.

Or using one for each fan?

 

It may also be the 19V power supply that is overloading.
Look at the reg input volts to see if it is stable.
It may be easier to just get a 12V supply. Or mains fans.

 

It may also be the 19V power supply that is overloading.
Look at the reg input volts to see if it is stable.
It may be easier to just get a 12V supply. Or mains fans.

I was wrong on the amount of lm317s in my previous post. My earlier circuit worked well using 2 lm317s for 3 fans but I accidentally destroyed it but I rebuilt it using 4 just in case and the new one is causing this issue. And I'm starting to believe it could be the supply. Possibly dividing through more ICs is causing this?

 

Hello there! After reading this post in its entirety.
If you wish a solution with a high degree of success you must provide an updated schematic, wiring diagram, components used. Your avatar is rolling over in his grave!

 

You have been asked for a schematic a number of times but you have not provided it. All you have done is post a copy of part of the data sheet. You now say you have a number LM317s connected in parallel. We need to understand what precautions you have taken to ensure the regulators share the load equally. I think it would be better to use a single suitable rated regulator such as an LM338.

Les.

 

@crutschow @dendad to clarify this has worked well before. I can guess that the extra fan is demanding more current than is available but shouldn't that mean that it wouldn't work at all instead of working if it's slowly adjusted? Also I can't tell for sure how much the amp draw is. These motors have no easily discernable markings. They do run easily on nearly any voltage and even down to <5v@150ma that I've noticed. I had thought heating was an issue but I can absolutely rule that out.

What kind of heat dissipation did you have on the LM317? It could be it needs a larger heatsink to help it dissipate... avoid thermal lockout

 

Thanks all of you. I do have a multimeter. I will test it an supply the results. I can definitely understand that the demand spike is a root cause of the issue but forgive my ignorance as I thought the draw or amperage would always be the same since the lm317s simply control the voltage given the circuit I'm using? As a good guess is that the motors probably draw 1amp at most. I don't believe more is possible? And the motors are only separated from the mains by the heating coils resistance and 4 bridge diodes to rectify to dc. I'm only speculating. I'll attach a picture of the supply if it helps

If you're running the fans in parallel, the voltage may remain the same, while the current will be 3 times what 1 fan would use.

when people ask for a schematic- they want to see a schematic of your entire circuit, not just the regulator suggestion from the datasheet. It lets us evaluate how all the parts work together and we can offer improvements/corrections, etc.

Just draw a rough sketch, if that's all you can do with all parts labeled and specs provided (volts, current, rating, size- whatever they are (ohms, watts, etc).

 

As you reckon there are now more LM317 regs in parallel, it is sounding more like as in post #24, is the 19V power supply is not big enough to start the fans.
It could have been with 2 regs, they went into current limit so the 19V power supply held up ok, but now with more, the 19V power supply is overloaded because the LM317s do not limit the current.
Have you measured the 19V to see if it is pulsing yet?
You could try lowering the volts to 10V and see if that helps.

 

 

its something like this. i am very sure heating isnt an issue as the ICs and heatsinks are in the direct path of the air flow and touching them barely feels warm. a multimeter test didnt show any pulsating voltage with or without the fans attached. so maybe a complete dip in amperage?

 

The datasheet for an LM317 shows that its input must be at least 2V higher than its output.
The LM317 is linear therefore you cannot parallel them like that.

 

The datasheet for an LM317 shows that its input must be at least 2V higher than its output.
The LM317 is linear therefore you cannot parallel them like that.

I don't disagree but many previous times I've done so the fans power was dramatically higher than with a single IC. this is also the first time I've encountered this issue.

 

I thought you were having 12V out. The 18.46V ?????
Can you set 12V?

 

Thank you for posting your project.
Did you assemble your components by point to point soldering?

I did, with solid gauge copper wire, going the shortest distance possible. Everything checks out and it does work. Just not like it did previously(same type different seller). Deep down I'm thinking they are inferior ICs but I can't admit it since they were from a "reputable" dealer.

 

I thought you were having 12V out. The 18.46V ?????
Can you set 12V?

My apologies! That was a typo!

 

As mentioned, just having a number ot LM317 ICs in parallel is not the best way to go.
Still, I'm concerned the volts is not 12V.
Another thing you could try, once you fix the 12V, is to add the slow turn on option mentioned in the data sheet.

 

As mentioned, just having a number ot LM317 ICs in parallel is not the best way to go.
Still, I'm concerned the volts is not 12V.
Another thing you could try, once you fix the 12V, is to add the slow turn on option mentioned in the data sheet.
View attachment 242448

Slow turn on is the problem I'm having now. Unfortunately it needs to be quickly adjustable even from on to off and vice-versa.

 

To parallel regulators, you can't just put all pins together, you have to equalize the current from each regulator and adjust the drive/adjust voltage. This is a proper design: (and your input voltage/current must be high enough)