Actually, I revisited the handling of ChaDeMos in the optimizer and simplified it a bit, and now it can actually give you routes for Ottawa to Halifax including seemingly useful alternatives. Try it out.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Actually, I revisited the handling of ChaDeMos in the optimizer and simplified it a bit, and now it can actually give you routes for Ottawa to Halifax including seemingly useful alternatives. Try it out.
RTPEV
Active Member
This enhancement helped and hurt me at the same time!
I have a particularly challenging route in that the last Supercharger on my way home is around 250 miles away, somewhat of a stretch in a Model 3 LR. However, there are 2+ CHAdeMO opportunities, and a meal stop that I've been modelling as being able to do a destination charge at to make the trip doable. Despite the CHAdeMO opportunities, ABRP never seemed to pick them.
Now, under certain (unrealistic) conditions, it does at least select one of the 2 CHAdeMO charging stop all on its own (not the one I would pick--it picked the one ahead of the meal stop, rather than after it which would make more sense).
But the thing that makes no sense to me is that it will only find a route at all if I do the following:
- Set the consumption to an unrealistic 214 Wh/mile (rather than the default 247) (215 does not work!)
- Set the charge to at the meal stop to 90% or more.
Then it has me stop at the meal stop, arriving at 88% and charging for 48 minutes to 100%.
And then it has me arriving at my destination with 51% left (I have goal destination SOC set to 5%), so clearly charging to 90% at the meal stop is not necessary, and I suspect a consumption of 247 Wh/mile would work out just fine as well (it's not an issue with the route PRIOR to the first CHAdeMO--there are plenty of Superchargers before that point). And even though I can specify charge to 90% at the meal stop and it works (only requires a 9 minute stop since it only needs to charge 2%), when I change the meal stop from "charge to 90%" to "charge for 1:00", it doesn't work (even though I'm giving it 51 minutes than it needs!)
It's also doing a strange route that it didn't used to do. Despite plenty of Superchargers on I-81 in New York Pennsylvania, it has me getting off 81 to hit a Supercharger on I-80 and taking a minor road to eventually get back to I-81, probably about 90 miles out of the way. I can't make sense of this at all.
If you want to replicate, you can try this:
From: Hammond, NY
Via: Blue Mountain Brewery, Critzers Shop Road, Afton, VA (I set charge at 11kW to 100% here)
To: Del Webb Arbors Drive, Durham, NC
WannabeOwner
Well-Known Member
Perhaps because CHAdeMo will put on more miles-per-minute the greater the SOC - so, with CHAdeMO, you will charge faster earlier in your trip, rather than later - the converse of Supercharging.
Of course you do have to have used up the battery-space equivalent to what you need to reach your destination before you can take on that charge
Maybe that is what ABRP is doing?
Peter_M
Member
RTPEV
Active Member
If this is what ABRP is assuming, it's just plain wrong. CHAdeMO is no different than Supercharging, or any Li-ion battery charging system. All Li-ion charging will taper off as it gets into the CV regime and nears full.
Now if you are referring to when it's in CC regime, then yes, you get a very slight power increase as the battery charges because the pack voltage is going up slightly, but this is almost unnoticeable. And again, technically no different than Supercharging.
WannabeOwner
Well-Known Member
I'm non technical, so may be speaking out of my wrong end ...
Here are two recent CHAdeMO charges, the first during cold weather, the second warm weather.
50% 55.0 MPH
55% 67.7 MPH
60% 79.4 MPH
65% 85.9 MPH
70% 90.0 MPH
75% 93.1 MPH
80% 96.0 MPH
85% 97.5 MPH
90% 96.7 MPH
47% 105 MPH initially, then fell to:
54% 92.0 MPH
60% 93.6 MPH
70% 98.7 MPH
80% 101.9 MPH
88% 102.9 MPH
Figures from TeslaFi's logging of my charging session (maybe there is an issue with the data logged / presented?)
RTPEV
Active Member
Sorry, fairly long technical, and off-the-main-topic-but-a-continuation-of-prevoius-discussion post follows...
When DC fast charging (whether CHAdeMO, CCS or Supercharger), the car's battery management system will go through 2 phases: Constant Current (CC) and Constant Voltage (CV). At low state of charge, it's in CC mode. Basically the car commands the charging station to supply up to a certain number of amps (current). In the case of a Tesla, this might be something like 300A. Now a given charging station may or may not be able to honor that request. A typical CHAdeMO station, for example, maxes out at 100A so clearly it won't be able to supply that much, but it will do the best it can. Even a Supercharger may cut the max current due to either sharing the supercharger with a paired bay, or even weather-related issues (maybe the charging station is overheating and is throttling the max current). During this phase, as the battery charges, the pack voltage will slowly rise as a result of the charging (I don't have actual numbers, but for purposes of this discussion, let's say it goes from 350V at 0% to 400V at the cutover point. Because POWER is CURRENT times VOLTAGE, and CURRENT is constant, and VOLTAGE is slightly increasing over time, ideally you will get a slight increase in power (or equivalently MPH in your chart above) during this phase. At a CHAdeMO station that can supply 100A, it will start out at 100*350 = 35kW and as it nears the 400V threshold, it will be supplying 40kW. In terms of MPH, assuming 300Wh/mile, the 35kW equates to 116mph and the 40kW equates to 133mph. So it would be normal to see a gradual increase in power during this phase in that order of magnitude.
Once the pack reaches a certain threshold voltage, the car commands the charging station to switch to CV mode. In this mode, the voltage is held to a value slightly above the target pack voltage (let's say that's 410V for these purposes), and the higher voltage of the charging station causes a certain amount of additional current to flow into the pack, but basically the battery itself dictates how much it is going to accept. The current at this point drops pretty fast until it reaches another threshold, at which point the car declares the battery full and shuts down the charging station. This is the so-called taper region where the charging speed drops very quickly and should be quite noticeable to someone charging their vehicle (as opposed to the slow gradual rise in the CC phase). Here is a picture that illustrates the effect pretty nicely (for a Nissan LEAF):
In this chart yellow is the current, blue is the voltage, and green is the state of charge. For the first 29 minutes or so, the charger is in CC mode supplying about 120A. Once the pack voltage reaches about 395V, it switches the CV, and you can see the current drop very quickly. This happens at around 83% state of charge in this picture.
But I will say that every car is different. My 2016 LEAF seems to get up very close to 90% SOC before tapering off. My 2012 LEAF would start tapering off at about 70% when new, and as low as 40-50% when older. I think certain Teslas start to taper off at very low SOC, but newer ones don't. And as I said, weather conditions and what the charging station is capable of supplying all play a role.
Now it may be that the Tesla did in fact switch to CV mode at a fairly low SOC, but the battery was still able to accept the full 100A or whatever from the CHAdeMO station for quite some time after that (if the car was plugged into a Supercharger and switched to CV mode, the current draw would drop quickly from 300A, but it would still take some time to drop below 100A). So the charging power would still continue to climb until the current draw fell below 100A. But eventually it will start to fall off a cliff.
RTPEV
Active Member
Cool. So I suspect what is probably happening at CHAdeMO is that the car shifts to CV at around the 21% SOC (I'll take your word for it--the scale of the battery level line is not obvious in the graph), but the battery happily sucks in the max current from the CHAdeMO charger until it drops below the 120A point (approx 1:34PM in the Supercharger example). I would assume that once the CHAdeMO charger starts truly tapering off that the end of the charge would look alot like the Supercharger session (it looks like in this plot the Supercharger charging session terminated at a lower SOC than the CHAdeMO).
Peter_M
Member
In both cases, I stopped charging when I had enough; I didn't wait for it to finish.
WannabeOwner
Well-Known Member
Hadn't thought to post the TeslaFi graphs, herewith:
09:10 50% 55.0 MPH Outside temperature 0.5C, severe windchill
09:21 55% 67.7 MPH
09:29 60% 79.4 MPH
09:36 65% 85.9 MPH
09:43 70% 90.0 MPH
09:50 75% 93.1 MPH
09:56 80% 96.0 MPH
10:03 85% 97.5 MPH
10:11 90% 96.7 MPH
Charger had a one hour time limit. Restarted the charge 2 minutes later, but some ramping up does not make it possible to combine them, per se.
10:15 91%
10:20 94.x%
10:30 97.x%
10:40 99.x%
10:50 100.x%
11:15 stopped (one hour limit reached)
06:46 47% 105 MPH initially, then fell to:
06:57 54% 92.0 MPH
07:06 60% 93.6 MPH
07:20 70% 98.7 MPH
07:33 80% 101.9 MPH
07:45 88% 102.9 MPH one-hour limit reached
Outside temperature 10C
09:10 50% 55.0 MPH Outside temperature 0.5C, severe windchill
09:21 55% 67.7 MPH
09:29 60% 79.4 MPH
09:36 65% 85.9 MPH
09:43 70% 90.0 MPH
09:50 75% 93.1 MPH
09:56 80% 96.0 MPH
10:03 85% 97.5 MPH
10:11 90% 96.7 MPH
Charger had a one hour time limit. Restarted the charge 2 minutes later, but some ramping up does not make it possible to combine them, per se.
10:15 91%
10:20 94.x%
10:30 97.x%
10:40 99.x%
10:50 100.x%
11:15 stopped (one hour limit reached)
06:46 47% 105 MPH initially, then fell to:
06:57 54% 92.0 MPH
07:06 60% 93.6 MPH
07:20 70% 98.7 MPH
07:33 80% 101.9 MPH
07:45 88% 102.9 MPH one-hour limit reached
Outside temperature 10C
Not sure if you guys agreed on everything, but I think you got it right: ChaDeMo is specified to max out at 120A, which actually means that once the battery gets more charged, and therefore gets higher cell voltage, the max power drawn from the ChaDeMo will increase. (Power = voltage times current.) Once the power limit of the battery is met (given the battery type and SoC), this will limit the power, but typically for a Tesla battery and a ChaDeMo, this happens for pretty high SoC.
ABRP takes this effect into account.
ABRP takes this effect into account.
Optimizing routes with many charging options is a very hard problem to solve, and adding many ChaDeMos throughout an area certainly does create a lot of charging options. ABRP tries to find a reasonable solution with a limited effort when you enable ChaDeMos, but since the effort is limited the result is sometimes not very good. It works well when there are not too many options, like in the Nova Scotia case.
ABRP will always return the optimal solution (plus some good alternatives) in the case of only Superchargers, since by design the number of charging options is relatively limited. In the future, once there are plenty of superchargers or equivalent everywhere, optimal planning will be a hard problem to solve and at the same time, rather uninteresting
Has anyone tried out ABRP with the new MCU web browser yet? I'd love to hear if it is usably fast and if there are any issues with it.
HankLloydRight
No Roads
I haven't really looked a ABRP in a long while, but I'm preparing for a 2500 road trip this summer, and threw some things into it, and wow, I'm really impressed with the improvements, new features, better usability, and overall quality. It really does everything I used to do by hand when planning a trip (like calculating total trip time including charging stops, etc). I really like how I can just click on alternate routes just like google maps, and the trip details change instantly. Really slick. Thank you.
Hi,
I just started using your program. I took delivery of a new S75D with the new MCU (verified by teslamcu.xyz) last week and I am trying to figure out how to use your program in the car. It loads fine in the car's browser, seems snappy enough, and I can log in to my account but I don't think the car's location is being updated on the screen as the car moves. Is this normal? I have real-time navigation enabled. Am I doing something wrong or is the program not designed for real-time navigational use in the car?
Similar threads
- Question
