SKR Mini E3 v1.2 fans always on 100%

[galardao]

Hello,   Sorry about my bad English, I speak Portuguese. I recently purchased an SKR Mini E3 v1.2 card.   Everything was working fine, but yesterday the cooling fan did not comply with the chosen speed, it was 100% all the time regardless of any settings on the panel or in the gcode. Measured with multimeter on the fan0 connector and is always with 23.6 volts. I would like help in solving this problem.  Best Regards,

[ymurenko]

I think we should look for a way to prevent this issue altogether. Is there a reliable way to protect this mosfet from whatever plagues it? From what I understand, the PL4009 is just barely suited to running a 24v PWM fan, and any voltage spikes or potential noise can put it in jeopardy. The stock fans on the Ender 3 aren't exactly high quality, and neither are many of the 5015's people use, I take it that poor fan circuitry kills this frail mosfet?

@CRTified You and @DaveJS46 seem very knowledgeable with electronics, what do you guys think we can do to avoid having the mosfets die? DaveJS46 mentioned putting a cap between the grounds of FAN1 and FAN0, but I don't really understand how this would help, could you guys can explain the principle behind this solution? Wouldn't a bypass/decoupling capacitor on the FAN0 pins be a better a solution?

I was thinking of putting a buck style circuit between the fan header and the fan (a cap, diode, and inductor) in order to reduce PWM noise and also protect against voltage spikes (two birds one stone hopefully). Another thought I had was slapping a tiny heatsink on this mosfet, since its power dissipation is only 1.4w at 25c, which seems to be very small especially with dual 5015 blowers, or is power dissipation unrelated to this issue?

Thanks

[DaveJS46]

@ymurenko If you look at the diagram in my 28th September comment, the capacitor is in effect across the mosfet terminals Drain and Source. A capacitor reacts instantly to absorb very fast pulses - which is what my fans produce. Diodes my well do the same job but all have a switching time so react to the pulses less quickly. If you fit a 10uF 50V capacitor before you connect and run the fan(s) you will at least have some protection, but if your fans have different characteristics and noise generating parameters from mine then possibly the protection will not be sufficient. As the SKR earlier version boards used the same mosfet to drive FAN0 as the heating elements and it seems without failures then maybe the best solution is to remove the PL4009 and wire in its place a WSF3085 mounted on a small stripboard or similar.

[mikegleasonjr]

I did the capacitor trick to my V2 board, so far so good...

[ymurenko]

@DaveJS46 Ahh, that makes perfect sense, I guess I didn't look closely enough at the schematic. If I understand correctly, since FAN1 isn't controlled, you're just using its ground pin. I was confused for a moment because I didn't realize that this revision of the board has only one controlled fan.

I'll give it a shot, I'm using the stock blower that came with the printer, but if we wanted a universal solution for all fan configurations, would a schottky diode with appropriate specs be a better solution? (I've had some issues in the past where the PSU would click due to the inrush current current from a 10uf cap)

@mikegleasonjr I don't think that's a good idea on the V2.0 board, because it has two of these pl4009 mosfets instead of one, which means two controlled fan headers. You're not connecting the capacitor to ground by connecting it to FAN1 ground, because those pins go into the second mosfet's drain. If you wanted to do this solution on the V2.0 board you would need two caps, one for FAN0 and one for FAN1, and each connected to ground somewhere else on the board.

[DaveJS46]

@ymurenko I know this topic is under the E3 board but as the mosfet failure occurs on the SKR 1.4 boards as well, the problem ended up being included here and my comments do, in fact, refer to the 1.4 board and its layout. Anyone who has an E3 board needs to be aware of the differences that you outline above. With regard to using a fast acting schottky diode, well yes, it may well work for you - it is up to you to try that route. The problem generally is that to find a solution that works for all fans one needs to test them all and as hobbiests I don't think too many of us have access to more than one or two types of fan. We can but hope that BTT will eventually design the problem out on the next iteration of the boards. As an aside, your comment that you had a PSU that had problems charging a 10uF cap - all I can say is that I can't imagine that PSU being able to drive hardly any load and certainly not a couple of fans. Most DC PSUs have smooting caps (i.e. across the output) way in excess of 10uF.

[ymurenko]

@DaveJS46 Thanks for you help, I ended up using a 10uf 50v cap since I don't seem to have any schottky diodes. As for board layouts, the FAN0 configuration on the 1.4 is slightly different than on the 1.2, but connecting the capacitor between FAN0 gnd and FAN1 gnd is effectively still a connection between drain and source, so this solution still applies to the 1.2 board.

I had little room to solder the cap on (the back of the board sits very low), so I opted to crimp the cap's positive leg into the fan connector, and the other leg into an empty connector. I drive my board fans from the hotend fan terminals, so I didn't need FAN1 for anything.

It's been running well so far. I've tested the fans at different speeds using hardware pwm, and with marlin's SOFT_PWM and SOFT_PWM_SCALE at each frequency. Of course, I'll never know if my fan mosfet was destined to fail, but I can confirm that this solution doesn't cause any issues on my board.

As I've mentioned before, you shouldn't do this if you use a 2.0 board, because both FAN1 and FAN0 are driven by mosfets. Connect the cap's negative straight to ground elsewhere on the board.

[DaveJS46]

@ ymurenko You're welcome - I'm glad that you are happy with the mod. As you rightly say we don't know if a working mosfet is destined to fail or not. However, as I did run some tests on the fans connected to my bench PSU and observed the noise pulses on my 'scope and then saw how those pulses were greately smoothed (from 9V peak to about a volt peak) so I know that the cap keeps the Vds max in spec.

[mikegleasonjr]

@ymurenko

Thanks I removed the CAP from my V2 board. Everything seems to work fine, both fans are controlled normally.

[ChuckNoxis]

My part cooling fan is doing the same : going on when I power on my skr mini e3 v2 and staying always on at 24v :(

[mikegleasonjr]

My part cooling fan is doing the same : going on when I power on my skr mini e3 v2 and staying always on at 24v :(

There are 3 fans: one always on which is mounted facing us on the hot end. For me, this is the one that makes the most noise. It is not controllable since it is plugged directly to 24v.

Another is on the side on the hot end. It usually starts on the second layer of a print.

A third one is in the controller box. It starts when stepper motor are active.

Technically there's a fourth one in the PSU which starts when it gets hot.

Which one are you referring to?

[Nenzyz]

In my case I bought BTT SKR E3 mini and after shortage had to replace PL4009 mosfet to another with similar characteristics named AO3400 SOT23. This doesn’t help, after replacement FAN0 is always on as well as diode. I ordered another MB from MKS Robin E3D and same shortage issue happened with this one, but in this case replacement helped and MB is working now. This is weird but it shows the difference btw vendors.

[ChuckNoxis]

My part cooling fan is doing the same : going on when I power on my skr mini e3 v2 and staying always on at 24v :(

Another is on the side on the hot end. It usually starts on the second layer of a print.

I'm talking about this one. In order to hotfix it, I changed PC6 and PC7 ports in Marlin so I get back my PWM control on the part cooling fan (but lost the control for the Motherboard/controller box fan).

[blockfeed]

I had the same issue on SKR Mini 1.2 (Ender 3 Pro), no fan control or variable fan speed. Replaced Q2 with a SI3404-TP (from digikey, as posted by @sregister ) and everything is working as expected now.

[senojed]

I used an IRLML6344.

Hi, thanx for the post! Is posiible to use IRLML6344TRPBF? could not find the difference :)

[1CM69]

I thought that I was losing my mind until finding this issue thread.

I have successfully replaced the faulty MOSFET on my board now, I used a NEXPERIA PMV19XNEAR.

Tested the fan, same as I have been using prior to the swapout of the MOSFET, running from 100% down to about 10% before stalling, then it needs to be raised to 20% to startup again.

Before swap I'd killed 3 fans in a week just by dropping them anywhere below 75%, they'd never function after this.

;)

[CRTified]

Is posiible to use IRLML6344TRPBF?

@senojed

Yes, that is the same one. The TRPbF (afaik) only means that it's packaged on a tape and reel and has a lead-free finish, which makes no difference for you if you only order a few one (so it's only packaging and RoHS-conformity).

[senojed]

Thank you for explanation :)

[testack18]

Take care! If you replace the mosfet on your own, you will lose warranty... Hold on a sec! Warranty?? I just forget that bigtreetech doesn't really know what is warranty! I purchased an skr 1.4 turbo board on aliexpress on july 2020 which worked for 7 month long then I had the exact same mosfet issue. They told me that as the board was purchased last july they don't know if this is a quality issue or an operation problem. They only offers a three month warranty period in case of quality problem. This is terrible service! For a brand new electronic product has only 3 month? Furthermore in case like this? This is clearly a design issue as they using a junk mosfet.

So take care with bigtreetech's products.

[selalipop]

Replacing my mosfet with a AO3400 mosfet worked perfectly in my case, no need to replace the resistor here. When removed mine was visibly burnt through

[CoefficientofSquish]

Is there a firmware workaround for this? I'm using on MPCNC with no heaters, could I reassign one of the heaters to a fan in order to use my laser? If so, what pins do I need to change in Marlin?

[Skedar]

Someone know what mosfet i can use in Q1 on a SKR mini e3 v2? i make a shortcircuit on my hotend and my board stops work. i think is the mosfet and the regulator 3.3v, but i cant find any info about the code of mosfet in this board WSF3085 HK09F

[tazma]

Hi, I have the same problem skr mini v 2.0 board Burned fan mosfet is possible to change the burned mosfet for a new one I have no problem soldering the new one, but I would like to know what alternative or equivalence to change mosfet if someone knows equivalence please an explanation for fools, since I do not understand electronics nothing Thank you

https://fotos.subefotos.com/2d79aa3f16ed53ab2ec16630ba157ea7o.jpg

[Skedar]

Hi, I have the same problem skr mini v 2.0 board Burned fan mosfet is possible to change the burned mosfet for a new one I have no problem soldering the new one, but I would like to know what alternative or equivalence to change mosfet if someone knows equivalence please an explanation for fools, since I do not understand electronics nothing Thank you

https://fotos.subefotos.com/2d79aa3f16ed53ab2ec16630ba157ea7o.jpg

I change mine, and works fine.

model of Mosfet is A09t Ao3400 N-channel

[tazma]

ok thanks i'm looking for a component

Today looking for remains of motherboards, etc., etc. I see that there are many mosfets sot-23 but I do not understand the components and I wonder if any of these mosfets that I have about 20 that are different would some be compatible for skr mini v2.0

[xjustbmw]

So i just installed e3 v2 board on my ender 3 and it works great since a few days

I am planning to add a 5015 cooling fan soon.:)

Anybody made the capacitor mod successfully on their skr mini e3 v2 board with good results on bigger fans to avoid mosfet damage?

Please provide clear instructions so we dont have to replace mosfets

Many thanks

[radek8]

BTT seems to have learned and is installing a protection diode in the new boards.

This is the SKR Mini E3 V3.0

[Will-Hollingworth]

BTT seems to have learned and is installing a protection diode in the new boards.

That doesn't seem to do much. My new SKR Mini E3 v3 died with the fan stuck on. I had been playing with adjusting the fan speed. No response from BTT on getting any kind of replacement.

[duMagnus]

@javimagro I know it's been a while, but I'm having the same issue you were on my SKR mini E3 V3. My fan1 shorted out and since the heating elements for both the bed and the nozzle don't work. Did you ever figure it out? I saw that the fan1 mosfet got blown out in my board and replaced it, but this didn't fix my issue.

[cryptopivot]

BTT seems to have learned and is installing a protection diode in the new boards.

This is the SKR Mini E3 V3.0

That diode will only prevent other fans from getting bleed over voltage and spinning up , not a protection mechanism. hence the reason v3 fets continue to burn.

[OliverParoczai]

Thanks for the help! I used an AO3400 as a replacement on the SKR E3 DIP V1.1 board. After the replacement, the fan speed can be adjusted again. In my case, the original got blown because it couldn't handle the power draw of the laser module.