3KW / 8KW MXUS & Powervelocity 24S2P / 34S6P ?

BobBob

100 W
Joined
Sep 4, 2020
Messages
195
So I'm after a sanity check rather than showing off my creation for now, hopefully this is the right place.
I'm planning a build for bashing around off road, to keep up with my mates who are 10 yrs younger

What I think I need:
Something relatively low key, that won't upset the UK mountain biking fraternity too much
Future version with larger battery and controller is Version 2 (V2)

150 KG/330lb bike and rider, 20% grade, thinking that adding 150W for 3hours is about right to not be seen as cheating too much. Edit: Probably doubling this +

The plan is to start with a downhill or freeride bike, use a gearless hub so it's quiet and mechanically simple, have enough torque to tackle some pretty steep (20-25%) short hills and enough Wh to just provide some pedalling assistance on a day trip

Bike
Bike shortlist: Orange 5, Scott Nitrous, Transition TR450, Kraftworks FRM125 or Lapierre 718 Froggy, all around the £6-700 mark off ebay. I'm thinking I need a big cog on the front, perhaps a wide range front and back
Edit - Bought the Nitrous

Motor: - already purchased - MXUS 3K V3 Turbo XF-40-45H - 26" wheel, 10 Ga spokes

Controller: Powervelocity 18 FET 7Kw as the sensible pedal assist option
V2 upgrade for hooliganing - 24 FET 15KW but twice the size and 3X the weight so not so good for low key
http://powervelocity.com/home/55-24f-15kw-sinewavesilent-controller-android-ios-bluetooth-programmable.html

Battery - low key - 24S2P
Sony US18650 VTC6 30A 10Wh 3Ah cells seem to be the best spec, new cells cost a bit more but give fewer hassles.
A 24S2P Battery could produce 60A & 100V peak, 84V nominal, maybe around 5.4Ah, 440 Wh at 20A draw
Two of these batteries in parallel could be used for extended range - self built to fit frame with bespoke 3D printed box to look, um, less boxey
Sold out Hmm - may go for VTC5 2.6Ah instead or start looking again
Looking at the Samsung INR21700-40T 4000mAh - 35A it can put out 45A if limited to 80 Deg C,

Optional V2 higher spec battery 34S6P for use with larger 15 KW controller - future expansion of project

BMS - Daly - planning to check it regularly initially and fit a reset button. I have a cheap scope, clamp meters etc so will have an external balancing type cable for checking and perhaps paralleling batteries - maybe through ~30 Ohm resistors
May not need a second BMS for second parallel batt that way either
https://www.aliexpress.com/item/4001359842440.html

Charger
Cheap Current and voltage limiting charger [EDIT} power supply used as charger
2 x https://www.aliexpress.com/item/32953568490.html in serie
powered by one 600W supply, assuming they're floating outputs https://www.aliexpress.com/item/32850123866.html
I should add a spare thermal cutout (sensors in the pack, wired to a relay that is powering the charger)
TODO - proper risk assessment of charging FMEA etc

The result? 61% efficiency at 5MPH up a 10% incline or 50MPH for V2 - seems pretty good
The two configurations on a 10% incline - Grin simulation if the 4504 is the 45mm magnet and 04 winding - any thoughts on better settings for the V3 MXUS motor - can anyone confirm this is correct for this motor?
This shows likely pedalling and max output V2 options I think:
https://ebikes.ca/tools/simulator.html?motor=MX4504_SA&batt=cust_86_0.2_5.4&cont=cust_60_150_0.03_V&grade=10&axis=mph&throt=13.6&hp=200&mass=150&bopen=true&cont_b=cust_150_300_0.03_V&motor_b=MX4504_SA&batt_b=cust_122_0.2_17&mass_b=150&hp_b=50&grade_b=10&autothrot_b=false&autothrot=true&throt_b=100&frame=mountain&frame_b=mountain

So, questions are, what am I missing?
24S seems pretty good for the motor and 7 Kw controller combination. Q. Why don't people like higher voltages?
The motor seems to work well accross the speed and torque range I need with the two battery configurations

For the purists, yes, geared is more efficient but I want silent, I don't even like the click of freewheels.
Mid drive is better on steep hills and for unsprung weight, but I want simple and quiet and snapped chains seem a pain (putting 3KW through a system designed for 10% of that would seem to need more engineering consideration)

The battery is pushed to its limit, I expect to outgrow it quickly so it may have a 100 recharge life. Anyway, I have to find more projects for my modified £20 welder :) a 3P version may fit easily in the frame (CAD TODO) which would be more sensible

Debating using a battery box and removable battery for charging versus a battery box that clips round the bike or just wrap in fibreglass fire blanket or rig a smoke alarm to a hose for a hillbilly fire suppression system or something... TBC

Security
I've a couple of USB drive sized GPS trackers which I could dismantle to embed in the electronics then encapsulate in epoxy or something. Parasitic back up power from the controller or direct from battery should be ok, might need a buck converter from a couple of cells from the batt but expect a 5V rail on the controller so may run from a big Capacitor to make it difficult to find or remove. Once localised with GPS, the Bluetooth from the BMS should allow me to find it again

No photos yet because I haven't bought much or built anything but if anyone spots has any thoughts and suggestions please shout out. I'll edit this if I get questions to make it more readable / explain
 
Is 24S2P VT6 the total pack size? What max current to you plan on drawing from the battery pack? Seems very undersized if you plan on drawing more than 30A total batt current.
 
The max draw per cell is 30A so it's upper limit is 60A but there are none of these cells available at the moment.
The Sony US18650VT5 is a nominal 35A max though only 2.6 Ah instead of 3.0 so in theory 70A with 2P (briefly).
This is not going to be my only battery, just the first one - V2 refers to the opposite end of the spectrum
This small battery was for putting out 150W normally and keeping up with cyclists but able to put out 3KW briefly for the steep bits for a few seconds rather than stalling.
If I've got the simulation right, it will match my 200W pedalling and take me up a 10% grade at 5.7 MPH for around 9 miles which is how it's intended to be used.
If pushed, it could take me up a 30% grade or a 15% grade at 27MPH for a couple of miles at which point the battery overheats, the motor overheats and the battery runs out of juice at about the same time which seems pretty good for a small battery.
 
Ok so you are expecting ~35A max current from this 24S2P pack (3000W / 86.4V = ~ 35A). That seems doable for a very short burst with VTC6 or VTC5. Have you considered slightly larger 21700 cells like the 40T or P42A? They are more capable and cost less. I'm sure there are even better alternatives too for a small pack like that.
 
BobBob said:
So, questions are, what am I missing?
24S seems pretty good for the motor and 7 Kw controller combination. Q. Why don't people like higher voltages?
The motor seems to work well accross the speed and torque range I need with the two battery configurations
People here rightly love high voltage, but under 60V is safest, 80V is the spec for ANN fuses, 90V (21s) is the limit for common controllers, proper charging becomes problematic, etc.

BobBob said:
Charger
Cheap Current and voltage limiting charger
2 x https://www.aliexpress.com/item/32953568490.html in serie
powered by one 600W supply, assuming they're floating outputs https://www.aliexpress.com/item/32850123866.html
I should add a spare thermal cutout (sensor in the pack, wired to relay powering charger)
These are not chargers, and the charger is not the place to skimp on a high-voltage 24s pack. Every other thread in the battery subforum relates to a failed BMS -- don't rely on it to prevent a fire. If you can't bulk-charge properly, consider charging through the balance leads -- it would be trivial if you're building a pack yourself.
 
pwd said:
Ok so you are expecting ~35A max current from this 24S2P pack (3000W / 86.4V = ~ 35A). That seems doable for a very short burst with VTC6 or VTC5. Have you considered slightly larger 21700 cells like the 40T or P42A? They are more capable and cost less. I'm sure there are even better alternatives too for a small pack like that.

Nope, I had matched the max spec of battery to max input of the controller IE 70A and 100v from 24sx2p cells
I was planning on turning it down from 70A to 60 for the 30A cells.
I was thinking that in some circumstances the controller or the battery might be the weak link

I'll have a look at the 21700s Thanks
 
fatty said:
People here rightly love high voltage, but under 60V is safest, 80V is the spec for ANN fuses, 90V (21s) is the limit for common controllers, proper charging becomes problematic, etc.
At work they like us to stay under 48V, we have 240 mains and I've been picked up for having exposed 5v wires :) I'll be careful and it's only for me, never to be sold.
Controller was as recommended from ES, not sure how common it is, just tried to match its performance with choice of battery.

fatty said:
BobBob said:
Charger
Cheap Current and voltage limiting charger
2 x https://www.aliexpress.com/item/32953568490.html in serie
powered by one 600W supply, assuming they're floating outputs https://www.aliexpress.com/item/32850123866.html
I should add a spare thermal cutout (sensor in the pack, wired to relay powering charger)
These are not chargers, and the charger is not the place to skimp on a high-voltage 24s pack. Every other thread in the battery subforum relates to a failed BMS -- don't rely on it to prevent a fire. If you can't bulk-charge properly, consider charging through the balance leads -- it would be trivial if you're building a pack yourself.
Agreed - corrected :)
I don't think I will rely on being able to prevent a fire.
Even with the best charger in the world, a battery fault such as a dent could cause a short and thermal run away so I'm thinking I should only charge it or store it in a fire proof container, transport the battery in a fireproof container in the car and be ready to ditch it if it fails during a ride.
The fire proof container must be good enough, If it is, then the risk becomes cost of new battery vs cost of good charger.
This is possible with the battery in a Steel container with flame arrest and pressure relief - might do some calcs (Ex-D)
This planned battery should cost around £200 whereas some chargers seem to cost £3-500
I'll give this more thought - Thanks for the feedback I will take it on board and let it settle a few hours / days :)
 
BobBob said:
I don't think I will rely on being able to prevent a fire.
Even with the best charger in the world, a battery fault such as a dent could cause a short and thermal run away so I'm thinking I should only charge it or store it in a fire proof container, transport the battery in a fireproof container in the car and be ready to ditch it if it fails during a ride.
The fire proof container must be good enough, If it is, then the risk becomes cost of new battery vs cost of good charger
This planned battery should cost around £200 whereas some chargers costs £3-500
I'll give this more thought - Thanks for the feedback I will take it on board and let it settle a few hours / days :)

:thumb:
I'm not a fearmonger saying everyone should charge LiIon in fireproof containers, or even outdoors.
But especially if you're playing with >80-90V, a proper charger with safeties is a necessary investment -- not just to protect yourself and your house, but also to protect your battery.

LiTech and SuPower make them.
See also 24s lipo charger (100.8V) recommendations
 
Fire
Bouncing ideas around shouldn't be taken as advice.
My bike will live in a detached brick garage 20 ft from my detached brick house so minimal risk to anyone in case of fire.
If I put this relatively small battery in a fireproof container to charge, it protects my car, motorbike and ebike.
Failure modes
A BMS seems sensible because, I'm stretching this small battery regarding current and Voltage. I also like the BT monitoring option. I expect the majority of them work ok even if they fail quite often.
A failed open circuit charge balancing line should be obvious on the BT BMS and not immediately fatal - external balancing should be ok
A failed dead-short on the balancing line seems unlikey and would kill that cell or parallel cells regardless of whether external balancing is done, in other words a posh charger may not help and may not prevent a fire in this case.
A fuse on each balancing line turns the closed circuit failure into an open circuit failure which seems a good idea.
A low current leak would kill that bank of parellel cells but I can put another two cells in for £10 so it doesn't matter. Need to check before use to avoid reverse polarity on cells - BMS is BT so should be obvious.
More checks than this? Probably won't bother for a £200 battery, just build another.
Function
The DC PSUs I found (somewhere on ES) for charging have CC/CV and are adjustable to cover my 14S, 24S and 34S batteries and it's trivial to set them to a partial charge for longer term storage or put a resistive load (bulb) in parallel to discharge to a voltage.
Put them in a lockable box with a £5 digital timer and thermal cutout and they are fireproof & thief resistant and have more functionallity and safety features than a dumb charger at a similar cost
With a timer they perform in a pretty similar way to a normal charger and there are two sets of redundant circuitry to shut of charging - The charger timer or the BMS can stop the charging.
Constructive criticism always welcome.
 
pwd said:
Ok so you are expecting ~35A max current from this 24S2P pack (3000W / 86.4V = ~ 35A). That seems doable for a very short burst with VTC6 or VTC5. Have you considered slightly larger 21700 cells like the 40T or P42A? They are more capable and cost less. I'm sure there are even better alternatives too for a small pack like that.
Looking at the Samsung INR21700-40T 4000mAh - 35A it can put out 45A if limited to 80 Deg C, not that the controller could draw more than 70A but it makes me think that a 2P battery might be OK
Cost is similar now to the VTC6 from 6 months ago and there doesn't seem to be much better available
Energy density is not as good as the VTC6 but it's close.
I'm now thinking of a 34S 2P battery and a 14S1P "get me home" spare battery. total 800 Wh between the two
These equate to a 2L coke bottle and a 600ml standard bike bottle - either will fit in an adjustable cage for a low key appearance and I can ride in areas with more grumpy people with the smaller battery.
Smplistic CAD image
First Concept.JPG
 
Have you considered the voltage sag for your two pack options? Low parallel counts will sag a lot, even with good cells. If you haven't seen this already: you can compare discharge curves of difference cells IE the VTC6 vs 40T etc..:
https://lygte-info.dk/review/batteries2012/Common18650comparator.php
 
Great resource - thanks, I'd been hunting about piecemeal for info and the rate that the cells overheat is interesting.
The Samsung cells seem to overheat after around 2.5-3 Ah at 30A constant or 3.5 Ah at 20A which seems OK
Voltage sag is around 0.35V at 20A discharged which equates to a DC Internal Resistance (IR) of around 20 mOhms - OK
Double checked on the Grin simulator, yup, the battery's the limitation but seems usable,
I don't think it matters much whether it's parallel or series - low cell count will sag more as each cell supplies more power.
A 12S3P will sag as much as a 18S2P or a 36S1P if supplying the same motor output wattage. More P will give better torque alt low revs and more S will give better top end power at high RPM
2P36S gives more scope to increase using the 150V 15KW PowerVelocity controller if I can get one (out of stock)
936 Wh battery for around £280 / $380 seems ok for a start and another in parallel would achieve controller limits very briefly
 
34S lithium-ion is unusually high but certainly possible. :bigthumb: My only gripe would be that higher voltage components seem to be more expensive (charger, controller, any dc-to-dc converters, displays etc...) for the same wattage. What sort of top speed are you looking to hit?

FYI I just finished building a pack with Molicel P42A cells and they are working quite well and had very good consistency out of the box.
 
You may want to split the battery 34S pack and use two chargers. I do a split battery with 10S using tool pack batteries and brand name laptop chargers as they are dirt cheap.
 
BobBob said:
...and have more functionallity and safety features than a dumb charger at a similar cost
With a timer they perform in a pretty similar way to a normal charger and there are two sets of redundant circuitry to shut of charging - The charger timer or the BMS can stop the charging.

I'm not sure what you mean by "dumb charger". A thermal cutout is a great idea, but chargers, by definition, include other safety features that power supplies do not -- safety features a timer doesn't replicate.

Also, never plan for the BMS to stop charging, even as a backup. See the Battery forum for more info.
 
Without the motor kV, this is all mental masturbation.

Pack voltage is designed according to your desired speed. Any more is wasted, and in the case of ludicrous voltages like this, dangerous.
Desired speed in MPH * 1609344mm/mi = mm/h
mm/h * 1hr/60min = mm/m
mm/m / tire circumference = no-load RPM
no-load RPM / 85-90% = load RPM
load RPM / kV = Volts required

Then size controller and then pack current for desired torque.


BobBob said:
Great resource - thanks, I'd been hunting about piecemeal for info and the rate that the cells overheat is interesting.
The Samsung cells seem to overheat after around 2.5-3 Ah at 30A constant or 3.5 Ah at 20A which seems OK
Voltage sag is around 0.35V at 20A discharged which equates to a DC Internal Resistance (IR) of around 20 mOhms - OK
Double checked on the Grin simulator, yup, the battery's the limitation but seems usable,
I don't think it matters much whether it's parallel or series - low cell count will sag more as each cell supplies more power.
A 12S3P will sag as much as a 18S2P or a 36S1P if supplying the same motor output wattage. More P will give better torque alt low revs and more S will give better top end power at high RPM

2P36S gives more scope to increase using the 150V 15KW PowerVelocity controller if I can get one (out of stock)
936 Wh battery for around £280 / $380 seems ok for a start and another in parallel would achieve controller limits very briefly

You need to go back to the drawing board on this. If you got a 3T, your 36s pack on a 26" wheel would give 143MPH :lol:
 
pwd said:
34S lithium-ion is unusually high but certainly possible. :bigthumb: My only gripe would be that higher voltage components seem to be more expensive (charger, controller, any dc-to-dc converters, displays etc...) for the same wattage. What sort of top speed are you looking to hit?

FYI I just finished building a pack with Molicel P42A cells and they are working quite well and had very good consistency out of the box.
I have a shopping list and a plan (WIP)
Voltage and current limiting PSUs + timer, thermal cutout, BMS and a fireproof box total around $130 same price as an off the shelf charger but with more features as well as (I hopefully) safety - TBC / WIP
Controller - I talked to PowerVelocity about the 7 KW and they advised the 15 KK I was convinced, they now advise the 7KW but I want the 15 KW :roll:
It's not top speed on the flat, it's grin factor IE acceleration to a speed with 150KG minimum and up to 180 Kg if I'm carrying a glider up a hill, silly idea but not unlikely.
For example if I'm doing 20 and want to go vroom up a hill at 30 then power / torque peaking at the 20-30 range seems about right
I've relied heavily on the Grin / ebike simulator which roughly points at good acceleration up to 20 MPH for the 7KW and up to around 30 MPH for the 15 KW, around 45 up a hill and 60 (never going to happen) on the flat.
As I say, I'm not interested in getting into a tuck on a runway, I want to be able to accelerate up a hill when covered in mud, maybe with a rucksack on my back but only for a few seconds or maybe a minute. Low speeds and power most of the time with occasional bursts of fun.
Not supposed to be a sensible commuter
For performance I need a mid drive but I'm keeping it mechanically simple for this build and overpowering a hub motor instead.
Until the plan changes of course
 
markz said:
You may want to split the battery 34S pack and use two chargers. I do a split battery with 10S using tool pack batteries and brand name laptop chargers as they are dirt cheap.
I re-read my previous post and it came accross as rather too abrupt, sorry.
I would like to avoid having lots of connectors - particularly for power connections so physically splitting the pack isn't my first choice
I would like a BMS and I'm not sure that the BMS would balance if you charge a section at a time through tappings without disconnecting the sections. Would need to look at that
Simplest neat solution seems a correct voltage charger
Second simplest a correct voltage PSU with current and voltage limit and use the BMS for shutoff
in a fireproof box :)
 
fatty said:
BobBob said:
...and have more functionallity and safety features than a dumb charger at a similar cost
With a timer they perform in a pretty similar way to a normal charger and there are two sets of redundant circuitry to shut of charging - The charger timer or the BMS can stop the charging.

I'm not sure what you mean by "dumb charger". A thermal cutout is a great idea, but chargers, by definition, include other safety features that power supplies do not -- safety features a timer doesn't replicate.

Also, never plan for the BMS to stop charging, even as a backup. See the Battery forum for more info.
Dumb charger might be a bit harsh but a cheap chinese charger will provide one current limit and one voltage limit and one minimum current cut off if you're lucky and none of these are likely to be adjustable.
The primary safety features of a lithium charger when compared to a PSU are the voltage and current limiting function which my proposed solution has.
It also allows variable maximum voltage and variable maximum current so could be used on different voltage battery packs.
If I build all my batteries with an external temperature output I can use it to cut the motor when riding or cut charging as needed
If you know of any other safety features provided by the battery charger I'll give it some thought and maybe figure out a way to include them

Still planning on charging in a tin box in a brick garage so if it goes up it doesn't matter :)

Safe after two failures was acceptable for ATEX safety ratings inside a fuel tank (zone 0) if I remember right, so good enough for my bike I reckon.
I'd have a failure of charger, failure of BMS and failure of tin box to risk the garage which is not a risk to people
 
fatty said:
Without the motor kV, this is all mental masturbation.
Grin Simulator not stimulator, and I didn't mean it that way, not judging though :)
Motor part number is in first post, including the values used in the Simulation, data also in first post, Edit 9Kv 4T Thanks for ref fatty
fatty said:
Pack voltage is designed according to your desired speed. Any more is wasted, and in the case of ludicrous voltages like this, dangerous.
At, for example, 20% of max RPM; the controller and motor are pretty efficient. The exess voltage is not "wasted".
If wasted it would mean that a bike capable of 30 mph would be dissapating most of it's power as heat when traveling at 15 MPH
Low voltage will require proportionally greater current from the supply and visa versa but pretty much the same resultant power at mid speeds throughthe motor.
Efficiencies seem to be broadly similar for, for example, a 24V or 150V system running at 15 MPH with all other components the same.
https://ebikes.ca/tools/simulator.html?motor=MX4504_SA&batt=B2412SLA&cont=cust_150_300_0.03_V&grade=0&axis=mph&throt=100&hp=0&mass=150&bopen=true&cont_b=cust_150_300_0.03_V&motor_b=MX4504_SA&batt_b=cust_149_0.225_16&mass_b=150&hp_b=0&grade_b=0&autothrot_b=true&autothrot=false&throt_b=16.8&frame=mountain&frame_b=mountain
No such thing as a ludicrous voltage, this might be a lucicrous application for a 150V battery but I haven't seen why yet.
What bit is worrying you when you say dangerous? I'd treat it like 240V UK mains from before the days of RCDs.
fatty said:
Desired speed in MPH * 1609344mm/mi = mm/h
mm/h / 60min/hr = mm/m
mm/m / tire circumference = no-load RPM
no-load RPM / 85-90% = load RPM
load RPM / kV = Volts required

Then size controller and then pack current for desired torque.
OK, thanks, i'll try that and compare it back to the Grin simulator
Edit - tried your calculation but it doesn't work as it ignores torque.
fatty said:
BobBob said:
Great resource - thanks, I'd been hunting about piecemeal for info and the rate that the cells overheat is interesting.
The Samsung cells seem to overheat after around 2.5-3 Ah at 30A constant or 3.5 Ah at 20A which seems OK
Voltage sag is around 0.35V at 20A discharged which equates to a DC Internal Resistance (IR) of around 20 mOhms - OK
Double checked on the Grin simulator, yup, the battery's the limitation but seems usable,
I don't think it matters much whether it's parallel or series - low cell count will sag more as each cell supplies more power.
A 12S3P will sag as much as a 18S2P or a 36S1P if supplying the same motor output wattage. More P will give better torque alt low revs and more S will give better top end power at high RPM

2P36S gives more scope to increase using the 150V 15KW PowerVelocity controller if I can get one (out of stock)
936 Wh battery for around £280 / $380 seems ok for a start and another in parallel would achieve controller limits very briefly

You need to go back to the drawing board on this. If you got a 3T, your 36s pack on a 26" wheel would give 143MPH :lol:
OK, I'm at the drawing board which in my case is the Grin simulator.

Motor as noted above is a MXUS 3KW V3.
Official name of motor is xf40 45h which I believe is the MXUS 4504 on the simulator
The controller is as recommended by PowerVelocity after I forwarded this same info to them
Taking the motor (already purchased) and controller as recommended, I've simulated 2 battery configurations for discussion
A 100V nominal 24S6P and 150V 36S4P "ludicrous" with exactly the same number of cells
https://ebikes.ca/tools/simulator.html?motor=MX4504_SA&batt=cust_86.4_0.1_24&cont=cust_150_300_0.03_V&grade=20&axis=mph&throt=100&hp=100&mass=150&bopen=true&cont_b=cust_150_300_0.03_V&motor_b=MX4504_SA&batt_b=cust_129.6_0.225_16&mass_b=150&hp_b=100&grade_b=20&autothrot_b=true&autothrot=true&throt_b=100&frame=mountain&frame_b=mountain

So, with the same motor, the same load, the same controller and the same number of cells why would I want to have 5 Kw instead of 8 Kw?

24S6P Vs 36S4P.JPG

Performanc discussion
Looking at performance on an average 3.6V/cell and worst case discharged battery, steep 25% hill heaviliy loaded 180 Kg briefly
Nominal voltage IE 86.4V and 129.6V, this appears to give me a peak power output of 5 KW at 20 MPH for the 100V nominal system and 8KW at 30 MPH from the 150V system
With me unladen going up a 20% incline that gives a top speed of 31mph or 42 MPH respectively
The actual useful working range will be a bit less than the theoretical max speed, say 25 vs 35 MPH - 35 seems better.
Carrying my paraglider and kit the weight might go to 180 Kg and at the end of the day on 1/4 charge of, say 3.0V I get a top speed of 22 or 31 MPH repsectively. I think that as I'm using the same number of batteries why not have 30 MPH available?
https://ebikes.ca/tools/simulator.html?motor=MX4504_SA&batt=cust_72_0.1_24&cont=cust_150_300_0.03_V&grade=20&axis=mph&throt=100&hp=100&mass=180&bopen=true&cont_b=cust_150_300_0.03_V&motor_b=MX4504_SA&batt_b=cust_108_0.225_16&mass_b=180&hp_b=100&grade_b=20&autothrot_b=true&autothrot=true&throt_b=100&frame=mountain&frame_b=mountain

Motor control theory from the old memory banks
The Grin simulator gives the same efficiency for the same number of cells with lots in parallel and lots in series in mid ranges.
This is what I'd expect.
From some work I did a few years back on stepper motor controllers; the higher voltages are better at overcoming inductance in the coils and we used to hit them with a wider pulse at the start (similar to phase shifted sine wave controllers) to get the current flowing then drop the PWM back to give an appropriate current through the coil.
I'd expect some copper resistance improvements due to lower current and some flux / iron disadvantage but I wouldn't expect it to be too significant compared to pumping my tyres up or a headwind.

Let me know where my simulation is wrong and I'd certainly buy you a beer for avoiding my wasted effort :)
 
BobBob said:
markz said:
You may want to split the battery 34S pack and use two chargers. I do a split battery with 10S using tool pack batteries and brand name laptop chargers as they are dirt cheap.
I'm wondering why?
I'm considering it but haven't seen a good reason yet

Because with reasonable voltage, you can use a proper, quality charger with safeties.
 
BobBob said:
Dumb charger might be a bit harsh but a cheap chinese charger will provide one current limit and one voltage limit and one minimum current cut off if you're lucky and none of these are likely to be adjustable.
The primary safety features of a lithium charger when compared to a PSU are the voltage and current limiting function which my proposed solution has.
It also allows variable maximum voltage and variable maximum current so could be used on different voltage battery packs.
If I build all my batteries with an external temperature output I can use it to cut the motor when riding or cut charging as needed
If you know of any other safety features provided by the battery charger I'll give it some thought and maybe figure out a way to include them

Still planning on charging in a tin box in a brick garage so if it goes up it doesn't matter :)

Safe after two failures was acceptable for ATEX safety ratings inside a fuel tank (zone 0) if I remember right, so good enough for my bike I reckon.
I'd have a failure of charger, failure of BMS and failure of tin box to risk the garage which is not a risk to people

This has been discussed to death -- check my post history.
Even Chinese chargers come with:
1) severe overdischarged halt
2) overdischarged low-current pre-charge
3) charge termination

CC/CV is not a safety feature -- that's just characteristic of CC/CV PSUs and chargers alike.

Blind and unmonitored voltage and current variability with unknown environmental coefficients is a liability, not a feature.

Just get another charger if you build a different pack. They're cheap and that way you can use them both.
 
BobBob said:
Grin Simulator not stimulator, and I didn't mean it that way, not judging though :)
Motor part number is in first post, including the values used in the Simulation, data also in first post, Don't know KV, think 4T

There's too much in this wall of text to correct individually, but if you break it out into specific, individual questions, you'll get better feedback from technical users like myself.

Here's my general advice: the Grin Simulator is like cheating on your homework with Google Translate or Wolfram Alpha. It's a great (even invaluable) tool to learn from and a lot of people can skate by using it, like 48V or 52V or even 72V, 750W or 1500W or even 3000W systems.

But if you're pushing the envelope into dangerous voltage and power, you need to speak the language natively -- because when you don't, you can't identify errors. Hint: does the 4504 in the simulator have the same kV as your motor?

There's a reason automotive electrical systems are still on 12V, and 48V is perpetually a decade away -- because high voltage is dangerous, and it kills. EVs with high voltage take severe precautions to protect occupants from it.

So throwing out these silly numbers at the wall to see if they stick, like 8kW through a MXUS 3k or 150V without knowing your motor kV -- it's just that, silly. You need a hard sanity check.

You should be able to explain why a 4T is a poor choice for a 150V system, and (related) why higher volts at lower PWM is inferior to matching voltage to desired speed (another hint: you have a fixed number of cells).
 
fatty said:
BobBob said:
Dumb charger might be a bit harsh but a cheap chinese charger will provide one current limit and one voltage limit and one minimum current cut off if you're lucky and none of these are likely to be adjustable.
The primary safety features of a lithium charger when compared to a PSU are the voltage and current limiting function which my proposed solution has.
It also allows variable maximum voltage and variable maximum current so could be used on different voltage battery packs.
If I build all my batteries with an external temperature output I can use it to cut the motor when riding or cut charging as needed
If you know of any other safety features provided by the battery charger I'll give it some thought and maybe figure out a way to include them

Still planning on charging in a tin box in a brick garage so if it goes up it doesn't matter :)

Safe after two failures was acceptable for ATEX safety ratings inside a fuel tank (zone 0) if I remember right, so good enough for my bike I reckon.
I'd have a failure of charger, failure of BMS and failure of tin box to risk the garage which is not a risk to people

This has been discussed to death -- check my post history.
Even Chinese chargers come with:
1) severe overdischarged halt
2) overdischarged low-current pre-charge
3) charge termination

CC/CV is not a safety feature -- that's just characteristic of CC/CV PSUs and chargers alike.

Blind and unmonitored voltage and current variability with unknown environmental coefficients is a liability, not a feature.

Just get another charger if you build a different pack. They're cheap and that way you can use them both.
Thanks for getting back to me, It sounds as though the chargers you're suggesting are better than some of the ebay rubbish I've come accross. I shouldn't say chinese as being rubbish. Some are good/
I'll do some more reading on your post history - any in particular?
Current and voltage are output via bluetooth so not unmonitored if it's worth doing so. I was planning to be lazy but can look at this if needed
What environmental coefficients should I look at? do you mean values such as temperature?
 
fatty said:
BobBob said:
Grin Simulator not stimulator, and I didn't mean it that way, not judging though :)
Motor part number is in first post, including the values used in the Simulation, data also in first post, Don't know KV, think 4T

There's too much in this wall of text to correct individually, but if you break it out into specific, individual questions, you'll get better feedback from technical users like myself.

Here's my general advice: the Grin Simulator is like cheating on your homework with Google Translate or Wolfram Alpha. It's a great (even invaluable) tool to learn from and a lot of people can skate by using it, like 48V or 52V or even 72V, 750W or 1500W or even 3000W systems.

But if you're pushing the envelope into dangerous voltage and power, you need to speak the language natively -- because when you don't, you can't identify errors. Hint: does the 4504 in the simulator have the same kV as your motor?
That was one of my first questions "is the 4504 the right model of the 3 KV V3 MXUS"
fatty said:
There's a reason automotive electrical systems are still on 12V, and 48V is perpetually a decade away -- because high voltage is dangerous, and it kills. EVs with high voltage take severe precautions to protect occupants from it.

So throwing out these silly numbers at the wall to see if they stick, like 8kW through a MXUS 3k or 150V without knowing your motor kV -- it's just that, silly. You need a hard sanity check.
First line read "So I'm after a sanity check" seems sensible (previous quesion was what is best maths model for the motor)
This is perhaps what I would consider a soft sanity check, little cost, discussion of approach, thow ideas around and see what sticks
Thanks for helping
fatty said:
You should be able to explain why a 4T is a poor choice for a 150V system, and (related) why higher volts at lower PWM is inferior to matching voltage to desired speed (another hint: you have a fixed number of cells).
Why should I be able to explain why the Grin Simulator is wrong and why shouldn't I need help and advice?
The motor is bought so the choice is for the battery, IE the quesion is "for a 4T what is the best controller and battery?"
I explained why I thought it appeared to be a good choice with some pros and cons and with examples and simulation / graphs demonstrating and supporting this.
I also went into some of the theory behind this
The simulated data appears to show that the 150V system is the best match, for my application and for that motor.
Delighted if you knock holes in it. I am happy to be wrong if I learn something
The last line was "what's wrong with this model" I don't know, seems right, hence the question

I'll have a better look at your hints tomorrow and will try to come back with a better solution than the Grin simulator provides
Will give it a go, might have to remind myself how to use Wolfram Alpha
Thank you for taking your time to look at this
 
Back
Top