Battery Vehicle Society

I've been pondering adding an electric assist to my race car, probably for 2012. The idea is to use electric power alongside the ICE to boost power at full throttle. I want something nice and simple, so I was thinking about an Etek motor, which weighs 10KG and will take 96vdc @480A coupled to 25 Polyquest XF cells, which would weigh just 3.2kg and put out 92.5vdc @396A. Doubling up the cells is just 6.5kg and allows less tressing of the cells whilst being able to take full advantage of the 480A maximum of the Etek.

What I'm wondering is whether it's as simple as hooking up the cells to the motor via a contactor that's operated at full throttle. Is there something I'm missing? It all seems too simple for 13/16kg of kit (plus contactors, cable, brackets, etc.) to get me 48/59bhp. Obviously, I'm aware of the issue of pack balance when running no BMS and that planning to use 90C will shorten the life of the cells significantly. I spend ~45 seconds at full throttle at an event.

Mechanically, I've seen supercharger conversions for my engine that take power directly from the crank via a pulley. It seems the obvious way to transfer power directly to the crankshaft of the engine, using a 2:1 pulley ratio to allow for the red line of 13500rpm on the engine and the ~7000rpm limit on the Etek and making use of the gearbox at the same time.

Sounds like an interesting idea. The Etek has been out of production for a fair while now though, so I doubt you'll find one, although the Mars Electric "clones" are pretty close to the old Etek specs (and beware, there is a fair bit of "passing off" going on with non-B&S Etek motors being badged as Eteks by some retailers).

As for batteries, you can get a lot better cells than those Polyquest ones for a fair bit less money. One or two folk on the ES forum have tested the Nano range from Hobby King and found them to be superb for very, very high rate discharge : http://www.hobbycity.com/hobbyking/stor ... rentCat=85 Take a look at this thread to see just what the Nano cells will do: http://endless-sphere.com/forums/viewto ... 14&t=22464

Pack balance won't be an issue, just balance charge periodically and the cells will stay in balance just fine. These things are a heck of a lot more robust now than they are credited with, it takes serious abuse to kill them.

I'd suggest connecting the electric motor via a one-way clutch of some sort, so that you don't waste engine power turning the motor. The contactor idea has merit because it's simple, so is worth a try. My only concern is how long the contactor would last, but then again old controllers just used banks of contactors and they lasted OK. I have a feeling that some fork lift trucks just use big contactors, too, so they might be a fruitful hunting ground for parts.

Have you looked at the bigger motor that's similar to the Etek in many ways, the Mars ME0709? (this one: http://www.marselectricllc.com/me07091.html). It has a lower Kv, but a higher power rating than the old Etek, but it is a bit heavier. I have a new one that's spare if you're interested, complete with a new Alltrax 48V controller. Mars also make a higher Kv motor that is similar to the old Etek, the ME0708, and that is a direct replacement in terms of ratings: http://www.marselectricllc.com/me07081.html

Sounds like an interesting project.

Jeremy

Hi Jeremy,

Thanks for your reply - very detailed!

Those Nano packs are significantly cheaper than the Polyquest ones and very similar performance on paper. I wasn't aware the Etek was no longer in production. I've been scouring specs from sites selling replicas. The specs I've seen say it's 10kg and capable of 480A and 96v, albeit for just 45 seconds. That link to the Mars 0709 motor says it has a max of 300A @72v and it's over 16kg. That takes the power/weight down dramatically and at under 30bhp, it's not worth the effort. The 0708 is even lower powered. What was so great about the replica Etek specs I found was that it made over 6bhp/kg, assuming those specs are truthful. Do you think they are optimistic? Do you think the Mars motors could handle more? My usage is about 50-60% duty cycle for 80-140 seconds and on full throttle for no more than around 13 seconds at a time.

I'm not too worried about sapping power from the motor when I'm not at full throttle as I can simply press the throttle harder. Efficiency isn't really the goal here. Contactor life isn't too much of an issue either. As an example, I applied full throttle 16 times at my last race meeting for a total of around 180 seconds. Even with a life of only a few hundred applications, that's a season covered! Do the contactors take time to engage? I read about the series/parallel switching of some controllers taking a couple of seconds as the controller waits for the contactors. When I'm only using the throttle for 13 seconds, losing a second or two isn't ideal.

Of course, it also means the cells will need replacing (or recharging) between runs as at 90C, that's only 40 seconds run time, but at £250 for a set of five 5Ah 5S packs, that's not an issue. If I could last the season (~10 events) on a couple of packs, I'd be happy.

Cheers,
Andy

Glad the info was useful.

I don't believe that a genuine, well-made, Briggs and Stratton Etek would take anything like 480 amps at 96V (over 46kW!!!), so I'd treat the claims for the Chinese clones with a very large pinch of salt. IIRC, the Etek was prone to fly apart at around that voltage (the Etek rpm limit was around 4000, and 96V would push it to nearly 7000), and with that sort of current flowing through it I think you can be pretty sure that it'll start spitting solder out fairly quickly.

The Mars ME0708 is almost identical to the original Etek, made by a chap, John Fiorenza, who worked on the Etek project at Briggs and Stratton, so my guess is that his specs are going to be close to reality. Who knows what the real specs are for a Chinese clone of a Lynch design? I doubt they'll be as good as a genuine Etek, yet from the spec on that web site they look to be much better than both the Perm PMG132 (7.2kW, 72V, 110A, 11kg) and the Agni 95 (12.5kW cont, 22kW for 5 seconds, 56V, 250A, 11kg). The Agni 95R is probably the most powerful PM DC motor in this weight category, yet even it can only deliver 16kW continuous at 78V and 30kW peak for 5 seconds. Expecting anyone to believe that a lighter Chinese copy of an Etek can deliver 46kW for 45 seconds is simply beyond belief, I'm afraid.

I'm reasonably sure that Briggs and Stratton built the Eteks under licence from Cedric Lynch, using some of his patented ideas. The Mars motors are as close as you'll get to a real Etek and John Fiorenza will happily let you know just what you can stuff through one and how they compare to the Etek, he's a pretty helpful chap in my experience. My guess is that the Chinese copy will be like the vast majority of Chinese motors - pretty poor quality. They can often be turned into reasonable items with some work, but my overwhelming experience with buying Chinese made motors has been that you need to treat all claims as being highly optimistic and be prepared to do some serious work to get the motor working well (I've had several bearing failures, windings have been notoriously loose and of poor quality and the standard of assembly is often nothing short of appalling). They are cheap though, when compared to motors made under better controlled conditions.

If you want reliable high power then I'd suggest sticking with motors that we know are excellent, like the Agni 95R; it's probably the only motor that will come close to doing what you want. They are more expensive, but at least you can trust the specs.

I think you'll find that contactors will switch within maybe 100mS, so they should be OK.

Jeremy

It's as I kind of expected, then. The specs did appear to be much betteer than anything else out there. Nothing else in that class really makes enough power to be worth using, so I guess my best bet is to go up to an ADC 6.7" motor, which I have seen in a 144v variant, that allegedly takes 550a. That should be enough, but it's 25kg, which is a significant increase in weight. The whole car weighs 480kg (hopefully under 450kg by the end of the month), so to add 40kg in motor, batteries, brackets and sundries is a big ask.

I don't see the point in adding less than 50bhp to the car and 40kg is on the upper side of what would be acceptable. Is the ADC my best bet?

As for contractor life, I clearly forgot about gear changes, so you can triple that count at the very least. 100ms could make gear changes tricky if it takes that long to release the power. It's a dog box and up changes are clutchless in under 100ms when lifting off the throttle, which could be scuppered by an extra 100ms of power and an extra 100ms before it comes back in again. I will have to have a good think about whether it is really feasible.

When looking at extra power, it's probably worth looking at the torque figure, rather than the absolute power figure (the product of torque and RPM). In essence, it's the extra torque that is going to give you the acceleration boost you're looking for, not the ultimate power of the motor as it hits it's rev limit. Given that PM electric motors have a nearly flat torque curve (they give maximum torque pretty much from zero to maximum rpm) you may find that you don't need anywhere near as much power to give the performance boost you're looking for as you think.

Most electric cars, and some hybrids, have to actively restrict motor torque simply to retain traction. Apparently, if the traction control is disabled on the Prius, for example, it becomes pretty much undriveable because of wheel spin at speeds up to around 25mph, yet that only has around 40hp of electric power (albeit with well over 300 lbs ft of torque). It might be worth modelling what the extra torque would do at the operating point where you think that you would use electric boost. The results might be quite surprising, as I rather suspect that your race engine will have dropped off the maximum torque point of it's power curve at gear change rpm, so delivering a healthy dose of extra torque from an electric motor might give you more of an edge than you might at first think.

The Agni could probably chuck in around 40 to 50Nm of extra torque at full throttle, which might be a welcome boost even if the actual hp figures look low. My guess is that the whole system needs looking at carefully to see just how you might get the most benefit. It may be that you could use a relatively low power, but high torque, boost system to make up for the torque roll-off that any race engine will have beyond the max cylinder filling efficiency RPM, giving you a much broader effective torque curve. This might then mean you could hold a taller gear for longer, or keep in a lower gear and still keep accelerating beyond the rpm where you'd normally shift. Saving one gear shift would be worth, what, half a second, maybe? I know that when I was hillclimbing an extra shift because of picking the wrong final drive ratio for the hill would often cost me over a second and put me well out of contention.

I've no experience of the ADC motors, I'm afraid.

Jeremy

Hi Jeremy,

I really appreciate the time you spend replying and the quality of your posts. Thanks for sparing the time.

I tend to think in HP as it's the same, irrespective of speed, but you're right that torque at the wheels is the most important thing. My engine produces around 90Nm @13500rpm. The Agni revs to 6000rpm, so would need gearing at 2.25:1. 40-50Nm then becomes 17-22Nm or to put it another way, that's an increase in torque (and therefore power at a static RPM) of 20-25%. Of course, the increase is most important towards the rev limit as you rightly pointed out that torque drops dramatically as it approaches the limiter. That does make it look more appealing, as you rightly estimated it would.

I believe I spend most of my time in this low-torque area of the curve (>10krpm), although I've been deriving RPM from speed, making assumptions about where I change gear, because the logger I was using did not log RPM. This season, I have a new datalogger that does RPM, so I should be able to get some more meaningful data.

Cheers,
Andy

I think the net is probably at saturation point with various torque vs power arguments, particularly as so many have such firm views on the subject!

Being an analytical sort of individual I did do some calculations years ago of just how much power the 998 Hartwell engine was really delivering in first and second gear when I was hillclimbing my old Clan Crusader. The results were not that surprising, even without a datalogger (they didn't exist when I was racing!). It was fairly easy to determine that the engine was rarely delivering the 120hp maximum that was on the dyno plot. Much of the time it was putting out maybe 60 - 70% of this at best. I think I worked out that it spent less than 10% of the time delivering maximum power, really just the point before the first to second gear shift and a couple of places on the hill where the car hit max revs.

The most effective thing I did to my engine was swap out the old R23 "race" cam and replace it with the R22 "rally" cam. This lowered max usable rpm from around 12,000 or so to around 10,500 and similarly reduced maximum power by around 15bhp (a heck of a lot when you've only got 120 to start with), but it broadened the torque curve and gave far more usable power lower down the rev range. The net result was that the car had markedly less power but went up the hills faster (and broke more drive shafts.........).

I suppose it all depends on the sort of racing you do and how much time you really spend at maximum power. I only did sprints and hill climbs, primarily in the South West, so optimised the car for short windy hills like Tregrehan. I once did a sprint as a guest of Harrow Car Club on the club circuit at Brands and was hopeless, as the car just wasn't geared anywhere near fast enough for the straights - I hammered down them on the rev limiter practically the whole way.

I have a feeling that adding an Agni 95R would give you a very healthy boost and probably be worth the weight and effort of fitting it. If you decide to do it, keep us posted, as it sounds like fun.

Jeremy

You've clearly illustrated the point that peak power is unimportant. The average power over the time you're at full throttle is most important and this is always over a much wider rev range.

I too am doing sprints (but not hillclimbs) in the South West. I'll get some real data about engine revs and see what I'm really doing, rather than estimating. I'll also have a chat to the person preparing the car for this season and see what he thinks about the logistics of hooking it all up - I have no intention of doing it myself! First test of this year is booked in March.

By pure chance, someone's just posted the spec for the original Briggs and Stratton Etek motor on the ES forum. As I thought, the rpm limit is fairly low. Here's a cut and paste of the spec:

12 HP Peak, 2.7 HP continuous
Voltage........................50 VDC
Max Nonload Current.....6 AMP
Max Nonload Speed........3600 RPM
MIN Nonload Speed.........3300 RPM
Min Speed at 160 Lbln.....3200 RPM
Max Current at 160 Lbln...150 AMPS
Voltage Constant: 72 RPM per volt
Torque Constant: 1.14 in lbs/Amp
Continuous Current: 300 A 30 Sec.
Weight: 20.8 lbs
Motor Diameter: 7.91"
Motor Length: 5.64"
Shaft Diameter: 7/8"-or 3/4"

If the original motor was only able to deliver this sort of power, then I think it's a safe bet that the Chinese clone wouldn't be able to match it, let alone better it.

Jeremy