My XM-5000li Experience

227 posts / 0 new
Last post
garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
Cooling the XM-5000li Motor

An attempt to cool the 5000Li motor better:
Add some "fan blades" around the outer edge of the motor.

I think that one can fairly easily make (or get made) 12 "fan blades" for the motor. There could be 6 left-handed blades for the left side of the motor and 6 right-handed blades for the right side. Each blade could be mounted on 2 of the 12 bolts that fasten the motor to the wheel.

Cutting out the blade shape from flat stock aluminum, drilling two holes, and then making one straight 90-degree bend and another bend at about 60 degrees would complete one "blade".

Here is a picture of a paper "blade" to better show what I am suggesting.

FinOnBack-Lo.JPG

These blades would be very easy to install, removing just two of the motor-mounting bolts at a time. It is possible that longer bolts might be required. If the blades are effective, a "cooling fan" upgrade kit could be produced inexpensively.

Other suggestions?

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
Measuring the XM-5000li Motor Temperature

I want a visible reaout of the motor temperature.

So, I just purchased some small, inexpensive infrared thermometers from Harbor Freight. Item 93983 on sale for a few days only at HF stores for $7.99 with coupon printed on-line, otherwise $9.99 or a bit higher.

I will try to take two apart and connect them so that only one takes the reading, but the other one (or both) will display the value.

If that works, I will fasten one to the swing arm where it can "see" the motor casing. The "remote display" will go to the handlebar area.

I will keep you posted.
Others trying the "same" thing, let me know so we can share ideas.

Other ideas, or any suggestions?

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: Measuring the XM-5000li Motor Temperature

Gary -

Are you sure it's the motor overheating and not the controller overheating? Do you actually have a temperature sensor in the hub motor with a wire going to the controller? I know the XM-3500 does NOT have a temperature sensor in the motor - don't have any direct experience with the XM-5000.

I'm sure the motor is getting hot, but I bet the controller is getting hot as well.

Just a thought.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

Iccarus
Iccarus's picture
Offline
Last seen: 6 years 9 months ago
Joined: Saturday, December 20, 2008 - 08:05
Points: 347
Re: Measuring the XM-5000li Motor Temperature

I've been wanting to install a temp gage also. I have driven 20miles at full throttle on an 80F day on my 3500li with 24cells. The motor did not feel very warm but I know that doesn't help much. When I get my gage I will take a reading after a long run with 28cells. I'll post it when I get it.

2008 XM3500li Mods/Kelly KBL12251/84v 28cell 40AH pack/ Variable regen brake trigger on left brake handle/Givi/Cycle Analyst/Homemade BMS

KMX Typhoon Home build (recumbent pedelec) with two Astro Brushless 3220motors/twin castle Phoenix ICEHV 160/ Cycl

garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
Re: Measuring the XM-5000li Motor Temperature

John,

The 5000Li has (so the manual says) a motor over-temperature switch that activates the brake/kickstand circuit and that turns on the Diagnostics #1 light. We know which connector (along the left swingarm) to unplug to disconnect (disable) the motor's normally-open sensor-switch.

The controller also has an internal over-temperature sensor that apparently turns on the Diagnostic #2 light.

The controller felt just warm but the motor was quite hot to touch.

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

Iccarus
Iccarus's picture
Offline
Last seen: 6 years 9 months ago
Joined: Saturday, December 20, 2008 - 08:05
Points: 347
Re: Measuring the XM-5000li Motor Temperature

John, I thought the 3500li has an overtemp protection for the motor and controller? No? I probably wouldn't trust it with what I'm doing anyway but I thought at least the controller had that.

2008 XM3500li Mods/Kelly KBL12251/84v 28cell 40AH pack/ Variable regen brake trigger on left brake handle/Givi/Cycle Analyst/Homemade BMS

KMX Typhoon Home build (recumbent pedelec) with two Astro Brushless 3220motors/twin castle Phoenix ICEHV 160/ Cycl

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: Measuring the XM-5000li Motor Temperature

John, I thought the 3500li has an overtemp protection for the motor and controller? No? I probably wouldn't trust it with what I'm doing anyway but I thought at least the controller had that.

On the XM-3500 I think it's just the controller that has over temp protection - not the motor.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: Measuring the XM-5000li Motor Temperature

Double post deleted.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: Measuring the XM-5000li Motor Temperature

John,

The 5000Li has (so the manual says) a motor over-temperature switch that activates the brake/kickstand circuit and that turns on the Diagnostics #1 light. We know which connector (along the left swingarm) to unplug to disconnect (disable) the motor's normally-open sensor-switch.

The controller also has an internal over-temperature sensor that apparently turns on the Diagnostic #2 light.

The controller felt just warm but the motor was quite hot to touch.

Sounds like you've got it figured out. :-)

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
Re: Measuring the XM-5000li Motor Temperature

I took apart the small, inexpensive CEN-TECH infrared thermometer (from Harbor Freight).

It appears to have just two possible places to "split" two units so that one would be the sensor, and the other would be the display.

1. The IR sensor has 4 short wires going to the PCB, but extending these might introduce too much noise ... but worth a try with shielded wire.

2. The display appears to be an LCD with 20 "connections". A 20-conductor flat cable might be used, but the "connector" from the PCB to the display is a little rubber-like strip that must have conductive "channels" in it. It appears to just make a pressure-fit between the "glass" and the contacts on the PCB.

I have not seen this kind if "connection" before. Removing the display and soldering a ribbon cable to the PCB is easy. Connecting the ribbon cable to the "display" ... seems more difficult. And, again, the "extension" might be too long.

I will have to look at the display's "contacts" with my microscope.

The device includes a 0.01 sec. stopwatch function. It displays an HH:MM clock and its ambient temp. when it is "off". I have measured things from -2 to 175 degrees F.

I think the last day of the sale is this Sunday (Easter).

Any ideas or suggestions?

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

Henry42
Offline
Last seen: 8 years 2 weeks ago
Joined: Friday, August 1, 2008 - 15:44
Points: 85
Re: Measuring the XM-5000li Motor Temperature

Thanks for the tip. I was going to get the themometer from AeroStich/RiderWearHouse http://www.aerostich.com/aerosearch.php?restrict=Static&q=4899 for $49.00, however, the $9.99 is a much better price (I guess the sale is still going on, because I bought one today.) I found a Harbor Freight on the other side of town (~15 miles away). I rode the XM-5000Li to work, then Harbor Freight, then home. The motor temperature (after ~15 miles) averaged about ~108 degrees, outside temperature was ~61 degrees.

I was looking for a speedometer http://www.aerostich.com/aerosearch.php?restrict=Static&q=2131 when I noticed the themometer.

Iccarus
Iccarus's picture
Offline
Last seen: 6 years 9 months ago
Joined: Saturday, December 20, 2008 - 08:05
Points: 347
Re: Measuring the XM-5000li Motor Temperature

Henry, Thank YOU for the tip I just ordered one of those keychain IR temp gages cause I can use it for other things too. I'll keep you guys posted on temps running 28cells in the 3500li.

2008 XM3500li Mods/Kelly KBL12251/84v 28cell 40AH pack/ Variable regen brake trigger on left brake handle/Givi/Cycle Analyst/Homemade BMS

KMX Typhoon Home build (recumbent pedelec) with two Astro Brushless 3220motors/twin castle Phoenix ICEHV 160/ Cycl

garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
Adding XM-5000li Control Functions

The area on the plastic cover in the center area of the handlebars looks like a good area for adding some function switches, status displays, and possibly a motor-temperature display.

1. Motor Over-Temperature Override.
I plan to install a red LED that indicates that the motor over-temperature condition is being detected. I also want to add a switch (maybe a miniture toggle switch, or possibly a rocker switch) to keep the over-temperature condition from stopping the controller's motor drive, so I can get safely off the road then the over-temperature condition is sensed. I want to add a yellow LED to indicate "caution" whenever I have the new Motor Over-Temperature Override switch on.

2. I hope to use the Infrared (or other) temperature sensor to give me an actual (maybe not perfectly accurate) temperature reading of the motor temperature, so I can have some warning that the temperature is getting too high, and that I should reduce power to the motor. With that, I would be able to ride more safely.

Also, I might be able to give better feedback about this over-temperature condition, and measure the effect of adding 12 cooling "fan" fins (6 on either side) around the outer edge of the motor housing.

3. I want to add a Hazard Light (both turn signals blinking) function, and perhaps an orange (or yellow) LED to indicate the function is ON, or flashing.

4. If I can find some small meters, I would like to add an automotive 8-to-16 volt voltmeter (and maybe an ammeter) to show the state of the 12v DC system.
Running with the headlights on all the time does put a substantial load on that system, and using the brakes while the turn signals are on ... is a "BIG" load.

5. I might make a circuit to show me approximately how many cells are under 2.7 volts, and a yellow "warning" LED if any cells are that low.

6. I might also add a red "alarm" LED if any cell goes under 2.1 volts. The actual voltage of the lowest cell would be nice, but, harder to get.

7. The Battery Pack amps would be nice. The Pack voltage is already (approximately) indicated in the Instrument Cluster.

8. Relocate the Economy Mode switch, perhaps as a rocker with a green LED for Normal and a yellow LED for the "more-limited" Economy Mode.

-------------
Can you think of Other Ideas for things that would be useful as one is riding?
Thanks

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
LVM for My XM-5000li

I intend to make a Low Voltage Monitor circuit that does several things.

1. Activates a Low-Voltage "Alarm" function (at least lighting a red LED) when any cell goes under about 2.1 volts. A series of red LEDS under the seat would show which cells are critically low. Most LVC circuits activate the motor-cutoff circuit.

2. Activate a Low-Voltage "Caution" function (at least lighting a yellow LED) when there are any cells below 2.7 volts, and attempt to display to the driver some indication of how many cells are below 2.7 volts. Perhaps a small ammeter showing 1 ma per low cell would be used, where 15 cells low would indicate 15 ma on a 0 to 25 ma scale. A series of yellow LEDS under the seat would show which cells are low.

3. When the 12-volt system is not active (the ignition has turned off the DC-toDC converter), the current flow to the underseat LEDs would be cut off, so that the low cells are not even further drained by the cell-status LEDs. Or, an underseat push-button would need to be used to light up the cell-status LEDs.

4. This LVM circuit could be unplugged from the cells when the bike is to be stored for longer periods of time, to cut the current drain (of this monitoring circuit) on the cells to zero.

5. The circuit would be independent of High-Voltage Monitoring, which is really only needed during cell Charging (or during Regeneration).

Any other ideas before I make the test circuit and lay out a PCB?

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

Henry42
Offline
Last seen: 8 years 2 weeks ago
Joined: Friday, August 1, 2008 - 15:44
Points: 85
Re: The XM-5000li Motor Overheat

Well the diagnostic #1 light (motor overheat cutoff) came on for me, yesterday. I was going about 45 MPH, fortunately, I was able to just pull off onto a side street. It was 75 degrees and I had ridden mostly continuously for about 19 miles (26 miles total miles for this charge) averaging 45 - 50 MPH. I immediately checked the temperature with the Harbor Freight thermometer which read ~126 degrees measured on the left side. The controller was ~ 106 degrees. Thanks Gary for finding this first and providing a method to check it. Rather than having a panic of "what’s wrong now", it was a "this is what Gary reported", I need to stop and check the motor's temperature. After about 30 minutes the temperature dropped to ~102 degrees and I was able to continue.

I agree with Gary's comment. This abrupt no-warning complete loss of power is a substantial SAFETY issue.

The intent of this ride was to get an indication of how far the XM-5000Li would go with two riders. I figure with two riders, the GIVI 470, the battery charger, etc., the weight was around 400 pounds (still below the 440 pound limit). At the end of 29 miles the pack voltage was 67.6 (after stopping). The under a load voltage was 64.5 volts. I think I could have gone at least another five or six miles, but did not want to take a chance. I will attempt the two rider range test on another day when it is cooler. (It may not be until next weekend since it is raining today.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: LVM for My XM-5000li

Any other ideas before I make the test circuit and lay out a PCB?

Gary - your ideas all sound great. Have you looked at the Gary Goodrum / Richard Fletcher design on endless sphere? Gary's company (TPpacks.com) sells an unpopulated board. They have also published their schematics. You might prefer to do it all yourself and you may not want a balancer. However, starting from the G/F board might also be a good option for you.

If I were you I would make your LVM a hard cut off - you already have the caution alarm so you can avoid the problem of surprising the user and causing a safety issue. With some empirical testing you could probably create a similar caution alarm for motor temperature.

As far as over temperature and under voltage goes here's what we're implementing on the REV:
1) We have a device we call the BCU (Bike Control Unit) which is a meta-controller (i.e. it monitors inputs and can control the inputs to the motor controller)

2) We are implementing the G/F BMS design using SMT to keep size down and to be able to mass produce them

3) The BMS will monitor for LV and will cut off when a cell goes below 2.1V - this will be a hard stop. At this point we need to stop everything to avoid damaging cells.

4) The BCU will monitor Amps in and out and will drive the fuel gauge (state of charge / SoC). When SoC drops below 10% (or some other threshold) we will warn the rider and we will restrict throttle to half input (or some other amount)

5) The BCU will monitor pack temperature, controller temperature and motor temperature. There will be two thresholds for each of those temperatures. The first will be a warning threshold and will restrict throttle input to say 75%. The second threshold will be a complete shut down.

6) All these conditions will likely be communicated via colored LEDs on the dash

7) We are also implementing a 4x20 character LCD display which will be available as an optional extra (at a pretty low cost). The LCD display will give far more information (form the same input sources). The reason we're going to make it an optional extra is because (a) not every rider cares to see this type of info (and in fact some might be intimidated by info overload) & (b) we don't have the resources available right now to integrate it into the dash in a way which satisfies us (but we realize a lot of folks won't mind adding their own LCD panel somewhere).

8) We are currently debating whether the BCU will perform data logging and whether that log will be stored on internal memory (requiring the user to download via a cable) or on SD card (allowing the user to pop out the card and take it with them to a PC).

I'd welcome any input.

Good luck with your mods - they all sound great to me...

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

Henry42
Offline
Last seen: 8 years 2 weeks ago
Joined: Friday, August 1, 2008 - 15:44
Points: 85
Re: Adding XM-5000li Control Functions

I would like to see something like:
LED #1 - Brake on/Kickstand Down (blue LED is fine)
LED #2 - Controller Overheat Protection - maybe yellow or flashing red for warning, red for shutoff.
LED #3 - Motor Overheat Protection - maybe yellow or flashing red for warning, red for shutoff.
LED #2 and #3 could behave the same way the LVC works. (At a certain voltage the light starts flashing, which give ample time to compensate)

What would be excellent would be a temperature gauge for the motor and a temperature gauge for the controller, much like the temperature gauge on a car.

Maybe the motor override switch could be a button switch located to the right of the high beam switch around the left brake reservoir. (That way you could press it with your left hand to get to safety.)

I was also thinking of adding an emergency flasher, like the one on a car. The switch would connect both turn signals. The indicator would be that both turn signal indicators are flashing (instead of just one if you are indicating a turn.)

Maybe the emergency flasher and the "Economy Mode" switches could be located relatively close together. There is an access door on the left just below the handlebars or maybe on the handlebars.

I was thinking about the possibility of being able to turn off one of the headlights and the two extra brake/tail lights. This would be useful in conditions when I would need to save as much energy as possible to get home.

garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
Re: LVM for My XM-5000li

John,
There is a low PACK voltage warning, but it is substantially ineffective at protecting the lowest cell from going too low.

Also, a 2.1 volt "cutoff" is not likely to stop a "slightly-too-slowly-responding" controller from causing voltage spikes that go substantially below the (perhaps) 2.1 volt "cutoff" level.

I intend to make a warning that grows in severity as any cell goes below a "warning" level (perhaps 2.7 volts), which could also be used to reduce the max controller current to (essentially) zero as any cell approaches the "cutoff" voltage (2.1 or so).
This would lessen the chances of higher controller currents producing low-voltage "spikes".

I would essentially retain the cutoff function at a lower "cutoff" voltage, but by the time the cutoff is reached, the load current would already be reduced from 100 amps to maybe 10 (for example), or less.

To achieve that drive-reduction control, I might just reduce the (max) throttle voltage toward zero.

Maybe the simple "cutoff" LVC is sufficient, but I have not yet seen any data that confirms that.

---------
I intend to do a similar thing with the over-voltage protection, reducing the charger's output as any cell goes above a high-warning (that also activates a shunt).

The ideal marriage for a "charger" is a standard (voltage and current limited), lower-cost, simple "power supply" that has an input that can reduce the output current (changes the built-in limit) of the supply to zero.
Most (many) DC power supplies already include a simple current-limiting circuit.

Then, simple shunting makes the charging process work well, and no monitoring of Pack voltage is needed.

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
Re: Adding XM-5000li Control Functions

Henry,
All very good ideas.

I intend to do most of them, mounting most lights, gauges, and switches in the blank, unused center section of the handlebars.

The Instrument Cluster has 3 unused LEDs, where the leads appear to have been cut off right at the base of the LED (so new LEDs are likely needed anyway).

I have not yet investigated other switch/indicator locations, since the "handlebar panel" can be fabricated and then replaced as a unit. Is there a similar (or identical?) panel on the 3500?

Presumably the new XM-4000li has a similar "unused" panel.

Definately "hazard" flashers, but I have not yet found an easy, effective way to add temperature measurement to the motor.

Now, AndyH, you, and I have all "suffered" the motor-overheat. How far would I have gotten today (our first 99-degree) "summer" day?

I have suggested adding cooling "fan" blades to the 5000li, but I do not know if they would help significantly. I have not gotten any real action plan from API in this matter.

Do we each need to open a "ticket" on this failure?

CAUTION: For some reason, X-Treme STRONGLY stated that switching the Economy mode switch must NOT be done while in motion.

I intend to replace the brake/tail (and probably the turn-signal) lights with LEDs, largely solving the power consumption (and over-loaded DD-to-DC converter) problem.

However, I have considered un-hooking the tail light function from the middle of the 3 tail/brake lights, so that the "brake" function becomes much more obvious. But, that will depend upon the brightness of the LEDs selected.

I just found some shorter (only 44 mm), much less expensive, "yellow" 1156 LED "bulbs" that might be suitable for the turn signals. Most likely a load-independent flasher module will be required, but the "stock" flasher in the 5000Li includes a "beeper" (that's frequency is too high for me to hear).

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: LVM for My XM-5000li

John,
There is a low PACK voltage warning, but it is substantially ineffective at protecting the lowest cell from going too low.

Also, a 2.1 volt "cutoff" is not likely to stop a "slightly-too-slowly-responding" controller from causing voltage spikes that go substantially below the (perhaps) 2.1 volt "cutoff" level.

Hi Gary,

Not sure which system you refer to when you talk about there already being a low PACK voltage warning? The stock system has pack voltage. Your proposed system, the G/F system and our system all work on the cell level.

I need to check on the specifics but I'm not sure what the exact characteristics of cell damaging voltage and time duration is. E.g. Does a few tenths of a second at 1.9 mean certain death for a cell? How much sag will be occurring under these load conditions? etc.

Our intent is to persuade the rider to never find this out. I.e. by starting to back off the throttle (and hence the load) when SoC is reading 10% available (and SoC is based on using the cells to 80% DOD) the user will likely want to charge before reaching the LVC situation.

BTW - although LVC will cause the controller to kill power it doesn't latch-off. In other words LVC won't be a safety concern because when the load is removed the cells voltage will recover above 2.1V allowing the on-again-off-again limping to the side of the road. Not ideal - but not stranded either. At this stage though it's not all that different than running out of gas. We gave the rider all sorts of warnings before it happened.

The ideal marriage for a "charger" is a standard (voltage and current limited), lower-cost, simple "power supply" that has an input that can reduce the output current (changes the built-in limit) of the supply to zero.
Most (many) DC power supplies already include a simple current-limiting circuit.

Then, simple shunting makes the charging process work well, and no monitoring of Pack voltage is needed.

Agreed. In future versions we are planning on integrating the BMS and the charger so that we can use a current-limiting power supply and rely on the BMS shunts. However, our current time-line makes sticking with a separate BMS and separate smart-charger a necessity.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: Adding XM-5000li Control Functions

Definately "hazard" flashers

When/if you switch to LED turn signals then, as you note, you'll likely need a new flasher unit. That flasher unit will likely include "hazard" functionality.

Now, AndyH, you, and I have all "suffered" the motor-overheat. How far would I have gotten today (our first 99-degree) "summer" day?

One of the key limitations of the hub motor design is heat dissipation. In a standard motor the hot bits (the windings) rotate and are on the outside edge of the motor assembly - thus it's fundamentally easier to cool them. On a hub motor the windings are stationary and are on the inside surrounded by a nicely sealed metal case (for environmental purposes). Your idea of fins may help some - but in the standard hub-design that we've seen there's no good thermal contact between the windings and the case so any cooling effect will be limited.

At EVMFG we're working on alternative hub-motor cooling designs.

BTW - one of the reasons the Vectrix has more power is because they use a traditional motor (not a hub motor). So they can cool it more easily. (Some folks claim the VX has a hub motor - not so, it's a traditional motor with a direct connection to the rear hub).

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

garygid
Offline
Last seen: 13 years 6 months ago
Joined: Friday, December 19, 2008 - 23:25
Points: 441
Re: LVM for My XM-5000li

John,
Perhaps I do not understand how you will determine an individual cell's SOC to be getting near 10%. Is it by measuring the cell's voltage? Or, if it is just by integrating the pack current, that might not be soon enough to help one low cell. If you want to chat privately, PM your phone # to me.

I am not aware of any good battery data either. The "low-spike" data might not even exist.
And, I suspect the control-input response-time of various controllers is not generally available. A slow response to an abrupt cutoff would generally allow a spike. Many cutoff/resume cycles per second could all generate potentially-damaging under-voltage spikes on the low cell(s).

I intend to try for a smooth cutback in "power" as any cells get near "low", much like I think you intend to do with your "10% SOC" control.

Some limited experiments using the 5000li's TS smart charger (in the normal, "stock", pack-connected way) and simple shunt levels of 3.6 (+/- 0.05), it appears that this charger never gets a high enough "pack-voltage" to quit charging.

I am going to add a small "pot" to each GF channel to be able to raise the shunt-ON point to 3.7 (or higher), and get all the channels set closer to the "same" voltage. Maybe then the TS "smart" charger will work as intended. However, we do not know what is "intended" for the "end-of-charge", do you?

Do you use a PakTrakr for gathering cell-voltage and pack-current data?

If so, I can send you my PTViewer program, if you would like. It graphs up to 48 cell's voltages, and pack voltage and current. I have found it VERY helpful.

Cheers, Gary
XM-5000Li, wired for cell voltage measuring and logging.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: LVM for My XM-5000li

John,
Perhaps I do not understand how you will determine an individual cell's SOC to be getting near 10%. Is it by measuring the cell's voltage?

Ahh, I think I see where I confused things. There are two separate concepts - both related to an enhanced riding experinece.

1) We do not measure individual cell voltage - but we do monitor low voltage and when any one cell drops below the LVC threshold this is signaled to the controller. So, we can't tell you cell #5 is at 2.6V and cell #21 dropped to 2.0V (or whatever). But we can tell you that one of the cells in the pack is below 2.1V. This is sufficient for battery pack protection.

2) To help the rider keep track of the state of charge (SoC) of the pack we're implementing an Ah counter for the pack. This will be far more accurate and informative to the user than the XM's and EFD's usage of a voltmeter as an approximation of SoC. When pack SoC falls below 10% (or some other threshold) is when we will start throttling back power usage.

If the rider continues to ride the bike until the fuel gauge is all the way empty and then further still it is possible that they will encounter LVC. However, we hope that by then they've taken the hint that it's time to charge. Thus the 10% SoC threshold is more about "rider behavior modification" than a system safety cutout.

Is that any clearer?

Thanks for the offer of the PTViewer - however, we're not using a PakTracker.

Note that in a future incarnation of a combined BMS/Charger/BCU setup we will likely go the extra mile and get individual cell voltages. For our first release we may decide to not only signal that one cell in the pack has hit LVC but also which cell. That way we can keep track and see if the same cell is always going to LVC first and way before any other cell (by comparing the average cell value to the one known cell value).

As far as charging goes - just like we know when the first cell hits LVC we know when the last cell hits HVC and at that point we can turn off the charger completely. The TS "smart" charger doesn't know anything about individual cells so it uses a conservative average cell value to get to a conservative pack voltage cut off point. Because it knows that the cut off is conservative it goes into a mode where it lets the cells naturally balance for a while and then switches back on and chargers for a little bit longer. It sort of "nudges" the overall pack voltage higher. That's my understanding. I can ask for more details from TS if you like but as we're planning on cell level HVC we're not particularly concerned with how the TS charger works (if we use a TS charger it will be tuned to the appropriate voltage levels for our application).

I think you and I are basically on the same track...

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

Mikie
Offline
Last seen: 1 year 6 months ago
Joined: Friday, July 25, 2008 - 15:55
Points: 114
Re: My XM-5000li Experience

John and Gary-- good conversation and good work. I think I can basically follow the concepts and where you are going. The most important question I have is can I buy one from you John and will there be layman instructions on how to install? I already have set up a 24 cell port for access to each cell so hopefully this could just plug into there? Best....Mikietd

mikie

Henry42
Offline
Last seen: 8 years 2 weeks ago
Joined: Friday, August 1, 2008 - 15:44
Points: 85
Re: The XM-5000li Motor Overheat

I went on another extended ride yesterday. The temperature was 85 degrees. Before the ride I used dish soap and water to clean the motor and controller. The color of the water was black at first. I rode ~16 miles, stopped for a few hours, then rode another ~30 miles (averaging 45 - 50 mph). The temperature of the motor where I measured (at the end of the ride) averaged ~125 degrees. I was going to go further, however, the voltmeter was reading 56 volts at full throttle and I thought I should stop. I started out with a charge of 72.6 volts. I wonder if the motor needs to be washed every 500 miles or so, maybe sooner if dusty conditions exists.

Is anyone else having display issues on this message thread? What was on the right side is now displaying at the bottom.

reikiman
reikiman's picture
Offline
Last seen: 12 months 2 days ago
Joined: Sunday, November 19, 2006 - 17:52
Points: 8447
Re: The XM-5000li Motor Overheat

Is anyone else having display issues on this message thread? What was on the right side is now displaying at the bottom.

Henry (and others) .. I've been doing some tweaking of the presentation of the site. I've moved some of the information from the sidebar to the bottom. I've been posting about this in the 'General Discussion' and 'Problems & Suggestions' forums. Please follow up in those areas with your concerns. Thank you.

Henry42
Offline
Last seen: 8 years 2 weeks ago
Joined: Friday, August 1, 2008 - 15:44
Points: 85
Re: Charging the XM-5000li

I was reading the Operation and Maintenance Instructions manual and was attempting to figure out what "70% discharge" means (page 9, Battery Longevity). I then went to the Instruction Manual for LFP/LCP Lithium Power Battery manual (Page 10) and see a chart indicating 70% is 3.6 volts. As I understand the chart 100% is above 4.5 volts and 20% is 2.5 volts. I'm probably reading the chart wrong. However, I do see where at the top of the chart is "LFP battery can be charged by fast-charging mode" So I have a few general questions.

What voltage is considered 100%

What percent is 2.5 volts?

If you let the charger continue and all batteries charge to 4.5 volts (94.5 volts total) which may take days, what will break?

The last question is about pulse charging. What happens if the charger outputs 15 amps until all batteries reach maximum. I have attempted to find the answer on other threads and have been unsuccessful.

Any help appreciated, including any links that have explanations.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: Charging the XM-5000li

Hi Henry:

1) 100% full can be considered 4.5V, but most folks I believe use 4.25V as 100% full. And most folks actually charge to something like 3.7V because you don't actually get much more energy storage in that last part of the curve and you just stress the batteries for little gain.

2) 2.5V is 20% SOC (state of charge) which is 80% DOD (depth of discharge). Above when you read the chart and say 70% is 3.6V that's 70% SOC and 30% DOD. SOC is a measure of how full a battery is, DOD is a measure of how empty (so the two always sum to 100%). To get the best life out of the batteries one would use a very conservative 70% DOD. Most folks in the EV world use an 80% DOD and still believe that to be reasonable. 90% DOD is starting to push things hard enough whereby you'll markedly shorten the life of the battery.

3) If you let the stock charger with no BMS charge all cells to 4.5V you are virtually guaranteed to overcharge some of your cells. If you have a balancer then you won't overcharge your cells. However, it's likely that you will trip the overvoltage protection in the controller - meaning that your bike won't go until the volts drop down below it's upper threshold.

Hope this helps.

BTW - if you look at the discharge curve for the batteries you'll notice a quick drop at the beginning and end of the cycle. This is why with LiFe a voltmeter is far less effective at telling you how much charge is remaining. A state of charge meter that actually monitors the power being drawn out of the battery pack is more effective.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

Iccarus
Iccarus's picture
Offline
Last seen: 6 years 9 months ago
Joined: Saturday, December 20, 2008 - 08:05
Points: 347
Re: Charging the XM-5000li

Well put John. What I do now is just charge till I get the two green lights unless I think it might be a close call getting to where I'm going. Then I will bump it up a to make sure. Your battery life will be longer if you don't push the cells to high or draw them to low, it's that simple. The more you stay in the middle the longer they'll last. John is right it seems like they come down so fast from that higher float charge (3.7v) it's hardly worth the trouble it takes to get them there. In my experience they come down to the 3.3Xs almost right away.

2008 XM3500li Mods/Kelly KBL12251/84v 28cell 40AH pack/ Variable regen brake trigger on left brake handle/Givi/Cycle Analyst/Homemade BMS

KMX Typhoon Home build (recumbent pedelec) with two Astro Brushless 3220motors/twin castle Phoenix ICEHV 160/ Cycl

Henry42
Offline
Last seen: 8 years 2 weeks ago
Joined: Friday, August 1, 2008 - 15:44
Points: 85
Re: Charging the XM-5000li

Thank you for your excellent explanation. It helps explain a lot. Averaging 40 - 50 mph, I am getting about 45 miles per charge (charging to ~76.4 volts). Last weekend I went for a six mile ride, then recharged. I only had time to charge the batteries to 72.6 volts. I was still able to go about 45 miles. The voltage at the end was ~55.0 volts ( at 30 mph, going up (in parts) what I think is a 30 percent grade). All the cells are within 0.04 volts of each other, both charged and discharged (at rest, no load). I was wondering if charging to 84 volts would produce a range of ~50 miles, however, I now understand it is not worth the effort, especially if I want the longer battery life.

Once again thanks for your excellent explanation.

Pages

Log in or register to post comments


Who's online

There are currently 0 users online.

Who's new

  • Juli76
  • xovacharging
  • stuuno
  • marce002
  • Heiwarsot

Support V is for Voltage