I've had a suspision since I installed this thing that it wasn't working... and I'm pretty dang sure now it's not.

Two reasons:
First, I took the empeg apart, and held the exhaust from my butane soldering iron in the area of the DS1721 temp chip on the I2C board to the point where it definitely was hotter than the threshold I've got it set at... 25 degrees C. No fan movement.

Second, When power is applied to the empeg, the fan momentarily oscillates and then goes still. The center of the spindle gets warm to the touch the longer it's on, and there is resistance to movement when I try and rotate it by hand, more so than when the power is off... like it is fighting same charged poles on a magnet, if that makes sense.

I've metered the obvious points... the board is getting 12V (11V to be exact). I am at a loss as to what else to test, and am a bit peeved I paid for this board when I could just hook the fan straight to 5V and be done with it.

Is there ANYONE who can verify that their board works?? And any help on troubleshooting this is GREATLY appreciated!!!

EDIT: Well... hmm... I put it all back together, and when I plugged the empeg in to AC the fan spun up and is working fine. Color me puzzled!

Mmmm.. I wonder if maybe some stray bits at power-on cause the init sequence to fail once in a while?

When that happens, Hijack outputs "Fan control error; disabling" to the serial console. I suppose it really should do a retry or two before giving up completely, though. Mmmm...

I have used the fan controller board for many months now. The hijack settings are such that the fan should come on everytime and the fan used allows me to hear it working when it is on. There is definitely a problem in that (what seems to be randomly) the hijack "fan control error" will show up and the fan will not cut on. I have to reboot a few times until the fan runs. Mlord: Please check that "Random bits at power up" problem you are mentioning. maybe delay that reading or something. (heh: i couldn't program a phone number)

woops

Hrm... I've never looked at the console to see if it spit anything out about the fan controller... would it help if I could also prove that is occuring? Thanks for thinkin' about it Mark!

Yeah, if you could prove it's getting that message, then perhaps I'll try a change or two to the code and then have you re-test to see if it helps.

I have a fan controller here, but it is not currently installed in any of my players.

cheers

Mmm.. actually, the existing code seems to try to use a front-panel pop-up message to report failure, as well as the serial console log.

Fan Control Error (displayed for five seconds).

Updated Hijack with fan retries is now here.

Please FIRST verify that you can get the current Hijack to fail with console or front-panel messages as noted in this thread, and only THEN try the replacement. Report back here afterward.

Cheers

Ben,

I have been curious to ask you: Have you never had any problem with the fan starting? No hijack "Fan Controller Error" messages?

I've tried the 435 hijack mlord just posted and am getting about the same number of random fan starts and fan not starts, i haven't seen the "Fan Controller Error" message though. But i've only started empeg about 12 times in car and @ home. I believe i would have seen that message in 12 starts though.

woops

Mark:
I've never received an error message pop up, but I just plugged it up via serial and received this error:

i2c: Failed to receive ACK for data!
i2c_write8: device select failed

Then, upon several reboots, it printed out

i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error; disabling

and showed the fan control error popup for the first time (that I've noticed at least, which given it's frequency now seems to point to the fact that I've never noticed it =), so it seems that sometimes it doesn't get to the stage of printing out that error and triggering the pop up.

I'm on Hijack 413 on a MkIIa (no idea why I haven't updated in ages). I've rebooted 15 times at least and it has printed that message every time, sometimes with the "Fan control error" line, sometimes not.

AFTER upgrading to hijack435+, I get the following:


i2c: Failed to receive ACK for data!
i2c_write8: command issue failed
Fan control error, retrying
i2c: Failed to receive ACK for data!
i2c_write8: command issue failed
Fan control error, retrying
i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error, retrying
i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error, retrying
i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error; disabling


So now I can't get it to work at all, after about 20 reboots. I guess that one time I noticed it working was a fluke?

Ben:
No worries, I'm sure we'll figure it out.

Again, thanks to both of you for working on this!

I just came back from a trip to town. I restarted the car 4 times and the fan came on every time. yesterday the fan worked every time (twice). These are all occurances where the ignition was turned on (not just turn ignition to accessory as when i tested before). Needless to say i'm starting to be very happy with ver 435.

Loren,
If your fan has never worked, then i suggest the wiring is not correctly installed.

woops

Loren,

Try some tests in the car. Can you hear your fan working if everything is quite in the car? Also, set hijack so the fan will ALWAYS cut on so you can get a better read on when and when it is not working. I will be interested in knowing what is going on that your fan is not working...so i hope you will keep us posted (pun?)

woops

Unfortunately I can't tell if the fan is on in the car as the fan I am using is whisper quiet. Only way would be if I pulled the sled to be able to visually confirm it spinning. I have the temp setting at min 15 and max 18 so it should be on all the time.

Loren,

Just for test purposes at least try the RadioShack Part #273-240 40mm 1 9/16" brushless fan.7.7 CFM and enough noise that when music isn't playing you can hear it. I use it and it works well.

woops

Mmm.. Ben'll have to help out with this one. If Hijack cannot talk to it to set the thermostat, then the fan isn't going to spin no matter which model/brand of fan is used.

Mmm...

mlord,

can you increase the loop time by a third for receiving fan data bits? I mean the fan works most of the time now and every time using hijack reboot. i can't tell you how much angst this has eliminated.

woops

I don't think the retry loop is useful at all.

But in the meanwhile, I have just now looked at the original i2c code that clocks the bits out, and found something suspicious there that might have an effect. So I've changed it slightly, and the resulting Hijack kernel is here.

Again, try this one out folks, and if it solves your fan problems then I'll incorporate the changes into a proper release.

-ml

with Latest Hijack:

i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error, retrying
i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error, retrying
i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error, retrying
i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error, retrying
i2c: Failed to receive ACK for data!
i2c_write8: device select failed
Fan control error; disabling
player.cpp : 385:empeg-car 2.00 2003/04/01.
Prolux 4 empeg car - 2.1434 Mar 26 2003
Vcb: 0x4089d000


I also wanted to note that yesterday I did a bit of driving and pulled the empeg a few times and never got the fan control error pop up, however I still have no way of knowing if the fan is spinning in car. Is there any reason Car vs. AC would make a difference?

Well, in the car, the player has to visibly pause to output those error messages, because the baud rate is so much slower in car (unless you've up'd it to 115200 in config.ini).

That's about the only difference I can see.

If you're in the testing mood, I'll spin another kernel for you here in a minute or two.. this one SLOWS down the I2C protocol immensely, BECAUSE TALKING SLOW WORKS SOMETIMES FOR FOREIGNERS

UPDATE: Okay, try this.

-ml

Heheh, I guess you weren't talking slow enough:


player.cpp : 385:empeg-car 2.00 2003/04/01.
Prolux 4 empeg car - 2.1434 Mar 26 2003
Vcb: 0x4089d000
i2c: Failed to receive ACK for data!
i2c_write: address select failed
i2c: Failed to receive ACK for data!
i2c_write: address select failed


Notice the different error message however... Thanks for plugging on this.

YOU NEED TO MAKE IT TALK LOUDER, TOO!

Okay, I think this proves that Loren's fan controller is either dead, or disconnected.

Cheers

Just for the hell of it, I tried again ... and it seemingly worked, no error in the terminal and no pop up.

Rebooted... no error message again. Hmm...

Pulled it out of the dock and stuck straight AC so I can see the fan... No error message, but the fan's not spinning. Same thing, multiple times. The fans jiggles a bit on boot up but that's it. Any further troubleshooting ideas?

Has anyone got an alternate email for Ben? I contacted him about having boards and he said he still had some and then when I asked where to send money all of the emails have bounced back as undeliverable. I hand wired up this circuit utilizing a smd proto board and just get constant fan control error so I think maybe I need to stick with his PC board... but I need to get in touch with him