Classic controller DEFCON 4: FET Drive Error

[Dmitry]

Colleagues, boB, it took quite a long time, I had the opportunity to come and to work with the controller, time was short, but I recorded a new video that you must see.

Clarify some points that are reflected in this video:
1. Finally I have bought 4 helium battery 100 ampere-hour 12-volt.
Currently, the total Bank is 100 ampere hour to 48 volts.
I think you have had doubts when I recorded the past video controller is connected to the small-capacity batteries.
Current capacity, I believe it is quite sufficient for objective assessment of the performance of the controller.
2. To avoid incorrect removal shown was used another digital voltmeter. In the very beginning of the video I dedicate the time of calibration voltmeters. Forgive note, the difference in readings between voltmeters is 0.2 volts. That speaks about the reliability of their testimony with regard to errors of measurements of each.
3. Like last video was a test charge the batteries. No load connected. Only solar panels, the controller, batteries, and voltmeters.
4. The weather at the time of the experience was cloudy with frequent change of illumination from maximum to almost zero.
5. Due to the complexity of the experiment, I had to ask to make a video of my friend, in the process video I gave him instructions on what to show. Unfortunately not ready to comment in English. As it turned out shooting on the phone is not very good, there are times loss of field and low-light. In the video I tried to photograph all that you might be interested in, look completely in choosing quality 720.
6. The controller is connected to the battery copper wire 6 square millimeters, 1.5 meters long.


Now the feature that has appeared during the experiment.
Unfortunately, this time, I failed to capture the moment when the controller starts not display the correct voltage. With small-capacity batteries that moment it was possible to catch fast enough. With the new batteries I wait for a moment longer hours and could not catch. As a result, the controller was left on, coming in a few hours I noticed the difference in 6 volts. Immediately after this was recorded video.
Note.
Here is a slightly different result. Because the batteries are high enough, the controller fails to turn up the voltage up to 60 volts, as it was at the last video. Instead, the indicator displays at 6 volts less than the current battery voltage.

At the moment I see no need to paint every minute that I did on the video. I think it will be clear. If necessary, ask questions, pointing out the exact time, I will answer.
Video.
http://youtu.be/4ITmfO89_TQ

[Dmitry]

Colleagues, boB,
On the same day at evening, accidentally recorded an error "DEFCON 4: FET Drive Error".
Like last video was a test charge the batteries. No load connected. Only solar panels, the controller, batteries, and voltmeters.
Video.
http://youtu.be/_B1Q4JO0HjI

[Resthome]

Colleagues, boB,
On the same day at evening, accidentally recorded an error "DEFCON 4: FET Drive Error".
Like last video was a test charge the batteries. No load connected. Only solar panels, the controller, batteries, and voltmeters.
Video.
http://youtu.be/_B1Q4JO0HjI

I have no idea what this reason for resting is said but here is the info for RFR 31 shown in your video.

ReasonForResting = 31   AD1CH.IbattMinus > 900   (peak negative battery current)

boB will have to answer this one.

Also looks like when on the RFR screen the BAT voltage in lower right corner with out the decimal is reading 548-552 or near the same as the meter. But the Status screen is displaying the incorrect voltage. Another one for boB to answer.

[zoneblue]

It is unusal for the battery voltage reading to be that far off. Do you have anything set in the voltage offset? You could try using that to correct it. There is also a potentiometer on the control board , which bob can advise on.

[Resthome]

It is unusal for the battery voltage reading to be that far off. Do you have anything set in the voltage offset? You could try using that to correct it. There is also a potentiometer on the control board , which bob can advise on.

Zoneblue. The video showed the offset at 0. The interesting one was the BAT reading that shows up on the RFR secret display showed the BAT voltage without the decimal and it is close to the reading on the voltmeter, it's the reading on the Status display that is way off. Would like to hear what boB thinks about that.

[boB]

Dmitry, thank you for this video.   The battery voltage is certainly reading wrong as this plainly shows.

Did you say that the voltage on the Classic ~sometimes~  reads correctly and matches your multi-meter ?

If so, it may be a bad connection inside the Classic on that top circuit board.  If it is ~always" 6 volts low,
the adjusting that top trim-pot variable resistor may fix the reading and make it correct.

If the Classic MNGP/Remote meter reading is intermittent and only sometimes wrong and sometimes
correct, then please try gently pushing on the top PCB to see if it might be a bad connection that makes
the Classic come back to correct readings.  You may want to set the Classic's mode to off when trying that
to make sure it does not affect the switching control of the power section during connections moving around.

[Dmitry]

Colleagues, boB,
As usual we will wait for the result, when I have the opportunity to come and work with the controller.

I have a great suspicion that the problem in the firmware or something boards are responsible for controlling tension and here's why.

In this post I draw an issue that is very important, as it can lead to the decommissioning of batteries because of their overcharging.
I actually wrote that the stresses are the same. This condition is observed only after downloading the controller up some time.

Clarification.
After adding new batteries, I found the difference of 1.0-1.4 volts between the readings of the controller and voltmeter in different modes. When you charge the controller tries to charge the batteries 1.0-1.4 volts more than it really shows on the panel and worth it in the settings. Mode ABSORB stand settings 57.2 volts, the controller gave 58.2 volts, in FLOAT mode settings are 54.4 volts, the controller gave 55.8 volts. All settings of the controller I showed in the video.
Just now, to focus attention on that did not. These 0.8-1.4 volts, you can try to adjust settings. Video did not even write. We will write.


After a while something happens, it starts correctly display voltage, only a reboot helps. I have repeatedly reset the controller at boot. When you tried to work off a solar panel, change the mode of charge, nothing helps. Also at your request, pushed charge in the area of the terminal strips and processor, also any results. On the top Board are all perfectly built, covered with a varnish, there is nothing Gorely, when the controller all elements warm.
I can try to parse the controller, but then write what and where exactly it show.

And the question of why the site is not protected old firmware on the device?
At least you could see the controller's operation with different firmware versions.
Now I have a laptop accidentally left earlier already have a desire to return back to the old firmware.

[boB]

Dmitry, I do not quite understand what is happening here but the firmware should not have this effect unless somehow
the Classic did not get programmed correctly ?

The reason you cannot immediately return to the older firmware version is because there is not enough
memory inside the Classic to store more than one version of firmware.  I wish it could do that.

I have every old version of firmware available though so any one of them you would like, just let me know
and I will point you to it...  (or anyone else that wants them)

I have a feeling that with the intermittent problems you are seeing, you have a bad control board but please
try pressing it lightly to make sure there is not some obvious bad connection where anytime you move or
jiggle the Classic, it changes its voltage reading.

I saw in your video the debugging screen on the display where the lower right number was showing the
correct voltage so something is working OK even though the main battery voltage is reading wrong.

boB

[Dmitry]

Thank you.
Colleagues, what are the considerations about my second video "DEFCON 4: FET Drive Error"?
Anyone able to explain in detail the error RFR 31?

[boB]

Thank you.
Colleagues, what are the considerations about my second video "DEFCON 4: FET Drive Error"?
Anyone able to explain in detail the error RFR 31?

Dmitry,   RFR 31  means that the Classic thinks that it sees more than 90  amps.

I would guess that with your other problem, (battery voltage0, the RFR 31  message may mean that your hardware has a problem or that the firmware update did not quite take correctly.

I think your idea of trying a previous or older version of firmware is a very good idea.

Let's see what happens when you try that.

[Dmitry]

Colleagues, boB,
Today we have an opportunity to work with the controller.
Simultaneously recorded some new video for you to be able to see some lost moments in my early videos.
All video was written by me, so there are moments when something is not clear or renewed orientation.

It was decided to make the most of the mechanical and measuring work with the controller. I hope that in the controller somewhere bad contact.
After verification, take a decision about the firmware of the controller in the early firmware.

In brackets shall specify the time for the video, which is reflected described the moment.
We hope the work done knowingly, perhaps my videos somebody will help.


1. The first video.
http://youtu.be/qH6js9XDrpU

Was Sunny weather, solar power batteries was enough, was recorded moment of raising voltage on the battery up to 61.6 Volta [00:08]that displays in this moment showed 55.8 volts [0:36]in the service information readings varied from 62.2 to 62.4 [0:40].
Note that the current mode “FLOAT”.
Make measurements second voltmeter [1:51], the application on a laptop also shows incorrect values [2:34].
Try to put pressure on the elements controller card no positive result [2:37].
Disconnect the cable panel [3:44], re-connect the panel [4:00]that no positive result.
Piratical panel in another port [4:54], a positive result not.
Do control measurement on batteries [5:50], voltage 61.6 volts.
Make the control current measurement [6:12], a current of 1 ampere.
Do control measurement voltage solar cells [7:03], the voltage in the range 130-137 volts.


2. The second video.
http://youtu.be/mJ_pusJkUVM

Previous state.
Try to disable the solar panel.
Note after turning off the solar panels voltage proportional decreases, the difference in 6 volt remains.

3. Third video.
http://youtu.be/N0baPVd0EKA


Took off the top management card. Visible damage on the power Board no, on the control card no injuries. All connectors good.

4. Fourth video.
http://youtu.be/yofD9ngITIw

Installing the control back. All the connectors are fully seated.

5. The fifth video.
http://youtu.be/pqerXnYHWsk

The controller is started in work with the initial discharge.
Please note a few points.
After running on batteries 54.3 Volta [1:00pm], scoreboards at this moment showed the correct voltage value 54.5 volts.
In the service information testimony I was somewhat surprised, there was an increase of voltage from 34 to 41 [1:07]. It appears I do not understand.
Next were connecting solar cells [2:32],
Pay attention to what voltage is charging mode “ABSORB” [2:53], the panel 58.3 Volta, voltmeter shows 57.8 Volta, in the service information readings from 58.2 to 58.6 [2:56].
In the settings voltage 57.2 Volta [5:55]. Correction voltage not.
Why classic hard fed to charge the batteries on 1.1 volts more unclear.

6. The sixth video.
http://youtu.be/erm2Lra482Q


The transition from the regime of “ABSORB” to “FLOAT”.
The controller is switched mode battery mode “FLOAT” [0:29], the voltage drops on the panel 55.7 Volta, voltmeter shows 55.3 Volta [0:38], in the service information readings 55.6 [1:02].
In the settings voltage 54.4 Volta [1:41]. Correction voltage not.
Classic also eats the batteries charge voltage of 1.2 volts more than worth it in the settings.


Today the work is complete, the controller in the work.
I will watch it tomorrow during the day.
It's unclear why in the modes “ABSORB” and “FLOAT” controller hard eats the batteries charge voltage at 1.1 - 1.2 volts more than set in the settings.

[dgd]

Interesting how the voltages differ.
Somethings munted, likely culprit seems to be the MNGP unless the local app is agreeing with the MNGP. Then its the Classic.

dgd

[Dmitry]

Colleagues, boB, all of you who follow my theme.

My hopes for bad connections inside the controller is not justified.
Today morning changeable weather, the morning was clear to dinner clouds appeared. The controller has worked properly for about three hours. While it was clear enough. In the end, as usual went to incorrect display of voltage. Catch the moment it is quite difficult, I noticed what happens when a small solar panels power the controller intensively seeking the maximum power point. I usually have fixed it in the end of the day. Today, clouds appeared fell sharply power solar panels. Ultimately disappointing result. The mode in which was the controller before was “FLOAT MPPT”.

Now it only remains for me to try to flash the controller older firmware and test its performance.

I expect all constructive proposals, what would you do another.

[boB]

I have another idea, Dmitry.

It may be that the wiper of the battery voltage adjustment trimpot (variable resistor) is not making a
good connection or has a bad solder joint.   This is the small square variable resistor
just below the bottom yellow inductor and just above the lower horizontal 12 pin connector.
There are 3 of these 12 pin connectors on the control board.  One of then is on the lower left
and is vertical and the horizontal one just to the right of the vertically oriented connector
is the one I am talking about here.  Just above that horizontal connector and below the
yellow inductor is that trimp-pot, VR1.

Try this....   While watching the voltage on the MNGP display, push on this trim-pot VR1
and see if the voltage reading goes correct and bad again when you stop pushing on it.
Maybe it is not soldered on one of its bottom connections ?

If that does not do anything positive, remember where it is set to now and with a
small screwdriver, adjust it one way and the other.  The battery voltage should
go up and down some.   This is one thing that could cause this behavior also.

boB