Electric Beemer

Open for any sort of non-technical discussion regarding EVs
User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Electric Beemer

Post by acmotor »

The ACP control would have been PWM sine ? to potentially only supply 86% of what a Voltage Vector Control or Flux Vector Control VFD could supply given the same battery voltage and motor.

DC bus voltage is not quite as simply linked to motor AC voltage as it may be with a pure sine wave. All modern VFDs use VVC or flux control.

TJ, is this why you expect to supply more like 200kW rather than the 150kW with the old controller with the same motor ?

Nice consideration of battery temperature rise weber ! Image
iMiEV MY12     110,230km in pure Electric and loving it !

BG
Noobie
Posts: 20
Joined: Wed, 04 Mar 2009, 17:54

Electric Beemer

Post by BG »

Hi James
    Just as well you have a mech eng on board.
Surely the power factor etc. is a result of inductance at a given frequency. Therefore Power factor is unknown if the frequency is unknown?
But, they don't supply a value for the inductance either.
   Ps you'd better leave that gearbox alone, because the highest listed power of that 911 variant was 132kW which means 158kW is the max legal power

Image

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Electric Beemer

Post by acmotor »

The first thing a VFD does is measure the resistance and reactance of the motor you have connected it to so its control software can adapt to the motor. This then covers most frequencies and loads as the coils are taken to full current / magnetisation etc.

The other way..
The pf is stated on the motor nameplate / data sheets for full load at the nominated frequency. Given this pf, the reactance at nominated frequency can be calculated. XL= 2pifL . Thus inductance is known.
This still varies with load and slip. I bow to TJ and Richo's superior knowledge here though !
Ok, simplified summary.

BG, TJ did set the condition of nominal operation ! (V,A,RPM) this sets pf at a known point.
My comment was that the pf created component of motor current is not seen by the DC bus as shown in red suzi data log.
Typically the pf is worse at lower loads.
iMiEV MY12     110,230km in pure Electric and loving it !

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Electric Beemer

Post by Tritium_James »

acmotor wrote: The ACP control would have been PWM sine ? to potentially only supply 86% of what a Voltage Vector Control or Flux Vector Control VFD could supply given the same battery voltage and motor.

DC bus voltage is not quite as simply linked to motor AC voltage as it may be with a pure sine wave. All modern VFDs use VVC or flux control.

TJ, is this why you expect to supply more like 200kW rather than the 150kW with the old controller with the same motor ?


The ACP drive is all analog, so I suspect it's sinewave out, not flux-vector control. It's incredibly complex in there though, so it is possible that it's doing something a bit smarter than just sinewaves. Having said that, I have not actually checked what it's doing.

I expect to get more power because we can put out higher peak current. Going off what Tesla is now managing to put into essentially the same motor (~800A), I think it's capable of being pushed a bit harder than the ACP drive in the Porsche is doing at the moment (400A bus = 330Arms in the motor). Approximately   Image   We can also go to a higher pack voltage.

User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

Electric Beemer

Post by Richo »

weber wrote:
I wonder if some might find it easier to follow if the calculation of nominal battery pack voltage and total internal resistance were separated out, and resistance given in ohms, and power in watts.


Examples are good Image
weber wrote:
And let's assume this 20 kW power dissipation was maintained for 40 seconds with no significant battery cooling. 20 kW * 40 s = 800 kJ.
In that case the average battery temperature could be expected to rise by 10 degrees Celsius, in those 40 seconds. That sounds quite acceptable.

Wow 10 DegC that's nice even for an extreme case. Image
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!

User avatar
weber
Site Admin
Posts: 2686
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Electric Beemer

Post by weber »

Thanks woody and acmotor.

It should be pointed out that the rule that says DC current is sqrt(3/2) times power factor times AC current, only applies when the 3-phase VF drive is putting out its maximum sine wave (between phases).

If it's running at a speed well below the motor's field-weakening point, it might only be putting out say 200 Vac or half the maximum AC voltage. In that case the "DC-transformer"-like action of the VFD means that DC current would only be half what you'd expect from the above rule.

The more general rule is

Idc = Iac * pf * sqrt(3) * Vac/Vdc

or a little more realistically

Idc = Iac * pf * sqrt(3) * Vac/(Vdc - 10 V)

ACmotor, I am aware of what I learned (from Ross Pink) to call "neutral wobble", whereby the waveforms on the VFD's outputs are not sine-waves when viewed relative to a DC bus centre-tap (real or imagined). They have about 11% of 3rd harmonic added so they are somewhat flattened on top to become round-cornered trapezoids. But the motor windings never see this 3rd harmonic because the _difference_ between any two outputs is still a pure sine wave. The 3rd harmonic has the same phase for all outputs (because the outputs are 1/3rd of a cycle apart) and so cancels out. Coulomb posted some graphs of this somewhere.

I also fiddled around in a spreadsheet some weeks back and found that if you also added a small amount of 5th harmonic to all outputs you could get a slightly higher amplitude of the fundamental again, but in this case the 5th harmonic actually appears across the motor windings. Presumably it would cause some noise and heating.

Are you saying that some (all modern?) VF drives actually do this, to make up for the 5% or so voltage drop in rectifying and smoothing the 3-phase mains, and presumably the IGBT's Vce(sat)? If so, that's good news.

User avatar
weber
Site Admin
Posts: 2686
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Electric Beemer

Post by weber »

Richo wrote:Wow 10 DegC that's nice even for an extreme case. Image

I worry that I've lost a zero or a decimal place somewhere. I'd appreciate an independent check.

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Electric Beemer

Post by Johny »

I know you have stated not to consider cooling or such on the pack but I believe you have to.
If you raise the pack temp. by 10 degrees in 40 seconds, how long will it take to cool it back to where it was (or near)?

My point is, there is now a heck of a lot of energy in heat, even though it's only 10 degrees, the actual thermal inertia of 200kg of metal/plastic is going to take some cooling before you use that kind of power again and again.

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Electric Beemer

Post by acmotor »

Re: temperature rise. Don't worry too much. The TS cells will most likely fail if you draw 5C for the 10 minutes until they are flat. Again and again is probably not a realworld option !
I would have thought if you wanted to discharge this fast then TS was not your baby ?

back to waveforms..
Yes, 3rd (and all the other odd harmonics) is what a square wave is all about !

weber, yep, Danfoss VLT for instance has run VVC and flux vector control since 1993 that I recall. I purchased one of the first 5000 series then.
People have called it Danfoss torque boost voodoo, but it is just cleaver software. It is more than just the 5% voltage makeup. It also provides full torque from zero RPM (by achieving correct magnetising)and other features such as open loop speed and torque control. It also allows the controller to run synchronous and PM motors.
BTW, I have no vested interest in Danfoss, they just have a good product that I use and they were the first to build an analogue VFD (SCR in 1950s ?), a digital VFD (1980s), apply VVC(1990s), flux vector, DC bus connect(1980s), grid connect regen (1990s) and they have one of the most electrically quite (RFI/EMI) drives available.
(probably helped by being European in origin (Denmark) where RFI standards are a lot higher than the US, Japan or China.)
Sorry if that sounds like an add !

Nearly all digital VFDs I have used over the last 10 years seem to have taken on these features as well. The limitation being the HVAC (heating ventilation air cond) ranges of drives that don't need to do a lot more than v/f control on a quadratic load curve.

I am certain TJ and Richo are designing with the VVC or flux vector software.
iMiEV MY12     110,230km in pure Electric and loving it !

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Electric Beemer

Post by Tritium_James »

Yes, the 3rd (and higher) harmonic stuff is not done by explicitly adding in the harmonics, it almost comes for free when you're doing space-vector PWM.

To actually handle the motor properly, you want flux vector control (this will be the torque boost voodoo), and for easy application you want sensorless flux vector control - no mucking around with encoders etc on the motor. Basically the motor controller runs a mathematical model of the motor and uses that, and info from it's internal phase current and/or voltage sensing, to get an accurate estimation of the rotor position. Then you can position the magnetic flux anywhere you want relative to the rotor, to give best torque, efficiency, whatever. It's not trivial to do, but it's certainly becoming more common. Having ever increasing processing grunt in low-cost DSPs helps too.

User avatar
coulomb
Site Admin
Posts: 3882
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Electric Beemer

Post by coulomb »

Johny wrote: I know you have stated not to consider cooling or such on the pack but I believe you have to.
If you raise the pack temp. by 10 degrees in 40 seconds, how long will it take to cool it back to where it was (or near)?

Excellent point. If you want to cool it back to where it was in another 40 seconds, you'll need a heat pump (like an air conditioner) capable of 20kW of cooling. That's massive. I'd guess wildly that a car radiator would be needed to get rid of that magnitude of heat.

Even if you only want to cool it in 400 seconds (over 6 minutes), you'll need to get rid of the heat at an average rate of 2kW. And that doesn't count the heating from cruising while this is happening, it's just from that one pulse of acceleration.

One nice thing is that cooling works best when there is maximum difference between the hot and cold bodies (here, the cold body is ambient air). That means that the battery will likely get hot (say 30-50 degrees celcius above ambient) and stay in that range. It so happens that TS LiFePO4 work better when hot (though perhaps like laptop betteries, they may age much quicker when hot).

antiscab
Senior Member
Posts: 2523
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Electric Beemer

Post by antiscab »

heat might be a problem for larger packs.
but you would have to work the pack fairly hard to get it to seriously warm up.
say you are drawing 20kw continuous from a 144v 200AH pack, and when cold you see 145A@137.25v.
that results in 983w (or there abouts) of heating within the pack.

however, when the pack gets warmer (say 40 deg C) to get 20kw you see 141A@141.75v.
that results in 317w of heating within the pack.

a pack that size weighs around 300kg.
unless you are drawing some serious amps on a 40deg day, heat wont be a problem with these cells.

If you are worried about heat, i would suggest using a phase change material, such as parrafin wax, in the groves between the cells.
this will limit heat rise, and is reasonably cheap to implement (and maintain) - i stole this idea from some guys on the EVDL.
parking the car out of the sun would also be advisable.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Electric Beemer

Post by acmotor »

I have some experience with the phase changers (latent heat energy banks). Problem is they only store energy not remove it all together from the pack. Try a fan ! Image

If the TS had a heat issue at DESIGNED C rates then they wouldn't have been put in their plastic lunch boxes !
iMiEV MY12     110,230km in pure Electric and loving it !

User avatar
coulomb
Site Admin
Posts: 3882
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Electric Beemer

Post by coulomb »

antiscab wrote: Say you are drawing 20kw continuous from a 144v 200AH pack, and when cold you see 145A@137.25v.
that results in 983w (or there abouts) of heating within the pack.

however, when the pack gets warmer (say 40 deg C) to get 20kw you see 141A@141.75v.
that results in 317w of heating within the pack.
The internal resistance reduces by a factor of ~3? Is that starting at -40 degrees?

I take your point that batteries with negative coefficients for the internal resistance will tend to limit their temperature rise to a degree, but I don't know that it's enough to not worry about cooling.

Also, the open circuit voltage (at least for lead acid) decreases by about 4mV/degree C/cell, which works against this effect (need more current to get the same power).

antiscab
Senior Member
Posts: 2523
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Electric Beemer

Post by antiscab »

coulomb wrote:
antiscab wrote: Say you are drawing 20kw continuous from a 144v 200AH pack, and when cold you see 145A@137.25v.
that results in 983w (or there abouts) of heating within the pack.

however, when the pack gets warmer (say 40 deg C) to get 20kw you see 141A@141.75v.
that results in 317w of heating within the pack.
The internal resistance reduces by a factor of ~3? Is that starting at -40 degrees?


the internal resistance does fall significantly with high temperatures.
so at 20 deg C, at 145A on a 200AH cell at 80%SOC, you get 3.05vpc.
at 40 deg C, at 141A on a 200AH cell at 80%SOC, you get 3.15vpc.

the pack on my scooter has a very low thermal mass, so it is affected *alot* more by ambient temperature, than internal resistance losses to heat. so if the scooter hasnt been in use for some time, its performance is affected by what the temperature has been for the past 15 hours or so.

a car sized pack has a large thermal mass, so if you havent used the car recently, you are more affected by what the temperature has been for the past few days, and how much you have used the car.
more use + higher temperatures = drastically improved performance.

the other thing about bigger batteries, is the bigger you go, the lower the resistance heat in kw for the same output.

Tuarn - latent heat storage would reduce the variability of the pack performance due to temperature. If high temperatures are an issue, a more effective cooling system would be needed.

I have only had one experience where the pack on my scooter was over 60 deg C. it was a 45 deg C day, and it was just off its 2nd ful charge (from 20%SOC). voltage sag had reduced so much that even at 2C, i was seeing more than 3.1v on every cell. 10C is entirely possible from these cells when they are hot, i just didnt have a controller powerful enough to make use for the extra available power at the time (a flaw ive since fixed :p).

ive probably gone off topic here, so if we want to continue on this line of thought, ill start a new topic.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

User avatar
weber
Site Admin
Posts: 2686
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Electric Beemer

Post by weber »

Richo wrote: The thermistors are standard in the 3-series for performance motors.
So you don't need to give the optional codes for it.
On the standard series it was I think 11kW and above.

Hi Richo,

I still have no quote for the ABB 22 kW 2-pole motor:

M2AA 132 SME 3GAA 131 008-JSE, 436

and no indication of why no quote is forthcoming. My reseller prompts ABB Australia every week or two. But I'm not quite ready to give up.

What page of what document (URL?) makes you think the thermistor variation code 436 is unnecessary?

M3AA are from the Industrial Performance catalog, M2AA are from the General Performance Catalog. The same motor often exists in both catalogs. I've looked at the ABB spreadsheets of variation codes for both of these catalogs and see nothing to indicate that thermistors are standard for 132 frame motors.

I've also changed my mind about the flange size. We definitely need feet _and_ flange, but I think that the absence of a gearbox mount in the MX-5 means we need the rigidity of the larger flange. So now I'm thinking we need.

M2AA 132 SME 3GAA 131 008-HSE, 436
or
M3AA 132 SME 3GAA 131 008-HSE, 436

Also, I'm a little concerned that the small flange may not be available for the 132 frame motors as there is no drawing for it here.

http://www.abb.com/product/seitp322/6d4 ... f527f.aspx

Why did you prefer the small flange?
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Electric Beemer

Post by Nevilleh »

Been a while since I last looked here and lots of interesting discussion has been going on.
But, to get back to my original subject, my BMW conversion is going along. The two Advanced L91 motors have arrived, also the LogiSystems controller, some contactors courtesy of ZEVA and a whole lot of other odds and ends.
I am still working out the motor mountings, but I have the coupling for the two motors made from a couple of BMW front driveshaft ends plus the rubber doughnut that normally couples the transmission to the driveshaft and I thought that might be of interest. All courtesy of the local wrecker for a grand total of $30. Cost me a case of good red to get a mate to do the machining and welding, but that was a good price.
Here it is loosely assembled into my MDF frame mock-up!


Image

BG
Noobie
Posts: 20
Joined: Wed, 04 Mar 2009, 17:54

Electric Beemer

Post by BG »

If you're running the coupling to 5500rpm do you think its worth getting it balanced?

zeva
Senior Member
Posts: 422
Joined: Sat, 15 Dec 2007, 05:09
Real Name: Ian Hooper
Location: Australia
MSN: sigmunky@hotmail.com
Contact:

Electric Beemer

Post by zeva »

BG wrote: If you're running the coupling to 5500rpm do you think its worth getting it balanced?

Definitely worth getting it balanced! At high RPM it really doesn't take much of an imbalance to generate nasty vibrations, and that coupler looks fairly sizable.
Ian Hooper
--
"Never doubt that the work of a small group of thoughtful, committed citizens can change the world." - Margaret Mead
http://www.zeva.com.au

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Electric Beemer

Post by Nevilleh »

Yes, I will get it balanced. In its original form it spins at driveshaft revs and the manual says to mark the way it comes apart so you can re-assemble it the same way, so I assume that means it was balanced originally.
Same goes for the coupler from the rear motor to the driveshaft.
I've been away on holiday for the last 2 weeks (in Oz!) so there has been no progress, but next week should see the mounting frame design finalised - I hope.
All the bits are here now, just a matter of enough work to get it together.

Post Reply