[telefonica]

Hello all! I'm new here and this is my first post. I'm thinking about doing a motorcycle conversion and have my eye on converting a Kawi KE100. I'm trained as an electrical engineer so I have ~ok base knowledge.

My question is this: I see a lot of people connecting batteries in series to get the required voltage, but why not connect in parallel and use a power converter to boost to the needed voltage? IMO this would allow more current draw and you wouldn't have to add batteries in sets of 4, 6, etc or whatever your voltage requirement is.

Thanks!

 

[rwaudio]

Hello all! I'm new here and this is my first post. I'm thinking about doing a motorcycle conversion and have my eye on converting a Kawi KE100. I'm trained as an electrical engineer so I have ~ok base knowledge.

My question is this: I see a lot of people connecting batteries in series to get the required voltage, but why not connect in parallel and use a power converter to boost to the needed voltage? IMO this would allow more current draw and you wouldn't have to add batteries in sets of 4, 6, etc or whatever your voltage requirement is.

Thanks!

I think it comes down to efficiency and practicality, higher voltage allows lower current which means smaller cables, there is only so much current that is practical before the losses or cost of parts becomes outrageous. You would also have another power conversion which might be 85-90% efficient requiring that much more battery power to get the same output power. Batteries are usually considered the weak link, so using them in the most efficient way is important.

 

[DJBecker]

The current required for a series connection is already very high. With a parallel battery connection, the cables would be massive and an impossibly large inductor would be needed for the boost converter.

We already have a inherent inductor in the motor windings when we use a buck converter converter configuration.

 

[Sunking]

Weight, efficiency, and cost.

The lower you go in voltage will add weight and cost, and lower efficiency. All of which are hurdles for an EV. You want to run as high of voltage as possible.

 

[telefonica]

All good points, but allow me to play devil's advocate:

1. Take the hypothetical that I can only add batteries in series of 6 because I require 72V (assuming 12V batteries). Now, let's assume that I have the space to fit 11 batteries - one shy of running two sets of batteries. In this instance, even with an 85% efficient boost converter, would not running 11 batteries in parallel and a boost converter to 72V be more beneficial than only 6 batteries in series?

2. Yes, the current demands on the low side of the converter would be great. According to my math, a 30kW peaking motor would require 2,500 amps at the low side of the converter (again, let's assume we're using 12V batteries; not including any efficiency calcs.). This massive current, however, is the sum of all of the current provided by the batteries. If I were to use 20 batteries, I would only need 125 amps per battery and thus 1 AWG to bring each one back to the boost converter - thus, solving the problem.

3. I disagree that it would require an impossibly large inductor. I've personally designed and built buck/boost converters that have transmitted 750 kW of power. The whole thing fit in a box no larger than 3 ft x 3 ft x 1 ft

4. I agree with the cost argument, but to the weight, efficiency, etc. arguments, I go back to my first point: if I could almost double the batteries, I think it would be worth it.

 

[Sunking]

1. Take the hypothetical that I can only add batteries in series of 6 because I require 72V (assuming 12V batteries). Now, let's assume that I have the space to fit 11 batteries - one shy of running two sets of batteries. In this instance, even with an 85% efficient boost converter, would not running 11 batteries in parallel and a boost converter to 72V be more beneficial than only 6 batteries in series?

Not IMO because that would tell me it was a poor design and used the wrong type/size of battery. Parallel strings are difficult to manage charge and discharge equalization. Besides how on earth are you going to arrange 11 batteries since 11 is a prime number? 6 in parallel with 5 will not work.

 

[tomofreno]

In this instance, even with an 85% efficient boost converter, would not running 11 batteries in parallel and a boost converter to 72V be more beneficial than only 6 batteries in series?

Seems it would. Say 100 Ah batteries, then 11 in parallel is 1100 Ah. Boosted by a factor of 6 in voltage I expect means about a factor of 6 higher current draw from them compared to the series case, effectively making them like 1100/6 = 183 Ah capacity. If the converter is 85% efficient, they effectively would have 156 Ah capacity, compared to the 100 Ah capacity of the series pack. Sounds like you have good experience in this area. Maybe you could get some old batteries to try a cheaper, scaled down version to do some tests for proof of concept and compare capacities for parallel and series? Of course if you had room for one more battery, you could run them 2p6s and have 72V/200Ah, a lot simpler.

 

[Qer]

1. Take the hypothetical that I can only add batteries in series of 6 because I require 72V (assuming 12V batteries). Now, let's assume that I have the space to fit 11 batteries - one shy of running two sets of batteries. In this instance, even with an 85% efficient boost converter, would not running 11 batteries in parallel and a boost converter to 72V be more beneficial than only 6 batteries in series?

Why would you have to limit yourself to mere 72 Volt? Why not go for 132 Volt with 11 batteries in series? If you're limited to exactly 72 Volt it means that you've already specified what controller you're going to use and then you're stuck with 6 batteries in series, if you're gonna change that to 12 Volt you can as easily change it to 132 Volt. Or rather, it'd be easier since then there's off the shelf controllers that already handle that voltage.

2. Yes, the current demands on the low side of the converter would be great. According to my math, a 30kW peaking motor would require 2,500 amps at the low side of the converter (again, let's assume we're using 12V batteries; not including any efficiency calcs.). This massive current, however, is the sum of all of the current provided by the batteries. If I were to use 20 batteries, I would only need 125 amps per battery and thus 1 AWG to bring each one back to the boost converter - thus, solving the problem.

You forget that you need transistors, contactors, capacitors etc that can handle 2500 Ampere too. There's a reason controllers that can handle 4 digit number currents are pretty expensive, switching 2500 Amps is gonna cost a buck or two...

3. I disagree that it would require an impossibly large inductor. I've personally designed and built buck/boost converters that have transmitted 750 kW of power. The whole thing fit in a box no larger than 3 ft x 3 ft x 1 ft

At what current? The current will be your bitch here.

4. I agree with the cost argument, but to the weight, efficiency, etc. arguments, I go back to my first point: if I could almost double the batteries, I think it would be worth it.

And again; why not simply add as many batteries as you can fit in series and be done with it? All the advantages, none of the disadvantages.

 

[major]

My question is this: I see a lot of people connecting batteries in series to get the required voltage, but why not connect in parallel and use a power converter to boost to the needed voltage?

Hi tele,

When using a buck converter as a motor controller one can get very smooth low speed control. If you were to use a boost converter for a motor controller, wouldn't you start the motor at a set voltage, ie the source? For your example, 12 volts. This would be difficult to drive.

I realize that problem could be solved by going to a buck-boost converter. But then you just doubled the controller cost, which typically is a high ticket item in conversions anyway.

This idea has been kicked around for years. Even to the point of a 3.2 volt battery having all the Lithium cells in parallel. Makes the BMS simple

I think there have been a few attempts to actually do a boost converter motor controller, with less than satisfactory results. But hey, give it a go

major

 

[telefonica]

6 in parallel with 5 will not work.

All 11 would be in parallel.

Seems it would.

Thanks for doing the math. I think this shows that in theory it is beneficial.

You forget that you need transistors, contactors, capacitors etc that can handle 2500 Ampere too. There's a reason controllers that can handle 4 digit number currents are pretty expensive, switching 2500 Amps is gonna cost a buck or two...

At what current? The current will be your bitch here.

Cost is one thing, I agree, but all the parts can be purchased; they are out there.

In my 750 kW experience, we were working with 800V, so the current was significantly less than with my EV example. IIRC, the parts cost us somewhere around $75k.

Why would you have to limit yourself to mere 72 Volt? Why not go for 132 Volt with 11 batteries in series?....

And again; why not simply add as many batteries as you can fit in series and be done with it? All the advantages, none of the disadvantages.

Good point - I think this is the argument killer. If everything is well thought out up front, one shouldn't find oneself in the position to have to make a parallel string of batteries. I hadn't quite understood why people chose their voltage; I figured there was some sort of limit. If it's possible that you can simply keep adding batteries in series, then that solves everything!