Tesla Parameters Thread

[catphish]

i means current.

This is to differentiate it from voltage AC and voltage DC. Not really redundant in this case. In my opinion while AC may literally stand for alternating current, it's more a description of a waveform these days than a description of current specifically.

[P.S.Mangelsdorf]

In short, the maximum current the inverter is allowed to pull from the battery (DC current limit), and the maximum current it is allowed to deliver to the motor (AC current limit).

[johu]

So now I am really anxious to hear from Damien what the difference is and will it hurt anything in single mode. I've been driving in single mode for weeks without knowing it so whether it is hurting anything is beyond me.

Just found this year old question when going over the thread, may be too late but I'll still answer it.

Single mode:
Uses just one single channel (either A or B) and counts pulses since last sample and in addition time between adjacent pulses. Because it uses only one channel and only the rising edges, the resolution is 4 times lower. On the other hand velocity is measured by hardware instead of interpolated by software which might explain the lower jerkiness at low speed. The biggest downside is that it can't determine direction. So when you're rolling backwards in forward gear it can't counteract.

A/B mode:
Like said uses both channels and uses bespoke hardware to count pulses. It is more EMI-tolerant because it will never count a transition twice and it gives you direction information. On the downside it cannot measure velocity directly as the timer is already used for pulse counting. Therefor every 10ms we count the number of pulses seen in the last 10ms. That comes with aliasing which is noticeable at low speed.

[muehlpower]

That explains that I see the speed on the CAN in steps of 41 RPM.

41 RPM/60 x 36 teeth x 4 Edges = 98,4 imp/sec, -> 1 imp/ 10ms

[Roadstercycle]

So now I am really anxious to hear from Damien what the difference is and will it hurt anything in single mode. I've been driving in single mode for weeks without knowing it so whether it is hurting anything is beyond me.

Just found this year old question when going over the thread, may be too late but I'll still answer it.

Single mode:
Uses just one single channel (either A or B) and counts pulses since last sample and in addition time between adjacent pulses. Because it uses only one channel and only the rising edges, the resolution is 4 times lower. On the other hand velocity is measured by hardware instead of interpolated by software which might explain the lower jerkiness at low speed. The biggest downside is that it can't determine direction. So when you're rolling backwards in forward gear it can't counteract.

A/B mode:
Like said uses both channels and uses bespoke hardware to count pulses. It is more EMI-tolerant because it will never count a transition twice and it gives you direction information. On the downside it cannot measure velocity directly as the timer is already used for pulse counting. Therefor every 10ms we count the number of pulses seen in the last 10ms. That comes with aliasing which is noticeable at low speed.

Thank you Johu, Yes, it is a bit late but still well received for others and my future builds. The Mustang that had that board in it is now in the hands of AEM-EV company for testing their EV products.

[Golf_mkIII_E]

Updated parameter set. Same car/setup, just improved performance and driveability.

Further details on the what/why behind the changes can be found here.



https://openinverter.org/forum/viewtopi ... 6&start=40



updated tesla LDU parameters.json

You seem to be pointing out that lowering your fweak would increase performance. Doesn’t your fweak describe the point where flux weakening begins, and therefor you would want that value to be as high as possible?

My fweak is currently at 305, but I would like to have more power at higher speeds. At about 100-120kmh I’m able to pull about 1100-1200A, but at 170kmh for example there’s only about 700A left at WOT. When pulling 1200A, pack voltage drops from about 375V to 335V with a cold battery. Haven’t tried with the battery at a comfortable temp yet.

I’m using the red PCB sport LDU.

[johu]

Lowering fweak means delivering MORE current EARLIER in the rev range. And that means more torque. Of course the drop-off past fweak feels harsher then. You can somewhat compensate with the fconst/fslipconstmax parameter pair, i.e. increasing slip past fweak.

That 100-180 time is quite impressive (and only legal in Deutschland)

[Golf_mkIII_E]

Ah, thank you! That makes sense.

I’ll first set my fslipmax a little bit higher, after which I’ll start some testing with the fconst and fslipconstmax parameters.

Lowering fweak isn’t useful for me, since I can’t put all of that torque down with my FWD. I’m not looking for more peak power, I’m only looking for keeping some more power at higher speeds. If increasing peak power is the way to go, I’ll try that.

I think I’m also gonna set my Fmax a little bit higher, because it decreases power very soon to not get any overshoots. I believe it’s at 625 at the moment. Top speed is 203kmh GPS.

[johu]

I'm amazed how far you guys push that stuff that I developed in the 35 kW Polo
So yes, for more top speed fconst and fslipconstmax are your friends (says he who didn't dare driving faster than 120 with his Polo)

[Golf_mkIII_E]

Awesome!

The Golf consumes a little over 400kW peak at the moment, according to my BMS.

My only concern at the moment is the maximum RPM the LDU can take. With my Fmax at 625 it’s at now, I should be pushing 18.750RPM if you don’t calculate the losses by slip. I want to set it to the same 650 Jon Volk is using, but I don’t really want to kill it yet.

[muehlpower]

I assume you have a wheel circumference of 1.8m, which corresponds to 18470 rpm at 205 km/h with no slip . No original tesla motor makes more than 18000 rpm.

Mount bigger wheels!

[Golf_mkIII_E]

I’m swapping to 16” rims because I’m upgrading my 288mm brakes to 312mm brakes. Wheel size will increase by about 4%, but I’m very limited by the Golf III wheel arches.

[jon volk]

Just saw this now. Running the lower fweak was required to maximize torque for a hard launch and its made up for on the top end by piling on a bunch of slip at fslipconstmax. For what it's worth, 1/4 mile trap speed did not improve with a higher fweak, only low speed acceleration suffered.

[Golf_mkIII_E]

I went on a small vacation with the electric Golf this weekend. I departed on my 250km (155 miles) journey with my untested higher fslipmax setting, but luckily I didn’t run into any problems or OC’s.

I can’t even remember when I tripped my last OC.

After those 250km, cell voltage had dropped from 4.05v to 3.50v. I can discharge to about 3.20-3.30v per cell. Since I didn’t even depart with a fully charged battery, I think I can do about 400km (250 miles) on a single charge… during winter! Not bad.

I do have to say, I’ve been very easy on the throttle during that trip. I also wanted to test what kind of efficiency the LDU can reach. Looks like it has been using about 17kWh/100km. It’s not bad, but my previous 110HP IPM motor could do about 12kWh/100km.

[johu]

I do have to say, I’ve been very easy on the throttle during that trip. I also wanted to test what kind of efficiency the LDU can reach. Looks like it has been using about 17kWh/100km. It’s not bad, but my previous 110HP IPM motor could do about 12kWh/100km.

Very impressive. Those 12 kWh were also during winter?

[Golf_mkIII_E]

It was during winter indeed! I haven’t tried what kind of efficiency it could achieve during summer. The car was already pulled apart by that time.

I’m currently testing some setups with very high Fweak and Fslipmax. I’m able to pull 1150A consistently now from about 95kmh to about 155kmh. Acceleration is really impressive.

[Golf_mkIII_E]

I've been thinking about something else;

As we all know, low speed open throttle performance is way better with low (~250) Fweak. I also noticed my LDU seems to be performing a lot better at higher speeds (above 150kmh/90mph) with higher Fweak settings (370-380), rather then by just using enormous amounts of slip.

Of course when driving with higher Fweak settings, low speed performance suffers a lot. How amazing would it be if Fweak could be made a variable value, which increases over increasing motor RPM?

[EV_Builder]

I've been thinking about something else;

As we all know, low speed open throttle performance is way better with low (~250) Fweak. I also noticed my LDU seems to be performing a lot better at higher speeds (above 150kmh/90mph) with higher Fweak settings (370-380), rather then by just using enormous amounts of slip.

Of course when driving with higher Fweak settings, low speed performance suffers a lot. How amazing would it be if Fweak could be made a variable value, which increases over increasing motor RPM?

Yes; this is something I plan to build but we should wait sometime until I get there.

Lookup table class and GUI is already somewhere on the forum.

[jon volk]

Making fweak variable with throttle was easy enough. No reason mapping it to RPM wouldn't work. Maybe Ill experiment when I get my car back on the road in a couple weeks. If you dont feel like waiting or changing the firmware then you can always alter it over CAN.

[Golf_mkIII_E]

I had a day off last week, and decided I wanted to test the Golf’s range. I live in the south of the Netherlands and drove to France and back without recharging. I did a total of 421km, which I’m very pleased with. On the first 150km I drove extremely cautiously and measured my “fuel” usage. It managed to do 12.67kWh/100km. Not bad!

[johu]

oh wow, not sure if any affordable OEM vehicle can do that. Hats off!

[jon volk]

It managed to do 12.67kWh/100km. Not bad!

What’s the full set of parameters for that economy? Even at light throttle I tend to see maybe 30kwh/100km or 2mi/kWh. I haven’t tried optimizing and that’s mostly doing 70-80mph on the highway in a soft top brick

[Golf_mkIII_E]

I will post the parameters when I get home. I do have to say I was only doing about 90-95kmh for achieving those figures, but it actually was in a very hilly part of Belgium. I haven’t tried optimising the parameters for economy either. Maybe it gets even better with lower Fslipmax and higher Fweak?

The Golf is pretty small and light for a car with a capacity of 66kWh. It weighs about 1520kg with driver at the moment.

When I really push it, I can consume 60kWh in 80km. But that’s only possible on open bits of German autobahn, with lots of accelerations and braking between 150 and 210kmh. The battery’s aren’t watercooled yet, but they seem to take it really well. Temperatures barely exceed 30 degrees celsius after those kind of punishments.

[Golf_mkIII_E]

Sadly, I have to report that the Golf has had an accident right after my last post over here.

The electric components weren’t hit and are still okay, but the chassis is bend which means I’m not going to rebuild it.

It truly has been a great adventure, and I learned a lot by doing it.

[EV_Builder]

Wtf! What a sad news... Are you gonna build another one?