Fan Fault

Hi all,

A couple of weeks ago (before I could actually see console output from my gear) I bought some lab equipment of which I got a 2950.

The switch was particularly noisy and the fan was obviously dying so I decided to find a replacement.

The one inside the switch was a SUNON KDE1204PKVX MS.M.B400.

Initially I wanted to get the same model in replacement but I just could not find it anywhere - I tried Digikey, Mouser Electronics, and Jameco. I found nothing.

In the end I figured that said fan was replaced by a new line introduced by SUNON so I went for the KDE1204PKVX MS.A.GN which is the same form-factor, and shares the same electrical operating characteristics as the other one (i.e. wattage, voltage rating etc).

The only difference is that the one I replaced had three wires, a +ve, -ve andYELLOW wire which judging by the DATASHEET (http://www.alliedelec.com/Images/Products/Datasheets/BM/SUNON_INC/997-0026.PDF) is for the frequency generation waveform. Now, I am not an electrical engineer, but if I am correct, this must be similar -if not the same as- Pulse Width Modulation (PWM) and must be a feature of the old fan to be controlled by the main board for controlling the fan speed in relation to internal temperature of the switch, yes/no?

Anyway, the only difference with the fan I replaced the old one with is that is does not have the third YELLOW wire with this feature.

Getting to the point now, when I switch on the 2950 I get the following console output:

--- System Configuration Dialog ---
 
 Would you like to enter the initial configuration dialog? [yes/no]: 
 
 [COLOR=Orange]00:00:13: %ENVIRONMENT-2-FAN_FAULT: System Fault: FAN FAULT [/COLOR]
 [COLOR=Orange]is detected.[/COLOR]
 00:00:14: %SPANTREE-5-EXTENDED_SYSID: Extended SysId enabled for type vlan
 00:00:18: %SYS-5-RESTART: System restarted --
 Cisco Internetwork Operating System Software 
 IOS (tm) C2950 Software (C2950-I6Q4L2-M), Version 12.1(14)EA1a, RELEASE SOFTWARE (fc1)
 Copyright (c) 1986-2003 by cisco Systems, Inc.
 Compiled Tue 02-Sep-03 03:33 by antonino

What I am worried about is the first message about the fan fault. Despite the fan actually operating as normal, I am just worried that this is something I should definitely be worried about?

Software isn't intuitive (we're not smart enough to create AI are we) and I am thinking that this could be due to the software on the switch not detecting the third wire or something similar, and is thus generating the error message, despite the fan operating normally.

What do you think?

Thanks.

#######
BTW - one of the reasons that I went for this fan as a replacement to the older one was because during my research for a replacement after learning that suppliers did not stock the older model was that I kept seeing information like this: http://www.thermalfx.com/fans/40/KDE1204PKVX-MS-AF-GN.html

As you can see, it is stated there that this newer model also goes by the same name of the older one. Interestingly, the image of the fan there (which is the newer one I replaced the older one with) is photographed with having a YELLOW wire, but mine does not. Maybe I should just go ahead and solder one on...

Find more posts tagged with

Save $250 on 2025 certification boot camps from Infosec!

Book now with code EOY2025

Button

Comments

chmorin

Well, your knoweldge on this is probably greater than mine, but if the fan is working and that is the only error message you get, I wouldn't worry about it. Putting a non-native part into a cisco switch having this result should not be surprising, especially if it is not fully connected.

But a fan is a fan is a fans fan. If it is blowing, I see no issues.

Netwurk

If this is an always-on fan, you can safely ignore the error message. The third wire probably monitored the fan. To me, if you rig the fan to add a third wire it may not send the correct signal.

PhildoBaggins

If the fan is running you are fine. If the unit isn't hot then your golden.

Its probably just the MB isnt detecting fan speed so it assumes the fan is dead.

qwerty_face

Cheers all, I'm just gunna go ahead and declare the issue solved.

Netwurk wrote: »

To me, if you rig the fan to add a third wire it may not send the correct signal.

I tried that in the end, I soldered the wire on to the third solder point but the 2950 still gave the error message. Perhaps the internal circuitry of the replacement SUNON lacks a particular IC on its *very small* circuit board. On second thoughts, if I was bothered enough I think I can see a small hack there!

PhildoBaggins wrote: »

Its probably just the MB isn't detecting fan speed so it assumes the fan is dead.

Dead on, I think that's it really.

chmorin wrote: »

But a fan is a fan is a fans fan. If it is blowing, I see no issues.

Ditto. My lab has cost me quite a bit so I wanted to make sure nobody had any anecdotes that would have otherwise saved me a lot of headache. I'm not sure how hot these pieces of equipment can get.

Anyways... It blows.

QF

tiersten

The third wire is to sense the speed of the fan and will cause the errors you're seeing if it doesn't detect it or it isn't within the defined range. You'd have to supply a square wave signal of the appropriate frequency range to fake it but as it doesn't actually cause the 2950 to power down, I wouldn't bother. The chassis based switches tend to force a shutdown if they detect a fault with the fans due to the possibility of damage due to overheating.

The datasheet is showing examples of the output of this extra wire. It isn't describing PWM.

The fan doesn't actually contain the hall effect sensor chip that is required to actually sense the rotation speed. That is why just adding the extra wire did nothing. There is the possibility that it does have one but the way the sensor is configured means that the output isn't correct i.e. it is counting more or less per revolution.

If you're buying a replacement fan, there are several criteria you need to check. You need to check that the voltage it requires is the same, the current consumption is the same or less, the CFM rating not into free air is the same or better, it mounted in the same way, same size, whether it needs a PWM input and that the sense wire if present has the correct number of pulses per revolution. For a basic replacement like in this 2950, the minimum would be voltage, current and CFM as it runs without the sense connection.