Can this Noctua 12v Fan be fixed?

[enut11]

My 120mm Noctua fan died. It was great while it was working, efficient and relatively quiet.
Since they are expensive, I decided to have a go at fixing it. There are lots of YT video on fan repairs but when I pulled mine apart I did not see the familiar 3pin Hall chip but rather an 8 pin IC.
There are no markings on the chip apart from an embossed square in one corner.

Not familiar with Hall chips so looking for tips on how to trouble shoot this circuit.
enut11

[enut11]

Powering the PCB from a current limited 12v supply shows up as a short or very low ohms circuit.
All the discrete components tested OK. The two coils both tested at around 36ohms so that seems OK.
Sounds like the 8 pin chip is dead but not knowing the pin-outs makes it difficult to test further.

Any suggestions before I pronounce it dead and unrepairable?
enut11

[RoGeorge]

Any suggestions

In case you don't find a replacement for the dead component, maybe try to fit an entire PCB taken from another cheap/used 4 wires fan.  The low noise of a Noctua comes mostly from the mechanical parts, coils might not be much different from the coils of an ordinary fan.  Chances are it might work, but I never tried, only speculating. 

[magic]

Sounds like the 8 pin chip is dead but not knowing the pin-outs makes it difficult to test further.

It's a single layer board, come on. I can figure out most connection even from your photo.

[enut11]

@magic, quite right, it would be relatively easy to trace out the circuit.
However, I just found out that Noctua use a proprietary chip in their fans.

[Runco990]

Probably the NE-FD1 IC designed by Noctua.  But where to get one?

I'd longshot it and re-flow all the connections first.  Seen plenty of surface mount bad solder, even if not visibly apparent.

Other than that....   

[fzabkar]

https://noctua.at/en/ne_fd1_pwm_ic

ISTM that Noctua's SCD (Smooth Commutation Drive) tech could do with a lot of improvement. Hard drives have been using STMicroelectronics' SMOOTH 3-phase DC motor controllers for decades. This approach is a lot smarter.

[floobydust]

OP the PCB trace along the outside edge looks quite carved up, make sure it's not open circuit.
The IC looks JEITA, it's not an SO-8 with the straight leads and pics of the NE-FD1 are gull wing.
Any idea what the SC-70 part number is, what it's doing.

[fzabkar]

Pic of the NE-FD2:

https://www.relaxedtech.com/reviews/noctua/nf-a14-industrialppc/noctua_fd2_ic.jpg

[enut11]

Probably the NE-FD1 IC designed by Noctua.  But where to get one?

I'd longshot it and re-flow all the connections first.  Seen plenty of surface mount bad solder, even if not visibly apparent.

Other than that....   

Yes, it is an FD1 chip. Will try your suggestion on re-flow.

[enut11]

OP the PCB trace along the outside edge looks quite carved up, make sure it's not open circuit.
The IC looks JEITA, it's not an SO-8 with the straight leads and pics of the NE-FD1 are gull wing.
Any idea what the SC-70 part number is, what it's doing.

The outer track in OK, just my ham-fisted approach to removing the PCB assembly.

'SC-70'? not sure what you are asking

[floobydust]

The outer track in OK, just my ham-fisted approach to removing the PCB assembly.

'SC-70'? not sure what you are asking

see pic

[thm_w]

This fan was released in 2018 or 2019, warranty 6 years, surely it was still under warranty?

[fzabkar]

@magic, quite right, it would be relatively easy to trace out the circuit.
However, I just found out that Noctua use a proprietary chip in their fans.

Let me try.

I think that IC1 is a 2-phase brushless DC motor driver with on-chip Hall sensor, PWM control input and tacho output.

8/ Config / Mode #1
7/ Config / Mode #2
6/ Ground / V-
5/ V2 -- motor phase #2
4/ V0 / V+ -- positive supply to motor and PCB
3/ V1 -- motor phase #1
2/ S -- speed sensor output ??
1/ P -- PWM input ??

I suspect that R2, R3, R4 are configuration or mode select resistors or shorting links.

[enut11]

This fan was released in 2018 or 2019, warranty 6 years, surely it was still under warranty?

I am not the original owner and do not have the receipt. Also, cost of postage from Australia could exceed its value.

[enut11]

The outer track in OK, just my ham-fisted approach to removing the PCB assembly.

'SC-70'? not sure what you are asking

see pic

Transistor ID is RKM40

[floobydust]

I meant SC-59 for as a package guess, it seems smaller than SOT-23?
SMT marking code "RK" is a hall sensor like AK1751 but "M" is not a valid year code.
Or RK7002 but SOT-23.
Just looking at other parts that might have died.

[fzabkar]

Transistor ID is RKM40

What is the circuit reference? Is it a Q or an IC?

Does it look like this?

https://ae-pic-a1.aliexpress-media.com/kf/S25333058a3724cc9823f87c4c5a7255cc/20PCS-50PCS-100PCS-1000PCS-New-original-RK7002T116-RK7002-MOSFET-N-CH-60V-250MA-SST3-silk-screen.jpg

[enut11]

@magic, quite right, it would be relatively easy to trace out the circuit.
However, I just found out that Noctua use a proprietary chip in their fans.

Let me try.

I think that IC1 is a 2-phase brushless DC motor driver with on-chip Hall sensor, PWM control input and tacho output.

8/ Config / Mode #1
7/ Config / Mode #2
6/ Ground / V-
5/ V2 -- motor phase #2
4/ V0 / V+ -- positive supply to motor and PCB
3/ V1 -- motor phase #1
2/ S -- speed sensor output ??
1/ P -- PWM input ??

I suspect that R2, R3, R4 are configuration or mode select resistors or shorting links.

@fzabkar, looks about right to me.
There is 5.6R resistance between 4 (+IN) and 6,7 (Gnd).
Pins 6 and 7 are connected together.

I found some brushless fan internal circuits to get a feel for how they work. The two 8-pin controllers below have separate hall sensors.

[enut11]

Transistor ID is RKM40

What is the circuit reference? Is it a Q or an IC?

Does it look like this?

It's a Q. Similar to your URL.

[fzabkar]

It's a Q. Similar to your URL.

Then it's probably an RK7002 by Rohm:

https://avtronic.eu/medias/rk7002/rk7002.pdf

[fzabkar]

There is 5.6R resistance between 4 (+IN) and 6,7 (Gnd).

ISTM that if V1 and V2 are shorted to ground, then the IC has a shorted internal MOSFET. Is that the case?

[magic]

The discrete circuit seems to short the negative terminal of the supply bypass cap (C1) to ground after a time delay determined by R5,R6,C2. Is it supposed to limit inrush current?

At any rate, this part doesn't seem particularly critical and most likely the IC is blown. If R5 measures 3k3 then the supply can only be shorted by the chip or by C1. Easy to check by desoldering C1.

edit
"Shorted by the chip" could either be internal short between VCC and GND or shorted power transistors in series with motor windings, as suggested by fzabkar. Maybe 5.6Ω is simply winding resistance and the chip has about 0Ω from V1/V2 to GND...

[enut11]

Both R5 and R6 measure 1.65K in circuit.
It is a nuisance when manufactures use custom chips, especially in what would otherwise be a valuable and repairable item.

[thm_w]

This fan was released in 2018 or 2019, warranty 6 years, surely it was still under warranty?

I am not the original owner and do not have the receipt. Also, cost of postage from Australia could exceed its value.

They are not going to ask you to mail it back.
https://www.reddit.com/r/Noctua/comments/st4vb7/noctuas_amazing_warranty_services_secondhand/