[FIRST DRIVE] StreetScooter Repair and Reverse Engineering

[Heye]

Hey everyone,

I almost can't believe it, I managed to wake the vehicle up from its slumber today! My hardware seems to be a bit different from yours, @gb993. Cycling the ignition didn't help, and I also didn't get a sound from the buzzer. What did help was the procedure from this video:

• Keep drivers side door open
• Keep brake and gas pedal depressed fully, hold gear selector on D.
• Cycle ignition on and of (I had to wait a few seconds after enabling the ignition)
• Keep pedals and gear selector held

You should then hear the contactors click and the VCU should be reset.

I had to repeat the procedure twice. I think I didn't wait long enough while cycling the ignition on and off the first time.

It drove fine, though the brakes sound horrible. Too much acceleration once I lift my foot from the brake for my taste.
Many other weird noises, I'm going to assume they are normal for now. The parking lock error has gone as well.

I don't know whether the charger works yet. The cooling system seems to be active, but the values from the CAN bus are inconclusive. If it's charging, it's charging very slowly.

One bad thing: the cell voltages have begun to drift again in the same pattern as before. I'm almost certain there's a software bug in the BMS. I hope that I can do a good top balance by keeping the car on a charger for ages, but for that to work I have to have a working OBC, since it will have to limit the charging current to what the balancing resistors can dissipate.

The next step for me is getting the car roadworthy, though and I am sure that I don't need a working charger for that.

@muehlpower, I wasn't able to lift the car today, but once I do, will the outer dimensions of the battery and maybe some rough distances between the screw holes be sufficient? I'm hoping that I won't have to drop the battery again in the foreseeable future...

[muehlpower]

For now, the external dimensions of the aluminum base plate are enough for me to know whether a newer StreetScooter is an option. My battery from the old model was 930 mm x 1365 mm.

Too much acceleration once I lift my foot from the brake

This acceleration depends on the brake pressure transmitted by the ABS. In our show car, we removed the ABS and transmitted a constant value of 1000. ID 0x258 ... byte[1] = 1000;

[gb993]

Congrats @Heye.

Yes yours appears to be I think a B16. Mine are I believe S17 vans.

Good work though.

[Heye]

Today, I had some time to work on the truck. Besides getting the battery measurements and re-torqueing the bolts holding it in, I needed to do some preparations for the general inspection for getting the truck road legal.
For documentations sake: The battery bolts seem to be DIN 6921 Hexagon flange bolts. M8x1.25, 40mm length. I measured 25mm of thread length, but that's likely wrong. They are made from steel of grade 8.8, so I torqued all 23 of them to 24Nm after applying a small drop of medium strength threadlocker.

The outside measurements of the battery base plate are 1360mm (in direction of travel) by 1175mm. Both measurements should be accurate to +/- 5mm.

I started investigating the weak E-brake. As it turns out, when applying the E-brake, the only resistance it creates is between the lever inside of the brake drum and the ABS ring. I'm not quite sure yet, but maybe this is a design flaw? The brakes (currently, TRW GS8639 brake shoes are installed) seem to be of an old design, so maybe there didn't use to be ABS. I will search for some documentation and check whether the currently installed brake shoes are still in spec. If they are, this would be incredibly weird.

Edit: They very much are in spec. The brake material is 4.8mm thick, that's a lot more than the required 1.5mm.

[Heye]

Quick notes for documentations sake (Purely mechanical):

I installed a complete brake kit (apart from the brake cylinders, since the current ones are still good and I was on my own, so I wouldn't have been able to easily bleed the brakes) made by LPR. It's identifier is "OEK 419" and I found a good deal for it that isn't (unlike most other offers) significantly more expensive than just the bake shoes and cylinders. It's made for 230mmx32mm drum brakes, which seem to be somewhat uncommon, so local parts stores didn't even have them in stock. The new kit seems to fit very well and it was easy to install. I found a simple method that doesn't require much effort or tools, but wasn't shown in any of the tutorial videos I watched:

• Remove the nut from the handbrake adjuster in the cab (optionally measure the length of the threads for readjustment later)
• Remove the lower spring from the brake kit, then bring both lower ends close together and you should be able to fit the assembly through the gap between the brake cylinder and the wheel bearing
• Put the adjuster side in its final position and install the retaining clip and spring
• Reinstall the lower spring and the handbrake cable while the handbrake lever side is still loose
• Grab the lower metal part of the lever-side brake shoe with a pipe wrench or with vice grips and move it into position
• Install the other retaining clip and spring

This method offers the advantage that you don't need to remove the wheel bearing or the brake cylinder (although I had to remove one of the wheel bearings later anyways).

Here's what I believe happened: The stamped pressure bar that was installed looked very similar to the one from my new kit. I think it's possible that LPR parts were originally installed, and at some point, only the shoes and cylinders were replaced with TRW parts, while the existing LPR accessories were retained. I believe the hole pattern between the LPR shoes and the TRW shoes is slightly different (I sadly forgot to check), so that the tension of the upper springs was mismatched and the adjustment mechanism didn't work. This led to the handbrake lever (which additionally has a larger cutout in the critical area than the TRW part) having a larger travel after some brake shoe consumption and hitting the ABS ring. The adjustment works now.

On one side, the ABS ring completely came off the wheel bearing (Febi 09521). I had to take a hammer and a welder to the wheel bearing, which felt quite weird! But it seems to work fine now. I will replace the wheel bearing soon, it seems to be very available as a part that was used on the Golf IV.

Two further interesting observations: While the brake is obviously a drum brake, the wheel bearing seems to be made for cars with disk brakes! Maybe, this is the reason why there was a thin distance ring (that I, again, sadly forgot to measure) between the wheel bearing and the axle flange? This might also have been a contributing factor to the issues I was having previously.

Anyways, I hope I can get my car inspected tomorrow or at some later point next week. I'm thrilled!

[Heye]

Passed the inspection on the second attempt!

A stabilizer coupling rod and both stabilizer rubber bushings needed to be replaced.
The part numbers are:

• Febi Bilstein 07989 for the coupling rod (I bought the same part)
• Eibach 45-85-008-01-01 for the rubber bushings

The bushings don't seem to be available separately. I have only seen them mentioned in this mounting manual: https://www.fk-shop.de/out/media/M40-85-008-01-VA.PDF Since that custom stabilizer was made for the VW Polo 9N with stock mounting hardware, I tried to get a rubber bushing for the Polo 9N suitable for a 22mm stabilizer bar (which is the diameter used in the StreetScooter). The data that's available online seems to be of little accuracy, but I got lucky and was able to make the part "Febi Bilstein 31350" work, although I needed to use a lot of grease for mounting it. I pressed it on using two longer screws.

Hopefully, I will get the CoC papers soon and I'm excited to get the car street legal. The few kilometers I have driven so far were an absolute blast!

In other news, I ordered an "OWON CMS101" clamp meter from China and I'm now searching for the cheapest charger I can find in Germany that can charge to 436.8V. Do you have any recommendations on that? In the long run, I would like to repair the original charger, but until then I need something to keep the car going and balance the cells fully.

Also, do you have any favourite DC HV connectors for cable to cable/housing connections? Ideally, they would be rated for at least 600V/50A and touch-safe.

[Heye]

More updates (but progress is slow):

• I got a CoC document in the mail. Yay
• I had an appointment at the motor vehicle authority where I found out that there is an error in the CoC document (wrong type number), so I had to mail it back and am currently waiting for new ones...
• I bought the cheapest VAG charger I was able to find, but I need to reverse-engineer it before I can use it. Progress on that is slow.
• The clamp meter arrived and it's quite nice that I have one now, but the accuracy isn't that great. I can always use more turns of wire if it matters, so that's fine. In any case, I'm glad to have one now!
• I decided on safety banana plugs. They are cheap, touch-safe and reasonably common. But they have no protection against shorts, wrong polarity and disconnecting while under load. Also, the current carrying capacity isn't that great.

I made this blanking plate for the VAG charger that can take 4 banana plugs and replaces the HV gland: https://www.printables.com/model/140760 ... lanking-pl

[Heye]

After my troubles with the motor vehicle authority, I had to make a new appointment. It was supposed to be in October, since nothing works in Germany. Last Thursday, I had the CoC document in the mail, failed to get a spontaneous appointment at the MVH on friday (the worker gave a me a lot of misinformation at my last appointment, including the horrible advice that I should show up immediately after opening). So on monday, in the early morning, i wrote a script to scrape the MVH website for the earliest appointment to notify me on cancellations (expecting to get a new appointment some time this week) and literally two hours later, I had my new appointment for 14:00 that same day.

Otherwise, the bureaucratic procedures were uneventful and I was now able to legally drive my truck to and park it next to my home. It reports 21% charge left, it's high time I get the charger going!

[marcexec]

Glückwunsch / Congrats!
The paperwork seems such a small task in comparison but it's a) annoying and b) super satisfying to have completed.
If you're getting stuck on the charger, maybe get a "known good" one in the interim?

[Heye]

Thanks, marcexec! I thought about getting an MG ZS charger as well, since those are supposed to reach the relatively high voltage of the 104S pack. They are expensive, though and I'm trying to keep this project as low-budget as possible. Currently, I'm pondering just switching the OBC I have on and off in the 10 second intervals that it can charge, for now.

In other news, progress is slow. While inspecting the CAN log from a charging attempt, I noticed that the OBC reports 0 Amps and 0 Volts, but also no errors. This makes me suspect that the charging contactor might not even be closing. I hear multiple contactors click when plugging in the charging cable, but I can't tell which ones those are. If the OBC can't see any voltage anyways, I don't think there is any harm in disconnecting the HV DC plug from it and checking it directly. Only issue is that I can't get the thing off.

Does anybody know how to disconnect a plug like this? It's likely a Kostal-made plug, but they all seem to be different. It would probably be a lot easier if I was able to see the plug, but it's quite hard to reach on the underside of the inverter. I wasn't able to pull out the red part.

I tried connecting to the PTC HV lines (which wasn't very hard), but I can't get them activated anymore, since the battery is so low the vehicle is in limp home mode and won't activate the heater. Trying it while having a charge lead connected didn't help either.

The only HV lines that certainly get activated are those going to the inverter. It's a Bosch InvCON inverter, as it can also be found in the e-Golf. I'm hesitant to disconnect the big DC input leads, but it has a smaller connector for an AC compressor. Does anyone happen to know whether it is permanently connected to the DC input or somehow switched inside the inverter?

If all else fails, I will find someone to take the EV in tow with my ICEd van to regeneratively charge to get over the limp mode threshold. But that would be quite the hassle and I want to avoid that if I can.

[Heye]

Another thing I'm slightly worried about:

I read somewhere that (at least for a different EV, though I can't remember which one) there should be absolutely no current flowing, when the BMS tries to open a contactor, because it will think that the contactor is welded and not close it again until reset using the servicing device I don't have.

Is this a common thing? Do I need to be especially careful here?

[marcexec]

According to the Wiki the Outlander Charger/DCDC stops DCDC'ing at 397V (~3.8V per cell for you) but not necessarily charging as it needs external control to sense voltage / stop. Even if it does, 60% is better than 25%, right?

First of all you have to get the connector off though, hopefully someone can recognise it?

[manny]

Does anybody know how to disconnect a plug like this?

if it look like this:

te 2103177.JPG (6.73 KiB) Viewed 1290 times

I think you push the green bit and pull the red tab out a bit (might be stiff). Than push the orange bit in frond, than the connector should come lose.

Good luck

[Heye]

Thanks, you two! Definitely, some charging would be better than no charging. And I love the simplicity of the Outlander OBC. In another thread, @arber333 found out that the maximum voltage of it is indeed 400V. I would like to be able to charge close to 4.2V, to force the BMS into balancing. So getting the Outlander OBC would be kind of the last resort for me, to get the car going for now.

In the spirit of getting the car going for now, I managed to disconnect the connector in the middle of the INVCON 3.3 Inverter that was originally intended for an AC compressor. The full inscription is "TE 1587819 17237 2103149-17258". The trick for getting it off was to remove the gray cap first (sadly, I couldn't take pictures, because my phone was empty). That was quite hard to do and I still mangled it somewhat. I finally managed to remove it by sliding in one end of a pair of bent snap ring pliers under the plastic clip on the bottom. Afterwards, the plug could be removed by pressing down the orange tab, pulling the plug out a few millimeters and then pressing the red tab in and sliding the plug out completely. Amazingly, it does seem to be a straight (fused) connection to the big DC terminals, at least the voltage came up immediately, as the main contactors closed. The plug does have an HVIL bridge, but I overcame my constant fear of bricking something and having to go to that expensive workshop and tried without the HVIL jumper and it still worked! I'm pretty hopeful that I will be able to use that connector to charge, at least temporarily. That misuse of the AC compressor HV connector for charging is what most (all?) VAG cars with that inverter seem to do, which I find quite funny.

@manny, thanks, that connector does look very similar! It does seem to be compatible with the other one on the inverter. I don’t get why they don’t all have the same locking mechanism. I will try that out, soon.

[Heye]

I managed to get the charger running. It's currently charging at 3 Amps, but I will have to interrupt it, because I currently haven't hooked cooling up to the charger. So far, the car hasn't locked me out. Knock on wood!

[Heye]

I've now started university in another city, but for practical reasons, my EV has to stay at home. So I haven't been able to work on it much. Charging it to 100% SoC works and I can even make it balance by plugging an EVSE into the charge port afterwards. So continuous balancing isn't possible yet. Also, I still have to fix the OEM charging circuit so my family can charge and drive the car as well while I'm away. Currently, I'm not trusting anyone else with charging, since it's somewhat sketchy.

It's possible that the contactors for charging never close, since the OBC doesn't report any voltage. Of course, the OBC voltage sensing might be broken, but that doesn't sound likely to me. I haven't yet understood how the BMS knows when to open the main contactor/close the charging contactor. I don't think that's controlled over the CAN bus. The BMS tells the OBC the maximum current available over AC, so maybe the PP/CP lines go to the BMS, which then uses that information to control the OBC? In the CAN logs, the messages coming from the BMS look just fine. So maybe the contactor went bad? It's a shame that dropping the battery and opening it is so much effort.

There's another thing that annoys me very much: In the CAN logs, I see the messages BMS_BCM_startClearanceRequest and BMS_BCM_startClearanceResponse with apparently encrypted contents. I very much hope that these are only required by the BMS and not by the BCM, because otherwise replacing the BMS with a DIY unit would be a lot harder that I initially expected.

[gb993]

It is possible to turn off the start Auth for the batteries.

This way there doesn't need to be a handshake.

[Heye]

That sounds great! I assume I would need the StreetScooter diagnosis device for that?

[gb993]

I'm pretty sure I have seen it in the documentation that it can be switched off - I also know that some of the batteries have been used for stationary energy storage and assumed the security would be switchable.

Peak CAN interfaces were the StS unit of choice from memory.

[Heye]

Noting these things down before I forget (also, because this doesn't seem to be documented anywhere else):

Sometimes, after unlocking the vehicle, the driver side door won't open. There is no resistance when pulling the door lever. A (measured) smack on the door near the lock helps. If you get in through the passenger door, simply driving a bit also usually works.

Much more important:

Today I wanted to drive home (I had parked at the underground car park at my workplace where it's not permitted to park personal vehicles overnight) and couldn't. The parking lock didn't disengage. Turning the car on and off repeatedly didn't help, at one point it went into the error mode I had struggled with earlier, so I had to reset it again. But it was still stuck. After some research, I was able to fix it by removing the hood, finding the parking lock motor on the gearbox attached to the drive motor. It's black and circular, almost looks like an oil filter. It has a rubber plug in the middle that, if removed, reveals a 4mm inbus screw that can be turned clockwise to release the parking lock manually. It needed quite a few turns, more than I expected. Having a small ratchet was nice. This seems to be very similar to the "Service PRND" error in the Fiat 500e, where the the "Park Pawl" motor fails. The Fiat 500e uses the same motor (Bosch SMG 180/220) as the StS, so I assume that it also uses the same gearbox/differential and parking lock motor as well (I haven't checked yet).
In Goingelectric forum (German language), it was mentioned that you need to take care not to use too much force, only 2.5N are permissible. You will also find a very nice PDF file (in English) there, with more details on the parking pawl.

I apologize for not having taken a picture, I was in quite a rush. The gearbox and parking lock motor is very easy to find, though, once you have taken the hood off. In case anybody else ever struggles with this, I will leave these keywords for the search engines: "Streetscooter Parksperre defekt".

[Heye]

The parking pawl motor saga continues. It remained locked basically every time I went to start the vehicle, which annoyed me very much. So I removed it from the vehicle (to unlock the parking pawl, the motor has to compress a spring, so it should be in the completely locked position for removing it) and opened it up, being careful to not let any dust/dirt/debris get into the motor or into the gearbox.

On the inside, it looks exactly like the description in the Fiat repair manual. A beefy DC motor drives a threaded rod with a weirdly sphaped nut on it over a 2-to-1 reduction gear. On the end of the motor is a magnet, that is used by two hall effect(?) sensors to register motor rotation.

The only thing I was able to do was to remove the old slightly crusty grease and replace it with some fresh grease. Afterwards, I ran the motor on a 12V battery for a few seconds, hoping that it might free up the bearings somewhat. After reinstalling it, I had to perform the reset procedure again, but it seems to have done the trick, as I was able to lock and unlock the parking pawl multiple times, without having any issues whatsoever.

Here's hoping that I don't have to touch that thing any time soon.

Notes:

• Against the light suggestion (the words they use are DO NOT and MUST) of Fiat, I reused the bolts. At those low torque numbers, I don't see why not.
• The geared motor (Nidec 405.712) also seems to be used in a Peugeot Partner Electric (Mk2)
• The motor has to provide more torque for unlocking than for locking (because it has to work against spring pressure for the former), which would explain why it would lock, but not unlock.
• Take note of the orientation of the nut on the threaded rod. If I recall correctly, the shorter flat side has to point upwards to the electrical connector.
• I'm not an engineer, but I don't like this design very much. This issue seems to appear quite frequently, so I would guess that the motor is likely still undersized (even though it is already really beefy and heavy). The calibration procedure of having the motor slam into the endstop and observing the current spike isn't very confidence inspiring to me. They already have to have two hall effect sensors and an H-bridge for controlling the direction of the motor, how much more effort/expense could a third half-bridge for a brushless motor have been?