Smart Car of America Forum banner
161 - 180 of 180 Posts

· Registered
Joined
·
11 Posts
It looks like very bad news - you definitely have a bricked HV pack.

The individual cell statistics (01 through 93) are showing very low numbers and some crazy high numbers.

To get the car running again, at a minimum you will have to drop the pack out of the car to try and individually charge the cells and send the BMS to get fixed (the code can't be cleared any other way).

Probably have some cells that would need to be replaced as well (the crazy high number ones - 14, 48 and 62), but that's just a guess from an amateur.
I have heard that charging low cells is a hazardous route. Is there a chance of getting a fire 🔥 in the future.
 

· Super Moderator
Joined
·
2,159 Posts
Couple of random observations:
Your line 6 shows that your car has not had the one computer update that came out for the 451. That should say SW:2014/18/0 and not SW:2013/29/1 like yours does. It also indicates that it has been 82 days since your battery has met the requirements to do a self test, but that's not news to you. The Ah numbers being over 50 are nice. However as DealMeIn says, it looks like you have bigger problems than these.

Len
 

· Registered
Joined
·
323 Posts
Can I get some help interpreting my arduino log?


-----------------------------------------
# ;mV ;As/10
01;42;18846
02;53;18759
03;50;18846
04;52;18802
05;51;18759
06;52;18759
07;43;18846
08;38;18672
09;41;18759
10;39;18802
11;28;18715
12;37;18846
13;57;19113
14;65527;18587
15;50;18978
16;47;18802
17;53;18802
18;53;18759
19;49;18759
20;51;18715
21;55;18759
22;34;18846
23;39;18890
24;49;18934
25;54;18890
26;48;18978
27;32;18846
28;20;18802
29;33;18759
30;29;18846
31;18;18759
32;5;18934
33;42;18890
34;0;18672
35;35;18802
36;42;18715
37;46;18846
38;53;18846
39;46;18846
40;48;18759
41;44;18802
42;39;18802
43;40;18759
44;48;18715
45;42;18672
46;40;18715
47;52;18715
48;62047;18934
49;51;18802
50;39;18759
51;41;18802
52;47;18672
53;42;18846
54;40;18802
55;34;18715
56;42;18846
57;48;18759
58;47;18759
59;39;18587
60;48;18759
61;33;18587
62;65530;18629
63;30;18802
64;24;18715
65;33;18846
66;31;18846
67;41;18890
68;33;18934
69;28;18846
70;31;18759
71;18;18544
72;36;18846
73;35;18890
74;32;18934
75;28;18934
76;29;18802
77;27;18846
78;32;18890
79;40;18978
80;36;18890
81;33;18890
82;26;18802
83;23;18759
84;30;18934
85;35;19023
86;29;18802
87;32;18802
88;43;18802
89;42;18890
90;15;18846
91;43;18802
92;42;18802
93;21;18802
As you got only some values starting with 65xxx it seems to me your cells are empty but the CSE PCB of the stacks still talk to the BMS. Means if you are fast (or in meantime disconnest the 12V battery) you could rescue your HV battery and car.

For the others, there is a software bug in the interpretation of the single value of the received voltage level in the BMSdiag software. If you read 65xxx it is a binary sign issue :cool:

So let's start the rumble and rescue your Smart :) Disconnect the 12V Battery and order a OpenPort V2 clone as you will need it for read out and reseting some of the failures.

In meantime all (and i mean all) parts of the HV battery are available :) i even own new old stock cells :cool:
 

· Registered
Joined
·
5 Posts
Thank you @Jmk2020 for your help!

I'm helping @walkersurf with his battery challenge. I have EV hacking experience (DIY and Tesla) and this is the first Smart car that I've touched.
My comfort level recovering cells that are over 1.7V is high. If any are below 1.7V then in my eyes the project takes on a different level of risk and may not be worth the effort.
At this time we are trying to decide if we want to commit to dropping the pack to manually measure cell voltages. If they are over 1.7V then we can justify diving into the full repair.

I understand little about the source of CAN values that the Arduino is sniffing. I noticed something interesting with the HV value that has me curious.
Last night it was 316.9 V and 19 hours later it has dropped to 316.3 V. I wonder if this is a value that is measured in realtime? If so, it would be very hopeful for the cells. I'm not thinking of any other HV source that it could be measuring that also drops slowly over time.

If it is measured in realtime, I imagine that the mv level cell readings (which are changing slightly) might be misleading due to bad boards or codes in the pack.

What do people think?
Is it possible that the 316.3V reading is true pack voltage?
If it is a real value then it's certainly time to rumble!


As you got only some values starting with 65xxx it seems to me your cells are empty but the CSE PCB of the stacks still talk to the BMS. Means if you are fast (or in meantime disconnest the 12V battery) you could rescue your HV battery and car.

For the others, there is a software bug in the interpretation of the single value of the received voltage level in the BMSdiag software. If you read 65xxx it is a binary sign issue :cool:

So let's start the rumble and rescue your Smart :) Disconnect the 12V Battery and order a OpenPort V2 clone as you will need it for read out and reseting some of the failures.

In meantime all (and i mean all) parts of the HV battery are available :) i even own new old stock cells :cool:
 

· Registered
Joined
·
323 Posts
Hi Omar,

A bad and a good message.
First indon't think the voltage level is live. But never the less we and the community already often rescued cells down to several mV.
The Li-Tec cells are really robust and could be not compared with the Tasla cells.
Only if a cell is not taking charge you run in the issue to maybe replace single cells. There in the past a break even point of several dead cells per stack/pack is a real limitation.

Only lessons learned here is tomstart slowly to pull the cells out of the low voltage area. I propose tomuse only about 50mA till the cells reach about 1V. After you could increase the current step by step to 100mA and 200mA till you reach 2,5V. After you could increase again to several amps if available and charge till then first cells reach 3,6V.
After check which cells are lower and asap single charge them also to 3,6V. This could easy be done within a few hours. Already if the lowest charged cell pass the 3,0V the system can catch it selve again.

Only issue are the maybe locked BMS and parallel triggered hw issue inside. Here Bill in this forum could help in the US.
The second issue could be a already died current sensor.
Here you should disconnect immediately the connector or disconnect the whole BMS or open the busbars. To prevent of further possible damage.
You could read out the bms but already if you turn on ignition to do so, when you tried to read out the stored failures you triggered the storage of the bad P18051C which is locked and need the manuell reset inside the BMS.

The most others disappear automatically. Only be aware the 12V system must never drift below 11,7V. Only if you disconnect it fully this area is allowed. Otherwise the HV battery is triggered by a side effect to recharge the 12V system which is no more possible. This effect a possible further damage of the bms.

So take the car in the shop and it can run already in a few hours or depending the equipment several days.

Best you post the already ready out failure of the whole system by a suitable tester supporting the electric Smarts. I propose direct to jump to a J2534 compatible adapter like OpenPort V2.0 or similar and use Xentry/Das or mor technical vediamo V4.2.2. But there much better and more expensive solutions available ;-)
 

· Registered
Joined
·
5 Posts
Update:
I've removed the pack from the car and checked voltages.

The good: The pack measures at 315V, so the Arduino sniffing setup did seem to be reporting actual voltage.

The bad: This seems to be a unusual failure compared to what I'm reading about here. 92 of the cells are between 3.330V and 3.452V. One cell in the middle of the pack, #48 from the bottom, is at 6mv. I don't yet know if it's an electronic failure of the balancing circuit or a bad cell.

My plan is to start a recovery process on that cell on Sunday and see how it responds. If it does not recover in good shape then I plan to learn how difficult it is to replace a cell. At that point I could look for a good cell in the USA. ps: I'm located in Corvallis OR.

I did immediately unplug the current sensor, and then unplugged the BMS master. Any cell recovery will only be tried at ultra low currents outside where a fire would not be a big problem. My "battery workbench" is a trailer that I can quickly roll out of the shop in case of a "thermal event."

Other thoughts: These are really cute cars! The battery is so compact! I find it's much easier to access than on a Tesla.
Many thanks to all the help on this forum for making this job relatively simple. I appreciate that so many others have documented a path making it much easier for me.

Electrical wiring Audio equipment Electronic engineering Engineering Electronic instrument
Tire Vehicle Automotive tire Automotive design Wheel


-Otmar


The car did already have a P18051C as well as many other faults.

I will go ahead and remove the pack, try bringing cells back and report back what I find.
Thank you!
 

· Registered
Joined
·
323 Posts
Hello Otmar,
I really wonder about the BMSdiag readout. But indeed you have the issue of one single died cell.
You could try to charge the single cell starting with 50 to max. 100mA and see if charge is taken.

Please check also cell 14 and 62 as the manual measurement of cell 48 fits to the BMSdiag readout.

If you measured the voltage via tester you could check the flex connection between the taps on the single cells and the CSE pins but don't make any shorts.

Never the less i fear you need to exchange single cells.
As the BMS need to be reset contact Bill maybe he can helpout with some reused cells.
The transport of the charged cells from Europe to US seems to me a nightmare by transportation regulations. Even they are available :) See picture below.

Please post a list of all failure you read out. Rectangle Material property Gas Font Gadget


Rectangle Material property Gas Font Gadget
 

· Registered
Joined
·
5 Posts
Hello @Jmk2020

I just rechecked 14 and 62 (# from + and also from -) and they are doing fine.
I don't know if cell #1 is most negative on a Smart, or if #1 is most positive as I remember it to be on a Tesla.

Now that I've disconnected the BMS, the pack voltage is holding steady at 315.4V. That's a relief.

I was planning to follow your suggestion to slowly charge the one low cell, but this morning when I came in the shop it smelled like bananas, that familiar smell of electrolyte. Because of that smell I believe the bad cell has vented and is not recoverable.

Since theBMS was not giving reliable cell values, I did check every cell using a DMM on the CSE taps. I also triple checked the low cell just to be sure I had that right. My records are here:

Font Parallel Symmetry Pattern Paper product


Due to the electrolyte smell, I do think I'll need to try to exchange the dead cell with a good (used?) one.
I've sent a message to Bill Kichman to see if he knows of any cells available in the US.

As for how to replace on cell in the module:
This video gave me some idea of what is inside the modules.


I'm concerned about removing and reattaching the CSE taps, and unsure how to reliably reattach the cells to the cell power interconnects.
I suspect replacing cells may already have been covered on this forum, but my searching is not showing good results. If anyone has a link or a description of how to change out a single cell, I would appreciate it very much.

Also If anyone knows of single good cells for sale in the USA, please let me know. I saw online that the part number may be a ICS 13/330/162.

Thanks again!

-Otmar

Hello Otmar,
I really wonder about the BMSdiag readout. But indeed you have the issue of one single died cell.
You could try to charge the single cell starting with 50 to max. 100mA and see if charge is taken.

Please check also cell 14 and 62 as the manual measurement of cell 48 fits to the BMSdiag readout.

If you measured the voltage via tester you could check the flex connection between the taps on the single cells and the CSE pins but don't make any shorts.

Never the less i fear you need to exchange single cells.
As the BMS need to be reset contact Bill maybe he can helpout with some reused cells.
The transport of the charged cells from Europe to US seems to me a nightmare by transportation regulations. Even they are available :) See picture below.

Please post a list of all failure you read out.
 

· Registered
Joined
·
323 Posts
Hi,

David Poz did a amazing job in these days to understand the Smart battery stacks and how to work on them.
Even in meantime the knowledge is improved and the approach of spot welding seems to me not the most favorite one anymore.

Also how to work on the stacks to replace a single cell could be done much more easy and more efficient.

If you turn the cell you may can see the valving are with the blow out.
The most efficient methode would be to clamp the stack wit a clamp and remove the long screws.

In parallel remove the taps of the dead cell and maybe some before and after to take of force.
The turn it by 90 degrees and slowly release the clamp.
The upper connection between the busbar is best sepersted with a cutter knife. You even could use a motorized cutter knife.
Then you could spread the bottom which is now accessable sidewards.
So you could take out the bad cell and insert the working one.
Afrer you beed to solve the upper connection by one of the different solutions and reconnect the sense taps by soldering or with a small shield metall screw.

So not as difficult when you have done it several times :)
 

· Registered
Joined
·
5 Posts
That sounds like great advice.
I can imagine replacing the one cell the way you describe without disturbing so many cell taps. Very elegant!
I'm willing to take pictures, or make a video when I replace the cell so other people can see one way to do it.

What methods are people using to connect the power tabs?
I have a spot welder, but I worry that it would not be reliable on the aluminum tab. I could test it on scrap parts.
I think some mechanical clamp might be good. Is there room to get in there with pop rivets?
Or maybe very carefully with screws and nuts?
Or maybe several strong metal clips to clamp the conductor to the bus bar? That might be the easiest. "Stainless Steel Spring U Clip" looks good on google if I can find small enough ones.

I hope that I can find a replacement cell for the pack.
Today I'll add a little charge to the rest of the cells and set it aside until I can find a replacement cell.

Thanks again for the great advice!

-Otmar

Hi,

David Poz did a amazing job in these days to understand the Smart battery stacks and how to work on them.
Even in meantime the knowledge is improved and the approach of spot welding seems to me not the most favorite one anymore.

Also how to work on the stacks to replace a single cell could be done much more easy and more efficient.

If you turn the cell you may can see the valving are with the blow out.
The most efficient methode would be to clamp the stack wit a clamp and remove the long screws.

In parallel remove the taps of the dead cell and maybe some before and after to take of force.
The turn it by 90 degrees and slowly release the clamp.
The upper connection between the busbar is best sepersted with a cutter knife. You even could use a motorized cutter knife.
Then you could spread the bottom which is now accessable sidewards.
So you could take out the bad cell and insert the working one.
Afrer you beed to solve the upper connection by one of the different solutions and reconnect the sense taps by soldering or with a small shield metall screw.

So not as difficult when you have done it several times :)
 

· Registered
Joined
·
323 Posts
In meantime there are several methodes to connect cells. Some solder the busbar edge to edge (not my favourite), some solder the busbar overlapping (seems to me give some issues with the cooling surface).
Others are really use screw/nuts or rivets.
Original it ultrasonic welding is used. Big issue seems to me the next busbar is very close and a lot of energy is there.
Some colleagues use a half busbar to go a way of clamping and parallel use high filled adhesive.
Other use a kind of U-Shaped clamp.
There are several methodes here shown in the forum.
I had a idea of a kind clamping by filling the empty area of the bus bar (opening with a kind micro clamp).

The big challenge is to find a way which overcome the issue of the high energy potential between the cells.
 

· Registered
Joined
·
531 Posts
I just watched several Dave Poz videos re the 2013 SmartED battery. He still sees no need to NUMBER his videos so you have to dig them out to follow properly. So I'm
finally seeing his video where he removes bad or excess cells from the pack. It clarifies the job one faces in doing the removal and replacement of a cell
but I don't think he does any spot welding like the factory does. Unless I missed it he relies on pressure alone to electrically connect the aluminum cell to the copper
top pieces. He tried to solder these together but of course found that to be impossible. And he is not using these in a car so wtf your mileage just might vary when you
least expect it.!!

As time marches on it seems unlikely that all this effort to use such old cells is going to be successful for any of us. I'd rather use brand new physically smaller cells
with open-source Toyota inverter ala Damien Maguire to get my ED back on the road. Smaller so that they might properly fit inside the battery box :)
--
 

· Registered
Joined
·
323 Posts
Seems you missed the videos from Dave where he worked on his spotwelder. But as i wrote, spotwelding is not the right choice. You need to dismantle the whole CSE before otherwise the current pulse could kill by overvoltage the CSE inputs.

And your proposal to replace in direction of other HV system by replacing battery and drive inverter is similar to replace the engine if you have some leakage on full tank. :) If you really struggle on fixing the hv battery to convert an overall powertrain seems to me very challenging and over the top.
 

· Registered
Joined
·
1,613 Posts
Seems you missed the videos from Dave where he worked on his spotwelder. But as i wrote, spotwelding is not the right choice. You need to dismantle the whole CSE before otherwise the current pulse could kill by overvoltage the CSE inputs.

And your proposal to replace in direction of other HV system by replacing battery and drive inverter is similar to replace the engine if you have some leakage on full tank. :) If you really struggle on fixing the hv battery to convert an overall powertrain seems to me very challenging and over the top.
JMK2020,

Regarding replacing the cells... Could at least 50Ah (and hopefully more) of 18650 cells in parallel fit in the space of the OEM pouch cells (i.e. to begin with, are the pouches no more than 18 mm thick?). If you could fit more Ah in the space of each pouch cell, would the BMS syatem "recognize" the extra capacity so it could be used?
 

· Registered
Joined
·
323 Posts
Yes it is possible. Bill already showed to replay bad cells against other with higher capacity. As you know the worse in the serial connection define the overall max capacity.
I just got a capacity measurement result of the original "new" cells from 2015 i found. The have 50Ah, so no need to go for others :)
 

· Registered
Joined
·
1,613 Posts
Yes it is possible. Bill already showed to replay bad cells against other with higher capacity. As you know the worse in the serial connection define the overall max capacity.
I just got a capacity measurement result of the original "new" cells from 2015 i found. The have 50Ah, so no need to go for others :)
Yes, but what I had in mind was building a whole new battery pack using 18650-sized lithium cells - about 20 of them in parellel replacing each OEM cell. Hey... it is what Tesla did or may still be doing.
 

· Registered
Joined
·
531 Posts
I carry a 33 gallon water tank in the rear of my ED so the extra 250 pounds of weight is really no big deal on city streets 30mph max speed.
My point is that the ED can indeed carry extra weight easily ie changing from Li-C-Mn (??) original cells to the heavier LiFePO4 cells
should not be a problem. Finding room for them is the problem of course. The ED will not be managing these new cells any longer so you
as the owner will have to devise new BMS and charging schemes. You will now enjoy the ability to charge them to 100% and discharge to
0% and leave them at either extreme for as long as you like. This gives extra operationg range compared to OEM stock assuming you
have matched the KWH battery capacity.

Tesla is still using 1865 cells for their 400v and 360v motors so perfect for Smart EVs. The cell-packs are very sturdy so no need to break
out the individual 1860s.
--
 

· Registered
Joined
·
323 Posts
Just remember the BMS and CSE measure the cell voltages and balance. Working range is from 3,4V to 4,2V. If you want to use the maximum low range, 3,2V is reached. If you could live with your cell chemistry the system could handle it and it would work.
 

· Registered
Joined
·
1,613 Posts
I carry a 33 gallon water tank in the rear of my ED so the extra 250 pounds of weight is really no big deal on city streets 30mph max speed.
My point is that the ED can indeed carry extra weight easily ie changing from Li-C-Mn (??) original cells to the heavier LiFePO4 cells
should not be a problem. Finding room for them is the problem of course. The ED will not be managing these new cells any longer so you
as the owner will have to devise new BMS and charging schemes. You will now enjoy the ability to charge them to 100% and discharge to
0% and leave them at either extreme for as long as you like. This gives extra operationg range compared to OEM stock assuming you
have matched the KWH battery capacity.

Tesla is still using 1865 cells for their 400v and 360v motors so perfect for Smart EVs. The cell-packs are very sturdy so no need to break
out the individual 1860s.
--
I carry a 33 gallon water tank in the rear of my ED so the extra 250 pounds of weight is really no big deal on city streets 30mph max speed.
My point is that the ED can indeed carry extra weight easily ie changing from Li-C-Mn (??) original cells to the heavier LiFePO4 cells
should not be a problem. Finding room for them is the problem of course. The ED will not be managing these new cells any longer so you
as the owner will have to devise new BMS and charging schemes. You will now enjoy the ability to charge them to 100% and discharge to
0% and leave them at either extreme for as long as you like. This gives extra operationg range compared to OEM stock assuming you
have matched the KWH battery capacity.

Tesla is still using 1865 cells for their 400v and 360v motors so perfect for Smart EVs. The cell-packs are very sturdy so no need to break
out the individual 1860s.
--
As a correction, LiFePO4 cells are lighter, but somewhat bulkier, for the same amount of kWh compared to NMC . Their permissible voltage range is different - never go above 3.7v, never go below 2.5v.

Their pluses are a very flat discharge curve staying right at about 3.28 to 3.32V under a 0.5C load from 95% DOD all the way down to 10% DOD, and better cycle life. Iron and phosphate are also, of course, far more abundant and cheap materials than nickel and cobalt.
 
161 - 180 of 180 Posts
Top