CCX 52V-19.2Ah BATTERY "Guard Voltage" & Other ANOMALIES"

[Ant~Phan]

Hullo All in Juiced Bikes (JB) Land. I have hesitated to do this post as, on the surface, it may well seem trivial to some, if not all lovers of JB e-bikes. But, it is equally about JB's customer support quality as well as batteries per se - so here goes.

As background I have a JB CrossCurrentX (CCX) with the 52V-19.2Ah battery pack. I enjoy the bike and am impressed generally, but as I load the bike close to its 275lb limit I am concerned with maximizing performance especially as I note that it starts to tail off approaching a 50% State of Charge (SoC). Given my weight, lack of fitness and other health issues - a key marketing demographic for e-bike buyers I suspect - I need solid assist - every pun intended.

The issue that I have tried to address, and failed, through JB support is that the standard charger results in a maximum SoC of 57.7V, ie, some 93%. "So what" some may say; what does it matter to miss out on the final 7% of charge. Well, given I see a lessening of assist at 50%, my Fun Range of charge is 50% to 93%, ie, 43%. Obtaining that additional 7% is a Fun Range percentage improvement of >16% and, within the best performance (voltage) range of all to include less amperage draw for any particular power (watts) requirement.

Please, also note, that we are discussing - given an extra cent - a $1,300 asset. I would like the option to decide what SoC I use for my battery and not be arbitrarily restricted by JB and for no reason given, indeed, resolutely refused. Please also note that while JB claim (see item #7 in article https://support.juicedbikes.com/hc/en-us/articles/360026938792) that the charger stops at 95%, the BMS puts out a spoofed voltage of 58.7 to 58.8 V thereby indicating a fully charged pack when it is at 57.7V [my standard JB charger is rated as and puts out an honest, ie, measured 58.8V]. That is, there is always a +0.90 to 0.99V Guard Voltage added to the pack's internal voltage. Please refer to 'attached' PDF file: Battery Voltage Levels.

The primary reason for this post is to bring to the attention of those that might be interested in such things not only of the spoofing a fully charged state and how to counter that to achieve a full 100% SoC; but, also, JB support staff's obfuscation on the matter in general; and, their complete intransigence - their utter refusal - to answer that simple question; in essence, if I obtain such as a GRIN Satiator can I program for 59.8V (58.8 +1 for the Guard voltage) and thereby achieve 100% SoC?

During the various communications with JB staff I received at least four separate and different explanations. Please refer to PDF file FOUR REASONS FOR GUARD VOLTAGE - EXTRACTS, 'attached'. For a "...Chapter & Verse" documentation trail - and if so inclined - please refer to 'attached' PDF file JUICED BIKES FOUR REASONS FOR STANDING VOLTAGE.

Perhaps two or three of you in this community with the 52/19.2 pack will measure voltage at the discharge port (feed to Controller Unit (CU)) with the pack powered OFF as well as ON; for the latter also record, once the battery is remounted, the voltage shown by the CU via the handlebar display?

Perhaps I do have a bad pack? But I suspect not within the general scheme of things.

Do JB have something to hide? What? Why? Especially the refusal to even discuss how to charge to 100%? Pre-emptive strike against users leaving packs at 100% and making warranty claims. Weak architecture? Older kit press ganged into a pseudo 52V world that (may) exceed their (original) design limits? Why the "...secrecy"? Who Can Say?

 

[Bruce Arnold]

Let's cut to the chase. Get the Grin Satiator and don't look back.

 

[Ant~Phan]

Bruce;

...and what would you set the charge complete voltage to. If 58.8V, notionally maximum for a "52V" (14S) pack, then no change from using the JB-supplied charger. If to 59.7 or 59.8V to allow for the guard voltage then what consequences say at the BMS and/or the controller. If there were no issues around this pack and charging above 57.7V why JB so resolutely refuse to answer my question in effect asking "...is it OK to set the charge cutoff at 58.8V +0.9V?" The fact that they have consistently refused - not carelessly or negligently forgotten - to answer that seemingly simple question after some three presentations of it and multiple follow ups should concern us all. It seems that JB are (much?) concerned at folk charging past 57.7V - why else go to the trouble of setting up the guard voltage hold? And, then refuse to answer in any understandable let alone meaningful way any question related to this (measured) observation. Granted I know little about matters electric - except my science master banging on about "...it's the volts that jolt, the millis that kills" - maybe all battery packs hold such a voltage, but I have just the one...…………...

 

[Explorer-1]

Just to clarify a couple of points in your communication, the BMS of course can be configured to balance at any voltage, also in general their last bit of advice to leave the charger connected after the light goes green is generally good advice for many BMSs. However; depending on how far the pack might be out of balance 30min may not be enough as most balance at an extremely low current, but it depends on what approaches they use.

It's possible the voltage drop is due to the resistance in an electronic switch/mosfet used for the power button or BMS output control. Most likely the voltage at the charge port is the real battery pack voltage, but the output voltage reflects the voltage drop across the BMS output.

 

[jharrell]

Trying to understand what you mean by "guard voltage", are you saying there is a boost converter in the BMS somewhere upping the battery voltage 1 volt to fool the charger? I see a similar discrepancy between my multimeter and the display, but occams razor says either the display is off of my multimeter is off, most likely the display due to line losses from battery to sensor and probably calibration as it shows 57.8 at full charge. I am actually looking into a accurate voltage source to check my multimeter for accuracy.

 

[Bruce Arnold]

Bruce;

...and what would you set the charge complete voltage to. If 58.8V, notionally maximum for a "52V" (14S) pack, then no change from using the JB-supplied charger. If to 59.7 or 59.8V to allow for the guard voltage then what consequences say at the BMS and/or the controller. If there were no issues around this pack and charging above 57.7V why JB so resolutely refuse to answer my question in effect asking "...is it OK to set the charge cutoff at 58.8V +0.9V?" The fact that they have consistently refused - not carelessly or negligently forgotten - to answer that seemingly simple question after some three presentations of it and multiple follow ups should concern us all. It seems that JB are (much?) concerned at folk charging past 57.7V - why else go to the trouble of setting up the guard voltage hold? And, then refuse to answer in any understandable let alone meaningful way any question related to this (measured) observation. Granted I know little about matters electric - except my science master banging on about "...it's the volts that jolt, the millis that kills" - maybe all battery packs hold such a voltage, but I have just the one...…………...

Here's what Juiced is doing. They don't make the chargers. They source them from a company where they can get them cheap. The charger only goes to 95% because it's not a great charger, not because Juiced is up to something nefarious that they don't want to explain. If you think about simple marketing, that's probably not something they are going to say, but there it is. I'm guessing -- educated guess -- that they are okay with this situation because (A) it keeps the price of their bikes down and (B) it's pretty well accepted that charging to 100% all the time is not good for the battery's longevity.

So, if you're more interested in getting that last 5% of performance than battery longevity, get the Satiator. It's really that simple. Go on the Grin site, absorb the information, ask them questions.

 

[Ant~Phan]

Thanks to you all for your input; as each correspondent's is sufficiently diverse I will attempt specific and, hopefully, clarifying responses in turn.

CityExplorer: (1). Balancing: I agree with you; a badly "...out of balance" pack may need quite a while ensure that all 84 cells are in balance. (2). "...Voltage drop across 'x', 'y' or 'z'" is JB's entire message. Specious at best and attempting to remain civil if not polite!

jharrel: (1). "... (BMS somewhere) upping the battery voltage to fool the charger?" is exactly what I am saying or trying to, at least. (2) Poor ol' friar Occam - sometimes too blunt an instrument, methinks. Even a Harbor Freight special (multimeter) is accurate enough for our needs; from a reputable auto parts store very much so. If any doubts, one can perform basic cross reference checks eg, measuring the voltage on brand new D, AA and AAA cells; comparing the readings of a common output with another meter, etc. And, further validating by cross reference across scenarios, eg, see first attachment above, last two lines show complete conformance with my measurements thus conclusions. Both your multimeter and the CU via the handle bar display are providing accurate (but different) readings because the voltages are different; the former at the discharge port (I assume), the latter by the CU. Indeed the difference is always near enough 0.90 to 0.95V is it not? NOT COINCIDENTALLY that 0.09-0.95 holding or guard voltage can (only) be measured at the discharge port ie the battery pack's current output to the CU while the pack is powered OFF. Unless my pack is a one-off with a fault, you will see said Guard Voltage at that output port. And, with the pack (still) powered off your will see a full voltage of say 55.0V at the charge port. And, further had you taken a near to concurrent reading off the handle bar display (CU measurement), ie, before removing or immediately after reinstalling the pack I suggest that you would have seen a measurement of 54 to 54.2V, ie, 55V MINUS the guard voltage (allowing for rounding and timing induced errors), See below.

Bruce Arnold: (1) You malign the charger (maker), Sir. They may be cheap but used by most ebike vendors. As I have already noted my insulted charger (it's probably in a huff now and won't work) puts out a very steady and honorable 58.8V (be careful when you measure your charger's output, there's a risk of the proverbial blue flash if you touch the shroud; try holding needle probes like chopsticks approach each pin from the middle, ie, opening the chopsticks outwards), That exact 58.8V is what is needed exactly to bring a 14S pack to a 100% SoC (assuming genuine cells). So when presented with 58.8V - zero difference - the charger says "...job done" and reverts to (say) 0.2A current for balancing purposes. (2) I grant that charging to 100% all the time and/or leaving the pack idle at 100% is not best practice; but that is not the issue we are discussing here, is it? That is, JB for reasons they resolutely refuse to discuss has jimmied the BMS to fool the charger to stop charging with spoofed 58.8V when the pack is at 57.8V 95% as claimed for the clever charger.

Happy voltage measuring - please let me know your results, especially discharge port/pack OFF. Cheers All.

 

[Explorer-1]

I don't believe they are fooling the charger in anyway. They either just have the BMS set to cut the charging before 58.8V or their BMS has a xV drop between input and output due to the design. This is pretty standard for inline circuits and could be be related to reverse polarity protection, or other circuit design issues. If i open one of their batteries one day i'll be sure to check.

 

[Ant~Phan]

CityExplorer: thank you for your input; clearly you have an electrical related background - certainly more than mine, I am sure. And, YES, if one day you open up your 52V-19.2Ah pack let me know what you find especially in relation to such as resistors, capacitors, etc between the BMS, cells matrix and ports, charge and discharge.

In the meantime - and not wanting to mess with the internal hardware of my singular pack all I can do is reiterate what I hope you will agree are logical interpretations of my measurements and observations and to which I would appreciate your further analysis and comments.

1. The standard, ie, JB-supplied charger holds a measured 58.8V across its two output pins; in that sense it is a dumb device
2. Once the charging process is flagged complete by the charger indicator light changing from red to green, the voltage measured at the pack's charging (input) port is 58.7 to 58.8V; hence the charger's light going to green.
3. At the same time as #2, the discharge (output to bike) port shows the same 58.7 to 58.8V, while the pack is powered on (pack power button, circular green light);
4. At the same time as #3, with the pack powered off the discharge port now shows 0.893 to 0.990V as measured over several trials.
5. [quickly reinstalling the pack aand powering up the bike], thus almost coincidental with #2-#4, the CU through the handlebar display shows 57.7-57.8V (92.5%SoC).

I find it beyond sheer coincidence that Charge port Voltage minus Guard Voltage equals CU displayed voltage. JB's literature remarks that the charger (rather internal process) ensures that the SoC never exceeds 95% - in practice 92.5-93.0V. I have taken measurements at various SoC always with the same arithmetical results.

That is, the honest element in all of this is the CU and its display of the (net) SoC of the pack. If you have another explanation for the regularity of the above arithmetic, please enlighten me?

I can only suggest and ask that those (still) interested in this matter would:

1. Take the mentioned measures on their 52V-19.2Ah packs and post them here
2. Take the similar measures for their 48V packs - is there a measurable voltage at their discharge ports when the pack is powered off;

I also ask that those fortunate folk with several e-bikes and from other vendors measure their batteries similarly to see if they have different discharge voltages powered on vs powered off?

Finally, as a final plea would those established e-bike and battery gurus and pundits, eg:

• Thomas Jaszewski;
• Ravi Kempaiah;
• bikerjohn
• et al

Would kindly offer their considered views?

Cheers All;

 

[Ant~Phan]

Here's what Juiced is doing. They don't make the chargers. They source them from a company where they can get them cheap. The charger only goes to 95% because it's not a great charger, not because Juiced is up to something nefarious that they don't want to explain. If you think about simple marketing, that's probably not something they are going to say, but there it is. I'm guessing -- educated guess -- that they are okay with this situation because (A) it keeps the price of their bikes down and (B) it's pretty well accepted that charging to 100% all the time is not good for the battery's longevity.

So, if you're more interested in getting that last 5% of performance than battery longevity, get the Satiator. It's really that simple. Go on the Grin site, absorb the information, ask them questions.

Bruce Arnold: you are a stalwart of these forums so I was hoping for further enlightenment from you. The standard charger goes to 95% not because it's a bad unit but because that is how JB has arranged matters. Refer to item #7 in https://support.juicedbikes.com/hc/...-the-voltage-on-my-LCD-not-accurate-sometimes- .

Have you measured the voltages on your 52V-19.2Ahr pack; especially at the discharge port, pack powered on and OFF? What readings did you record?

And, as I asked before, how, using your Satiator in getting that last 5% SoC, ie, charging to a genuine 58.7 to 58.8V as shownby the CU at the handlebar display.

Thanks again for your contributions and help.

 

[joshuaho96]

When I had an HF1100 I could get it to 58.2V or so with the cheap charger that they shipped. What's probably happening is that the line calibration is imprecise because the stock charger is cheap so they target 95% SoC max, with say a 5% deviation in SoC either way. Some people get chargers that do a full charge, others may only get a 93% charge as you have seen. You can either try and do a better line calibration on your charger which is usually achieved by shaving a resistor or adjusting a potentiometer, or you can get a Grin Satiator which has proper line calibration out of the box.

Edit: To be clear, your problem is assuming that because the open circuit voltage on the charger is 58.8V that it will charge to 58.8V. That's not how that works. The charger is passing current at a low level until max charge voltage is reached, it has to "know" the resistance from the voltage output sense to the battery, otherwise voltage is lost in the wires and you undershoot the target voltage.

 

[Ant~Phan]

When I had an HF1100 I could get it to 58.2V or so with the cheap charger that they shipped. What's probably happening is that the line calibration is imprecise because the stock charger is cheap so they target 95% SoC max, with say a 5% deviation in SoC either way. Some people get chargers that do a full charge, others may only get a 93% charge as you have seen. You can either try and do a better line calibration on your charger which is usually achieved by shaving a resistor or adjusting a potentiometer, or you can get a Grin Satiator which has proper line calibration out of the box.

Edit: To be clear, your problem is assuming that because the open circuit voltage on the charger is 58.8V that it will charge to 58.8V. That's not how that works. The charger is passing current at a low level until max charge voltage is reached, it has to "know" the resistance from the voltage output sense to the battery, otherwise voltage is lost in the wires and you undershoot the target voltage.

joshuaho96: thank you for the considered input. I am not enough of an electrical engineer to debate your points in detail other than from a (layman's) common sense perspective. Juiced Bikes (JB) are specific that they have arranged for their charger to only charge to a max. of 95% - in practice 92-93% (57.6V) - see referenced article two posts above. What I would appreciate your further input/comment on and as a summary of my OPs:

1. Whenever I measure the charge port (CP) voltage at the end of charge (light red to green) it is at 58.7 to 58.8V ie, 100% SoC); concurrently,
2. The controller unit (CU) through the handlebar display show 57.6V, ie, -1V approx.; and, as concurrently as I can achieve removing pack ASAP;
3. The discharge port (DP) also shows a max charge voltage of 58.7/58.8V while the pack is powered on; and
4. When powered OFF is shows 0.900 to 0.990V (some dozen measurements). This reading I refer to as a Ghost or Guard Voltage (GV)
5. So YES it may be pure coincidence that voltage readings CP (DP on) minus GV = CU reading; but I don't believe that.
6. It is my Ahem, Theory that JB has set that GV to spoof a full charge voltage of 58.8V when the pack is at 57.7V approx.

And, as a cross reference check for plausibility, why do JB over months of asking refuse to answer (a) how to charge to 58.8V (100% SoC) and (b) say using a GRIN Satiator if it is safe to do so, ie, without causing failure of (say) an underspecified CU, etc?

 

[joshuaho96]

joshuaho96: thank you for the considered input. I am not enough of an electrical engineer to debate your points in detail other than from a (layman's) common sense perspective. Juiced Bikes (JB) are specific that they have arranged for their charger to only charge to a max. of 95% - in practice 92-93% (57.6V) - see referenced article two posts above. What I would appreciate your further input/comment on and as a summary of my OPs:

1. Whenever I measure the charge port (CP) voltage at the end of charge (light red to green) it is at 58.7 to 58.8V ie, 100% SoC); concurrently,
2. The controller unit (CU) through the handlebar display show 57.6V, ie, -1V approx.; and, as concurrently as I can achieve removing pack ASAP;
3. The discharge port (DP) also shows a max charge voltage of 58.7/58.8V while the pack is powered on; and
4. When powered OFF is shows 0.900 to 0.990V (some dozen measurements). This reading I refer to as a Ghost or Guard Voltage (GV)
5. So YES it may be pure coincidence that voltage readings CP (DP on) minus GV = CU reading; but I don't believe that.
6. It is my Ahem, Theory that JB has set that GV to spoof a full charge voltage of 58.8V when the pack is at 57.7V approx.

And, as a cross reference check for plausibility, why do JB over months of asking refuse to answer (a) how to charge to 58.8V (100% SoC) and (b) say using a GRIN Satiator if it is safe to do so, ie, without causing failure of (say) an underspecified CU, etc?

They "arranged" for the charger to target 95% SoC because they don't want to risk overcharging the battery even with a worst case, potentially out of spec charger. Overcharging can lead to internal short circuits and fires so it's extremely dangerous to let that happen.

My advice is to disconnect the battery from the bike after a full charge, then probe the battery output directly with a multimeter. Be extremely careful to not cross the leads and touch them because you will cause a dead short that is guaranteed to destroy the battery connector, the fuses just don't blow fast enough on these battery packs to guarantee protection in these situations.

If you get 57.6V at the battery output then what's most likely is that the BMS is drawing enough current to create the "guard band" you're talking about. Either because it's trying to balance your battery cells which may be unevenly aging or because the BMS is not very good and has high standby draw.

If you get the full 58.8V then the answer is pretty simple, the display itself either has bad voltage sense calibration or it's drawing enough power to cause a full volt of sag just by turning on the bike.

 

[Ant~Phan]

They "arranged" for the charger to target 95% SoC because they don't want to risk overcharging the battery even with a worst case, potentially out of spec charger. Overcharging can lead to internal short circuits and fires so it's extremely dangerous to let that happen.

My advice is to disconnect the battery from the bike after a full charge, then probe the battery output directly with a multimeter. Be extremely careful to not cross the leads and touch them because you will cause a dead short that is guaranteed to destroy the battery connector, the fuses just don't blow fast enough on these battery packs to guarantee protection in these situations.

If you get 57.6V at the battery output then what's most likely is that the BMS is drawing enough current to create the "guard band" you're talking about. Either because it's trying to balance your battery cells which may be unevenly aging or because the BMS is not very good and has high standby draw.

If you get the full 58.8V then the answer is pretty simple, the display itself either has bad voltage sense calibration or it's drawing enough power to cause a full volt of sag just by turning on the bike.

joshuaho96: thank you much sir; all makes good sense. I suspect your first paragraph is what this may well be about. That said - and in spite of my repeated requests - not one other fellow forum member has offered measurements of either the 52V packs or other JB packs, nor other vendors' packs (to see if the problem is with the JB 52V pack only - Oh well).

As to the display itself I doubt that it measures voltage nor current internally; would it not just be displaying the appropriate data as formatted and sent to it by the CU?

As to voltage measurements I am not sure what you mean for me to try. How do I "...probe the battery output directly" (other than at the discharge port). I am having trouble picturing the "...disconnect the battery from the bike..." instruction. If I am to access the power feed from pack to controller I will have to dismount the pack. I apologize if I am being especially hard of thinking - but as noted earlier I am no electrical whizz kid.

No rush to reply today needed - time to kick back and watch TV! Good night.

 

[joshuaho96]

joshuaho96: thank you much sir; all makes good sense. I suspect your first paragraph is what this may well be about. That said - and in spite of my repeated requests - not one other fellow forum member has offered measurements of either the 52V packs or other JB packs, nor other vendors' packs (to see if the problem is with the JB 52V pack only - Oh well).

As to the display itself I doubt that it measures voltage nor current internally; would it not just be displaying the appropriate data as formatted and sent to it by the CU?

As to voltage measurements I am not sure what you mean for me to try. How do I "...probe the battery output directly" (other than at the discharge port). I am having trouble picturing the "...disconnect the battery from the bike..." instruction. If I am to access the power feed from pack to controller I will have to dismount the pack. I apologize if I am being especially hard of thinking - but as noted earlier I am no electrical whizz kid.

No rush to reply today needed - time to kick back and watch TV! Good night.

You would be probing the battery at the discharge port with it switched on. I would do this experiment myself but my HF1100 is long gone.

 

[Ant~Phan]

You would be probing the battery at the discharge port with it switched on. I would do this experiment myself but my HF1100 is long gone.

joshuaho96: I apologize if my earlier posts were unclear; probing the battery at the discharge port is what I have done and where I first measured the Ghost/Guard Voltage (GV). Some photo's 'attached' of readings when at a (partial) SoC of 70%.

CU (53.9) = Charge/Discharge ON (54.8V) minus Discharge OFF (0.97V) allowing for drinking tea at the time!

 

[jharrell]

Ok I see what you mean by "guard voltage" although I would call it just some voltage leak when the BMS is off, probably the switching mosfets are leaking a little when off. When on the charge port and discharge port have the same voltage which reads about 1 volt higher than the screen. So again occams razor there is just some voltage drop to the screen and why the screen shows 57.8 when the charger says full (which should be 58.8).

To actually "fool" the charger into shutting off early the BMS would have to have a some sort of boost converter and you would see a 1 volt difference between charge and discharge and I see no point in that at all nor any readings that would confirm that.

 

[Ant~Phan]

Ok I see what you mean by "guard voltage" although I would call it just some voltage leak when the BMS is off, probably the switching mosfets are leaking a little when off. When on the charge port and discharge port have the same voltage which reads about 1 volt higher than the screen. So again occams razor there is just some voltage drop to the screen and why the screen shows 57.8 when the charger says full (which should be 58.8).

To actually "fool" the charger into shutting off early the BMS would have to have a some sort of boost converter and you would see a 1 volt difference between charge and discharge and I see no point in that at all nor any readings that would confirm that.

jharrell: so you believe that it is just coincidence that the difference between CU display and charge/discharge ports equals the pack-off voltage at the discharge port?

Please also note that when the 52V pack charger shuts down - charge complete - (red to green light) the charge port presents a measured 58.8V, ie, the max voltage for that 14S pack. No one has presented any data that any other JB or other vendors' packs exhibit the same characteristic. One other - Bikerjohn (if I remember correctly) - also noted that his 52V pack stopped the charge at 93% SoC - net 57.7V (58.7V at the charge port).

JB's documentation specifically states that it is intentional that the pack is never charged above 95% (93% - 57.6V in practice); so, how do you think they are managing to make a dumb charger (certified labelling at 58.8V and 2A), stop at that 57.6 to 57.7V level? It seems to me the only honest element in all this is the voltage read out at the CU display.

See item #7 in https://support.juicedbikes.com/hc/...-the-voltage-on-my-LCD-not-accurate-sometimes

That is, the way that JB stops the charge process early - to avoid going past SoC of 95% - is to spoof up the pack full voltage pf 58.8V by adding the ghost voltage to the pack voltage. Indeed, isn't that the only way that can achieve their declared intent?

Perhaps you would like to comment on why JB absolutely refuses to answer the simple question: "How does one charge the 52V pack to 100% SoC"? (EG by using a GRIN Satiator and setting the charge to voltage to (say) 59.7V for pack voltage -0.9V = 58.V. What do you they think they are frightened of?

 

[jharrell]

Yes its probably coincidence, again the battery on the charge and discharge port read the same and it is the expected voltage at 100% SOC. Whats more likely the battery is only 93% charged based on the screen voltage or a 100% based on a multimeter reading? Why would you go through the trouble of putting some kind of boost converter (on both charge and discharge mind you) in the BMS and have the screen calibrated to show 1 volt less??? Just change the voltage adjustment in the shipped charger and and have accurate voltage readings throughout?

Honestly I'm not even sure you could integrate a boost converter in such a way to fool a charger if the battery was connected, you would have to pull current from the battery to boost voltage then feed it back and the battery would just eat it up again, even if some circuit could be designed why all that expense and complexity, just drop the trim on the charger if you want under 100%, I think there is even a video around showing which pot to turn inside the stock charger to tweak it.

 

[jharrell]

So just tested the off voltage on the discharge port with a 1k resistor in parallel with multimeter and voltage dropped from almost 1v down to .04v and with a wire shorted across .01v this confirms to me its mosfet switch leakage on the BMS, there are no relays in there and mosfets will still let a little leakage through even if off, as soon as you give it a load the voltage drops because its so low power being let through. Be sure to not turn on battery with these test connections in place!

You can read another account here: https://electricbike.com/forum/foru...y-when-battery-is-switched-off-it-reads-44-5v although the voltage reading is higher probably do to better mosfets in our BMS.