Issues with IDEX 4028 part cooling fans quickly changing speeds
Hello all,
I have some weird issues with my 4028 12v Part cooling fans connected to my Octopus MAX EZ board. Im in the process of setting up my VC4 IDEX, everything worked fine in CoreXY and Hybrid but since putting in a second toolhead I'm now seeing some issues. Both fans are connected to either the 12v pin on the 2 pin fan port OR the 12v pin on the 4 pin fan port, both work independently (only 1 fan connected to power at a time) and I can ramp the speed up from 0 to 100 without issue. As soon as I introduce the second fan I can only gradually increase the speed of the fans, as soon as 1 fan has a quick increase in speed BOTH fans will stop working. You try to adjust the speed on the sliders and nothing happens, the only way to the fans working again is to reboot klipper.
I've tried a different combination of 12v 4028 fans (Nidec, Sanyo & Delta) but they all act the same way. I've tried an independant 12v PSU to connect the fans and that didn't work. Both fan are connected as per this guide: https://os.ratrig.com/docs/guides/4028/ using the gnd on the heater ports. Any help/suggestions would be very much appreciated.
Connecting a 4028 fan | RatOS
4028 server fans are becoming a popular option for part cooling because of their light weight and great pressure and flow rates. In this guide we'll go over how to connect one and use it for part cooling. 4-pin fans usually aren't used in 3d printers, and many boards only provide 2-pin ports, but fear not! There's always a way.
46 Replies
So you have the pwm line from both fans going to the same port on the octopus?
No, the PWM line is on the individual toolboard t0 and t1 respectively. 2 heater ports for GND and then 2 12v pins for the power
so the only thing they have in common is the power supply. If you slowly get them both to max power, it's fine. If you rapidly put either one to max power, it's fine. But if you try to rapidly set both to max, both fans stop running?
Yes, they don't even need to get to max power. Just a substantial change in speed on either fan seems to be enough to kill the power to both fans... However If I completely disconnect 1 of the fans from the 12v power I can change the speed as much and as quickly as I like without issue.... The issue seems to be when the second fan is introduced into the mix.... It's an odd one and I'm pulling my hair out.... Not that I have much. 😂😂
If you change it quickly lower does it still break?
If I go slowly like 5 or 10% at a time then it's fine....it seems like the issues is a sudden large change in speed.....
large change either direction ?
Yep....
baffling
This is the kind of thing I'd want an oscilloscope to investigate further
I had thought about trying to run both fans back to the Octopus using the full 4 pins to the 4 pin port on the board..... I don't see that making a difference though
I'm starting to think there might be an issue with the octopus
Any errors in klipper?
You might try enabling debug mode, getting it to happen, and post your debug.zip (I don't know exactly where those are done, just seen it mentioned in here)
You're using 4 pin 4028s?
Nothing reporting in the console..... I will have go at enabling debugging mode
Yes.... 4 pin 4028s but only using the 3 pins as per the VC4 build guide
Okay, and when the problem happens, the fan speed goes to 0, or full blast?
PWM to the toolboard and then the love and ground back to the octopus
The fans stop but the sliders stay at the speed originally set
Adjusting the speeds after they lose power makes no difference. They just won't start spinning again
Okay, that is extra weird. The 4028s are server fans and they default safe, as in if they stop getting a pwm signal they turn on full blast to keep stuff from melting.
if you disconnect the pwm wire, do they go to full blast or not do anything?
When do you mean, while I have them running or before I turn the printer on?
at all. Safest would be disconnect while printer is powered off, then power it on and see what the fans do
You may have the "wrong" 4028s and their behaviour different
Unfortunately my printer is at work so I'm unable to test tonight but can grab debug report and test pwm in the morning
Cool. This is a fun puzzle, hope we can find the root of it soon 🙂
I have the Sanyo 9GAX0412P3S001
The ones recommended in the how to.connect a 4028 guide
I've also tried the nidec one that comes with the VC4 kit and I have a delta one which I've not tested yet
I also have another Sanyo but a slightly different model number
They ALL act the same way
that 9GAX should be the right one yeah
if they don't get a pwm signal, they should go to 100%. If they stop spinning, then they are getting a pwm signal, just not the one we want.
Ok. That makes sense. But why would they lose a signal and not be able to establish it again? I have the EBB42 1.2 version
Both toolboards plugged into the waveshare USB pi hat
I have that toolboard, but my 4028 is being run all the way back to the octopus (set it up before I got the toolboard and didn't feel like changing the fan)
Do you have the Pro or MAX EZ board
Pro
the octopus doesn't have 4 pin fan ports?
The MAX EZ does.... 3 of them
ahh, when most people say octopus they mean the regular one that is in the "standard" config
No.... The issue is with the MAX EZ board
Attached are some debug logs regarding the fan issues.... not sure what I am looking for
Ok, so when I disconnected the PWM pin from the toolboards and started the printer the FAN didn't start spinning at 100%, they didn't spin at all. I then increased the fan speed on t0 to 1% and the fan started spinning at full power. I then changed the fan speed of t0 to 0% and the fan stopped. I then repeated this on t1 and the same happened. I then set t0 to 1% (full power) and did the same with t1 and both fans instantly turned off!!!!
Wtfff
How is 1% full power?
because he disconnected the PWM wire
I then set t0 to 1% (full power) and did the same with t1 and both fans instantly turned off!!!!this is powered through the board right? Can you try connecting positive and ground directly to your PSU and then attach the PWM back and control it? i.e. I wonder if the board is trying to protect itself from too much current draw at once
Yes this is powered through the octopus Max board, just the pwm wire on the toolboard. I've tried rewiring the fans up as 4 pin fans going back directly to the MAX board but i have the same issue. seems like they aren't receiving enough juice or is being throttled back. I will try re wiring a 12v psu tomorrow and report back
If it is just the pwm wire on the toolboard, then it shouldn't be the toolboard's over-current protection kicking in.
Hi, if I'm connecting both the positive and negative to the PSU it obviously won't use any enable pin so should run at full power until Klipper boots??
I have max ez, and have both fans working properly.
I assume you tried a seperate at least 3A psu for them?
@michaelaswood And why isn't your PWM on the toolboard?
I'm going to try and wire the fans up to a 12v PSU tomorrow.... But I shouldn't be needing to as the MAX should be able to cope with the 12v fans. The PWM has always been on the toolboard. It was only take off for testing purposes. I tried wiring the fans back to the 4 pin fan ports on the MAX back that made no difference.... Still experiencing the same issue when suddenly increasing or decreasing the speed of one of the fans while the other is running.....
Yes but not that amperage
It has the e fuse as well which is funky.
For max ez I recommend a step down converter
That supports the amperage of 2 fans
Which is around 2.8 max I think.
Depends on the fans in reality.
These fans are a beast, so 12V rail for them is important.
1A continues
They are all above 1A
4028s we use
So you basicly trigger the protection every time you hook them like that.
So how do you suggest that they should be wired up? If this is the case regarding the amperage then why are others not experiencing the same issues? Sure with IDEX and both 4028 fans spinning up all IDEX users will be running into a similar situation?
Could you please explain what you mean by a step down converter? Is that what they also call a BUCK converter?
You are running the Ez, not sure how many people do for idex, and how they wire the fans with the z, but since the ez has that hard limit and the E-Fuse this is what worked for me.
I have my 4028s connected to the heater grounds, PWM on the Toolboard, and 12+ wire, was converted from my 24v to 12v with a step down driver.
You have for instance these, but I recommend getting a higher quality ones:
Because they are at that 3amp range but are very small, so they are usually running at 2A which means if you go this route, I suggest getting 2 of them (one for each fan)
Or buy a quality 3A one.
Something like this
What you are looking for:
Input power (based on your psu, but at least 24v or more)
Output power (12V)
Max Amperage (A): 3A (at least)
This again depends on your Fan Amperage
Sanyo Denki 9GAX0412P3S001: 1.3A (x2 is 2.6A)
DELTA - FFB0412UHN: 1.9A (x2 is 3.8A) (In this case oyu need a coverter that has at least 4 but always leave 10% margin)
NIDEC W40S12BS2E5: 1.6A (x2 3.2A) at least 4A converter
Etc.