EBM Papst computer fan problem

[najrao]

I picked up two of these large axial blowers, hoping to run them with speed control. Attach 1.
EBM Papst do not offer enough data for this custom product on their website. It shows a Sun microsystems SN.
The label itself says 'speed control VT',  necessarily to mean  the 'violet' lead.
The EBM general catalogue suggests different speed control modes, including PWM as input. See attach 2.
The violet lead shows 6.0k to black in both polarities, and 0.2V or less with 20V on the red power lead. So, not an open collector, nor base.
The fan does not run without the speed input, and draws no current.
Suck-and-see with some PWM? What frequency, amplitude?
Any user experience to share?

Forum refuses attach 1, even resized.

[MAS3]

Was attachment 1 a photo shot with your phone or some other device adding GPS location ?
If so, convert it to *.PNG, which removes such privacy sensitive info from the picture's EXIF data (because it doesn't support EXIF).

Attachment 2 shows that there is no difference between the potmeter-wiring and the PWM wiring, it does show that the PWM signal or potmeter level is filtered to a non PWM signal of which the voltage is controlled by the PWM or potmeter and based on an internal 5 volt level.
I read that picture (which could be clearer) as R1 and R2 are internal together with the capacitor, the potmeter or the transistor are external so to be provided by you.
Based on your description, the fanspeed will increase while the input gets pulled down from the 5 volt level, so the stronger it will be pulled down the faster the fan will run.
If you do not have a PWM controller at the moment, put a potmeter between the purple wire ("input") and GND, and if you don't have a potmeter either, put a resistor in that place.
The lower the resistor, the higher the fanspeed.
I'd start with a 4K7 resistor or a 2K7, and would expect half fan speed.
Because of the internal resistor / capacitor, the PWM frequency doens't really matter, 1 - 10 KHz is mentioned, so you should be fine if you keep it within that range.
If you happen to have a 4 wire interface (you didn't mention that) the frequency should be 25 KHz.

[najrao]

Thanks, MAS3. There is a storm raging this PM and a power shutdown. I will try tomorrow.

[MAS3]

Can't read all of the sticker in your photo, but it seems to be a type ZP.
That type is not on page 263 of that datasheet, types A and P are.
But worse: the photo shows a completely different version from that in the datasheet, so now i'm wondering whether this thing is in any way similar to that page you've shown.
Can you post a clear photo of the sticker, so that all info from it can be read ?

[najrao]

The violet lead shows: 6k to ground, linear to 2.5mA, and NO voltage on it with red lead powered upto 20V. In this same condition, grounding the violet lead with 5k to zero external resistance does NOTHING. Indeed applying a positive voltage up to 5V, ditto.
We cannot assume that the diagrams and explanation in the (screenshot download) 'general' catalogue will apply. This custom variant may have altogether different control specs given by Sun Micro.
It is also entirely possible that units I have are indeed shot and kaput. They came from a not reputable source.
Opening the motor requires a very small circlip remover, and even doing this is not likely to expose the control components on the PCB (stator hides).
Will try to upload pics again.

[MAS3]

So, it is a 4 wire version.
Copyright printed on it is 1998-2002.
Unlikely to find anything worthwhile from today's EBMPabst site about some 20 year old fan.
However, there´s also 19.06 printed on the sticker, which could mean production date week 6 of 2019, week 19 of 2006 or June 19 in some unknown year, your guess is as good as mine.
Because of the 4th. wire, i´d investigate the Intel 4 wire specifications including the speed signal divided by 2 ( this may help you on your way (Click !))

[Andrew LB]

That fan connector is made for hot swapping fans in servers. I believe this is the pinout.

[najrao]

I managed to get to the circuit board inside the motor. Attach1, subject to forum not rejecting my photo.
Further disassembly is invasive and destructive too. The 'stator' core pack  hides the board under it, and has been fitted into position on a hydraulic press.
Probing where I can, I see that the main +feed is interrupted by a blown SMD fuse, which presumably should kill all circuit functions. Probably about 10A, can't see.
Temporarily feeding +15V to the load side of the blown fuse draws 20mA or so, and, more importantly, sets up the 5V reg as detected on the violet control wire. I dare not increase the input further, the motor is still disassembled, no rotor. This tells me that simply replacing the fuse (somehow, in situ!) gives me a sporting chance of repairing the blower.
So I wired up the second unit to a controlled and limited supply. There are no mysteries at all: starts up at 16V and goes to highest speed directly as limited by the voltage. At the rated 24V, it draws the full 4.5A, and, boy, does it kick up a racket! The control wire needs just a pot to ground to pull the speed down, and works to very low speed. No doubt PWM at any frequency will work too, not tried. So far so good.
'Hot plugging' feature just disconnects or connects the power ground through a loop within the motor, so two wires, blue and black.
My thanks to all who responded.

Edit, 4hrs later: Well, what do you think! Alas, I couldn't get an SMD fuse, but had a PTC self-resetting job rated 5A, which could fit snugly. I did not remove the blown SMD, but shunted it by the PTC which had just the right lead spacing to solder on. No prize offered for the right guess, but it is working perfectly now!