Spare tank

[llile]

Range anxiety trigger warning! Yep, Range Anxiety is a big deal. I have a battery that can go about 40 miles, and I regularly go 43 to 45 miles. Somewhere in between I have to beg, borrow or steal a few amp-hours worth of juice to get home. Usually I am downtown halfway through the trip, and there are a lot of plugs around. Guarded by officious assholes that don't want anyone to charge.

Now this has been no problem for a while, but recently several of the places I used to charge on the regular route have dried up. Some officious jerk shut off the free juice in a public park. Other places are the same story, either the outlets fail (outdoor GFI outets actually do fail a lot, cheap junk) or my permission to get a sip of juice has evaporated.

Well what If I had a spare tank? Current battery is 40 amp-hour, 24S4P, 72 Nominal Volts and I run at 35 amps for the most part, and eat up an amp-hour per mile. It would be very expensive to carry two such 40AH batteries. Heavy too. But a smaller amp-hour rating, would not produce so much current?

I have a couple of 10AH 36V batteries. I could hook them in series and have a spare 72V tank. One of them is a Ping, most likely with specs of 20 amps continuous, the other I bought from a guy and have no idea, but it is probably about the same. 20 Amps/72V I could limp home, and I very well could just hook these batts in series (with a couple of Schottky diodes* just in case) and turn the cycle analyst down to 20 amps, limping home at half speed. Plan A: Switch to the 10AH battery and limp home.

But what if I hooked these batteries up to the charging port?
Well if the 40AH battery was mostly charged, nothing would happen - there would be no more voltage in the spare than in the main battery. No current flow.

I believe there isn't a problem with charging while driving, so as the main battery waned, the spare would slowly add some juice. Would that be like having another 10 AH? Probably not but it might just work. It is certainly the simplest scheme and I might just try hooking it up and running that way. If I go 45 miles and still have juice, it's a win. I'll make sure I have a sag.

But if the 40AH battery was more than 75% discharged, and the spare was full, one might get half the charge out of the spare battery before the voltage in the main battery rose so far as to eliminate any voltage difference. This might be a reasonable emergency scheme - the few times I've run short I've been less than 5 miles from my destination. Take a look at the attached graph for reference. Plan B: Hook right to the charging port. Simple but cheap.

OK, what if there was a DC-DC converter on the spare batt? Small DC-DC converters are actually surprisingly cheap these days - boosting 72V to 80V doesn't waste a whole lot of power or tax the DC-DC converter much. What if there was a little black box that took whatever voltage from the spare and boosted it to 80V or so? In that case, one could suck all the available juice out of the spare battery. The main battery, being nearly spent, is down around 3.0 or 3.2V per cell, would accept juice from an 80V source without any problems. No risk of overcharging, there isn't enough capacity in the spare battery. I'd even have fun building such a DC-DC converter, went through the design math already. Except for a chip from Maxim I probably have the rest of the parts here already. Plan C: Use a DC-DC converter.

Plan D: Replace the wimpy BMS in these two 10 AH batteries with good BMS's that can handle 35 amps. Switch to the spare battery in case of emergency. Simple, but costly.


What do folks think would any of these schemes work? The simplest one, just hooking the spare into the charging port, would that not give me an extra 10 AH capacity? Or even 75% of that?

*Why Schottkys? Why diodes at all? If you don't know ask and I'll give the longwinded explanation.

 

[amberwolf]

Just parallel a smaller capacity battery of the same chemistry and number of cells as and with your existing one, to increase it's capacity to sufficient to get you the range you need, plus a little extra for detours, bad winds, further aging of the pack, etc.

It's the simplest way to do it, both for charge and discharge.

No need to worry about whether the new pack can *by itself* handle the current, because it doesnt' have to. No need to switch over, etc etc.

It gets more complicated every other way than that.

 

[dogman dan]

Seems like you could just "limp home" on the 10 ah pack. bear in mind, 20 amps of 72v will get you to around 35 mph, it's 1500w. I wouldn't call that limping much.

Or parallel them at the mains, at least part of the ride. If the chemistry is dissimilar, you will have to juggle some when you connect, and make sure the pack that needs to stop at a higher voltage gets disconnected in time. Kind of a pain in the ass unless the chemistry is the same.

You can't just hook to the charge port when they are not at the same voltage. That would just send a huge surge through your bms. You'd need some kind of device to limit the amps going into your charge port, much like your controller limits current going into the motor.

But what I do when I'm similar situations, is just ride a bit slower the whole ride. This is why I love my cycleanalyst so much. I know to go X miles I need to hit an average of Y watt hours per mile. So on that 5-10 mile longer ride, just ride slower enough of the ride, to make sure you hit that magic number by the end of the ride. If your CA is the direct plug in type, you can also limit amps when you are on the weaker battery. VERY handy.

It's not range anxiety, once you know you will make it. Slow down some a bit, early in the ride. Once you ride 5 miles slow, you know you just doubled your range for that 5 miles. After that, you will know you will make it letting it rip. Instead of worrying at the end of the ride, you just take care of it early in the ride. Presto, no range anxiety the whole rest of the ride.

But if you are going to carry a reserve, this is where the RC lipo can shine. Four 5s 5ah packs of 30c lipo is still a pretty small, not too heavy bundle. And it can plug in when you need it, and pull 40 amps with no problem, giving you that 5 miles at full speed.

 

[llile]

Now about that efficiency when slowing down: Yeah, I understand the physics, but there's something going on here that didn't read the book apparently.

On my old bike, a two-wheel mid-drive, there was a very direct relationship between speed and efficiency. I'd normally use about 25 Wh/Mile, but I could dial that down predictably by slowing down. Top speed on that bike was about 22mph. At 12 mph I could see 15 wh/mile, predictably getting much better efficiency at lower speed limits. 60% better. 22^2/12^2 = 3.36

This current bike is a two-motor three wheeler hub motor tadpole, averages 80 WH/Mile/35A. Never knew why, but I have to slow way, way down to notice much better efficiency at all, and it never gets very low. At a CA limit of 20 MPH, I only dial down to 65 WH/Mile/25A. Only 20% better. 35^2/20^2 = 3.06, should be a similar square law reduction in friction. My buddy, also with a tadpole trike using my old mid-drive motor I sold him, is going the same speed at three times the efficiency right next to me on the trail. No difference in tires, weight, or aerodynamics.

Part of the issue is control. At those lower speeds, the Cycle Analyst loses good control of the powerful motors. Although at 35 mph, everything is pretty much flat-out, at 20MPH CA Speed Limit , the system surges - vrooming up from 19.5 mph to 20.5 mph then coasting down. It is annoying on the bike trails, which are legally limited to 20mph, vroom, coast, vroom, coast no matter where I set the throttle whenever I get close to the CA speed limit lock I've set. I think the efficiency is terrible at those conditions. I've tried dropping the amp limit, the throttle limit, and even messing about with some of the integral terms in the Cycle Analyst setup, but haven't come up with anything that makes a difference.

Part of the issue could be two motors - I can turn one of them off but it didn't make any difference and makes the bike steer crazy. I'm aware that two hub motors probably isn't the most efficient setup. Nothing like a mid-drive for efficiency, where you can always set the gears so the happy motor is at the top of the torque curve.

And, these are just powerful motors and their peak efficiency is way up the curve, at these speeds I'm way down on the worst end of the efficiency curve. Either way, from a practical standpoint slowing down doesn't gain me much efficiency, not as much as the physics text says it should. Yes, the wheels are aligned, yes, the bike coasts well,Yes the tires are pumped up nice and yes there isn't a brake rubbing or a dead body being dragged along behind it. My biddy's trike and mine tie in a coasting race, we've tested it. It's control or motor efficiency I'm convinced.

If I take off in the morning with a headwind I'm really needing that spare tank battery.

Well for now, with the same chemistry, I'll just hook the extra 10AH in parallel and quit messing around with crazy schemes.

Anyone have a good theory why my bike is flunking Physics Class?

 

[llile]

Seems like you could just "limp home" on the 10 ah pack. bear in mind, 20 amps of 72v will get you to around 35 mph, it's 1500w. I wouldn't call that limping much.

But if you are going to carry a reserve, this is where the RC lipo can shine. Four 5s 5ah packs of 30c lipo is still a pretty small, not too heavy bundle. And it can plug in when you need it, and pull 40 amps with no problem, giving you that 5 miles at full speed.

Well, 20A at 72V might go 35MPH on a two-wheeler, but this tadpole trike has more boots on the ground. I like the tadpole for a lot of reasons, not the least of which it can't tip over, but efficiency isn't one of them.


Hmmm. I keep wrestling with the idea of lipos, so far I've been a-feared. Maybe I'll get some bricks and build a charging shed.

 

[wineboyrider]

This is where a spare lipo range booster pack might have some huge advantages for you. They usually have a higher discharge rate and lower life cycle anyways. When I first started using lipo I started out with a few packs in conjunction with my ping. I had better luck using them at first by using them in series with my ping and monitoring the 6s packs with a cell log while riding.

 

[FluxZoom]

I tried that cycle analyst speed limit long ago with the same results, tried messing around that and just gave up on it. You mentioned throttle limit, I don't know what that entails but that is how I do it when I don't want to muck around modulating the throttle. It's not perfect, as in you might go slightly faster or slower but it's not the end of the world. How I do this is using a 3 speed switch and an infineon controller. You can tune the throttle percentage through trial and error so the throttle percentage setting for each switch setting is to your liking through the controller settings.

Are you sure you have the three phases connected correctly? Connected in the incorrect order could mean that the motor still spins in the right direction but it takes insanely more power because the motor is kind of fighting or whatever. https://endless-sphere.com/w/index.php/Determining_the_Wiring_for_a_Brushless_Motor What is the no load speed and power consumption? Maybe one of the hall sensors are dead or have a poor connection? Just shots in the dark.

Maybe it's just that the motors are wound for wildly higher speeds and slowing down even more just brings you to such an incredibly low point in efficiency that it doesn't gain you anything.

 

[wineboyrider]

CA calibrated right?????? Make sure the shunt value is set to the proper amount for your controller?!

 

[dogman dan]

I see, I just assumed you had a one motor, something like a muxus or cromotor on er.

Getting way above my head technically, maybe there is something going on with two motors, something causing waste. I've built some pretty big, heavy, cumbersome, distinctly un aero bikes, and never saw more than about 40 wh/mi at 20 mph ish. I was definitely thinking you could cut your wh/mi in half easily, just riding the first five miles slow. Then have plenty for the remaining part of the ride.

Even if your CA is wrong, it should still show a big improvement if you ride half speed.

Definitely, RC stuff is something to consider, if you want to carry a 5 ah spare tank. But you have the other batts, so paralleling them for the longer rides could work best, despite any hassle caused by differing chemistry and LVC's.

Best approach suggested right away by Amberwolf, get 5 ah more of the same chemistry cells you have now. (or close enough to same) then you can parallel them in and full charge and end of discharge voltages will match fine.

 

[llile]

CA calibrated right?????? Make sure the shunt value is set to the proper amount for your controller?!

Well that's something to check. I suppose an ammeter in series with the whole shooting match could tell the tale. I'm sure I set this up in a thorough manner but it was a couple of years ago and I don't remember what I did.

 

[llile]

I tried that cycle analyst speed limit long ago with the same results, tried messing around that and just gave up on it. You mentioned throttle limit, I don't know what that entails but that is how I do it when I don't want to muck around modulating the throttle.

Aha! So I;m not the only one. The "Throttle limit" does the same thing as backing off on the throttle - normally it is set at 4.5 volts or something meaning if the CA sees 4.5 V it goes wide open.

But if you set the throttle limit to 75% of that, then the CA never sees 4.5 volts, so it only goes 75% of wide open, even if you hold the throttle all the way. This is only useful if you have to dial down the speed, and get bored trying to hold 20 MPH after the first ten miles.

 

[llile]

Made a little network that plugs two batteries in series. I've scrounged/bought used two 10 AH 12S/36V Lifepo4's. I could just plug them in series, but Li Ping warns us that some of his BMS's might fail if you do that. His V5 BMS will handle two-battery stacks, earlier versions maybe not.

Although in theory plugging two 36V things in series results in 72V, with each battery seeing only 36V, in practice things could be different. The big battery could theoretcially overcharge the little battery if they aren't isolated by a diode. Also if the big battery is trying to charge the little battery, the BMS should try to trip out on overvoltage. If the little battery only has 60V FETs, and the big battery can provide over 80V full charge, the BMS fets could blow. This is one reason to use this diode network. The other reason is that the manufacturer says it works and it might fail if you don't.

One battery is a Ping of unknown vintage, the other is even more unknown.

Ping's solution is a stack of Schottky diodes that help protect the BMS's and prevent back-feeding the smaller battery.

I'm using a 100A schottky diode that has a convenient hole in each tab that fits a 10-32 bolt, so the whole thing went together with ring terminals and bolts. No soldering required, although I do love me some soldering.

Now there are folks on this board that argue this is nuts and they've never had problems with series batteries. If they happened to pick Ping's V5 BMS off the shelf and series'd two batteries, there should be no problems at all, except that the big battery might tend to overcharge the small battery if one is full and the other spent. Would the BMS prevent this? Hopefully, if the Fets can take the voltage. This could be overkill, and creates a slight power loss with series diodes, but I'm figuring it is more robust.

 

[llile]

This is where a spare lipo range booster pack might have some huge advantages for you. They usually have a higher discharge rate and lower life cycle anyways. When I first started using lipo I started out with a few packs in conjunction with my ping. I had better luck using them at first by using them in series with my ping and monitoring the 6s packs with a cell log while riding.

Well, Boeing can't even prevent Lipo fires, and they really ARE rocket scientists. I've decided Lipo is above my pay grade. Good luck for those who use it, I'm sure you guys are all fine except the ones who've been burnt out of your houses. Justin says he keeps eleven foot poles laying around the shop in case someone brings in a Lipo - he wouldn't touch them with a ten foot pole so that's why they are eleven feet.

 

[llile]

I see, I just assumed you had a one motor, something like a muxus or cromotor on er.

Getting way above my head technically, maybe there is something going on with two motors, something causing waste. I've built some pretty big, heavy, cumbersome, distinctly un aero bikes, and never saw more than about 40 wh/mi at 20 mph ish. I was definitely thinking you could cut your wh/mi in half easily, just riding the first five miles slow. Then have plenty for the remaining part of the ride.

Perhaps the motors are fighting one another? Could one controller go into regen while the other was powered? I don't believe this is possible - I have a separate E-Brake terminal on each motor controller, and I do have an E-Brake, but I never feel the motors dragging unless I engage the brake switch. But perhaps there is still some excess drag associated with moments where one motor is off and one on. I don't really notice this happening much, but every once in a while there is a little tug right or left.

Crap if I could get 40 WH/Mile I'd have an 80 mile range.

 

[dogman dan]

What does this thing weigh? It just seems really unlikely that you'd be getting such high wh/mi if you really are going 20 mph max. For a heavier ride, sure, it could be as high as 40wh/mi at 20 mph. But more likely 30-35 max. Again, some of this oddness about the numbers can be a mis calibrated CA. But even if that is the case, you should still see a very substantial range increase if you travel 20mph, vs 30 or whatever.

Way above my pay grade, but I'd look at how each motor runs by itself. One motor could be coasting and you don't know it, like a bad connection at the wire, or a halls blown. That's the only simple answer to why one motor could be fighting.

It might seem to run smooth, with the other motor pushing it. Once I had a cut wire on a motor, but two tiny strands still connected. So in a no load test, or riding up to 5 mph, smooth as silk. But under load, or any faster than 10 mph, it would start to stutter. The motor ran, but essentially had power on only two phases.

So look at your motors, make sure both seem to have full power when under load, running the vehicle by itself. The push from the good motor may be masking a problem with the other. And then you are running with one motor dragging without knowing it.

 

[Hwy89]

I'm late to the party but it occurred to me that 80wh per mile would have to be generating a lot of heat. Would it be worthwhile to try infrared photography to look at the system and see where the hot spot is? If one of the motors is hotter than the other it would be a clue as to where the power is being consumed.

 

[Ykick]

80Wh/mi is nuts. Something's very fubar either measuring or motor/controller operation.

This current bike is a two-motor three wheeler hub motor tadpole, averages 80 WH/Mile/35A. Never knew why, but I have to slow way, way down to notice much better efficiency at all, and it never gets very low. At a CA limit of 20 MPH, I only dial down to 65 WH/Mile/25A. Only 20% better. 35^2/20^2 = 3.06, should be a similar square law reduction in friction. My buddy, also with a tadpole trike using my old mid-drive motor I sold him, is going the same speed at three times the efficiency right next to me on the trail. No difference in tires, weight, or aerodynamics.

 

[llile]

What does this thing weigh? It just seems really unlikely that you'd be getting such high wh/mi if you really are going 20 mph max.

OK, I am running 80 wh/mile at 35MPH, people seem to think I'm using that much juice at 20 mph. Just to straighten that out.

The bike, batteries, and me on it weigh 500 lbs. It's a tank, and most of the weight that is not me is batteries. It is essentially a motorcycle with some bike parts bolted onto it.

I do happen to have a thermal cam, maybe I'll take a peek at it sometime.

I am quite sure both motors are running all the time, I can hear them. Once or twice I've left a motor unplugged, and I hear the difference immediately. THey aren't cutting in and out at full speed (35 mph) I can feel the torque and hear them running.

Long time ago I learned that if your bike makes a different noise, something is wrong, so I listen to it pretty carefully.

Both motors make the same noise, and on a stand they both go the same no-load speed at the same throttle setting. I've tested them with a stroboscope just to see.

Early work with Justin's simulator showed that I should be expecting 65 wh/mile at 35 MPH, so 80 wh/mile isn't a shock.

THere's a coast-down test I would like to look up, that can calibrate the drag coefficient to put into Justin's simulator. That would be a worthwhile experiment to try.

 

[llile]

I'm late to the party but it occurred to me that 80wh per mile would have to be generating a lot of heat. Would it be worthwhile to try infrared photography to look at the system and see where the hot spot is? If one of the motors is hotter than the other it would be a clue as to where the power is being consumed.

I didn't mention the toaster? I get three slices every two miles

 

[amberwolf]

1 ) The CA speed limit problem is probably caused by one or more of the ramping values, based on stuff I've seen in various CA discussions. Best recommendation I have is to look up Teklektik's UUG either here on ES (in the CAv3 beta thread) or on Grin Tech's site, and follow it's complete step by step setup procedure, without skipping anything. (after first noting down all of your original settings, just in case).


2) The efficiency problem could be a few things.

On my SB Cruiser trike, I'm using mismatched motors and controllers, but it doesn't matter if I use just one motor, or both, or which motor, I get about the same efficiency. So it shouldn't matter even if your motors were different windings, and one was being "overrun" (causing generated voltage that might cause drag/regen even without applying ebrake) at a higher speed than it's no-load max speed at that voltage...which it almost certainly isn't.

It could be aero, if it's a really terrible shape, super boxy, etc. I suspect it isn't that, because there isn't that much difference between your 20MPH and your 35MPH wh/mile. My SB Cruiser is not very aero, but it doesn't really make much difference to the power usage, compared to the next thing.


It could be weight, if you either have sufficient grades to go up for long enough, or if you have lots of stops and starts. That's what really kills my efficiency on SB Cruiser, which can be as bad as 50-60WH/mile under "normal" conditions and cargo loads, for my short work commute with a dozen-ish complete stops and starts in 2.5 miles (each way), and 70-80wh/mile with Yogi or Tiny in the big trailer doing the same commute.

It's a lot better on much longer rides with significantly fewer stops/starts, even with the same cruising speed of just under 20MPH.


With the weight thing, it can also be rolling resistance, either if your tires aren't at the right pressure, or if they're just generally harder to roll.


It could also be something like the phase/hall combo being wrong, but if that was the case, your motors would be getting pretty hot, most likely, and possibly the controllers too. You can test no-load current fro each motor, in the air, to possibly eliminate that.


Could be something really mundane like brake rubbing, bearing binding, etc. You can test that by spinning the wheels in the air by hand, and if they don't freely spin you found a problem.

WHat is your continous current at 20MPH on a flat-level run?

What is your peak curent for startups, same run?


SB Cruiser has two "40A" controllers that really only are 33-35A controllers, but I do get peaks of 80A+ on some startups. Typical cruise continuous current (using just one motor normally), is 10-15A depending on conditions, IIRC. (800-1000w, with the trailer). Little less with just the trike.

 

[dogman dan]

I was thinking 65 wh/mi was crazy high usage, for 20 mph unless your rig has the weight and the drage of the SB cruiser. That SB is essentially a small electric pickup truck, with capability similar to a golf cart, which gets pretty poor wh/mi compared to a typical bicycle or trike.

Re your motor testing. That was exactly my point, one of your motors can be running poorly and still seem fine no load. I'm saying try running it under load on one motor, then the other. If one is wonky, you'll know immediately.

But likely, the reason is just weight, and lots of starts and stops in city riding. Lots of power that gets going fast is great in the city, but it costs you. If you and your rig are 400 pounds, then your wh/mi is not really so high. But even on this, distinctly un aerodynamic bike, loaded to 400 pounds, I could get 40 wh/mi riding 20-25mph. One big motor on it. But that is riding in the burbs, not city center.



Re the CA limiting. I just limit amps when I need better range, not speed per se. But you need all the amps on any big hill of course. But if you can limit amps on any long, flatter runs, that may help tame the wh/mi better than speed limiting, because you will take off more slow on any stops.