I have in my posession two LX204-24-USB units. They both exhibit the EXACT same problem... and it's the *same* problem as in this thread:
http://www.lcdforums.com/forums/viewtop ... light=gpo3
Bringing the total number of units to 3 that are defective in this manner.
When anything is plugged into GPO3, it bounces the units. It looks like a brownout condition, which leads me to believe there is a short in the GPO3 traces somewhere. I spent a good 6 hours tonight debugging what exactly the problem was, and I've come to two conclusions:
1 - GPO3 is faulty from the factory on a lot of units. There probably isn't a lot of reports because not as many people use GPO3 as they might 1 and 2.
2 - The entire unit is *NOT* grounded properly. I found this out accidentally while trying to figure out why I had this problem. If the unit is not grounded explicitly, it will not function properly. If it's mounted in a bay, it's grounding through the case, and not through the USB or Floppy ground. This is a REALLY REALLY bad thing. I suspect EVERY unit suffers from this flaw.
As I said, I have two LK204 units, one direct from Matrix Orbital and one from a 3rd party vendor, so they aren't even from the same lot. They both suffer exactly the same problems under exactly the same conditions. Failure scenarios can be duplicated on both units reliably and repeatedly.
RMAing one or both units, I do not believe, is going to solve the probelm. So I'm wondering if anyone has any suggestions on what to do? MO seems to be a great company, and the functionality of the unit is what I'm looking for (well, almost, I wish it had a 2000mA+ rating, 1000mA is too limiting) - so I really want to keep the MO unit in the machine, but it's far too unreliable right now.
For the record, I have hooked up to it:
3x Panaflo 120x120x38's (2 in parallel, 1 single)
2x Generic 120x120x25's (Both single)
2x D1W probes
Any combination of the above fans works fine, so long as nothing is plugged into GPO3. GPO's 1,2 and 4 seem to be working just fine.
GPO3 & unrelated grounding problem
-
Paradigm
- Matrix Orbital
- Posts: 255
- Joined: Thu Sep 13, 2001 6:00 pm
- Location: Calgary, Alberta, Canada
We've been hearing more and more reports of certain fans triggering a reset on the MX series, especially with low PWM values. We didn't discover this in our initial testing because it seems that only some fans will do this. However after going out and buying around 40 different types of fans we found a few (2 so far) that cause problems. And yes GPO3 seems to be a sensitive one.
We have addressed the issue last week with a new PCB rev (Rev 1.5 on the 2x20's and a Rev 1.1 on the 4x20's). The new PCBs should be coming in in a couple of weeks and they will be introduced into the main production line shortly there after.
A couple fo things you can try are:
- Try different fans. As I've said so far we've found very few fans that have exhibited this problem.
- Keep your PWM value up. The problem only seems to come up with very low (sub 128) PWM values.
As for grounding issues: As the one that laid out the PCB I can assure you that it is hevily grounded through both the USB cable and through the floppy connector. The only reason it wouldn't be is if the cables had their ground wires severed.
We have addressed the issue last week with a new PCB rev (Rev 1.5 on the 2x20's and a Rev 1.1 on the 4x20's). The new PCBs should be coming in in a couple of weeks and they will be introduced into the main production line shortly there after.
A couple fo things you can try are:
- Try different fans. As I've said so far we've found very few fans that have exhibited this problem.
- Keep your PWM value up. The problem only seems to come up with very low (sub 128) PWM values.
As for grounding issues: As the one that laid out the PCB I can assure you that it is hevily grounded through both the USB cable and through the floppy connector. The only reason it wouldn't be is if the cables had their ground wires severed.
James McTavish, P.Eng
Director of Engineering
Matrix Orbital
Director of Engineering
Matrix Orbital
Sorry it took me so long to respond to this. I had tested out your suggestions, but then promptly forgot to post the results.
I've basically given up on GPO3. If it's connected, I have all sorts of problems. Doesn't matter if it's connected to a generic .25a fan or my bad boy Panaflo .45a fans, either in series or singly. Behavior is identical on both units.
Both units are v1.1 revision, and thus the problem has not been addressed.
As far as grounding goes, I assure you there are huge grounding issues with both units. I find it hard to believe I would get two units from two separate sources (one of them being directly from MO) that have the exact same problems if it's not a design flaw. While I do suppose it's possible, the changes of it happening are fairly remote, as I suspect the manufacture date is different as well. If it is not a design flaw and it turns out to be a manufacturing defect, I guess it doesn't much matter, as the result is the same. The units MUST be grounded to the chasis to function properly.
As far as PWM values, I don't have any direct control over those, as I am using LCDC, and there doesn't seem to be a setting to adjust PWM. I realize it can be adjusted through the power % output, but I have no idea what those values might be, other than a percentage value.
Anyway - what prompted me to return here and remember I failed to report back was further issues with the unit(s). I have not swapped up the unit to see if they both behave in this way, but the unit I have in the machine will switch off randomly after a bit of time (which is random). I can get the screen back by stopping and starting the display in LCDC, but it makes the unit pretty useless for continuous unattended visual monitoring of information.
I really like the concept of the MO fan controller, and the idea is great on paper, but I'm really unhappy with the reliability of these products. They just plain don't function as advertised. If you guys want to send me another unit with the "updated" PCB (although both of my units have v1.1 on them), I'll be happy to try it out. However, make sure you guys send the right screen color - black background, green text.
I've basically given up on GPO3. If it's connected, I have all sorts of problems. Doesn't matter if it's connected to a generic .25a fan or my bad boy Panaflo .45a fans, either in series or singly. Behavior is identical on both units.
Both units are v1.1 revision, and thus the problem has not been addressed.
You do have more than one GPO not working - you have the whole unit failing in interesting manners.There is absolutely no grounding issues with our units. If there was a grounding issue, there will definitely be more than the GPO not working.
As far as grounding goes, I assure you there are huge grounding issues with both units. I find it hard to believe I would get two units from two separate sources (one of them being directly from MO) that have the exact same problems if it's not a design flaw. While I do suppose it's possible, the changes of it happening are fairly remote, as I suspect the manufacture date is different as well. If it is not a design flaw and it turns out to be a manufacturing defect, I guess it doesn't much matter, as the result is the same. The units MUST be grounded to the chasis to function properly.
As far as PWM values, I don't have any direct control over those, as I am using LCDC, and there doesn't seem to be a setting to adjust PWM. I realize it can be adjusted through the power % output, but I have no idea what those values might be, other than a percentage value.
Anyway - what prompted me to return here and remember I failed to report back was further issues with the unit(s). I have not swapped up the unit to see if they both behave in this way, but the unit I have in the machine will switch off randomly after a bit of time (which is random). I can get the screen back by stopping and starting the display in LCDC, but it makes the unit pretty useless for continuous unattended visual monitoring of information.
I really like the concept of the MO fan controller, and the idea is great on paper, but I'm really unhappy with the reliability of these products. They just plain don't function as advertised. If you guys want to send me another unit with the "updated" PCB (although both of my units have v1.1 on them), I'll be happy to try it out. However, make sure you guys send the right screen color - black background, green text.
I had problems with lockups too. Very random, like ones in about 8 ours.
This problem started when I attached 4 temp probes.
Fixed it by using another USB port
Now I have another problem, as off yesterday I got spontaneously key clicks. Once in about 10 minutes or so I hear the key click sound, sometimes assigned buttons and sometimes buttons that do not even exist.
Very irritating, sound level that
This problem started when I attached 4 temp probes.
Fixed it by using another USB port
Now I have another problem, as off yesterday I got spontaneously key clicks. Once in about 10 minutes or so I hear the key click sound, sometimes assigned buttons and sometimes buttons that do not even exist.
Very irritating, sound level that
To state for the record, we do not have any grounding issues. There will be four external ground wires that are incorporated into the displays ground plain.
1. The ground wire of the USB cable.
2. The shielding on the external USB cable is also grounded and incorporated into the ground plain.
3. Two ground wires from the unmodified floppy cable connecting straight to our display, which is also connects to the displays ground plain.
We found that certain fans would cause the pic microcontroller to reset, sometimes only on a particular GPO. As a resut we have placed more filtering around the pic microcontroller itself on the newer revision (204 R1.2/202 R1.5).
Please email me at tlam@matrixorbital.ca for some options on resolving your display issues. Our products are very reliable, and I would like to prove this to you.
Best Regards,
1. The ground wire of the USB cable.
2. The shielding on the external USB cable is also grounded and incorporated into the ground plain.
3. Two ground wires from the unmodified floppy cable connecting straight to our display, which is also connects to the displays ground plain.
We found that certain fans would cause the pic microcontroller to reset, sometimes only on a particular GPO. As a resut we have placed more filtering around the pic microcontroller itself on the newer revision (204 R1.2/202 R1.5).
Please email me at tlam@matrixorbital.ca for some options on resolving your display issues. Our products are very reliable, and I would like to prove this to you.
Best Regards,
I apologize for the post above, but R1.1 is the newest rev we have for the LK204-24-USB .We found that certain fans would cause the pic microcontroller to reset, sometimes only on a particular GPO. As a resut we have placed more filtering around the pic microcontroller itself on the newer revision (204 R1.2/202 R1.5).
Can you email me at tlam@matrixorbital.ca to further discuss this.
Best Regards,