Theoretical power Vs Actual power

[Blunderbuss]

I have been doing some checks on what the potential power of system should be and what the display reads:

Build 1 = 13s6p – 2200MaH cells – 20a BMS – 20a controller – 350W 36V Bafang hub motor.
Theoretical – 52v off the charger = 1040W (52vx20a)
Actual reading on the LCD was 820w

Build 2 = 13s6p – 3000MaH cells – 35a BMS – 35a controller – 500W 36V Bafang hub motor.
Theoretical – 52v off the charger = 1820W (52vx35a)
Actual reading on the LCD was 1580w

All my connections are soldered or XTC60’s, I keep the wires as short as possible and I use a fatter wire than what is needed. Nothing is getting hot (or even warm) after a good few miles so it may be that the inefficiencies are just that.

Does anyone know if the actual readings I am getting are “par for the course”?

 

[wheeliepete]

Are you testing peak watts with WOT? Maybe your display is inaccurate. You could try wiring in a cheap watt meter between battery and controller and compare the results. I've seen 1750w peak draw with a 52v, 58.8v off the charger, battery with a 30 amp controller with one of these, which is about right.

https://www.ebay.co.uk/itm/12V-150A-Digital-Volt-Watt-Meter-Car-Caravan-4x4-Camping-Solar-Power-Voltmeter/132681383982?hash=item1ee46d402e:g:dh4AAOSwiu5bNMLC:rk:13:pf:0

 

[anotherkiwi]

Is this just WOT or with load on the rear wheel? The LCD reading looks like real Watts delivered to the rear wheel i.e. -20%

 

[Blunderbuss]

Is this just WOT or with load on the rear wheel? The LCD reading looks like real Watts delivered to the rear wheel i.e. -20%

It's the reading from the LCD (KT-lcd6) with load - full throttle from standing start

 

[wheeliepete]

The LCD reading looks like real Watts delivered to the rear wheel i.e. -20%

Good point AK, not sure what reading the KT display gives.

 

[anotherkiwi]

Mine displays Controller Watts and is spot on when compared to the Wattmeter I have inline between the battery and the controller.

 

[Blunderbuss]

Mine displays Controller Watts and is spot on when compared to the Wattmeter I have inline between the battery and the controller.

OK, so first a Wattmeter to check/confirm the Watts (that is now on order), then I need to understand the -20% bit? Is that linked to voltage sag once the system is under load?

 

[anotherkiwi]

No losses - lost in the wiring, motor efficiency etc. etc. Former member tech guru said he estimated it at 25% but between 20 and 25% is about right.

 

[Blunderbuss]

No losses - lost in the wiring, motor efficiency etc. etc. Former member tech guru said he estimated it at 25% but between 20 and 25% is about right.

Go it, thanks. Build 1 is about 26% and Build 2 about 16% on the figures from the LCD - awaiting the delivery of the Wattmetter to confirm

 

[sjpt]

Efficiency varies a lot with speed. https://www.ebikes.ca/tools/simulator.html has lots of measurements on different motors/batteries etc under different conditions.

 

[Nealh]

Difference on lcd and actual controller watts output is as said due to efficacy /watts waste from hub,battery and wiring.
Any wring that gets warm is inefficient and too small in mm2 size, this is bottle necking power, thicker wire is the answer.
Wasted watts can also be lost from battery series interconnectors (bus bars) to thin, to narrow and not enough series connections cause poor amp flow and watts wasted.
This is often pretty typical of China made batteries sold by reputable sellers of parts and complete bikes, often series bus bars are not capable of carrying the full series amp load with out heat loss/ wasted watts.
Typical optimal ampacity of nickel is 4.5a 1mm2 or 6.6a is acceptable with some warming, 0.15mm x 7.5mm nickel bus bar is 1.125 mm2 . O.K for crappy mediocre 4- 5a rated cells but not for 10a cells or more, often the nickel isn't beefed up enough so cells aren't being optimally used. The answer ( for those who know how) is to solder/add 0.8/1mm copper wire to each of the existing series bus bar connectors, this adds superior ampacity flow across series connections making better use of power flow. Another area of battery building to look at is the pack main supply wires often they terminate at from the first and last cell, better power delivery would be if the wiring picked up off every cell in the first and last parallel group.

 

[Blunderbuss]

Typical optimal ampacity of nickel is 4.5a 1mm2 or 6.6a is acceptable with some warming, 0.15mm x 7.5mm nickel bus bar is 1.125 mm2 . O.K for crappy mediocre 4- 5a rated cells but not for 10a cells or more, often the nickel isn't beefed up enough so cells aren't being optimally used. The answer ( for those who know how) is to solder/add 0.8/1mm copper wire to each of the existing series bus bar connectors, this adds superior ampacity flow across series connections making better use of power flow. Another area of battery building to look at is the pack main supply wires often they terminate at from the first and last cell, better power delivery would be if the wiring picked up off every cell in the first and last parallel group.

Excellent bit of info - Build 1 has 7.5mm x 0.15, Build 2 has 12mm x 0.15 Nickel. Also for build 2 I soldered the main supply wires (45amp rated wire) down the whole nickel strip (so covering all six cells) before spot welding the strip onto the cells. It does not look the prettiest thing but if it has helped in a move from 26% to 16% loss then I am happy with that.

 

[Nealh]

Good to hear Bb, build 2 nickel bus's will have ampacity of 1.8mm2 so each series one can handle 8a optimal or 11.9a which is acceptable. Some times soldering isn't always pretty to look at but sounds like you are for sure improving the battery ampacity supply.
Soldering all the way along the final P strip is the way to go before spot welding to the cells, H strip with slots is good to use.
I solder the supply wire to one side then bend the H strip i90 degrees before spot welding to cell ends, on the V+ take off insulate the V- cell case against the nickel to prevent shorting if the cell sleeve insulation fails. a layer or two of Kapton tape works well or a small strip of thin glass fibre sheet.

 

[anotherkiwi]

The answer ( for those who know how) is to solder/add 0.8/1mm copper wire to each of the existing series bus bar connectors, this adds superior ampacity flow across series connections making better use of power flow.

Do you have a link to a HOWTO or a pic Neal?

 

[Nealh]

Best way to build packs for optimal best current share is to use even number of cells like you have for P groups 4, 6 or 8 etc.
That way you don't have to solder or spot weld series strips to cell ends, only the P strips need spot welding and if necessary in two layers to increase ampacity flow side ways. Then for series bus's either spot weld two layer of nickel between each pair of cells on to the P bus strip or solder copper wire for better power sharing. That way on a 6p build each series of cells only needs three bus's, each series bus will draw power only from a pair of cells and will draw equally from each flowing directly to the next group down the line. Another key for equal power sharing is to keep any bus or wire length exactly the same size.

 

[Nealh]

I book marked these a couple or years ago from ES and always use them as a guide.
Well worth reading the threads from start to finish with pics and illustrations.

Common pack design mistakes.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=84412

ultimate to spot welding.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=68005

 

[Blunderbuss]

Good to hear Bb, build 2 nickel bus's will have ampacity of 1.8mm2 so each series one can handle 8a optimal or 11.9a which is acceptable. Some times soldering isn't always pretty to look at but sounds like you are for sure improving the battery ampacity supply.
Soldering all the way along the final P strip is the way to go before spot welding to the cells, H strip with slots is good to use.
I solder the supply wire to one side then bend the H strip i90 degrees before spot welding to cell ends, on the V+ take off insulate the V- cell case against the nickel to prevent shorting if the cell sleeve insulation fails. a layer or two of Kapton tape works well or a small strip of thin glass fibre sheet.

I spent ages trying to bend a crease 2mm in from the edge of a 10mm wide nickel strip so i could lay the wire into it and solder it - i never thought of using the H strip!

 

[anotherkiwi]

I spent ages trying to bend a crease 2mm in from the edge of a 10mm wide nickel strip so i could lay the wire into it and solder it - i never thought of using the H strip!

Drill holes near the edge of 10 mm strip, lace the wire through the holes, flatten with hammer, hold down with clothes pegs during soldering.

 

[Nealh]

I tin/pre-solder the nickel and the stripped bare wire, then add heat and extra solder to aid flow of solder. Using a wider flat chisel type tip is better for heat distribution then a silly pointy one.

 

[Blunderbuss]

I did the same approach as Nealh, hoping that the crease/bend i put in the strip would act as both a little "reservoir" for the solder and hold the wire in place - it has provided a very good joint but was so fiddly i won't do it again. Both your ways look much better