FEATURED Building PWM Controller for 4 wires PWM fan

[txus.palacios]

Replace the 470pF capacitor with a 680pF one, IIRC.

 

[Bobnova]

I haven't noticed anything like that in my oscilloscope testing. I won't say it is impossible, but I will say it is unlikely.

 

[Jasonaaronfox]

I haven't noticed anything like that in my oscilloscope testing. I won't say it is impossible, but I will say it is unlikely.

Its not something you could see on a scope, it's something you could only see with a high speed camera haha.

 

[Bobnova]

You're not going to see anything physical, far too much inertial for that.
If you're talking physical wear, I say no way.

 

[Jasonaaronfox]

You're not going to see anything physical, far too much inertial for that.
If you're talking physical wear, I say no way.

Excellent point, thank you for clearing that up. I am more used to combustion based motors than dc.

 

[inVain]

Do you happen to have a drawing of that setup available? Currently I have the dual 555 circuit setup with reverse Schmidt buffer (from here) on my breadboard and my 260 cfm San ace 12v 4a is of its way in the mail. I'm curious as to what I have to change to be able to run from 0-100 at 16khz as you are doing.

Thanks so much for the information!

forget building the 2-98% if you're aiming to control the SG...
it's whining noise was unbearable, even if you set it at 16Khz

On my experience the san ace 12V 4A 260cfm works well even at 25 Khz. No difference between 16 Khz and 25 Khz signal noticeable.

For the circuitry schematic have a look right here.

yep, I'm now running the 0-100% with multiple 555 and no more noise issues

he're what I've got to tame the puppies

for 16Khz signal, try to use 820pF to substitute the 560pF C1 on the 0-100% schematic,
as in my case, I use the 560pF in parallel with the 220pF to get the closest result (I'm out of luck to get the 820pF )

 

[Bobnova]

That looks amazing. I continue to be impressed!

 

[Jasonaaronfox]

forget building the 2-98% if you're aiming to control the SG...
it's whining noise was unbearable, even if you set it at 16Khz




yep, I'm now running the 0-100% with multiple 555 and no more noise issues

he're what I've got to tame the puppies

for 16Khz signal, try to use 820pF to substitute the 560pF C1 on the 0-100% schematic,
as in my case, I use the 560pF in parallel with the 220pF to get the closest result (I'm out of luck to get the 820pF )

Wow... The 0-100 dual 555 with opamp works amazing in my simulator. Is there any reason a 556 couldn't be used in place of the two 555 ic... I have one laying around. I'm no expert in electronics so please forgive any stupid questions I may have.

 

[bing]

Wow... The 0-100 dual 555 with opamp works amazing in my simulator. Is there any reason a 556 couldn't be used in place of the two 555 ic... I have one laying around. I'm no expert in electronics so please forgive any stupid questions I may have.

Here you go.

Note : For 16Khz, replace the C1 with 820pF.

View attachment 112674

 

[Jasonaaronfox]

Here you go.

Note : For 16Khz, replace the C1 with 820pF.

View attachment 112660

Thank you so much! This will help me make the controller small enough for my custom itx build.

 

[inVain]

Wow... The 0-100 dual 555 with opamp works amazing in my simulator. Is there any reason a 556 couldn't be used in place of the two 555 ic... I have one laying around. I'm no expert in electronics so please forgive any stupid questions I may have.

well,
I can't tell you correctly what's happening either, sorry
they look identical on their datasheet too

I'm really having a hard time with the SG until I build a controller with multiple 555,
but I can tell you that having a 0-100% with the C1 "near" to 820pF (this all I can get), is better than anything else (except for 2x 555 )

since you've already got the part, you should give it a try....
just buy some spare 555 if by any chance you need to buy more components

 

[Jasonaaronfox]

Here's what I'm planning on building when I get home from vacation. Works flawless in the simulator at around 19k frequency. Close enough I suppose. The labels saying 10nf should be 100nf. I was very tired when I drew these up.

 

[Bobnova]

I'd bridge the two ground wires just for good measure. I've found as much as a 0.13v difference between a Molex ground pin being used by a Fan and one not being used.
That shouldn't be enough to cause issues, but you never know.

Do you have PCB manufacturing supplies or do you have it done somewhere? I've always thought it'd be fun to do yourself.

 

[Jasonaaronfox]

I'd bridge the two ground wires just for good measure. I've found as much as a 0.13v difference between a Molex ground pin being used by a Fan and one not being used.
That shouldn't be enough to cause issues, but you never know.

Do you have PCB manufacturing supplies or do you have it done somewhere? I've always thought it'd be fun to do yourself.

They are pretty easy o make yourself. My wife went to a technical school for engineering during high school, before she went to university and became an English professor. She helps me make all of my electronics. Thanks for the tip on the molex ground wires.

I may go with this design instead to save a little space. It places the fan pins closer to the molex. The labels saying 10nf should be 100nf. I was very tired when I drew these up.

 

[bing]

They are pretty easy o make yourself. My wife went to a technical school for engineering during high school, before she went to university and became an English professor. She helps me make all of my electronics.

Really envy you on the wifey thing , anyway, hope everything will run smoothly and expecting some pics for the finished circuit.

 

[Jasonaaronfox]

Really envy you on the wifey thing , anyway, hope everything will run smoothly and expecting some pics for the finished circuit.

Thank you, I truly am blessed of God to have a wife who loves me.

Here's hoping I can source all my parts tomorrow here in Costa Rica!

I'm also going to try to set up a seven segment indicator to show 0-9 as I turn the pot.

 

[Jasonaaronfox]

Well gentlemen, I have hit a snag. My local electronics shop is out of 9.1k resistors and 10k potentiometers. They have smaller resistors and 5k pots. I'm using the 0-100 556 schematic I showed above (bread boarding until I find the parts,) and I have the circuit working... "kind of" off hand all I have available is a number of 10k resistors, 500 resistors, and a few 5k pots. I obviously am having trouble getting a reasonable sweep from high to low pwm. Any suggestions?

 

[Bobnova]

4.7k resistors and 5k pots oughta work if they have 4.7k resistors.
Pair the 10ks up to make 5ks (two 10k resistors in parallel makes 5k), that'll work decently.
I don't know if you'll get all the way to 0 and all the way to 100, but it'll be close.

Failing that you can chain four 1ks together in serial to make 4k resistors.
Messy, be functional.
Add a 500 to the chain and you have 4.5k.

 

[Jasonaaronfox]

Well, looks like I have to go to another city to find parts. Just as well, they did not have double sided pcb in town either. At least the circuit works in the streamlined layout.

 

[oscarbravo]

Hi im trying to get my Delta's to work ive a zalman PWM mate on a loom ive rigged up so powers from molex for 12v and 5v etc but won't respond.
Tried grounding across from black to blue with a wire which had no result so wondering
what or how does the delta pwm circuitry work/configure? as tried an old intel stock cooler and that worked ok chopped that wiring/plug off to use to fan as know plug pins make contact ok and connected to fan wiring.

I like that what looks like a bay controller on InVains posts as thats the sort of thing im looking for as going to run 3+ sets of fans.

Anyway guy's keep up the good work.