Tesla ModelX Gen2 iBooster Hardware and Software

[alexbeatle]

Hello, folks.
Does anybody know what these ibooster statusi mean (see attached) - Failure, Standalone, Reduced, Diagnostics, Not_init_or_sup, maybe there're others?
I only have .dbc file from Model S that can decode the messages. Other Tesla .dbc I've found cannot decode

It seems that when my ibooster is attached to a pedal on startup it goes into Failure mode. If I detach, it goes into Standalone->Diagnosis->Reduced. As I push the rod without pedal it goes into Standalone->Diagnosis->Not_init_or_sup->Reduced->Standalone->, etc. Not much pressure on the rod when the pedal is connected, and the maybe only 5degrees of angle difference on the rod.
The rod extension values also do not start from 0. When fully extended it shows around 38mm, then as pressed goes to 0 and then increases again until fully pressed.

I used to have CANbus disconnected, but ibooster stopped reliably working so trying to troubleshoot. This may have happened after I was bleeding the brakes and pumped the pedal without powering the ibooster, but not sure.

Also these CAN messages are from the vehicle CAN bus. I see the same on the YAW can bus.

Thanks in advance.

[alexbeatle]

From trying to make adjustments I believe the ideal scenario for the ibooster to report Standalone status and when the pedal is fully released for the ibooster to report 0mm rod extension.

Not_init_or_sup happens when power applied, but ignition not yet, and! the rod is within certain range! If not - Failure mode.

My ibooster reports 38mm when the rod is fully released, so I pushed the rod in until it reports 0mm and secured to the pedal that is fully released. Looks like the rod has 50mm travel range with overtravel, i.e. if rod goes over 50mm, the value will jump to 0 and then increase; if the rod goes over 0mm mark, it will report 50mm and then decrease in values. With my adjustments it starts at 0mm goes to 50mm and slightly overshoots. ibooster seems to work ok. But! Having it pressed so much in fully released pedal engages my brakes and wheels cannot turn. I have disc on the front and drums on the rear of the car. I used this setup that Lars provided - https://www.evcreate.com/installing-the-ibooster/ . Thank you, Lars.

I tried adjusting the pedal so that it doesn't start at 0mm in fully released, but 49mm, 45mm, still stay in Standalone, but still brake engagement in pedal release. Around 40mm mark, ibooster starts not in Standalone, but Reduced mode. I think Reduced means it will assist with smaller force - required more force from the foot. I did not get steady results though, need to play with the minimum values else it sometimes goes into Failure mode. The issue I also found that in Reduced mode, if I slam on the brakes, the rod doesn't fully extend back - wheels have some brake drag.

I'm also currently investigating whether iBooster doesn't like battery to go below 12.5VDC and goes straight into Failure mode.
Also whether, when it's connected to CAN bus, another device must read its data. Only my laptop with Savvy was reading the CAN data from the ibooster, while other devices weren't. When my laptop ran out of juice, after some time, ibooster went into Failure mode.

[jrbe]

I would imagine the ibooster is reporting how much travel is left from the starting point / brake pedal fully up.

When it gets to 0 I'd suggest this is being considered as the bottomed out point.

Past that 0 point I'd expect the booster to report something along the lines of a brake failure.

You do not want to push the brakes past where the brake master cylinder is connecting the caliper feeds to the reservoir at rest. Past that point you will trap pressure in the braking system that will increase with heat which pushes the brakes more and more.

[alexbeatle]

I would imagine the ibooster is reporting how much travel is left from the starting point / brake pedal fully up.

When it gets to 0 I'd suggest this is being considered as the bottomed out point.

Past that 0 point I'd expect the booster to report something along the lines of a brake failure.

Seems to be the opposite for my ibooster. Numbers increase as the rod gets pushed in - bottom out 50mm.

You do not want to push the brakes past where the brake master cylinder is connecting the caliper feeds to the reservoir at rest. Past that point you will trap pressure in the braking system that will increase with heat which pushes the brakes more and more.

This would explain why in some cases the pedal doesn't come all the way back after slamming the brakes.

[jrbe]

My ibooster reports 38mm when the rod is fully released

This is without anything pushing on the booster, correct?

so I pushed the rod in until it reports 0mm and secured to the pedal that is fully released.

This sounds a lot like you're pressing the booster in. If the brake master is there too it's definitely blocking the reservoir and applying pressure to the brakes.

Looks like the rod has 50mm travel range with overtravel, i.e. if rod goes over 50mm, the value will jump to 0 and then increase;

So you're leaving 12mm of brake travel? 50-38=12. It should allow full brake master cylinder travel.

ibooster seems to work ok. But! Having it pressed so much in fully released pedal engages my brakes and wheels cannot turn. I have disc on the front and drums on the rear of the car.

Sure sounds like you're pressing the master cylinder in when it should be fully out. But I might be misunderstanding your wording. A picture might help people understand what you're doing and what the struggles are.

Sounds like you're doing this to get the booster to start in standalone mode. Seems like doing the wrong thing to make something else work.

[alexbeatle]

I see your point.
Let me shoot some videos and post for better understanding.
Looking at the Tesla service manual, nothing mentions that disconnecting the pedal from the clevis will make the rod spring out - guess no need to preload the booster.
https://service.tesla.com/docs/ModelX/S ... 252D4.html

[alexbeatle]

This is my setup. When brake pedal gets pressed, the pushrod engages the ibooster.


The little threaded aluminium cylinder allows extending the rod for adjustment. The rod is OEM from the car.

Videos start with just power on, ibooster in Not_init_or_sup mode. Then I flip the Ignition switch.

Here's what CAN bus looks like when the adjustment is such that the rod is nearly all the way in the aluminum cylinder. Pedal is to its upper limit.

Here the ibooster status is not constant in the begining.

Here I extended the rod out of the little aluminium cylinder, since the pedal is at the upper limit this pre-loads the ibooster more

The ibooster status here is pretty much Reduced throughout the range of motion. (Some minor surface rust on the discs lets you hear when the brakes get engaged ) close to 50mm mark.
This is about the last step of pre-loading the ibooster before the front discs brakes start to drag.

I realised it's important to do a complete cycle of power on the ibooster to get consistent results. That is, even switching ignition off and waiting for ibooster to stop talking and then making adjustment, isn't the same as completely power cycling it for it to realise new encoder values.

All these are done with me applying breaks by hand. The disc brakes in the front engage vey well - though still need to take measurements.
Naturally, forces are much greater when using the foot and given the long range of the pedal travel, which is what I think may be causing problems.

[jrbe]

I have manual brakes in my race car (it's not fancy.) It starts off at 6:1 to take up the play then ends up about 14:1 at full braking and achieves full travel of the brake master cylinders with a full swing of the brake pedal. It's a bit odd to drive at first but after a few times you hit the brakes it gives amazing feel at the limit and has great granularity.

I mention it because it uses a bellcrank similar to yours to achieve the ratio tricks. Depending on where the swing is on the bellcrank it's changing the ratio through the travel. You might be having some unexpected ratios from your bellcrank or possibly more travel than a common brake setup. This ratio changing effect is more dramatic the more travel it has and the more the angles change at the bellcrank.

I'm not sure how much booster assist your vehicle used to have but it likely has a lot now with the ibooster. I'd expect you want around 5:1 - 6:1 ratio feeding into the ibooster to have a normal feeling pedal.
This also depends on the master cylinder bore and calipers.

Basically it might be worth measuring the travel to the booster vs travel at the pedal to figure out your pedal ratio to see if it's sensible. It might be really challenging to get a reasonable amount of assist if things aren't setup well stock.

*Edit
I can't help much / at all with the failure mode besides assuming there is a window of acceptable range for the ibooster at startup that it's happy with. Outside this it might hunt around looking for certain things.
Is there a 12v wire that should have constant power to keep some memory alive that's getting disconnected at key off?
I thought standalone mode was pretty much figured out for these.

[alexbeatle]

Ok, perhaps this is considered a Gen3 iBooster. As I understood Tesla used vaccuum pump first on MS and MX and then switched to prev. gen ibooster and then to the version I have.
However, I found the latest Tesla Model 3 DBC - attached - and it decoded quite different results! The rod mm is very linear, 0mm at full release and to approx. 10mm fully pressed. It *correctly* shows when brakes are pressed and when not as well as much more understandable statusi - "ready", "actuated" when pressed, "transition to idle" when not touched for some time.

I think therefore this is the correct DBC decoding to be used for this ibooster generation p/n.

[alexbeatle]

I have manual brakes in my race car (it's not fancy.) It starts off at 6:1 to take up the play then ends up about 14:1 at full braking and achieves full travel of the brake master cylinders with a full swing of the brake pedal. It's a bit odd to drive at first but after a few times you hit the brakes it gives amazing feel at the limit and has great granularity.

Yes, I need to see how to change the ratio on my setup. With the current ratio the brakes are hard, but touchy - small range of pedal.

[alexbeatle]

...
I'm also currently investigating whether iBooster doesn't like battery to go below 12.5VDC and goes straight into Failure mode.
Also whether, when it's connected to CAN bus, another device must read its data. Only my laptop with Savvy was reading the CAN data from the ibooster, while other devices weren't. When my laptop ran out of juice, after some time, ibooster went into Failure mode.

I added 0x39D message to my VCU reading, yet after 15+ mins of inactivity ibooster goes to Fail mode. Cannot be restarted through ignition, must be power cycled. To restart, can either be powercyled or after switching ignition off and waiting for the ibooster to stop communicating via can bus, flip ignitions back on. Battery was at 12.3VDC, but when I did power cycle it started no problem - not power issue.

From my VCU log it went
From:
19:26:50.646 -> BRAKES_NOT_APPLIED
19:26:50.646 -> IBOOSTER_READY
19:26:50.646 -> NO_MODE_ACTIVE
To:
19:26:51.630 -> FAULT
19:26:51.630 -> IBOOSTER_FAILURE
19:26:51.630 -> NO_MODE_ACTIVE

Anybody is aware of the keep-alive message required? Or it needs YAW CAN bus connected for it not to die.

[jrbe]

From:
19:26:50.646 -> BRAKES_NOT_APPLIED
19:26:50.646 -> IBOOSTER_READY
19:26:50.646 -> NO_MODE_ACTIVE
To:
19:26:51.630 -> FAULT
19:26:51.630 -> IBOOSTER_FAILURE
19:26:51.630 -> NO_MODE_ACTIVE

Is the 0-10 somehow really 0-100?

From those messages it seems it's looking for a mode to run in. Have you found a list of commands for different modes?

I wonder if there is a command to reduce the assist level. If you could reduce the assist enough you might be able to leave the bellcrank / pedal ratio as is.

It's a very odd choice by Bosch to not go into a default assist mode instead of booster failure, I wonder if this is part of the ev braking issues that keep popping up.

And if you find this really is a Gen 3 it would be good to change the subject line.

[alexbeatle]

Was tuning the drive - start\stop with brake - i.e. no opportunity for the ibooster to go into sleep mode or what not. Still it went into failed mode after some time and stopped assisting.

[alexbeatle]

Fingers crossed... I think the issue was due that ibooster spring isn't strong enough to retract the pedal of my vehicle all the way out due to my pedal setup. Sometimes it did, other times it didn't causing the rod length to report different value from that during "ignition", causing the failure. Essentially as if the brake is always pressed a little bit.
Hopefully this is gone by me fixing the original vehicle pedal return spring, to return the pedal predictably.