Classic controller DEFCON 4: FET Drive Error

[Dmitry]

Please explain the warning "DEFCON 4: FET Drive Error" that I found today.
I do not understand why this warning.
The classic controller 250ks used the least, no overloads not.
After this warning, it ceases to charge the battery after about 10 seconds the charge begins again.
In the description of Local App is said about the failure of the processor or the software, the more information I have not found.
Maybe it's simple overheating keys or failure of internal fans? What would you say Guru.

Yes, and more.
Is there somewhere a list of errors controller with minimal description?
Suddenly, there was an error "104: Unknown" , because of what is also unclear.

[Westbranch]

here are 2 other posts on Defcon 4

http://midniteforum.com/index.php?topic=1568.msg13412#msg13412

http://midniteforum.com/index.php?topic=1448.msg12049#msg12049

hth

[Dmitry]

Westbranch thank you.
I briefly read pracitces all the posts under The "Classic" charge controller. But mostly, questions or are solved simply, or hang with no solutions and no one really to say nothing.

In these positions answer as such, no. Only General considerations.
All I could find is a current overload.
But how can the current overload when classic 250ks holds 50A at 48v the battery, and I have solar panels at the peak 15A.
Over voltage can not be.

Appeared only new questions:
It is overheating, the refusal of the fans or they to this warning do not have a relationship?
If you encounter this error, it will reset or so and will hang?
How to reset this error if there is no direct access to the controller (I can't completely remove power and then turn on)?
What and where else can you look to understand the cause of this error?

[boB]

Defcon 4 is not a heat problem.

Defcon 4 may mean Over Current Protection (OCP) with your Classic.

If OCP, then it may be from the fast PV array MPPT sweeping.

If so, try this...  Go into the MODE menu and change the mode from
SOLAR to Legacy P&O instead.  This sweeps slower.

Also, we do not use the message "Defcon 4" any more so maybe your software
for the Local App status panel is older and needs to be updated ?

Older Local Application may also be the reason you see "error 104 unknown"...
newer software should know what the errors are and give you some meaningful
information.

Either way, look to see what software versions are in the Classic and MNGP Remote control
and also how old the Classic itself is, maybe.

First thing to try though is to set the mode to Legacy P&O and see if this helps
the errors and stopping of the charger.

[Dmitry]

boB thanks.
Today the problem has recurred. On your recommendations remotely changed the controller's operating mode to Legacy P&o, Nothing has changed. The controller is started, you receive the warning "DEFCON 4: FET Drive Error" after that the charge is terminated and a new error 33, a screenshot. And so again and again. Also tried mode Dynamic Solar Tracking, all the same, the same 33 error.
Drew attention to the fact that the time of occurrence of the problem persists. They now have an array of 8 units, 230 watts, 24 volt solar panels. Four parallel chains of two in each. The problem starts when the sunlight gradually reduces the angle of incidence of the rays. Assume that at the time the problem began sun's rays fall on the panel at an angle of less than 30 degrees and angle all meniayetsia at this point, the partial shading panels from the surrounding trees. Not all simultaneously and consistently moving shadow. Darkening partially, not fully, but enough panel lost power by half or more.
Classic brand new, bought 3 months ago. For simplicity, until all assembled on the test stand. So connection errors are excluded.
Current firmware version.
Application Version 0.3.60
Firmware:
- Classic Rev: 1779
- Network Rev: 1674
While altering can't, because there is no direct access to the controller.

The questions now.
According to the description which I have given you if you can receive this error message because of the blackout panels or small beam of incidence of sunlight?
Is there a need in the firmware the latest version of firmware?
If Yes, will it be possible to get any new data from the controller when the error occurred?

[boB]

Reason for Resting (RFR)  33  means OCP or Over Current Protection.

This could happen because something is wrong with the Classic itself but normally means
that the Classic has some kind of large load switching on on its battery side OR what can
exacerbate  that issue is when the battery lines are fairly long and shared with an inverter.
That is because the inverter can draw its load from the Classic rather than the battery.
So you might want to check and/or report that information.

The Classic made some energy so it's not "broken" per-se but do you see this error
more towards the end of the day and not during the bulk of the day ?  That would
be interesting.

[dgd]

Dmitry,

I hope that battery Ah rating is incorrect.
a 48v 10Ah battery will be a major problem if you have any inverter connected to it as well as the Classic.

dgd

[tecnodave]

Dimitry,

I think I see your problem,  p.v. Volts in is too close to battery volts. In your last post you said you had 4 parallel strings of 2 "24 volt" panels in each string. If these panels are 60 cell panels they would be about 30 volts each or 60 volts for the string. And in the last screen shot shows p.v. Input volts is 56 volts , that is with the panels unloaded!   You do not have enough volts to charge a 48 volt battery!
Generally minimum charge voltage will be something like maximum charge voltage of the batteries plus about 12 volts "headroom" for the controller to work with. Try wiring 4 panels in each series string. This will help with this and especially at low light conditions that you described in an earlier post.

I cannot comment on the error codes but I have had an issue with my Classic undercharging a 24 volt bank when solar panels were wires 2 in series "12 volt panels" for 34 volts total voltage on the string. I re wired to 4 panels in series (68 volt strings) and output increased considerably, expesially at fog, low light etc.

tecno

To all in the forum : if I am wrong about this shout out, this is a Classic 250 ks , designed for high voltage input!

td

[vtmaps]

I think I see your problem,  p.v. Volts in is too close to battery volts.
<snip>
Try wiring 4 panels in each series string. This will help with this and especially at low light conditions that you described in an earlier post.

Tecnodave is correct... you can't (under most conditions) charge a 48 volt battery with a Vmp of 60.  If you do have 60-cell panels, the sweet spot (for controller efficiency) is three in series for a 48 volt system.  In most climates, four panels in series will require one of the higher voltage Classics. 

Of course, if you have 8 panels, you cannot make three strings of three panels per string.  Can you add one more panel?

--vtMaps

[tecnodave]

Vtmaps,

I think Dimitri has a Classis 250 Ks.........no problems with 4 panel stacks at 120 volts.....efficiency would probably a bit better at three panels in series but the 4,in series was recommended as a quick solution.


td

[Dmitry]

Colleagues, thank you for clarification.
My answer is a large and complex, what would you have had the impression that I don't understand what you are doing. It seems to me that the occurrence of this error depends on the incorrect operation of the controller, firmware or partial damage of the electronics.
As I have said now, the controller runs on a test stand. The battery consists of 4 12 volt 32 amp hour batteries. Don't hit me, I understand that it is bad, but so long as there. Due to limited at the moment of the budget system is going sequentially. In the nearest plans purchase of batteries normal capacity. While the controller only works in order to understand in its work and in order to properly build up the system. No inverters to the battery is not connected. I took the simple static load of 200 watts (current consumption from battery power about 3.5 - 4 amps depending on the current battery voltage) and picked her up through relay leaving AUX 2. No other loads are not. The controller is almost always keeps the battery in FLOAT mode. I think the current 3-5 amps are issued by the controller may not cause such a problem. I understand you now think that I am here hammer head is unclear what, collect something is not clear how. Why bother to help me. I love to get a system for 3-5 thousand dollars, but understand I do not earn so much that could afford it immediately. Your product is a revolutionary and he has not much competition. The more you sell it only in the American market and officially will not deliver. I think that I am one of the first customers in Russia who decided to try your revolutionary product. And not bought on our market Xantrex XW-MPPT, OutBack FlexMax or often offer me sellers TriStar-MPPT.
Now for the bad.
Yesterday I had the opportunity to work with the controller.
I have tried to fulfill almost all your suggestions for alteration of the system.
First, it was updated controller software.
Firmware:
- Classic Rev: 1849
- Network Rev: 1839
Was made reset at boot keys, right left, as indicated in the instructions for the firmware off the outputs AUX1, AUX2.
After that the controller is running, but strangely behaved. Almost always in the RESTING mode: 5 featuring low input voltage, although the difference between the power was more recommended 12 volts. At the entrance 62-63 volt, battery about 50 volts. Randomly switched in charging mode. Assuming that there is not enough input voltage connection of panels was converted into two chains of 4 pieces. Eventually voltage input array panels was about 130 volts.

And so then the problems started.
For the purity of the experiment were off both comfort AUX1, AUX2, was also disabled all the load on the battery. Simple battery charge controller. Periodic off (disable) the controller, the battery discharge to the load.

After turning on the controller, it is after initialization is included in the work and depending on the voltage of the battery starts one or another stage charging. All as I first thought was normal. Goes statistics kilowatt, ampere hours. After a while, he suddenly begins to restart. The exact time can't say, but it's about 1 time in 30-50 minutes. All statistics kilowatt, ampere hours is set to zero. Then again starts charging.
I am attaching first screenshot with the data after spontaneous reboot the controller.
But this is only half the problem, the second I noticed when we decided to measure the current-voltage batteries. And here I saw that the controller shall issue to the battery voltage 5.9-6,0 more volts than it really shows. I thought I lost calibration device. Immediately checked settings compensation they were 0, I tried to change them, but they have a limit of only plus minus 2.5 volts. So to align the voltage did not succeed. Ultimately left to 0.
Colleagues, I normally versed in electronics and able to measure correctly any tension. I make a small difference in measurements of the order of 0.1-0.3 volts of precision instruments. So I went ahead to understand and found the following.
Fully abastecibe controller. Now if you turn on and connect the solar panels, he begins the boot process. After full loading it starts to give data about the voltage of the battery. Measure the control voltage tester almost identical, the controller shows 0.2 volts more than I measure. Mode of charge and after about 60 seconds I see that the control voltage of the batteries has increased by 6 volts.
I am attaching the second screenshot with the data after the tension rose.
The controller I see that charge voltage 57,2 Volta actually measure of control tester on batteries issued 63,1 volts.
In fact, the controller fry batteries and it is unclear for what.

Conducted another experiment.
Fully abastecibe controller. Now if you turn without connecting solar panels, it begins the boot process. After full loading it starts to give data about the voltage of the battery. Measure the control voltage tester also coincide, the controller shows 0.2 volts more than I measure. Leave the controller in this mode, after a while (already about 5-10 minutes), you see that the control voltage of the controller, sharp drops on the same 6 wills.
I am attaching the third screenshot with the data after a voltage drop.
The controller I see that the voltage of the batteries 45,5 Volta actually measure of control tester battery 51,4 volts.

To connect with this problem solar panel to the controller, I was afraid, for now he is connected to the battery.
What to do next I do not even know.
Help, can give access to a laptop with Midnite Application via TeamViewer.

[Halfcrazy]

If I am understanding you right the controller is reading the battery voltage to high when charging and to low when resting? Sounds like a bad connection somewhere. Forgive me if you said but I did not see it. Did you verify on the blue terminal block with a volt meter if the volt meter agrees with the classic say within a volt?

So try this test for me:
1- Classic on no solar panels on. Display on classic shows what voltage? (      ) now volt meter on the blue terminal block shows what voltage?(      )
2- Classic on and Solar panels on and charging. Display on classic shows what voltage? (      ) now volt meter on the blue terminal block shows what voltage?(      )
3- Also in Step2 does the Classic say Bulk MPPT or Absorb?

[Dmitry]

I don't quite understand what you want from me.
All measurements which I do, same if I measured at the terminals of the battery or on the blue terminal Board controller.
Then the problem is in the controller.

Halfcrazy please call us on skype login: gr.solar
I right now you can show them to you.
Set up which takes calls from any user.

Now the picture in Skype you can see the data controller and parallel data from the tester is connected to the blue block controller.
Unfortunately I have no tester that would show the measured voltage, the one I use is turned off after 10 minutes of non-use, the other at the moment I don't have.

[Dmitry]

If to say briefly:
During the charge controller charges the batteries with voltage of 6 V more than he shows.
If solar panels are disabled, the tension which shows the controller
exactly 6 volt lower than in reality batteries.
The connections and wires do with it.
I can get you to record video where all the measurements I do on the blue block controller.

[Halfcrazy]

Ok let me ask this:

Does the volt measurements on the blue block agree with the Classics display (Or pretty close to it)?