Another Rebooting Problem

[MobileMe]

I just got a new board last week running the latest firmware(3.02.16c). The board seems to be rebooting randomly about every 6 to 8 hours. There is two temperature sensors and a light sensor on that board. I setup the board to send me an email every time the Plc starts over, so when I upload new code or power cycle the board it sends me an email.

I've eliminated a sensor problem by putting the temp and light sensors on another board running firmware 3.02.16b. The board ran fine without any reboots.

There are two variables that I think might be the problem. The first is its location. I mounted the board in my basement on a rafter by the furnace. It does get warm, but definitely not warmer than 90 degrees at any given point. The second is the power supply. It's a 9v 650mA power supply. The other board is running a 9v 1A power supply and it seems to be running perfectly. Both boards are running on the same power strip and are only separated by 6ft or so.

I do plan on swapping the power supplies and moving the board to the same location as the other board. My question is would these two things cause the board to reboot like that? Knowing what happened with firmware version 3.02.16a, I'm gun shy about the latest version 3.02.16c.

Thoughts?

 

[pittom]

Ya try power supply,it might not have enough amps when you add sensors that is what happened to me

 

[CAI_Support]

Like Pittom recommended, try to get a more powerfrul power supply. If your power supply labeled 9V 650mA, that is the max power it can provide.  For normal running, you want the power supply max rating twice higher than required current or more. Dallas Semiconductor published a testing result back few years ago, in that report, most power cubes, even from Panasonic and Sony are not able to provide enough voltage at the labled current.
 
3.02.16/a firmware had compiler problem, that is why we actively recalled those two version firmware. Any version after that does not have that problem.  Both 3.02.16b and 3.02.16c are stable release.
 
Another test would be swapping power supply between those boards, it will probably show you the weak power supply cause the baord rebooting.  Keep us posted on you swapping test result.

 

[MobileMe]

I kind of figured it was the power supply. Unfortunately, I have everything disassembled so I can remount the boards in a better location. I have another power supply on order. Thanks for the second option. Since I'm putting my furnace on one of the boards, I wanted to make sure.

 

[CAI_Support]

Sure, it is always better safe than sorry.
When control furnace, please make sure your PLC code properly handle the power loss situation.  After power restored, it should check to make sure everything is okay before turn on the electricty or oil burner.

 

[MobileMe]

Sure, it is always better safe than sorry.
When control furnace, please make sure your PLC code properly handle the power loss situation.  After power restored, it should check to make sure everything is okay before turn on the electricty or oil burner.

I'm not sure what you mean. How can the board check if everything is okay before it turns the furnace on? The furnace runs on natural gas and is five years old. It basically runs on a contact sensor, the furnace does the rest.

 

[CAI_Support]

Before sending any command to operate gas valves, please always make sure temperature range is okay. Make sure the logic will remain those valves staying in  shutoff state if power lost.

 

[MobileMe]

The furnace takes care of that so the only thing I have to worry about is setting the temperature.  Thanks.

 

[MobileMe]

I changed out the power supply to a 1A and the board is still rebooting every few hours or so. Where do I go from here?

 

[CAI_Support]

Can you measure the 5V regulator output on board, is that stable 5V or varies?  If you don't have anyway to measure it, you can send the board to us, we will check it out for you.  Please make sure to send to the address on our web site:
 
CAI Networks, Inc.
4790 Irvine Blvd, Suite 105-203
Irvine, CA 92620
 

 

[MobileMe]

I did some more testing and found something very interesting.  I have two temp sensors on the board.  When only one is plugged into the board, it works fine.  Only when both are used at the same time, the board reboots randomly.  To make this more confussing, I have another board that has another temp sensor like the other two.  When I plug either of the temp sensors into this board, it works fine.  Just when I plug both into the same board, it has a problem.  I don't understand why, but at least I know how to fix it. 

 

[pittom]

it must be the board

 

[CAI_Support]

If you could please use a volt meter to measure the 5V power, that may provide clue for what was going on. We noticed the on board LM2940S 5V regulator can not take any over current. If by any chance the 5V touched ground, or anywhere causing over current, LM2940S will burn out and need to be replaced.  When using voltmeter measure it, the voltage will be below 5V and up and down.
 
If that 5V is not stable, then the only choice is to replace it.  We do test that voltage for each board during QA. If your board is new and having problem, we will replace that 5V regulator for you. For boards shipped out for sometimes, it is hard to tell how the regulator burnning out.

 

[MobileMe]

I did finally figure out that when I was hooking up the temp sensor, I used some wire out of a cat 5 cable. When I ran the wire, I used staples to fasten the wire. Why, I was either lazy or stupid. The vcc was shorting out on the staple. It was intermittent so it was very hard to find.

Would that have damaged the board?

 

[CAI_Support]

Normally it would not hurt anything else but only need to replace LM2940S.  That is not hard to replace.  Another choice would be using Ross's board with 7805 to provide power to WebControl. 7805 can handle larger current than onboard LM2940S, because additional heatsink can be added on Ross' board.  Also, 7805 internal voltage drop is 2V compared to LM2940S' LDO 1V, that 2V internal voltage drop can limit short circuit current, too.