Smart Car of America Forum banner

521 - 540 of 559 Posts

·
Registered
Joined
·
213 Posts
When considering buying a wreck at an auction Co-Part shows photo of dash which would ideally show the miles remaining on the hv battery, or 'mechanical limits' meaning you have the dreaded 'smart bricking syndrome', or whatever. The best is to see the miles which means the batt is good. My (#1)Spark_EV showed the miles so I grabbed it (even though 'airbags deployed' meant a whole 'nother mess to deal with) since I was primarily after a good batt. My #2 Spark_EV had 'no airbags deployed' and was listed as 'runs & drives' - this is the best auction car to purchase!! This one I took home and drove it off the trailer and use it daily on the ranch. Go to copart.com and knock yourself out. 8^)

LKQ knows how to spot a good hv batt, mine is described a 'grade A', so I am hopeful.
--
 

·
Registered
Joined
·
213 Posts
Thinking of the future:

This thread shows the endless battle ahead for owners who run afoul of the security constructs of Daimler, Inc. The answer here is to create an 'open-source' inverter control board that can run the smart inverter directly, no 'permission' from Daimler needed. Ok, the o/s inverter control boards already exist so we just need to choose one that would best run the smartED MOSFETs in the smart inverter. Or you could use a Prius inverter/open source board instead but this is much more powerful than we need for our little 80hp motor.

I have not reached this level of desperation where I would remove my smartED inverter and replace its logic board with an open-sorce board, but that day may not be far off. The smartED is such a unique ev that we should work to perfect this solution. One benefit would be that you could use any hv batt that you wished: lead-acid or any Li-Ion chemistry you choose. You would be free to choose a voltage range and run whatever BMS would match your series/parallel cell arrangement.
--
 

·
Registered
Joined
·
27 Posts
In salvage auction photos you (typically) CAN'T see if the smart HV battery is still holding a charge. Unfortunately they usually take a photo of the odometer only, and on the smart ed that part of the dash tells you nothing. Occasionally you might see the dreded hv system workshop message on the dash, but usually the photo is zoomed in on just the odometer. Either way, you can't see the battery charge indicator area (range remaining) .You'd have to see the part of the dash further towards the center, which I've never seen in Copart dash photos... The only metric you can go by is "runs & drives" , or go look in person if/when possible... BUT You are still taking the big risk that it dies while sitting (several month auction process). Most if not all airbags deployed smart ed's are going to have a dead hv pack..from what I can tell. Not comparable to most other brand ev's.

I would caution against Junkyard packs in general, unless you can get some kind of guarantee that it came out of a still moving/driving/etc car! "Grade A" may be based on appearance only in my experience. Crossing my fingers this is not the case for Vincent Rapide
 

·
Registered
Joined
·
213 Posts
After a recycler (LKQ) buys a wreck they will give it 12v and read the miles off the dash. If they see an empty hv batt they will not waste any more time on it except to remove it and send it off to battery heaven. They do report how many days the hv batt has sit on their shelf, but 3 or 4 months means nothing to a good hv batt. They will then strip car of components to put on their shelves to sell for cash. Why is this not an acceptable business model?

'Runs & drives' is what to look for if you want to buy an auction car yourself. Most users here will only consider local auctions anyway so go visit the yard before you bid and check out the vehicle. I've never visited but I think Copart will allow you to apply 12v from your buzz-box and turn the key ON to read the dash, especially if they failed to make a photo themselves. Call the yard to verify this. Then you can bid with confidence.
--
 

·
Banned
Joined
·
19 Posts
GoSmart Service, am I understanding you correctly that if we get a "bricked" battery, we can simply remove the battery from the vehicle, re-charge the individual cells, re-install the battery in the car.

Then remove the BMS, send it to you for clearing of the error, re-install the BMS back into the car after we get it back from you, and then everything is working again?
Yes
1. Need to charge cells to 3.7v
2. If u have error P18051C, just send me BMS, I will fix this error, and u will have a running car 👌
My solution is plag&play, but don’t forget about current sensor “Essential to unplug/remove before attempting anything“

before disconnecting or doing something, you need to remove the connector from the current sensor, and connected back on the last step before closing the battery.
 

·
Registered
Joined
·
29 Posts
Time to speed up the discussion a little bit again and unhide some information :cool:

This is the 32 pol CMC (Molex) connector of the BMS. (by the way nice breakout adapter )

64499


You need to connect GND on H1
+12V on H2 (KL15) and H3 (or H4) (KL30)
+12V on G4 (KL30C) as this pin is the supply of the upper pins of the contactor coils coming indirect via pyrofuse. By this way in case of emergency the BMS is locked down. By the way you can override the Pyrofuse to charge the battery. It only open the KL30C loop and also the Interlock loop.

Connect CAN Bus on EV_CAN Low and High and add a 60 Ohm resistor as the Multiplexer (or OpenPort 2.0) and also the BMS on the EV-CAN pins do not own the necessary 120 Ohm termination resistors inside.

By the way the idea of junk yard mode is rubbish, because the CRASH_IN pin es even not connected at all on ED3 :) There is a hidden electrical damage, which need to be corrected to be save on this.

Also the Current Sensor which need to be disconnect is a need by saving it for becoming destroyed by the issue of the brownout or better coming back of the battery stacks during charging.
Also the P18051C is not related to the stacks and 0V cells at all :) The situation of a not defined status by discharged batteries and cells cause an issue on a internal power supply (and read out of the system) as soon you power up the main part of BMS in the unhealthy condition again.

I can tell you also the HV measurement part is supplied by the whole serial voltage of all 3 stacks. The minus is the connection on the current sensor. The plus is direct connected to the + terminal of the upper contactor.
The main controller part of BMS switch on the relays, which power up the isolated HV measurement part. Also the current sensor is switched on by one of these relays.

I know and understand the solution from our friend of the Ukraine, but i don't like to use the sledgehammer instead of the fine skalpell. ;-) And yes even single cells could be exchanged and stacks could be repaired.

So lets dig in step by step and please report you success :) More information on Quid pro quo :)
 

·
Registered
Joined
·
213 Posts
After a recycler (LKQ) buys a wreck they will give it 12v and . . . [Vincent Rapide]
Let me update my post from 2 months ago: The battery I got from LKQ by the time I got to install it did not work in the car and proved to have 2 or 3 bad cells in each module. Only a 30 day warranty so my bad, I should have rejected it since they had it on their shelf for 133 days, the only info they provided. I charged each module from ~11vdc with 180 ma using a current-limited power supply as per the github pdf and also @mine_cookie and things were looking good, the empty cells were showing voltage but not as much as the good cells. The next day the bad cells were right back to o.oov while the majority seemed to hold most of what they showed when I turned off charging 8 hours earlier. So I can continue this slooow process but the few bad cells will have to be removed or replaced and this is likely to render them useless for running a Smart car. These batteries are factory built in non-modifiable form, they must be completely rebuilt or you have a mess on your hands trying to rejoin the foil connections between cells. You can't even solder over the bad cells or solder in cells you physically replace. All you can do is pop rivet or rely on crimping of the foil, a fool's errand.

So I'll convert this car to run a Prius gen3 inverter/converter which can run on any Li-Ion battery in the 300v range. I'll fill the Smart battery box with new cells and possibly add another box on the back floor for additional cells. Link: >openinverter.org<

Which Github pdf ?
The one where he pleads with the user to apply low current, like 200 milliamps, so as to not blow up the cells depleted to such low voltages. The entire exchange was wasted effort since the user had no idea how to use a current-limited power supply.
--
 

·
Banned
Joined
·
19 Posts
This all information is useless, there are no info about how to repair P18051C
I understand that until now you have not found how to subtract the processor and, most importantly, how to delete the fault code
it was easier for you to send me that unit and I would repair that code for u and you would have received a working one.


Time to speed up the discussion a little bit again and unhide some information :cool:

This is the 32 pol CMC (Molex) connector of the BMS. (by the way nice breakout adapter )

View attachment 64499

You need to connect GND on H1
+12V on H2 (KL15) and H3 (or H4) (KL30)
+12V on G4 (KL30C) as this pin is the supply of the upper pins of the contactor coils coming indirect via pyrofuse. By this way in case of emergency the BMS is locked down. By the way you can override the Pyrofuse to charge the battery. It only open the KL30C loop and also the Interlock loop.

Connect CAN Bus on EV_CAN Low and High and add a 60 Ohm resistor as the Multiplexer (or OpenPort 2.0) and also the BMS on the EV-CAN pins do not own the necessary 120 Ohm termination resistors inside.

By the way the idea of junk yard mode is rubbish, because the CRASH_IN pin es even not connected at all on ED3 :) There is a hidden electrical damage, which need to be corrected to be save on this.

Also the Current Sensor which need to be disconnect is a need by saving it for becoming destroyed by the issue of the brownout or better coming back of the battery stacks during charging.
Also the P18051C is not related to the stacks and 0V cells at all :) The situation of a not defined status by discharged batteries and cells cause an issue on a internal power supply (and read out of the system) as soon you power up the main part of BMS in the unhealthy condition again.

I can tell you also the HV measurement part is supplied by the whole serial voltage of all 3 stacks. The minus is the connection on the current sensor. The plus is direct connected to the + terminal of the upper contactor.
The main controller part of BMS switch on the relays, which power up the isolated HV measurement part. Also the current sensor is switched on by one of these relays.

I know and understand the solution from our friend of the Ukraine, but i don't like to use the sledgehammer instead of the fine skalpell. ;-) And yes even single cells could be exchanged and stacks could be repaired.

So lets dig in step by step and please report you success :) More information on Quid pro quo :)
 

·
Registered
Joined
·
29 Posts
So where is the real problem?
To access the BMS in bootstrap mode to read out the whole tricore?

Or your last approach to use the MPPS ?

Or to delete after searching for the signaturen the entry and correct the checksum?

I told the mean reason for the dead of the current sensor.:)

i had no need to buy as i just have a BMS to play but no dead battery which need it :)

As i told i know to do it by the sledge hammer (bootstrap mode) approach but i am even work on a solution to do it instant and remote if the rootcause for some of the failures, which a real hardware issues is not present. :)
In this case to open and solder on the BMS is necessary.
 

·
Registered
Joined
·
213 Posts
Curious innocent minds want to know: just what is blowing up that Current Sensor and why is it so sensitive?

I charge only per module so I'm not sending any (miniscule) current thru the cs.
--
 

·
Registered
Joined
·
29 Posts
Short:
It is a overvoltage on the supply pin of the current sensor. This is caused by the fading out and special fading on during discharge and charge process.
In parallel by the hard switching on and off of the relays, the Viper which is responsible for the supply of the isolated part cause high voltage spikes.

Much more details are there.

So only reset the software or delete the failure is sometimes not enough. We found additional issues on several BMS. This would not direct cause an problem but it is one reason of the discharge of the HV battery and higher leakage currents.

So start to think.

I looking forward on the upcoming warrenty issues on again broken BMS repaired in the Ukraine. I never read about the HW issues on the BMS and what would be the additional work necessary to save it from the issue again. I got a first hint in this direction and could not believe till we measured it in the BMS HW.

I can tell you by validation on several BMS it is sometimes not enough to work on the software only :) it may would work but it would have some side effects.
 

·
Banned
Joined
·
19 Posts
until you find a solution, these cars will already be in recycling

You better don’t spend your time for write post, better find a solution. it's just very funny that you are a super programmer at Infineon who cannot find a solution.

good luck, I am confident in your abilities 💪🏻

Which approach you mean?

I gave you the option to hold your business and make a deal. You take the infos from others and optimise your approach in meantime. In the same time you hide your questions and the option for the others to understand the way of solution :)

So where is the real problem?
To access the BMS in Bootstrap mode to read out the whole tricore?

Or to do this with the miniWiggler? Which request to open the BMS.

Or your last approach to use the MPPS V18 clone without to open the ECU?
Or to delete after searching for the signaturen the entry?
Or to crash the memory to reinstall the whole firmware?

I told you the mean reason for the dead of the current sensor. Your info is the result why it die, but not the rootcause for the burst :)

You also seems not to understand i have no need to buy as i just have a BMS to play but no dead battery which need it :)

As i told i know to do it by the sledge hammer approach but i am even work on a solution to do it instant and remote if the rootcause for some of the failures, which is a real Hardware issue is not present. :)

Your can help and save a little of your business or you can let it be and everybody ask will get it much cheaper (maybe free) than your offer without to send it around the world :)
 

·
Registered
Joined
·
29 Posts
Don't think so. ;-)

To all who want a cheap solution please send PN.

And i am not a super software guy. This is not necessary.
But you need electronics skills.
 
521 - 540 of 559 Posts
Top