1.35bar isn't 3.5psi
Just checked in matchbot @8500rpm 105VE and 3.5psi@ 76% = 8.73hp = 6.5Kw
So @ 48v = 135 amps.

 

Perhaps in theory but my experiences tell me something else.

One key difference is the lack of a BMS built into the battery. Most Li-ion packs are built with one (not all), and same with other kinds of packs made for automotive use (e.g. Antigravity). It's almost a base requirement. Since the chargers in a car are usually simple (float voltage) and not LI specific.

But also, from personal experience as a RC hobbyist, Lipo batteries fail much, much faster (echoing another's experience above). Typically, they'll start to balloon up. And that's not from over charging or over discharging or out of temperature range ... etc. Once that starts the pack is dead.

 

How did I mix that up…

Ok, 3.5psi/0.24Bar are close to yours by my approach, two different approaches give the same numbers, hopefully good news. But I assume you’ll lose some % in motor efficiency as well.

I quit digging because high RPM motors was unobtainable for normal consumer at the time, very exciting that it’s now getting there, fantastic that someone with skills and experience are exploring it.

Further down the line: it may be an idea to let the motor “idle” from 12V to keep some speed normal cruise, to increase response and reduce peak mechanical stress. Just a thought.

 

Opposite of lead acid batteries, BMS is crucial for Lithium battery longlivety, not having BMS is asking for short life span - which is OK on RC cars etc where cost is limited and you want to squeeze all performance possible. So if I understand you correctly, your experience is not really relevant for what I’m trying to say; you need a BMS that is set up conservative for max life span, then, and only then, batteries can last for a very long time. At least lithium ion. LiPo seems related in link I referred to further up.

 

Great idea. The control I have has provision for two positions so maybe I could look at that.
Wouldn't want to use more than about 50w or the charging will be adversely affected but even that might make a big difference to spoolup/stress

Wasn't able to collect Amps data today but did get minimum voltage at end of 2nd gear pull to 9000 = 40v ......eek! . So, batteries are def struggling. Will pay to wait till new ones arrive before doing more testing.

 

Hm, not what I was saying no. Lipo batteries typically come without a BMS because their usually used by hobbyists' that purchase special chargers. Those chargers use the balance leads to ensure the pack's cells are getting charged evenly. Such as the ones Brett is using. It is important to keep the cells within a pack balanced, which is one of the functions of a BMS, or the special charger (in case of Lipo). The BMS is extra insurance as well against overcharging and over-discharging. So again, I don't think Lipo is a good long-term option here.

 

Scroll down to bottom in the mentioned link:
https://www.researchgate.net/publica...y_A_case_study

But yes, it seems like LiPo is even more sensitive, but if cycled between 45-75%SOC(and other parameters healthy), you still have 87.5% capacity after 6000 cycles. Not bad.

Another interesting point, if 100% is used, estimate is 600 cycles, but if 10% is used, estimate is 15000 cycles.

This indicates that increasing the capacity of battery exponentially increases life. Personally I would prefer to double the battery cost/weight if it gives me 10x lifespan, but thats a priority each and everyone can make.

 

Brett. Can you please stop experimenting with power upgrades on your RX8s. I have accumulated most of the bits I need for a turbo conversion including a spare motor to build specially for it.
Now you do this and i am frantically scouring websites getting the best deals for this setup! No wonder I don't have any spare cash!
On a serious note, nice work, ignore the haters, I love this idea mate!

 

Thanks Goandy. The first thing I'd say to that is ...this isn't as good as a the turbo setup I have sitting next to it. Not even close. It was never meant to be and it never will be. I think of it more as an equaliser in the sense that it now matches or betters most cars that it gets compared to for straight line speed. I'm hoping to improve it to the point where it becomes quite a desirable option for people not wanting the expense and complexity of a turbo.

 

One thing I can see that's different to how a hobbyist would use the batteries is that they never stay fully charged for any length of time. I discharge down to below 80% after every use and they only ever reach 96% (my max. setting) during a drive - which might only last a few hours. Hopefully this plus using the biggest lipos I can get (9500ah x2) will make battery life tolerable.

 

This is the first step towards your future lag free turbo setup with boost from idle

 

96% will decrease life considerably… When you receive the new batteries, I’d try to stop charging at 80% if possible and see if that’s ok for everyday driving. They should be able to give plenty of juice at that level as well.

 

You want to keep your battery charge between 80% and 20% to keep max life out of them. No more than 80%, and no less than 20%. Ideally.
This is how Tesla and other OEM car packs last a long time. Their BMS doesn't even let you charge the packs or discharge the packs all the way - and they actually increase this tolerance over time in order to make it seem like you haven't lost range.

 

That sounds a bit like too much of a compromise for the batteries I'm using.
MCSGUY on RCgroups.com recommends the following for LIPOs:

To Maximize Cycle Life and Performance:
1. Never fast charge cold packs, it causes permanent damage
*restrict charging temperature to between 70F/21C - 86F/30C degrees
*reduce charging rates to1C or less when near bottom of this temperature range (regardless of label specs)
*optimum charging temperature is > 77F/25C degrees
2. Never discharge at high amps when packs are at low temperature, it causes permanent damage
* optimum discharging temperature is > 77F/25C degrees
* real "C" maximum discharge levels should be avoided until pack temperature is > 86F/30C degrees
3. To enhance life even more, charge below 4.20v/cell and limit depth of discharge (DOD) as much as possible
* charging at 4.10v/cell doubles cycle life
* limiting DOD <70% doubles life
4. Maintain pack running temperature below ~131F/55C if you wish to avoid accelerating the aging process
5. Immediately charge a pack to 3.7v if any cell has been discharged to 3.0v (or near)

 

Twincharged RX8 hey? Hmmmm. 🤔

 

People online seem to think I should just raise the boost and make more power so just wanted to write a short explanation on why I don't do that for this setup:
1 3.5 psi is already heating the air to approx. 25C above ambient - This (I believe) is within the boundaries of what 98ron pump gas will handle. I feel I'd be asking for a detonation event going any higher.
2 Would need an intercooler plus higher octane gas or a w/m setup.
3 More boost means more electrical power (goes up exponentially with boost)which means the batteries are stressed more and the amount of driving I can do under boost goes down. So - would need bigger heavier batteries that would have to be boot mounted.
4The charger would need to be bigger which also means I'd need a bigger alternator. OR just remote charge the batteries meaning there is a finite life while driving.
5 Would need to upgrade the fuel system with bigger injectors/pump.

So ... I believe that around about this boost level is the sweet spot for a practical setup with minimal modifications and best drivability.

All that doesn't mean I won't build a more powerful setup . In fact.............. I've already got a plan for that. Mu...haha!!!!!

 

Didn’t think about the temp rise, good point.

Side question, how is motor controller given a input? On/off with a pressure setting, or is it progressive from a certain throttle command? My vote is for the latter, say “12V idle” up to ex 60%, then ramp linear up to 3.5psi at 100%. Giving 1.75psi at 80% if I explain good enough..? I imagine that would feel like a larger NA engine.

 

I initially thought I'd see the biggest gains down low. Not so , it's kinda underwhelming till 4500 at least, but up top it totally transforms the car. If you think about a turbo at 3.5 psi boost - it's not any different down low to that, but up top with a turbo you'd see some back pressure starting to affect power and it's pretty underwhelming there as well. With the eSC however, and zero backpressure it's making the power of a 5psi turbo setup up top.
So ...if I wanted the hit down low of a typical Greddy turbo setup with similar top end, Id set this to 5psi at low rpm and taper it down to 3-4 at the redline. That may still be doable within the boundaries of what I described above. What you describe might be effective if the setup did make 5 psi initially - but at 3.5 it would be a wasted effort.
For starting it I'm using rpm(vfad wire) and a vacuum switch combo . Works great, just like you would want it to.

 

I personally thought a kit like this was perfect for anyone wanting more low end power and the substantial increase Brett has already achieved is beyond encouraging imo. My hopes are that something like this can become another available FI option for the RX8. In the interest of rotary spirit and all.

+1 progressive throttle command.

Per Eicma:
"the electric supercharger is able to control intake air pressure independently of engine speed, resulting in high torque delivery from low rpm. In this configuration, the supercharger does not need an intercooler. In addition, the new ICE Concept has extremely compact dimensions, both longitudinally and transversely, and gives the designers ample freedom in the positioning of all components in the limited space of a motorbike, allowing centralisation of mass.

I came across this very informative article as well and I think the members who have the experience with this knowledge base will find the topics potentially relatable and inspiring. I pulled a few excerpts: https://www.cycleworld.com/blogs/ask...ept-explained/

"The Mitsubishi Technical Review reveals that from stopped to 90 percent of maximum rotational speed requires 0.3 second or less, with the electric motor drawing up to 5kW from the car’s electrical system. One automotive system described could operate for 30 seconds at the 5kW level, but could operate continuously at about half that level. The insulation on motor windings is temperature limited, which is why this car application requires 48V power. Electrical heating (as in a kitchen toaster) is proportional to resistance, and to current, squared. By quadrupling the voltage, current and therefore heating are reduced 75 percent.The type of electric motor often referenced for this application is the switched reluctance type, or SRM, which has no windings or permanent magnets in its multi-pole rotor—just iron laminations. This gives them great durability at high speed, and reduces cost. With all windings in the multiphase stator, liquid-cooling becomes easier.

"How much extra electric power will such a system require? This will scale with the smaller airflow required by a motorcycle engine. If an automotive unit sized for 3 liters requires a peak of 5kW, a system for an 850cc bike engine might need only a third of that, or 1,700W. That doesn’t look out of reach, considering the 1,550W output of late-model Honda Gold Wing alternators.How much weight will be added? On one Audi model the system added 22 pounds. This too should scale with the smaller airflow needs of a motorcycle engine.

Because of the smaller-diameter compressor rotor required by a motorcycle electric supercharger, it should spin up at least as fast as the 0.25 second Audi claims for its systems."

"To sum up, sudden throttle opening on a turbocharged engine places it at a disadvantage, because the small initial exhaust volume being produced is unable to rapidly accelerate the rotor. But in an electrically driven system, full motor torque is available even from zero rotor speed."

 

Then I did not explain well enough I think.

I was thinking about pressure setpoint as function of throttle position instead on/off. Pressure setpoint as function of RPM is another thing - that I support if doable to get some low RPM grunt. I think I'll even could live with some charge cooling to make that happen.

If you combine the two, max pressure of 5PSI and taper down to 3PSI(your example), then constant 80% throttle will give a setpoint of 2.5PSI, tapering down to 1.25PSI at high RPM.

 

The same effect is achievable with the throttle itself via the throttle maps. I'm already doing this with my turbo setup and it is very effective.

 

So 2.5PSI before partially open throttle, 1.25PSI after throttle?

If so it does not sound like a optimized approach, since you both use extra power from motor that is transferred to heat/temp rise(no clue how much). For your turbo setup I definitively see the point, having pressure on tap. But this setup should have excellent transient response(?).

Easy to say behind a keyboard - but when I think about it, the best might be to have open butterfly at the mentioned 60% throttle and above for no loss, then gradually increase pressure setpoint from 60 to 100% throttle? In other words, SC pressure setpoint is really controlling air mass from 60 to 100% throttle control, instead of butterfly valve. Just thinking out loud here.

 

I think if you drove it with only 3.5psi boost you'd realise that there just isn't anything to gain by complicating it like that. If we were talking 10psi ...then yes.

With my turbo setup which is currently running 12 psi with exceptional spoolup .... I'm able to modulate the throttle down to 5psi and that makes a huge difference to drivability and cornering. AT 3.5 you just feed it (like you would an N/A) there are no car control problems.

 

Update :
Just came back from a test drive on a windy road.
Wanted to test viability to see if the eSC is actually worth having in this scenario.
1/ Still on cheaper batteries so I expect improvement once the new ones arrive.
2/ Was able to drive hard on and off the boost for about 10 mins before battery was drained and starting to overheat.
3/ Max IAT recorded 52C at ambient temp of 24C.
4/Max battery temp recorded 51C (high temp charge cutoff activated) but was in low 40s most of the time. Battery seems to be getting good cooling at speed where I have it, but starts getting hotter when stationary.
5/To get from 'storage' voltage to full charge seems to only take 5 -10mins but from low charge to full it takes a LOT longer . Need to define this better with accurate times.

Overall , the extra performance was nice but I was spending too much time monitoring stuff to really enjoy the drive. Was encouraged enough to believe that this setup could indeed be viable for this type of use.
The suggestion to have the eSC partially spooled might improve the experience further. I have the eSC acceleration at quite a low setting ATM so might experiment more with that first.

 

Are you monitoring motor temp as well? The permanent magnets cannot take too much heat, something to be aware of. Ask the supplier about the limits to be sure, seems to differ greatly with what material magnet is made of.