ScanGuage AE Pig-tail, code flasher, and TechStream

Thanks to w2co and friends, my new OBD pig-tail adapter works:

Features

• Passes only ISO 9141 signals K and L, ground and B+ to scanner
• Connects Ts, Tc and chassis ground to miniature audio plug to trigger two-code flash or error code clearing

ISO 9141 Signals

The pig-tail keeps any extra signals from the NHW11 diagnostic plug from the OBD scanner. This reduces noise to the OBD scanner and protects the NHW11 from OBD signals that might spike or confuse the vehicle. The following table shows the signals passed:

	Column 1	Column 2	Column 3	Column 4
0	pin	signal
1	4	chassis ground
2	5	signal ground
3	7	ISO 9141	K-line	data signal
4	15	ISO 9141	L-line	data is active signal
5	16	battery voltage	12 V

No other signals pass between the scanner and vehicle.

This pig-tail solves the problem of Auto Enginuity inducing false codes. It also stopped the mystery hangs.

Since the ZVW30 uses a different pair of signals, I'm planning to add a 4-pole, double throw, slide switch so it can work with both the NHW11 and ZVW30. It is very likely that the ZVW30 and NHW20 are using the same signals.

Code Flash and Clear

There are two signals used to trigger flash of the two-digit error codes; Tc, triggers flashing the error codes, and; Ts clears the code when tied to chassis ground. The mini audio jack connects to a momentary toggle switch, (on) off (on). This allows reading major error codes went an OBD scanner is not available or does not have enough support to read all codes:

	Column 1	Column 2
0	pin	signal
1	4	chassis ground
2	13	Tc tied to ground causes ECUs to flash two digit codes
3	14	Ts tied to ground causes ECUs to clear codes

These two signals are documented for the ZVW30 and probably work on the NHW20. This may be a universal code reader for all North American Prius.

PARTS

• Cable, J1962M to 2-J1962F, Splitter Cable, 1ft (145802) - I cut off and used the second J1962F as the female, pig tail. Estimated cost, $13 w/o shipping.
• J1962 Male Connector - bought on Ebay. Estimated cost, $2 w/o shipping.
• mini stereo audio connector - junk box. Estimated cost, $2 (option 1)
• mini stereo audio jack - junk box. Estimated cost, $2 (option 1)
• momentary toggle (on) off (on) - Estimated cost, $5 (option 1)
• 4 ft., three conductor wire - Estimated cost, $0.40 (option 1)
• housing for switch - Estimated cost, $5.00 (option 1)
• multi-pole, double throw switch - Estimated cost, $5.00 (option 2, untested)

I've listed the parts with two options:

1. Tested, this option brings Tc and Ts to a switch so the connector can use the built-in, code flash and reset function.
2. Untested, this second option allows Tc and Ts to be flipped back as well as any other signals that might be needed for other cars (aka., ZVW30).

I built the prototype from junk box parts but if anyone would like one, send me a PM and based upon current parts costs, I'll work up the price. Currently, these look to be ~$50-75 or just 'roll your own.'

Bob Wilson

 

Thanks! It was very early in the morning when I got back from the first test run.

I'm not sure. Out of 16 pins, now there are just 5 pass through. But having Ts and Tc exposed to an OBD scanner that might consider them 'fair game,' these would be my leading suspects.

Since it looks like the ZVW30 uses OBD bus signals, I'm thinking of adding a 4-pole, double-throw slide switch so it can be either an NHW11 or ZVW30 pig-tail. I suspect the NHW20 and ZVW30 use the same set of pins.

Bob Wilson

 

Ts is on pin 14 and pin 6 is tied to the air bag ECU:


Bob Wilson

 

I just passed through "Prius Electrical Wiring Diagram 2003 Model", EWD493U, to find all signals connected to the 'Data Link Connector', which we've been calling the OBD connector:

	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6	Column 7	Column 8	Column 9
0	pin	label	ECU(s)	Comments
1	4	CG	J11	chassis ground
2	5	SG	EB	signal ground
3	6	AB	Airbag
4	7	SIL	HV	Engine	EMPS	Airbag	Brake	K-line	ISO 9141[/B]
5	8	OP3	HV
6	11	OP1	Transponder
7	12	OP2	HV
8	13	TC	HV	Engine	EMPS	Airbag	Brake	flashes two-digit codes
9	14	TS	EMPS	Brake	clears codes
10	15	WFSE	HV	L-line	ISO 9141[/B]
11	16	BAT	OBDI	7.5A fuse

The bold text are passed by the pig-tail. The blue text are accessed via the audio connector.

The pig-tail does not pass four signals but apparently the Auto Enginuity doesn't care. What these four signals do is not yet documented.

What is interesting is pin 15, WFSE or L-line ISO 9141, was necessary to get Auto Enginuity to work. Yet this signal connects only to the HV ECU. In contrast, the K-line signal connects to five ECUs. The battery ECU communicates only via the HV ECU.

Bob Wilson

 

As I look at the signals, I'm wondering if 'OP2' and 'OP3' might be CAN? Since these are open in the pig-tail, it might be worth while to experiment and jumper them to:

• pin 6 - CAN high - pin 8 - OP2
• pin 14 - CAN low - pin 12 - OP3 (or reverse)

In a world of GOOD, FAST, or CHEAP, pick two, FAST and CHEAP win often enough to keep me busy.

Bob Wilson

 

Very likely. I'd sure like to find out if OP2 and OP3 are CAN. Of course they could be a second ISO 9141, K and L lines:

• CAN - 120 ohm termination resistors
• ISO 9141 (?) - need to check

I wired jumpers from the HV OP2 and OP3 signals out to the AE CANHi and CANLo lines. Before 'wrap' splicing, I measured the vehicle and ProLine on voltages and saw a wandering, ~30 mv. on each. So I powered down the car and tested:

• OP2-CANHi and OP3-CANLo :: "Auto detect" called it ISO-9141 and connected.
• OP3-CANHi and OP2-CANLo :: "Auto detect" called it ISO-9141 and connected.

I'll fold the jumper lines inside the hood in case someone has another test they'd like to see. Right now, OP2 and OP3 remain undefined.

Bob Wilson

 

I have 'lessons learned' to share:

1. Mac polling is fast enough - it samples at 44.1 kHz and the data bits are 10.4 kbs

More details. Found this:
Second article for Evaluation Engineering

Both the K and L lines were captured. Only the K-line, pin 7, provides data. Operation requires the ProLine side to assert the L line:

It looks like a request and response takes about 300-350 milliseconds.

What we appear to see are individual bytes with a considerable amount of dead-space between. We can see six bytes followed by a ten byte response:


Here is the last byte followed by the first byte of the response. Now the first series of bytes probably came from Auto Enginuity and I have no interest in these values. I am perfectly content to let Auto Enginuity trigger the Prius response. However, the Prius response is interesting:


So when we expand the Prius response, we find multiple sample points:


By laying this over a graph showing the Mac sampling points and Prius data points, it becomes entirely possible to use the Mac audio recording to extract the bytes:

Based upon the earlier URL: 0001 0110 or 0x16:

xxxx xxxx START - first line
xxxx xxx0 START - Bit 0
xxxx xx10 START - Bit 1
xxxx x110 START - Bit 2
xxxx 0110 START - Bit 3
xxx1 0110 START - Bit 4
. . .
0001 0110 START - Bit 7
​

So it looks entirely feasible to decode the Prius responses and see if we can figure out the disconnect between the Prius response and the Auto Enginuity text values reported. Hopefully, we can send them an exact description of the problem.

Bob Wilson

 

I'm still not happy with this result:

	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6	Column 7	Column 8	Column 9	Column 10	Column 11
0	time	d7	d6	d5	d4	d3	d2	d1	d0	func
1	0.032539683	1	0	0	0	0	1	0	0	format	0x84
2	0.043219955	1	1	1	0	0	0	1	0	dest	0xE2
3	0.053469388	1	1	1	1	1	1	0	0	source	0xFC
4	0.063809524	1	1	0	0	0	0	0	0	cmd 1	0xC0
5	0.074331066	1	1	1	1	1	1	0	0	cmd 2	0xFC
6	0.084580499	1	1	1	0	1	0	0	1	crc?	0xE9
7
8	0.127210884	1	1	0	0	0	0	1	0	format	0xC2
9	0.132494331	1	1	1	1	1	1	0	0	dest	0xFC
10	0.137777778	1	1	1	0	0	0	1	0	source	0xE2
11	0.14276644	1	1	0	1	0	0	0	0	cmd 1	0xD0
12	0.148049887	1	1	1	1	1	1	0	0	cmd 2	0xFC
13	0.153809524	1	1	1	1	1	1	1	1		0xFF
14	0.158888889	1	1	1	1	1	1	1	1		0xFF
15	0.163968254	1	1	1	1	1	1	1	1		0xFF
16	0.169160998	1	1	1	1	1	1	1	1		0xFF
17	0.173582766	1	1	1	1	0	0	1	0		0xF2

Audio recorded, software and spreadsheet decoded.

The first byte of the first command looks OK. The 0x84 corresponds to the correct format and byte length. But the response looks more like an error being returned. The 0xC2 is is a functional response. Off hand, it looks like an error response more than anything else.

Bob Wilson

 

We don't yet know all of the answers and in many respects you are reading our 'lab notes.' But I'll try to help:

You'll be able to use it but the Scangauge works a lot better with just four of the eight, RJ-45 signals:

1 - ground
6 - K-line, the command and data bus line
7 - L-line, status from Scangauge to vehicle
8 - B+, 12 V
​

I took an old ethernet cable and cut it in half. Then using an RJ-45 female, punched down just four wires, verified continuity, and used JB Weld to anchor the cable in the connector:

Thanks to w2co and vincent1449p who provided initial data.

Right off the bat, you'll be able to read the error codes but to get detailed data about your traction battery, you'll need to program some XGAUGE values using vincent's table:
http://priuschat.com/forums/generat...-classic-prius-scanguageii-4.html#post1161266

Don't let the table intimidate you. The programming is fairly straight forward but a pain in the nice person because it has to be done in the car . . . unless you add a trick vincent came up with.

The Scangauge has to initialize the ISO-9141 bus before it supports programming. What vincent did was put a blocking diode on a 9 volt battery tied to the B+, pin 8. He can go to the car, plug in and start the car. The Scangauge initializes the ISO-9141 and then he connects the 9 V battery. At that point he can remove the scangauge powered by the 9 V battery and take it into the house to program up as many XGAUGEs as he needs.

I've got the parts to make another one. If you'd like, send me a PM with an address and I'll fabricate, test and mail one to you. Once you're sure it works, send me $10 or the part and we're even. I can get it in the Saturday mail so it should be there early next week.

The fancy box I described in the "Scangauge" thread is designed to let me use my laptop as an ISO-9141 analyzer and investigation tool.

Bob Wilson

 

Just a well stocked junk box. <grins>

I've got to run an errand but let me know. I have one already and I can add the diode and 9 V connector. But if you are interested, I'll make another and photograph the assembly. You could then round-up the parts and make a functionally identical part.

Bob Wilson