144v or 72v for motorcycle?

[Sunder]

Almost 6 months ago, I started a project to convert a Ninja 250R to a Ninja 230H (The H is for Hectowatt ).

Back then, I built the batteries to 3 x 48v for a 144v system. I figure with a 100A sustainable current with higher peaks, going 72V wouldn't be that smart.

The parts are available at 144v - just. And they dont seem to be that much more expensive.

But there seems to be so much more choice at 72v, and i read some of the race bike builders here use 72 or 96v.

I am wondering what the benefits of a lower voltage/higher amperage is, and whether I should take the effort to rebuild my packs.

I am hoping after 6 months of hell, and almost complete absence from this forum, I can get back into this project, but would love some opinions to get me back into the game.

Thanks.

 

[Dauntless]

Hopefully hell is no longer in session. So are you only hoping to have people throw out ideas? I don't picture you needing help making a decision such as this. I'm one who says when you have a perfectly good direction, just keep going. What do you really see as the advantage of switching to the 72V?

 

[liveforphysics]

You would get a performance and efficiency improvement from going to 100V mosfets over 200V mosfets, but if you already have the battery and parts to do 150Vdc, and you understand the balancing and corrosion mitigation needs, may as well do 150Vdc.

 

[LockH]

I vote for more amps.

 

[Philaphlous]

Definitely 144V. You'll get a much higher top speed. With 72v and that heavy of a motorcycle...you'll be pushing insane amps. 144V is going to be much better. What's a Tesla? Like 360v or something like that...

 

[billvon]

But there seems to be so much more choice at 72v, and i read some of the race bike builders here use 72 or 96v.

I am wondering what the benefits of a lower voltage/higher amperage is, and whether I should take the effort to rebuild my packs.

Only benefit is what you mentioned - availability of ebike type parts. But other than that, 144V is the way to go, from the point of view of power, conductor size, cost of protection etc.

 

[nicobie]

and you understand the balancing and corrosion mitigation needs, may as well do 150Vdc.

Listen to LFP.

Myself I'm liking 100V (24s hobby king or similar).

 

[Sunder]

Thanks guys:

@Dauntless. Thanks for the encouragement. I'm in the eye of a storm. I've taken 5 weeks leave to get away from it a bit, but I know what's waiting for me mid Jan. It's a bit of first world problem. I'm not starving, not divorcing, or anything like that, but it's sucked all the joy out of my life for a long time now.

@Liveforphysics. Thanks for the feedback. I think you were one of the people who said you prefer lower voltages and monster currents, and I was hoping to hear from you. Thing is, I only have a few of the cheaper parts - Cells, BMSes, Chargers. (60S LTO cell BMSes are impossible to come by, so I managed them as 3 x 20S protected charge, and 1 x 60S unprotected discharge, knowing that I'd either have to take a risk with regen, or disable it).

In terms of efficiency/performance loss, what are we talking about? 5%? 10%? More? The way you've phrased it, I'm getting the feeling it's not that big a deal, so leaning towards finishing the project as designed.

Also, what special corrosion control is required at 144v but not at lower voltages?

@lockh @Philaphlous @billvon @nicobie thanks for your opinions. I am leaning towards staying at 144V and finish the project as I originally intended, unless LFP says something like I'm going to take a 20%+ efficiency hit. I've estimated this build to give me a 100-120km range, which in the last 3 years, I've only needed one mid-trip top up, so I think this is perfect for me. If I were to go to 80km, about 1/4 of my trips would need a mid-trip top up.

 

[amberwolf]

(60S LTO cell BMSes are impossible to come by, so I managed them as 3 x 20S protected charge, and 1 x 60S unprotected discharge, knowing that I'd either have to take a risk with regen, or disable it).

FWIW, if the 20s BMSs still have the discharge protection (LVC/etc) but aren't used because of the limits of the voltage the output FETs can handle, then you could still use that protection by wiring up the outputs of eac of the BMSs to an AND gate taht requires all of them to have their output-enable lines engaged to let the gate turn on a contactor (or other switching device) that enables power on the bike/etc. (could be something as simple as a relay or FET that turns on the ignition/keyswitch line of the controller, if you don't need it to physically disconnect the battery from anything).

 

[liveforphysics]

The hit is small, just the power density difference in 200Vdc mosfet dies to 100Vdc mosfet dies. It's going to be single digit% at most. The other aspect is cell string having few cells to manage and corrosion effects being milder and shocks being milder.

 

[Sunder]

FWIW, if the 20s BMSs still have the discharge protection (LVC/etc) but aren't used because of the limits of the voltage the output FETs can handle, then you could still use that protection by wiring up the outputs of eac of the BMSs to an AND gate

I had considered that. FWIW, I think with LTO it's not worth it. They don't suffer damage as badly as LiPo or LiFe from high current discharge, so I figure a fuse will be good enough to prevent shorts, and the controller programming should "In theory" be good enough (Yeah, I know, temperature moisture, and failures will cause issues, but I think the risk is low enough. If we got below 0*C in Sydney, it might be worth reconsidering, but 10-40*C, is a fairly balanced temperature range).

I'm also less of a fan of discharge protection after I overtook a truck, pulled back in, pulled on the throttle a little harder to put a bit more space between me and the truck.... And the BMS decided to cut to protect the cells. Damn the cells, I had 10 Tonnes coming up on me fast!

The hit is small, just the power density difference in 200Vdc mosfet dies to 100Vdc mosfet dies. It's going to be single digit% at most. The other aspect is cell string having few cells to manage and corrosion effects being milder and shocks being milder.

Thanks mate. I'm sticking with the 144v then. As for shocks being milder... Good reminder: I've shorted 100v before. I was so damn lucky. Holes in shirt, scorch marks in the work bench. I wasn't wearing either gloves or safety glasses, but nothing on my hands or face. Not making that mistake again:



I'm away on holidays at the moment. I'm hoping by the time I go back to work mid Jan, I will have a finished and managed battery, and have ordered the controller and dash first. I am going to rig it up to a small 3kw motor first, and make sure that the signals from the original motorcycle can be used with the controller (E.g. the brake sensor can drive a moderate level of regen), and vice versa. (Hall sensor can drive the speedo).

Controller I was planning on getting was a Kelly 144v 600A (Phase amp) controller:
http://kellycontroller.com/kls96601-8080ips24v-96v600asinusoidal-bldc-motor-controller-p-1386.html

Once I am confident it will all work, I will order a QS 273 50H 8kw continuous, 21kw peak motor, and inject some Statorade. Targeting 23kw peak (Or jokingly, 230 Hectorwatts), since this will keep me under the OEM, and make it easier to pass registration. Area under the curve, obviously, looks extremely different.

http://www.cnqsmotor.com/en/article_read_836.html

 

[amberwolf]

I'm also less of a fan of discharge protection after I overtook a truck, pulled back in, pulled on the throttle a little harder to put a bit more space between me and the truck.... And the BMS decided to cut to protect the cells. Damn the cells, I had 10 Tonnes coming up on me fast!

Yes, this is a problem with any form of limiting on a vehicle, and why I have never used (until now) the speed limiting/etc on the Cycle Analyst. (and had problems with testing the Fusin 1000w kit on a regular bike becuase it had speed limiting that just cut power.)

On my SB Cruiser I run without a BMS at all, because so far the EIG NMC cell type I"ve been using for several years on CrazyBike2 and now SB Cruiser seem to be stable and very consistent in quality and properties between cells, so I havent' had imbalance problems that would require a BMS to ensure correct operation.

But if I had one, I'd also have an override "button" that I could easily squeeze on the grips that would engage a bypass (via relays, or whatever) of *all* limiting of any kind--BMS LVC/etc., controller or CycleAnalyst current or speed limits, etc. That way in an emergency I could safely continue to throttle up if necessary, to get the heck out of the way of...whatever.

(this is planned for the trike now that I'm using the CA for PAS control of it, with various limits in play).