My turn for the Check EV message

Toyota Rav4 EV Forum

Help Support Toyota Rav4 EV Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I'll replace the speed sensor, clear all the clearable Alerts (DTCs), then generate a fresh batch if problems persist.
Questions. Can you make and publish voltage oscillograms on the contacts of the motor speed sensor with the rotor stationary and when it rotates (before and after replacing it)?
 
Last edited:
Maybe. Timing of events plays a part.

The sensor "seems" to warm up and malfunction, based on the requirement to leave the system OFF for two+ minutes before a successful drive can occur.

---

Some random notes:

The Tesla part No. for the "SENSOR ENCODER" (speed sensor) is 1002722-00-C ("-C" for the version that's rated for >130 MPH, see SB-15-40-001; the "-B" version is only good for 125 MPH). A version seems to be widely available for less than $20 from China (and via eBay).

1723737932446.png

That's a heck of a lot less expensive than Toyota's $300. Counterfeit? Maybe. There sure are a lot of sellers of this part. They're cheap enough; I've ordered (2).

---

From http://www.turbo-electric.org/teslaresearchevcontrolskit.html
1723738470015.png

From the Toyota Service Manual, Powertrain Wiring Diagram (Tesla bits only)
1723738693393.png

The connector to mate to the sensor is TE/AMP 444046(-4).

LDU_Speed_Sensor_Encoder_Connector_TE_AMP_444046-4_01b.png LDU_Speed_Sensor_Encoder_Connector_TE_AMP_444046-4_02b.png

The "-1" version has an orange plug-to-receptacle housing seal; the "-4" has a green seal. Mouser.com lists TE & "-1" version as "Obsolete".

From what I can tell, ordering that part gets you the plug housing, the seal, and the yellow (or violet, for "-3") "spacer" that positions the terminals. When sourced from TE/AMP, the terminals and wire seals are not included and must be ordered separately.

LDU_Speed_Sensor_Encoder_Connector_TE_AMP_444046-4_03b.png

The "-1" version w/orange-ish seal is widely cloned and available, sometimes as a kit that includes terminals and wire seals.

This is a "Wire to Board/Device" connector and as such there is no mating receptacle housing offered, so a breakout harness would likely have to destroy a speed sensor in order to obtain a receptacle. Mating male terminals are available, as the terminals (receptacles/female and tabs/male) are used in other AMP products; table of terminal part Nos. for various wire sizes is here.

---

There are no fast/cheap sources for delivery to the US.
$44 will get me one clone piece in a week; less expensive options would be 2-8 weeks.
I found that $70 will buy me parts to assemble (5) from genuine TE/AMP parts, in about ten days, so that's what I ordered. From three different sources :rolleyes:. Thanks, octopart.


Date
Description
Amount
Vendor
Notes
15Aug2024​
SENSOR ENCODER 1002722-00-x, (2) @ $15.78; 2.84 tax​
43.40​
via eBay​
Housing, Plug, TE/AMP 444046-4 (“-4” comes with green housing seal, no terminals or wire seals). $3.96 ea. but MOQ of (5) = $19.80; $8.99 UPS; $1.29 tax. Update: was charged another 0.49 for unknown reason.​
30.57​
OnlineComponents.com​
Terminal, Receptacle (female), TE/AMP 282438-1, 17-20 AWG. $0.13 ea. but MOQ = (100) = $13.00. Shipping = $9.99. Tax = $2.07. From UK.​
25.06​
Newark.com​
Wire Seal, TE/AMP 963531-1, for wire OD .074”, Gray. $0.104 each, (20) = $2.08. Shipping - $12.90​
14.98​
TME.eu​
20ug2026​
Terminal, Tab (male), TE/AMP 282440-1, (50), 17-20 AWG. $.238 ea. = 11.90; shipping = 7.99; tax = 1.79​
21.68​
Mouser.com​
(ref: https://www.te.com/en/product-282440-1.html)
Total​
$135.69​
for materials to build (5) breakout harnesses.​
 
Last edited:
The sensor "seems" to warm up and malfunction, based on the requirement to leave the system OFF for two+ minutes before a successful drive can occur.
Maybe part of the story, but not the whole story. I found it was easiest to trigger the Check EV issue after the car had been sitting for a while. On my first investigative run, the Check EV message came up after about 3/4 mile, then it took probably no more than 30 seconds of sitting turned off for it to be drivable, I drove the car around the block for ~10 miles without the problem reoccurring. At that point I parked the car and left it sit for ~2 hours. Then it again took about 3/4 mile for the Check EV to reappear. However this time it reoccurred while driving the 1/4 mile home. It also seems to mostly occur when braking, and usually in the last little part of coming to a stop.

I was suspecting (without understanding exactly how it all works):
  1. The sensor was not picking up the motion of the "teeth" as well at low speed, which depending on the sensor arrangement would be expected
  2. Something to do with slowing down, not sure, it could be different magnetic fields as the motor goes into regenerative braking (thus affecting the sensor signal), or it could be something in the inverter circuit losing sync causing the whole thing to go into a funny state.
I did not get as far as understanding the wait for a few minutes recovery effect. In my normally very quiet neighborhood a car always seemed to appear as soon as the Check EV issue occurred!

I am realizing I should have tried driving without using the regenerative braking, using some gentle parking brake to slow at the corners.

Anyway the logical thing at the moment is to wait for the new sensor and see what happens.
 
On the single drive I did, it worked OK until I put it in 'B' mode, at which time it immediately errored and lost drive. Thus began my dance to get it home.

The Toyota sensor I'd ordered six days ago has been picked up from the dealer by a friend (the dealer is an hour away from me, but he lives much closer to it) and he'll bring it over tomorrow night.

The Chinese ones I ordered won't be here for about three more weeks.
 
Checked Tesla side via TPD: no Alerts (new or historical).
Firmware is at 1.3.101
Installed the new Toyota-sourced speed sensor. Old sensor was dry.
As soon as I put it in 'R' and pressed the accel, "Check EV".

Left system on, checked TPD: DI_f020

Temecula_20240819_1-2.png


The contactors remained closed, until I turned off the vehicle.
 
Questions.
1. Behavior* of this vehicle in Ready and position D and/or R, if after releasing the brake pedal without pressing the acceleration pedal when it is on level ground?​
2. in the same conditions, but with the drive selector in neutral, can you freely and without noise push the car on a flat surface?​
3. Please remind the insulation resistance value between the stator windings and its housing.​

*Notice. Of course, these checks should be done after resetting/clearing the Toyota code P312f.
 
Last edited:
1) Not sure. I don't have level ground. Not for at least 1/2 mi. around me. My driveway is sloped.

2) Yes. I had to push the car out of the road on my initial drive, and 'N' was able to be selected, and I did push the car by hand without issue.

3) ~2.8M for the entire HV system, contactors closed, according to TPD (BMS).

---

This morning, I took the speed sensor from the malfunctioning car and installed it to my daily driver, and drove ~ 20 miles, without issue.

Parts for the speed sensor breakout harness will arrive next week, and then I'll do some monitoring of the encoder. I have had poor results when backprobing an active (moving) connector, and I am loathe to use my wire piercing tools on the OEM harness; I don't want to re-seal the insulation piercings, and I am not in a hurry.
 
Thanks for the answer, but I was asking about a completely different parameter.
Not about the value of the Isolation Resistance parameter (between +HV Bat. and GND, -HV Bat. and GND.
https://alflash.com.ua/2019/to_rav4ev/isol_resist.png
isol_resist.png


I was asking about the isolation resistance between the motor/stator housing and the stator windings, which is supplied with 3-phase alternating voltage from the inverter.
 
Thanks for that clarification. I will obtain that value.

For others following along: until the inverter attempts to feed power to the stator windings, the BMS doesn't "see" the insulation isolation resistance of the stator windings, so TPD won't show it.

I would expect (ha!) a different Alert if the iso of the HV system dropped significantly when feeding power to the stator windings, but we all know how accurate TPD is with its warnings and Alerts.
 
Will the measurement tool be a regular/generic DVOM with a measuring bridge voltage of 5÷9 volts or a megohmmeter with a test voltage of 500÷1000 volts?

As I understand, violations of the insulation resistance between the stator and its housing are detected by the inverter.
And in the event of an unacceptable violation of this insulation, the inverter stops supplying alternating voltage to the stator windings and notifies the system about this with a Faulty signal.
And as a result, the BMS switches the contactors to the Open state.
 
Will the measurement tool be a regular/generic DVOM with a measuring bridge voltage of 5÷9 volts or a megohmmeter with a test voltage of 500÷1000 volts?

I used a Fluke 1507 Insulation Tester, set for 500v testing, and I obtained a steady 4Mohm (manual test, holding the 'Test' button for as long as I like).

As I understand, violations of the insulation resistance between the stator and its housing are detected by the inverter.
And in the event of an unacceptable violation of this insulation, the inverter stops supplying alternating voltage to the stator windings and notifies the system about this with a Faulty signal.
And as a result, the BMS switches the contactors to the Open state.

Except . . . not always, according to Matt's testing.

---

To my surprise, this has the QCC LDU drain kit installed.

IMG_8818.jpg IMG_8819.jpg

All four drain tubes are dry.

---

I found a couple of tools . . . are these yours, Richard? I can mail them back to you.

IMG_8807.jpg IMG_8815.jpg
IMG_8817.jpg

I have several magnet-on-a-stick tools, which are invaluable for retrieving tools that gravity assists, but which do not find the ground.

The socket is, of course, 10mm. Stanley.
 
Last edited:
I used a Fluke 1507 Insulation Tester, set for 500v testing, and I obtained a steady 4Mohm (manual test, holding the 'Test' button for as long as I like).
Except . . . not always, according to Matt's testing.
---
As far as I understand, there (https://www.myrav4ev.com/threads/fs-2012-rav4-sf-south-bay-san-jose-area.2660/#post-31368) describes checking the insulation resistance between high voltage DC* (+ or -) and GND.

I asked and described the resistance between the alternating voltage control of the stator windings and the vehicle frame ground.
https://alflash.com.ua/2019/to_rav4ev/inverter1.png
inverter1.png




*Note. BMS_f035 mean there is an isolation breakdown between the HV lines of the electric vehicle battery and ground.

Please show or describe the connection points during this check. Unfortunately, I didn't understand which electrical circuits you checked, HV DC + and - after contactors or HV AC motor control electrical circuits...
 
Last edited:
I found time to glue two CAN transceivers to an Arduino Due, tested via loopback and a test script, then built an OBDII<->CAN transceiver cable for the DLC No. 2 diagnostic connector; good thing that I have a lot of those OBDII connectors around! I was able to take a log via laptop (running old Ubuntu) connected to the Due via USB, which also powers the Due, and I sent that log to Matt for perusal.

The recipe for turning a Due into a tool for CAN sniffing and reverse engineering (or just logging), using off-the-shelf hardware, a few jumpers, a Due, a CAN transceiver, and free software by Colin Kidder: GVRET & SavvyCAN:

https://openinverter.org/wiki/CAN_bus_with_Arduino_Due

You have to load the Arduino IDE (currently v2.3.2) which hopefully is easier on Win than Linux, because Arduino's install guide seems years out of date; it pretty much requires installing libefuse2, a shim that's been deprecated for some time -- replaced by libfuse3, with has a different API and is therefore not backward-compatible. libfuse2 is no longer shipped with Ubuntu as of 23.10 (certainly not 24.04, which is what I just got working here), and I was really reluctant to install it, but found no workaround that actually worked. That IDE is required to upload the test program (loopback) and later GVRET.

Then you need to configure the IDE for GVRET, downloading a bunch of libraries, and the recipe worked OK for that.

SavvyCAN also wants libfuse2, but you'll have already got that sorted by the time you get to installing it.

Only one CAN transceiver is needed for logging, but having two and connecting them allows you to test your work via loopback, so I highly recommend buying and installing two, it's worth the extra cost & time.

Damned breadboard jumpers drive me crazy, but for proof-of-concept and prototyping, it works OK.
 
Checked Tesla side via TPD: no Alerts (new or historical).
Firmware is at 1.3.101
Installed the new Toyota-sourced speed sensor. Old sensor was dry.
As soon as I put it in 'R' and pressed the accel, "Check EV".

Left system on, checked TPD: DI_f020

View attachment 658
The contactors remained closed, until I turned off the vehicle.
Please, any updates on the cause of this code?
 
Back
Top