Help! Switching between 2 batteries

JermaineW93

Just Joined
Oct 17, 2019
1
0
Hi all!

I have recently completed my own e-MTB conversion and I have both a rack battery and a downtube battery. This is necessary as I am a courier for a living and do 8-10 hour shifts, sometimes more. The regular switching between batteries is leading to quite a fast degradation in the wires/terminals between batteries and the controller.

I am wondering whether there is some kind of a device or way to keep both batteries’ wires plugged in to the controller, but to be able to flick between which is active? I am guessing I need something which would be able to ensure that the battery that is off does not loop in with the rest of the circuit?

I am a rookie with electricals so any advice at all would be great and much appreciated.
I have included a picture of the terminals that I currently use.

Thank you,

Jermaine. 32490
 

Nealh

Esteemed Pedelecer
Aug 7, 2014
8,181
3,096
56
West Sx RH
Wire them in parallel and have one big ah/wh battery.
 

vfr400

Esteemed Pedelecer
Jun 12, 2011
2,858
827
Basildon
You can only bring the two batteries in parallel when they're at the same measured voltage +/- 0.5V. You must disconnect them when charging.

The batteries won't spark and destroy the connector if they're switched off when you connect to the controller.
 

Sturmey

Pedelecer
Jan 26, 2018
82
35
63
Ireland
I have used a 40 amp dual Schottky Diode (20 + 20 amp) mounted on aluminium (approx 2 x 1 inch) and put in small plastic box for the last year and it seems to work well. It means that the batteries are isolated from each other which is useful in my case as one battery charges to 42 volt while the other is set up for 41 volt, so I never really never have to worry about battery voltage etc. I have the same fitted now to second bike and have a portable battery that I can swap between bikes.
There is a small penalty of about half a volt lost. However this is more than made up for with using parallel batteries (with or without diodes) as the sag is far less and batteries run cooler. I have actually fitted three batteries in parallel using diodes with no problem (other than battery weight) for a long trip.
Anyhow the diode I used is listed below which seem suitable as controller is 15 amps .
https://www.ebay.ie/itm/Vishay-MBR4045PT-E3-45-Dual-Switching-Schottky-Diode-Common-Cathode-45V-40A/173385936086?hash=item285e9b6cd6:g:v8MAAOSw~jpbNlOW
 
Last edited:
  • Agree
  • Like
Reactions: ebiker99 and Woosh

ebiker99

Esteemed Pedelecer
Feb 17, 2019
510
191
The disadvantage of connecting 2 batteries in parallel is that if they're not exactly matched in voltage and discharge characteristics one battery can be supplying charge to the other. This is wasteful of energy since there will be a heat loss in the process.
With Bosch ebikes that use 2 batteries the controller alternates between them, when one is a few volts below the other it switches and similarly when charging on the ebike.
This avoids the problems of using them in parallel and also helps balance the wear on the batteries.
 

Woosh

Trade Member
May 19, 2012
11,962
9,116
Southend on Sea
wooshbikes.co.uk
The disadvantage of connecting 2 batteries in parallel is that if they're not exactly matched in voltage and discharge characteristics one battery can be supplying charge to the other. This is wasteful of energy since there will be a heat loss in the process.
With Bosch ebikes that use 2 batteries the controller alternates between them, when one is a few volts below the other it switches and similarly when charging on the ebike.
This avoids the problems of using them in parallel and also helps balance the wear on the batteries.
Sturmey's implementation is what most people would use, and I suspect that Bosch e-bikes do too.
The Schottky diodes ensure that the batteries can only discharge, never getting reverse charged.
The forward voltage drop when each branch conducts is about 0.2V-0.4V, the heat loss is much less than 1%, about 1W on average when you ride.
 

Sturmey

Pedelecer
Jan 26, 2018
82
35
63
Ireland
The disadvantage of connecting 2 batteries in parallel is that if they're not exactly matched in voltage and discharge characteristics one battery can be supplying charge to the other. This is wasteful of energy since there will be a heat loss in the process. .....
The diodes act as one way valves and prevent flow of (parasitic?) current between one battery to another .
Batteries heat in large part due to internal resistance and this heat is proportional to the square of the current. Two similar batteries in parallel can potentially reduce the heat in each battery by 75% and the overall heat loss and sag by 50%.
 
Last edited:
  • Agree
Reactions: ebiker99

ebiker99

Esteemed Pedelecer
Feb 17, 2019
510
191
Sturmey's implementation is what most people would use, and I suspect hat Bosch e-bikes do too.
The Schottky diodes ensure that the batteries can only discharge, never getting reverse charged.
The forward voltage when each branch conducts is about 0.2V-0.4V, the heat loss is much less than 1%, about 1W on average when you ride.
No, the Bosch system switches the batteries in and out as they discharge and charge, I guess they're using SCRs or FETs to do the switching.
Sturmey's approach is the right way to go for people who want to add a second battery to their ebike.
 

Benjahmin

Esteemed Pedelecer
Nov 10, 2014
1,426
888
65
West Wales
I use one rack battery and a soft shell in a pannier. I charge them seperately and connect them together with a Y lead that I made up with xt60 connectors. Each battery voltage is ALWAYS measured before cross connection, at a max difference of 0.5v there is no spark. Connection to the controller does cause a spark but it occurs at the tip of the connector not where they end up being connectrd, there is no degradation after about 3000 miles.
As said it is kinder to the batteries because current draw is shared. I get no noticeable sag on the steep hills around here.
I used decent guage silicone cable to make the Y lead and kept the arms the same length to keep resistance the same.
The two batteries are supplying the same load and, having measured voltages at the end of the ride, do not seem to go unbalanced. I would think that if either battery was even slightly higher than the other (say 0.1v) then there would be a resulting slightly higher current from that battery until balance - probably instantaneous - under load it's a self balancing system surely. The only opportunity for reverse flow is if they are at different voltages when first connected together.
 

Woosh

Trade Member
May 19, 2012
11,962
9,116
Southend on Sea
wooshbikes.co.uk
No, the Bosch system switches the batteries in and out as they discharge and charge, I guess they're using SCRs or FETs to do the switching.
Sturmey's approach is the right way to go for people who want to add a second battery to their ebike.
they may have used FETs for switchung but the junction loss on FET is about the same as on a Schottky diode, about 0.2V. FETs present no performance advantage, only extra flexibility.
With the Schottky diodes, the power is drawn from both batteries, the voltage drop increases with current intensity.
the one with slightly higher voltage supplies a bit more than the other, that spreads the load and reduces sag.
 

vfr400

Esteemed Pedelecer
Jun 12, 2011
2,858
827
Basildon
That reminds me, here's a fact that many don't know. Even geared hub-motors with clutches regenerate a small amount of charge back into the battery. When the controller cuts power when you stop pedalling or let go of the throttle, the motor spins on for a bit because it has momentum (kinetic energy). While it slows down to zero rpm, that energy is transferred to the battery via regeneration.
 

Nealh

Esteemed Pedelecer
Aug 7, 2014
8,181
3,096
56
West Sx RH
For sparking I use a parallel Y lead using XT90 anti spark connectors, would be nice if they also did this to the XT60.

Like Benjamin I use batteries of differing capacity and now and then during a ride I check them during a stop, voltages are usually within 0.1v.
 

Woosh

Trade Member
May 19, 2012
11,962
9,116
Southend on Sea
wooshbikes.co.uk
When the controller cuts power when you stop pedalling or let go of the throttle, the motor spins on for a bit because it has momentum (kinetic energy). While it slows down to zero rpm, that energy is transferred to the battery via regeneration.
the FETs will be off when the controller cuts power though.
They need to be on for regen.
 

ebiker99

Esteemed Pedelecer
Feb 17, 2019
510
191
they may have used FETs for switchung but the junction loss on FET is about the same as on a Schottky diode, about 0.2V. FETs present no performance advantage, only extra flexibility.
With the Schottky diodes, the power is drawn from both batteries, the voltage drop increases with current intensity.
the one with slightly higher voltage supplies a bit more than the other, that spreads the load and reduces sag.
I can see from my Kiox display when riding my bike that the capacity on one battery won't change until that that on the other battery drops below 5% of it at which point it will be switched in and the other battery switched out.
FETs allow the battery to be isolated whereas diodes don't and for sure the batteries are not active at the same time.
 
  • Agree
Reactions: Woosh

Sturmey

Pedelecer
Jan 26, 2018
82
35
63
Ireland
the FETs will be off when the controller cuts power though.
They need to be on for regen.
The mosfet often contain 'body diodes' which often work act as flyback diodes and allow reverse currents (even when mosfet is turned off) and you can get a bridge rectifer effect. But the voltage would not normally exceed the battery voltage ( and hence no current would flow in this case) until you go beyond the normal max rpm(RPM/volt).
https://en.wikipedia.org/wiki/Power_MOSFET#Body_diode
 
Last edited:
  • Like
Reactions: Woosh

vfr400

Esteemed Pedelecer
Jun 12, 2011
2,858
827
Basildon
If you put a wattmeter between the controller and battery, you can see the effect of the regen. It's measurable, regardless of what any theory says. That's how I discovered it.
 
Last edited:

Woosh

Trade Member
May 19, 2012
11,962
9,116
Southend on Sea
wooshbikes.co.uk
If you put a wattmeter between the controller and battery, you can see the effect of the regen. It's measurable, regardless of what any theory says.
this is what Sturmey talked about in the previous post about body diode effect that can conduct current in reverse when the FET is turned off. After you turn off the FETs, a small amount of energy left in the U, V, W coils will try to dissipate - they can enter into adiabatic oscillations as there is only a 10 kOhms damping resistor across the FETs when the FETs are turned off.
it's totally conceivable that the voltage from the coils exceeds momentarily the battery's voltage. It's the same phenomenon that is used in boost converter except in the latter case, the oscillations are maintained.
 
  • Informative
Reactions: danielrlee