SOC Meter stopped working

[TickTock]

Coming up on the end of day three with the fix and not a single issue. Powers on perfectly every time - used to experience it almost daily. So if anyone is experiencing some flakiness, measure the voltage on VR1. Just need to open the back and put your meter across the lower (inside) pin on the two pin power connector in the upper right and the bottom, wide terminal on the 5V voltage regulator in the lower right. If you are not seeing a full 5V when the unit is powered on, that could be the problem - it appears to have been the problem with mine.

This post may be a little pre-mature, but I might have solved the temperature? sensitivity issue. Mine, too, recently would come up in a strange state with garbage on the display about half the time. Toggling power multiple times would usually get it going eventually. While investigating, I noticed that the output of the 5V regulator was only 4.5V and the input was only 6.5V. The datasheet says the input must be 7-12 so first I replaced the zener with a 3.3V zener which raised the input to >8V. However, this didn't solve the problem. I still was seeing 4.5V on the output and encountering the malfunction. So next, I soldered a 1.5K surface mount resistor on top of the 220 Ohm R49 to bring it's effective resistance to 192 which, by my math should raise the output 500mV. That seems to have worked. It's only been installed 24 hours but so far no sign of the funky poweron state and the output voltage is 5.0V. I am not sure the zener replacement was necessary but I mention it in case it was. I recommend anyone experiencing this issue measure the output of VR1 on the back of the AVR-CAN board and see if they are getting 5V. Datasheet says it should work at 4.5V, but my guess is some may be marginal there.

I didn't have the right size SM resistor but the capacitor next to it shares a connection so I was able to solder from the cap to the far 220 Ohm R49 terminal. If you are handy with a soldering iron and long nose pliers, you may even manage to make an axial resistor work. Disclaimer: Proceed at own risk. If you screw it up and short something that shouldn't be you could destroy your AVRCAN. Measure the output of the Vreg even if it looks like it is working to make sure you are not over-stressing the chip.

 

[garygid]

Very helpful, thanks.

I got my abdominal (abominal) drain and urine-drain tubes pulled out yesterday, so it will still be a while before I can get back into this stuff.

I had been thinking that the Power-on-Reset might be the problem, since the POR chip is in the schematic, but not actually on the board.

However, 4.5v is definately marginal, and I had not seen that.

The 5.1v Zener was to cut down the 12v to 14.5v DC that the LEAF provides.

Again, thanks for your help!

 

[TickTock]

Doesn't sound fun at all. Here's to a speedy recovery!

I, too, was considering the power-on reset. But when I attached a temporary pushbutton to the reset pin and tried manually resetting the part after power was applied I was still getting the weird behavior. I think the 5.1V Zener was a good choice - I don't think replacing with the 3.3V as I did was necessary. It just happened to be my first experiment and I was too lazy to put the 5.1V back in when it didn't solve the problem. According to the LM1117 datasheet, their resisor choices (220 & 660) should result in 5V (=1.25*(1+660/220)). So my fix of changing the 220 Ohm to 192 Ohms *should* have raised the votlage to 5.55V but I ended up with 5.05. I think Olimex may have gotten the bottom of the barrel on that regulator order.

Very helpful, thanks.

I got my abdominal (abominal) drain and urine-drain tubes pulled out yesterday, so it will still be a while before I can get back into this stuff.

I had been thinking that the Power-on-Reset might be the problem, since the POR chip is in the schematic, but not actually on the board.

However, 4.5v is definately marginal, and I had not seen that.

The 5.1v Zener was to cut down the 12v to 14.5v DC that the LEAF provides.

Again, thanks for your help!

 

[highcountryrider]

I got my abdominal (abominal) drain and urine-drain tubes pulled out yesterday, so it will still be a while before I can get back into this stuff.

Abdominal surgery sucks! I used lots of bags f frozen peas.

 

[garygid]

Any further word on the effectiveness of the LOW-VCC fix?

Is the VCC low only in cold temps?

The regulator itself gets warm quickly, so it might be a bad temperature coefficient of the reference resistors?

After the fix, in warm weather, is the VCC too high?

Any additional data will be helpful, Thanks!

--------
I will be checking all the new-builds that I send out.

Perhaps I can check the AVR-CAN boards in the kits also,
including loading them with a display board.
Or, maybe just provide instructions for the kit-builder.

However, we do not yet know what the right, all-temperature fix is.

Cheers, Gary

 

[TickTock]

Not sure if it's temperature related, but I did check the voltage again a few days ago and read 5.11. Got 5.05 when I first made the change so there is something causing minor fluctuations but nothing like 0.5 V. Still working great.

 

[garygid]

Maybe you can measure the two "reference divider" Resistances
(with the power off), at room temp and at some "cold" temp?

 

[garygid]

I fixed one person's Kit-build today where the Zener diode
was "blown" (open circuit), so the Meter got no power.

The shrink tubing was covering about 1/4 of the Zener,
so that might have made the Zener get too hot.

Maybe I can find 2-watt Zeners for future builds.

The same unit also has a bad DPDT switch, possibly due to overheating when soldering on the EV-CAN inputs (white wire and gray wire). We replaced the switch.

 

[garygid]

SOC-Meter NOTICE GG12-01:

In some SOC-Meters, the shrink tubing (on the Red wire) was placed
(or later moved to be) TOO CLOSE to the Zener Diode (which is
soldered to pin 9 of the 9-pin "D" connector.
The Zener needs good air circulation to cool sufficiently.

Reported Cases:
The three cases of Zener Diode failure (open circuit) were all caused
by the shrink tubing actually touching the glass case of the diode,
shrinking the tubing to be an even tighter fit, causing
OVER-heating and failure of the diode.

Symptom:
No lights in the display.

Cause:
No power to the AVR-CAN circuit board,
resulting in no power to the display.

------
If your SOC-Meter has failed in this manner:

1. Contact me (PM or email).

2. If a competent person is available to replace the Zener Diode,
I will send replacement parts (a 5-watt 4.7 volt to replace the 5.1v 1-watt).

3. Or, we will make other arrangements to have your SOC-Meter fixed.

------
If your SOC-Meter is still working:

1. If your SOC-Meter was an early build, or a kit, you might want
a competent person to remove the back (4 small screws) and
examine your SOC-Meter (just look) to see if the shrink-tubing
is too close to the Zener Diode.
2. If the shrink-tubing is too close, a competent person could
remove some of the tubing with a sharp eXacto-type knife.

3. Take care to NOT cross-thread the screws when replacing the back.

------
NOTE: The shrink-tubing (usually un-shrunk) should be
at least 1/4 inch from the glass end of the Zener Diode.

Note: An upgraded Zener Diode is available to any SOC-Meter owners (see next post).

 

[garygid]

As of last week, Kits and Built SOC-Meters will include
the new Zener Diode Upgrade:
a 4.7v 5-watt Zener (instead of the original 5.1v 1-watt).

Hopefully this Zener will work better:

1. providing a LITTLE bit more voltage (0.4v) for the on-board
regulator to use when the LEAF's 12v battery is a bit low.

2. being a LOT more tolerant to the heat that it generates.
The heat generated will be slightly less (about 8%), but the
heat-handling ability should be several hundred percent better.

-------
If you want one of these new Zeners AND you can get your
SOC-Meter competently Upgraded (soldering required),
I will send the parts if you email a request to:

plugin2 at mboxes dot com

including the following:
1. "SOC-Meter Zener Upgrade" in the Subject line,
2. Your Name, Mailing address, and phone number,
3. A request for a Filter if you want one (specify Blue or Red).

NOTE: The Shrink-Tubing should still be kept about 1/4 inch
away from the Zener diode.

 

[garygid]

CAUTION:
I general I would not recommend replacing the Zener diode in an SOC-Meter that is working.

However, if someone has sufficient skill, one might remove the lid and CAREFULLY look to see if the shrink tubing is near (or touching) the glass body of the Zener. The tubing should be about 1/4 inch away from the Zener's body.

IF the shrink tubing IS TOUCHING, there is a significant possibility that the Zener will overheat and "blow" (go open-circuit). The 3 that I have repaired ALL had the shrink tubing touching, or actually on, the glass body of the Zener.

With CARE, an eXacto-type knife could be used to remove the offending tubing, with no soldering necessary.

If replacement of the Zener is necessary (or sufficiently strongly desired), the original 1-watt Zener could be removed and the new 5-watt Zener installed. However, one needs to do this skillfully, and then "stuff" it all back in the box in such a way that the Zener body AND its bare leads do NOT TOUCH ANYTHING else (like another wire's insulation).

So, it might be best to remember:
"If it ain't broke, don't fix it!"

 

[garygid]

Apparently the little glass Zener diode can begin to fail in such
a way that the "2011 model" SOC-Meter (with that 1-watt Zener)
works sometimes, and not others, and ultimately fails to work at all.

Three more have failed in this manner. Replacing the Zener has "cured" the problem in most cases.

Two people had problems with the LED display "going crazy",
and they were able to get a new part from SparkFun.

Anybody with a pre-2012 SOC-Meter (who still has the older
"little glass Zener diode" in their Meter) should consider upgrading to the "new" much larger, ceramic, 5-watt Zener.

If you can do it safely yourself, or get somebody competent locally to help you, I will send you a "2012" Zener if you request it. Maybe several people locally can "upgrade" together at a workshop before or after a local LEAF-Gathering.

USPS shipping, with tracking, but not insured, is just under $6 (one way), I believe. UPS Ground (Insured) is between about $15 to $20, depending upon your location.

If your SOC-Meter is "2011" and having problems, please let me know.

Cheers, Gary