Bafang Display stopped working

[Cyclomania]

So this time I have an issue with a display on a bike I have.

I have stored this bike, with the display, in a storage room up til now. Now, I often check battery-charge through this bike when I am unsure of how much level is still in a particular battery, since the Bafang displays gives very exact percentage charge. But all of a sudden now when I was going to see how much charge was left in a 48 volt battery I have, the display has stopped working, on this bike.

Now I have another exact same display so I plugged that in and everything works fine when that one takes over. So it has to be the display that is the problem. Also the button-picker connected to the display also works when plugged into the new display. So it is only the display part.

Now this display is a bit worn but how can this happen all of a sudden? It has been working up until now. So this is kind of strange.

Any idea what could be wrong with this old display? Should I open it up somehow and check inside? I think it is strange that this happened without me going out biking with this bike even. It was just a sudden shut-down.

Also when I check the pins in the male and female part where it connects to the rest of the bike I don't find anything out of the ordinary .

 

[fechter]

The display is glued together and doesn't come apart nice, but it can be done if you're careful.

No idea why it would fail like that, but most failures of things like this in general are caused by bad solder joints. Could also be a broken wire in the cable.

Since it's dead, you don't have much to lose by attempting to take it apart.

 

[Cyclomania]

The display is glued together and doesn't come apart nice, but it can be done if you're careful.

No idea why it would fail like that, but most failures of things like this in general are caused by bad solder joints. Could also be a broken wire in the cable.

Since it's dead, you don't have much to lose by attempting to take it apart.

Yeah precisely. I will open it. Perhaps screws in the back I guess.

 

[Cyclomania]

I have opened it. But it looks kind of like a circuit card or something. Not quite sure how to get into the actual glass display since there are no screws there. Perhaps bending out the glass or something?

 

[amberwolf]

The black case halves are probably either ultrasonically welded or glued together. Sometimes cooling or heating it can help separate them, there have been threads about taking apart and repairing various displays that may help you with this one if you poke around for them. (especially in the open-source firmware threads, since some people also modify them while flashing them).

One thing that has been found in some displays is the internal regulator that converts battery voltage to 5v/etc can just fail, probably from voltage spikes during turn on, or being spec'd to run right at the edge of it's capabilities (either voltage at input or current at output; it's very common for manufacturers to cheap out like this).

Sometimes it fails because of a downstream failure, wiring damage shorting something out, or another component (especially the external lighting control) putting too high a load on it.

But the most common problem with anything is wiring faults--usually broken connections, poor solder joints, or insulation failures allowing shorts--these usually happen anywhere within a cable or connector but more often where wires exit a connector shell or electronics casing, and sometimes on the PCBs.

 

[Cyclomania]

One thing that has been found in some displays is the internal regulator that converts battery voltage to 5v/etc can just fail, probably from voltage spikes during turn on, or being spec'd to run right at the edge of it's capabilities (either voltage at input or current at output; it's very common for manufacturers to cheap out like this).

Hmm I will try to find this thing. I will try to open it today. Perhaps I will destroy something. But what do I have to loose.

 

[Cyclomania]

Looks like this inside. What am I looking at?

 

[Cyclomania]

Looks like this now.

Any idea what might be wrong or what part is the internal regulator?

https://www.veed.io/view/44db4973-5864-4c3d-b8b2-60030d4a0095?panel=share

 

[amberwolf]

I don't specifically, but there are some tests you can do.

With it connected to the controller and battery, and the battery turned on, use the display power button to turn the system on. If the system does turn on and allow motor operation (even though the LCD is blank), it means the display is working except for the LCD. That might mean it's just the backlight on the display--if so you can put a bright light up against the LCD face and faintly see the stuff it would normally show--just no backlight. But at least it tells you the display is getting battery positive and negative, and is also sending battery positive back to the controller.


With it connected to the controller and battery, and the battery turned on, you can connect the black lead of your multimeter to battery negative, and then use the red lead on a few points to see if voltage is getting to them correctly, to narrow down the fault. Set the meter to 200VDC.


First, at the top edge with all the wires, one of those will be the battery voltage input--I can't read the labels on most of them (teh wires are in the way), but one is likely labelled B+ or something similar. Connecting the red lead to that solder pad, you should get battery voltage at that point. If you don't, it probably means something is wrong in the cable or connector, either with battery negative or positive.

Next, being careful not to short anything with it (touching only one pin at a time), connect the red lead to each of the pins on each part with a red arrow below (except the little round battery, which you will only connect to the top of with the red lead). Note down the voltage at each pin. Then try turning it on (even if the display doesn't light up), and retest all the voltages and note them down.



I don't know for certain, but suspect the transistor on the top left is the pass-thru for "keyswitch" voltage to the controller (turned on only when the display is turned on). The multipin "transistor" near the lower right is probably the main regulator controller. The tiny battery probably holds settings, and runs the real time clock if the display has one. (on some devices with a battery for this, the battery being dead can cause the device to malfunction).

 

[Cyclomania]

I know the bike works when the display is switched to another one so it has got to be something inside the display I am thinking.

I don't specifically, but there are some tests you can do.

With it connected to the controller and battery, and the battery turned on, you can connect the black lead of your multimeter to battery negative, and then use the red lead on a few points to see if voltage is getting to them correctly, to narrow down the fault. Set the meter to 200VDC.

You mean to have the negative probe on the battery and the other positive probe in one of the areas of the circuit card, right? Or how do you mean? And this should be done when the battery is turned on connected to the bike.

First, at the top edge with all the wires, one of those will be the battery voltage input--I can't read the labels on most of them (teh wires are in the way), but one is likely labelled B+ or something similar. Connecting the red lead to that solder pad, you should get battery voltage at that point. If you don't, it probably means something is wrong in the cable or connector, either with battery negative or positive.

Next, being careful not to short anything with it (touching only one pin at a time), connect the red lead to each of the pins on each part with a red arrow below (except the little round battery, which you will only connect to the top of with the red lead). Note down the voltage at each pin. Then try turning it on (even if the display doesn't light up), and retest all the voltages and note them down.

You mean all those areas where you did put arrows, correct? To have the red probe there and the black on the actual battery(?)

I don't know for certain, but suspect the transistor on the top left is the pass-thru for "keyswitch" voltage to the controller (turned on only when the display is turned on). The multipin "transistor" near the lower right is probably the main regulator controller. The tiny battery probably holds settings, and runs the real time clock if the display has one. (on some devices with a battery for this, the battery being dead can cause the device to malfunction).

I put the red end on all of those three parts individually right? And the black one on the actual battery in the other end of the bike?

 

[amberwolf]

I know the bike works when the display is switched to another one so it has got to be something inside the display I am thinking.

It does not have to be inside the actual display casing itself. It can be anything from the connector where the display plugs into the controller to something inside the display. That is why I suggested the various tests.

You mean to have the negative probe on the battery and the other positive probe in one of the areas of the circuit card, right? Or how do you mean? And this should be done when the battery is turned on connected to the bike.

Correct.

You mean all those areas where you did put arrows, correct? To have the red probe there and the black on the actual battery(?)

Correct--the arrows don't point to the actual points to touch with the red probe, just the components to test. The points on the board with the soldered-down leads of the parts (or the leads themselves) are the points to actually touch with the red probe.

I put the red end on all of those three parts individually right? And the black one on the actual battery in the other end of the bike?

Correct, as described above.

If you prefer you can connect the black multimeter lead to any wire that is definitely known to be a working battery negative / ground wire, but the best place is the actual battery negative, so you can be certain it is connected there.

 

[Cyclomania]

It does not have to be inside the actual display casing itself. It can be anything from the connector where the display plugs into the controller to something inside the display. That is why I suggested the various tests.



If you prefer you can connect the black multimeter lead to any wire that is definitely known to be a working battery negative / ground wire, but the best place is the actual battery negative, so you can be certain it is connected there.

Perhaps these pics and vid can help you understand better what I am looking at before the testing begins?

Also, should I look for some particular voltage or reading or something on the multimeter on the different areas of the card? Or should they all display the same thing? Like "5 voltage" or something like that?

https://www.veed.io/view/674b32e1-cead-4e8e-8d96-e3b6312f1ab9?panel=share

 

[fechter]

If you connect the battery, you should see pack voltage on the orange wire (measured against battery negative). When the display turns on, the brown wire gets connected to the orange wire to feed power to the controller. You could try (carefully) shorting the orange wire to the brown wire and see if it lights up.

The board looks clean. I don't see any bad looking solder joints or burned spots.

You can also remove power and measure continuity between the connector pins and the wires on the board to make sure none of the wires are broken inside the cable. (do this first)

 

[Cyclomania]

ts.

You can also remove power and measure continuity between the connector pins and the wires on the board to make sure none of the wires are broken inside the cable. (do this first)

Ah, the connector pins where the display connects to the rest of the system and the connector pins where the display connects to the on and off buttons? While looking at continuity between these on one side and the circuit board wires on the other correct?

 

[Cyclomania]

Ok now I have done the test. I did not quite understand how to read the continuity. But found a lot of continuity between these 15 or so wires, on the board, and the connection pins coming out of the display cable. In the way I believe you suggested although I perhaps misunderstood it.

When I read the voltages as Amberwolf said I got the same voltage, as the battery is giving to the overall system right now(48,5 volt), for the brown wire on the left side below. And the red wire on the right side in the picture below.

Those two got 48,5 volts. But the other cables seems to be completely dead.

Any idea what his means?

 

[fechter]

Seems like the cabling is OK, but can't be sure. Each pin on the connector goes to a wire on the board. Measuring continuity checks to see that none of the wires are broken inside the cable. It's hard to put the meter probe on the connector pins.

You could try attaching the battery and connecting the brown and orange wires with a screwdriver tip and see if anything lights up. This feeds power to the controller.

At this point it seems like something on the board failed but it would be hard to locate and replace the bad part.

 

[Cyclomania]

Seems like the cabling is OK, but can't be sure. Each pin on the connector goes to a wire on the board. Measuring continuity checks to see that none of the wires are broken inside the cable. It's hard to put the meter probe on the connector pins.

You could try attaching the battery and connecting the brown and orange wires with a screwdriver tip and see if anything lights up. This feeds power to the controller.

At this point it seems like something on the board failed but it would be hard to locate and replace the bad part.

You mean to have everything connected right? Battery up and going. And after that to connect the two wires in the pictures below to see if there is a spark?

How should I do this to not put current in myself? Two wooden matches perhaps and then something metal in between and then touch these two?

 

[Cyclomania]

You could try attaching the battery and connecting the brown and orange wires with a screwdriver tip and see if anything lights up. This feeds power to the controller.

You mean lights up as in a spark or short cut, right? Not as in the display screen starting and get up and going?

 

[fechter]

I guess making that connection won't power up the display, but the motor might actually run if you hit the throttle.