KS16s shows only 63% charge

[fryman]

App is not showing the accurate battery percentage or voltage, 63% and 60V. The wheel shows all but one green led is lit up and the ewheels charger indicates 67V. 

Went for a short ride after this pic and then plugged charger in and showed proper voltage at 60v and slowly started climbing and then charger shut off 25 min later showing 66.9 voltage. 

 

[mrelwood]

What does the official KingSong app show for voltage? Perhaps the issue is with the 3rd party app.

[fryman]

5 hours ago, mrelwood said:

What does the official KingSong app show for voltage? Perhaps the issue is with the 3rd party app.

It shows the same.

[fryman]

Did another test. Lifted the wheel a couple times , then I plugged charger back in and it read 62v then the charger turned off at 63v. Tried KS charger and light is green.

[mrelwood]

Unfortunately it does have the symptoms of a dead battery cell. Is the wheel still under warranty?

Do you have a multimeter? First thing to do by yourself would be to measure the charger output voltage.

[fryman]

Well, I took a ride yesterday and decreased the battery below 10%. Today I was able to charge the battery to 100%. Have no clue what’s going on. 

[mrelwood]

That does sound hopeful though! Hope it stays that way.

[Lutalo]

On 8/5/2018 at 12:23 PM, fryman said:

Well, I took a ride yesterday and decreased the battery below 10%. Today I was able to charge the battery to 100%. Have no clue what’s going on. 

The very same issue happened with my ks18. I found that the fault was with the charger brick. I had an extra brick so I plugged it in and everything was good. @esaj suggested that I could use a multimeter to measure voltage once I opened the faulty brick.

He mentioned three screws inside the brick regulate voltage flow and they sometimes lose tension. He suggested using a screwdriver to adjust the screws inside the brick and then retesting voltage with the multimeter until it read correctly.

Haven't yet gotten around to trying the repair because I discovered that I didn't have a multimeter laying about the house (they are inexpensive great tools; I should have one), I bought an 18L (different charger), and my extra 67 volt charger is charging my previous wheel just fine. 

[esaj]

On 8/7/2018 at 12:36 PM, Lutalo said:

The very same issue happened with my ks18. I found that the fault was with the charger brick. I had an extra brick so I plugged it in and everything was good. @esaj suggested that I could use a multimeter to measure voltage once I opened the faulty brick.

Just for checking the voltage, it's not necessary to open the brick, you can just plug the probes into the charge socket coming from the charger. If the voltage's off, then you need to open it (or get a new charger, if you think it's faulty / don't dare to go poking inside, the basic bricks aren't that expensive).

 

Quote

He mentioned three screws inside the brick regulate voltage flow and they sometimes lose tension. He suggested using a screwdriver to adjust the screws inside the brick and then retesting voltage with the multimeter until it read correctly.

Usually there's three "trimmer potentiometers" or similar inside the chargers, they (mostly) look something like this:

There's three to control three things:

- Output voltage

- Output current limit

- Threshold where the led changes from red to green

If you have wrong output voltage and want to try adjusting it, first disconnect the brick from mains and the wheel, then open it up. Be careful, the mains-side capacitors may hold a large voltage, don't poke your hands in there, just use a screwdriver to turn the trimmers (also be careful not to short anything with the metallic screw driver). To adjust the voltage, you then need to connect it to mains. At this point, the thing's really dangerous, there are live mains voltages there (110V / 220-240V RMS AC, depending where you live, and then depending how things are done, bypass caps after rectification might even contain full peak voltages from the mains?).

Monitor the output voltage with the multimeter while you try turning each of the trimmers. say, half or one full turn. If the output voltage doesn't change, turn the trimmer to back where it was, it's not controlling the voltage. If the voltage measured from the output changes, turn the trimmer until you reach the full voltage for the battery (67.2V for 16S, 84V for 20S). Once you get it set right, unplug it from the mains. If you have hot glue gun or some silicone/glue/whatever, you can put a dab on the screw to prevent it from getting turned by itself (it shouldn't, but it could if the charger gets knocked around or such). Then just close the charger up (still being wary of not touching the components or traces by hands), and you should be good to go.

 

 

 

[Lutalo]

34 minutes ago, esaj said:

Just for checking the voltage, it's not necessary to open the brick, you can just plug the probes into the charge socket coming from the charger. If the voltage's off, then you need to open it (or get a new charger, if you think it's faulty / don't dare to go poking inside, the basic bricks aren't that expensive).

 

Usually there's three "trimmer potentiometers" or similar inside the chargers, they (mostly) look something like this:

There's three to control three things:

- Output voltage

- Output current limit

- Threshold where the led changes from red to green

If you have wrong output voltage and want to try adjusting it, first disconnect the brick from mains and the wheel, then open it up. Be careful, the mains-side capacitors may hold a large voltage, don't poke your hands in there, just use a screwdriver to turn the trimmers (also be careful not to short anything with the metallic screw driver). To adjust the voltage, you then need to connect it to mains. At this point, the thing's really dangerous, there are live mains voltages there (110V / 220-240V RMS AC, depending where you live, and then depending how things are done, bypass caps after rectification might even contain full peak voltages from the mains?).

Monitor the output voltage with the multimeter while you try turning each of the trimmers. say, half or one full turn. If the output voltage doesn't change, turn the trimmer to back where it was, it's not controlling the voltage. If the voltage measured from the output changes, turn the trimmer until you reach the full voltage for the battery (67.2V for 16S, 84V for 20S). Once you get it set right, unplug it from the mains. If you have hot glue gun or some silicone/glue/whatever, you can put a dab on the screw to prevent it from getting turned by itself (it shouldn't, but it could if the charger gets knocked around or such). Then just close the charger up (still being wary of not touching the components or traces by hands), and you should be good to go.

 

 

 

Thanks for the clarification. I am sure that a lot of folks may find it useful. Just paid $67 (USD) for an additional brick for the 18L; not cheap. So, if a DIY is doable, its worth checking out. However, not worth getting shocked, so thanks for helping the electrical adventurers among us to stay frosty and safe ☺?

[esaj]

3 minutes ago, Lutalo said:

Thanks for the clarification. I am sure that a lot of folks may find it useful. Just paid $67 (USD) for an additional brick for the 18L; not cheap. So, if a DIY is doable, its worth checking out. However, not worth getting shocked, so thanks for helping the electrical adventurers among us to stay frosty and safe ☺?

Yeah, at cheapest you could probably find the "bricks" for something like 20USD (best guess), although the quality may be anything (good or bad). "Even" at $67, it's not a bad price to pay if you're not confident to go poking around the charger by yourself, they do contain high enough voltages (especially since you need to connect it to mains to measure the voltage) to kill you (technically, any voltage can potentially kill you, it's the current that kills, but the higher the voltage, the easier it is to get a lethal shock).

[Lutalo]

1 minute ago, esaj said:

it's not a bad price to pay if you're not confident to go poking around the charger by yourself, they do contain high enough voltages (especially since you need to connect it to mains to measure the voltage) to kill you.

$67÷1(The Threat of Death) = $67 ??

[TomOnWheels]

2 hours ago, esaj said:

Just for checking the voltage, it's not necessary to open the brick, you can just plug the probes into the charge socket coming from the charger. If the voltage's off, then you need to open it (or get a new charger, if you think it's faulty / don't dare to go poking inside, the basic bricks aren't that expensive).

 

Usually there's three "trimmer potentiometers" or similar inside the chargers, they (mostly) look something like this:

There's three to control three things:

- Output voltage

- Output current limit

- Threshold where the led changes from red to green

If you have wrong output voltage and want to try adjusting it, first disconnect the brick from mains and the wheel, then open it up. Be careful, the mains-side capacitors may hold a large voltage, don't poke your hands in there, just use a screwdriver to turn the trimmers (also be careful not to short anything with the metallic screw driver). To adjust the voltage, you then need to connect it to mains. At this point, the thing's really dangerous, there are live mains voltages there (110V / 220-240V RMS AC, depending where you live, and then depending how things are done, bypass caps after rectification might even contain full peak voltages from the mains?).

Monitor the output voltage with the multimeter while you try turning each of the trimmers. say, half or one full turn. If the output voltage doesn't change, turn the trimmer to back where it was, it's not controlling the voltage. If the voltage measured from the output changes, turn the trimmer until you reach the full voltage for the battery (67.2V for 16S, 84V for 20S). Once you get it set right, unplug it from the mains. If you have hot glue gun or some silicone/glue/whatever, you can put a dab on the screw to prevent it from getting turned by itself (it shouldn't, but it could if the charger gets knocked around or such). Then just close the charger up (still being wary of not touching the components or traces by hands), and you should be good to go.

 

 

 

@esajas always, best comment, with details, and right from technical point of view !

[esaj]

Oh, about the charge level shown by app / wheel, I think it was @DaveThomasPilot who pointed out that many wheels use a "not-so-good" set up for measuring the charge; there's usually a voltage divider for dropping the full battery voltage to something tolerable by the MCU (typically a 3.3V device, so you need a big divider to bring up to 67/84V+ down to tolerable levels, that is <= 3.3V). There's usually a voltage divider with large valued resistors to bring the voltage down enough for it not to fry the MCU. The resistors have large values so that they don't pull much current and don't have to dissipate a lot of power (if you think about it, dropping something like 60V over a resistor at 10mA would already be 0.6W, a typical 1206 SMD resistor can take 0.25W max).  And then there's a capacitor in parallel with the lower resistor to filter out any spikes.

What Dave pointed out was that the resistors are (often/usually) "too large". Consider this:

 

Voltage source at 84V, a divider made of 750K (what Firewheel used) and 30k (I don't remember the lower value, the Firewheel was 16S = 67.2V wheel anyway), bringing the 84V down to

V_out = R2 / (R1 + R2) * V_in

In this case

V_out = 30k / (750k + 30k) * 84V = 3,230769... V

I didn't calculate the power dissipations here, the 750k alone would already limit the current to  around 0.1mA.

So, still below 3.3V and the MCU should be able to use the ADC's to read that just fine. The 100n C1 is there to filter out any spikes and keep things steady, and should be there (it could be smaller, like 10n or 1n too). But... the C1's got this "evil" parameter called... Equivalent Parallel Resistance (EPR, although usually there's only ESR, Equivalent Series Resistance, listed for the capacitors). What that means is that the capacitor has something like a resistor parallel to it, the value is usually big (like megaohms). What that means, is that there's leakage in the capacitor, it lets some current run through it in otherwise "steady" conditions (even if there's just a DC-voltage there, like "0Hz signal").

Normally, it's not of concern, the current is very, very small, nothing that could really deplete your batteries (unless you leave your wheel sitting for several years or decades or centuriis...). Again comes the but... there's lots of resistance in the circuit before the capacitor (R1). What then happens is that the parallel resistance in the capacitor is in parallel with R2. Consider:

Sooo... we have a 10 megaohm resistor in parallel with the R2 & and the capacitor. Doesn't matter much? How's the output voltage from the divider?

V_out = (10*10^6 + 30k) / (750k + 30k + 10*10^6) * 84V = 3,221...V

Not that big of a change really, but actually it's 3..4% (my TI-89's batteries just died, so I just looked at the simulation values ). Imagine that the leakage ("EPR") of capacitor is larger than that, and you have a situation where the wheel "thinks" (actually, measures, and it's "right", the thing is just that the EPR of the cap will cause the voltage to look smaller than it actually is) the battery is more depleted than it is, due to leakage current through the cap causing the voltage to drop more than expected.

In Dave's case, I guess the wheel was "thinking" that the battery was overcharged because of the voltage divider didn't drop enough voltage through the capacitor leakage? Basically the voltage reading from the battery was "too high". Could also be a wrongly calibrated firmware (I'd expect them to have some form of calibration in firmware for the voltage).

But if the leakage is high, a wheel saying it's "only 67% charged" or whatever, could also be caused by a using "too large" voltage divider and a cap which happens to have "higher than usual" leakage in the battery measurement circuit, downplaying the actual voltage. If the leakage is high enough, the difference between reality & measurement can be substantial (imagine for example up to 10% difference... if the wheel measures around 67V at full, the actual voltage could be closer to 10% higher = around 73V!). Luckily, the BMS is there (and should be correctly adjusted) to prevent overcharging from happening the first place. Although, similar thing could potentially happen with the BMS also (especially since you want to use high resistance in the BMS to minimize the battery depletion by the BMS by itself! It's powered directly from the battery and always draining it, albeit with a very, very small current).

[fryman]

On 8/7/2018 at 4:36 AM, Lutalo said:

The very same issue happened with my ks18. I found that the fault was with the charger brick. I had an extra brick so I plugged it in and everything was good. @esaj suggested that I could use a multimeter to measure voltage once I opened the faulty brick.

He mentioned three screws inside the brick regulate voltage flow and they sometimes lose tension. He suggested using a screwdriver to adjust the screws inside the brick and then retesting voltage with the multimeter until it read correctly.

Haven't yet gotten around to trying the repair because I discovered that I didn't have a multimeter laying about the house (they are inexpensive great tools; I should have one), I bought an 18L (different charger), and my extra 67 volt charger is charging my previous wheel just fine. 

I have two chargers.  ewheels quick charger and the KS charger. The quick charger would shut off at the 60V and i would plug in the Ks charger and the light was green.

[Lutalo]

41 minutes ago, fryman said:

I have two chargers.  ewheels quick charger and the KS charger. The quick charger would shut off at the 60V and i would plug in the Ks charger and the light was green.

Oh. I see. Ok then, well whatever @esaj is saying; 'daswutahm thinkin' ???

[esaj]

The above is just an example... sometimes it's the circuits acting up, often times (it seems) it's the "cheapo" charge brick having the wrong voltage. No way to be sure without precise measurements. If the wheel's still under warranty, the surest way is to replace faulty parts, starting from the charger. If the wheel's acting "funnily" otherwise than charging, first make sure that it's otherwise ok, if there are riders in your vicinity, you could ask someone with more experience to try to ride the wheel to see it acts "like it should", then try to pinpoint any trouble (usually it's then either the mainboard or the batteries) to get the correct replacement parts straightaway,

Not everyone knows electronics very deeply (not that I do either), and it shouldn't be required by the customers to start replacing the parts by themselves, but these things are still (relatively) new, only some years in the market, and if you bought from overseas, it may be required to do some troubleshooting and repairs youself (unfortunately). Local shop helps there if you can bring in the wheel for troubleshooting & repair.

Keep on wheelin'

[esaj]

I'm not sure what to make of it, people upvote my posts (like the one about the capacitor leakage up a couple of posts), yet I notice (sometimes serious) factual errors in it reading it back again, although I do correct them once I notice them... Do you trust me too much?  

[litewave]

29 minutes ago, esaj said:

I'm not sure what to make of it, people upvote my posts (like the one about the capacitor leakage up a couple of posts), yet I notice (sometimes serious) factual errors in it reading it back again, although I do correct them once I notice them... Do you trust me too much?  

I upvoted you for sarcasm and for being a superior being.

[Rehab1]

1 hour ago, esaj said:

I'm not sure what to make of it, people upvote my posts (like the one about the capacitor leakage up a couple of posts), yet I notice (sometimes serious) factual errors in it reading it back again, although I do correct them once I notice them... Do you trust me too much?  

 

35 minutes ago, litewave said:

I upvoted you for sarcasm and for being a superior being.

Me as well. Ok...2 current upvotes. Lock um in so they don’t vaporize. 

[Hunka Hunka Burning Love]

I saw some diagrams and formulas.  That’s good enough for me.    Add in a chart, and it would be elevated to gospel.  

[TomOnWheels]

7 hours ago, esaj said:

I'm not sure what to make of it, people upvote my posts (like the one about the capacitor leakage up a couple of posts), yet I notice (sometimes serious) factual errors in it reading it back again, although I do correct them once I notice them... Do you trust me too much?  

It is upvoted because it's interesting and it's facts... Personally as an engineering I like to promote knowledge. Then you know actually the only fact that you correct errors in your post prove that it's not just a copy/past, and that you actually understand what you are posting... that give more credits to that post :-) Are you a Nokia engineering ? :-)

[esaj]

31 minutes ago, TomOnWheels said:

It is upvoted because it's interesting and it's facts... Personally as an engineering I like to promote knowledge. Then you know actually the only fact that you correct errors in your post prove that it's not just a copy/past, and that you actually understand what you are posting... that give more credits to that post :-)

Yeah, that... plus I was pretty drunk last night, surprised I got it about right in the first place  

 

31 minutes ago, TomOnWheels said:

Are you a Nokia engineering ? :-)

No, electronics is just a hobby for me. I hold an engineering degree in software, which is pretty different (easier, if you ask me).

[TomOnWheels]

35 minutes ago, esaj said:

Yeah, that... plus I was pretty drunk last night, surprised I got it about right in the first place  

 

No, electronics is just a hobby for me. I hold an engineering degree in software, which is pretty different (easier, if you ask me).

Ha ha ! Brilliant ! Enjoy ! I'm hyper-frequency/RTS engineer, believe me software appears to me like a kind of magic that I can't understand most of the time :-)

[Lutalo]

12 hours ago, esaj said:

I'm not sure what to make of it, people upvote my posts (like the one about the capacitor leakage up a couple of posts), yet I notice (sometimes serious) factual errors in it reading it back again, although I do correct them once I notice them... Do you trust me too much?  

I thought the point of upvoting was to elevate the discussion. You always post useful technical info that is worthy of more discussion, and that more eyes should see. So what if you discover that some of it is incorrect later.

I am sure that you are not too proud to self-correct as I have witnessed it as recent as yesterday. As long as you can forgive yourself for an error the rest us can also forgive you ??