SaabCentral Forums banner

1 - 20 of 43 Posts

·
Registered
Joined
·
156 Posts
Discussion Starter #1
After having my Chinese tech2 fail to communicate with my car when I needed it the most I’ve had to set about repairing it. As part of the process I’ve found very little information about the internal workings of the tech 2 so fault finding is somewhat limited.
The purpose of this thread is to bring together what information I’ve found on the web and to add what I’ve managed to nut out. As it progresses I hope get my tech 2 working again and provide enough information to help others fix theirs.

As my fault is not communicating with my car (non CANdi 9-3 and 9-5) I’ll be starting with the VCI module.
 

·
Registered
Joined
·
156 Posts
Discussion Starter #3
I must apologize- I started the thread and have a lot more to post then my 2 year old broke his leg so the TeCh2 got out on the back burner for a couple of days.

Yes there is that Wiki and the DIY-TECH2 as well http://diy-tech2.blogspot.com. But they are a bit light on in the nitty gritty technical details. I hope that what is put together in this thread can be also added to the other sites, the more the information is distributed the more likely it will survive and we can keep our cars on the road :)

I’ve been working through the self tests and subsequent results. I have noticed a number of people comment that their tech2 fails the self test but still appears to work - there is a reason for that. Yes the tech2 is faulty - it’s just the bit that fails is not used in your car.
 

·
Registered
Joined
·
156 Posts
Discussion Starter #4
Just get things rolling I was beating my head against a wall as I thought I did not have a DLC Loop back adapter.

Turns out I do have the right bits just the numbers on everything are wrong!

DLC Cable correct Number 02002952
My DLC 02003214

SAE 16/19 pin adapter correct Number 02002955 (this is the OBDII connector)
My SAE adapter 02002953

loop back plug correct Number 02002953
My loop back plug 02002955

I suspect the rest of the adapters are also labeled incorrectly as well but Im not likely to use them. AND to add to teh confusion the Vertronix Manual I have has the serial loop back and DLC loop back adapters incorrectly labeled on page 3-5
 

·
Registered
Joined
·
156 Posts
Discussion Starter #5
Once you have the right adapter how to carry out the tests are outlined in the Tech2 Manual which is readily available from multiple places online. Once you carry out the tests and you find a failure the manual nicely describes the possible outcomes - usually pass or fail and followed by an action to take. for almost EVERY failure the solution is
For any failure, cycle Tech 2 and retest.
If problem persists, contact Customer Support.
So pretty useless especially if youve got a Chinese clone although I did contact the supplier of mine and they were most helpful and with no hesitation they suggested I return it to them and they would fix it. THough freight on a VCI from here to there was about $40 and Ive seen VCI's fro only about $70 so Ill try and fix it myself and when I get bored with it Ill just buy a new one - hopefully Ill have collected enough information to make repairs possible for others in the future.
 

·
Registered
Joined
·
156 Posts
Discussion Starter #6
Arrgghhh... Dont believe everything you find on the internet :eek:

spent a number of hours yesterday trying to trace back pins on the 26pin connector on the VCI back to the the MUX's and was not getting sensible results.

I ended up buzzing out the cable and adapter and found that the Numbering for the D26 connector I was referring to was WRONG!!

The pin numbering in this image from from http://diy-tech2.blogspot.com.au/ is incorrect. The numbering in the diagram is reversed. ie

Pin1 should be pin 9
pin10 should be pin 18
pin19 should be 26

The pin assignments next to the connector image are correct just the diagram pin numbering is back to front.

I should add there is a lot of good information on the Blog and I appreciate the effort the blogger has gone to
 

·
Super Moderator
Joined
·
19,050 Posts
Arrgghhh... Dont believe everything you find on the internet :eek:

spent a number of hours yesterday trying to trace back pins on the 26pin connector on the VCI back to the the MUX's and was not getting sensible results.

I ended up buzzing out the cable and adapter and found that the Numbering for the D26 connector I was referring to was WRONG!!

The pin numbering in this image from from http://diy-tech2.blogspot.com.au/ is incorrect. The numbering in the diagram is reversed. ie

Pin1 should be pin 9
pin10 should be pin 18
pin19 should be 26

The pin assignments next to the connector image are correct just the diagram pin numbering is back to front.

I should add there is a lot of good information on the Blog and I appreciate the effort the blogger has gone to
Good find. I take it you notified them of their error?
 

·
Registered
Joined
·
156 Posts
Connector drawings are a PITA they are very ambiguous as to the direction of view as well.

if you reverse all the numbering the view is from the wire side (rear) of the connector.

The only way I could see to contact the blogger was a reply to his blog which Ive done.

Now I need to go back and redo all the testing I was doing yesterday and hopefully it will all make more sense :lol:

I buzzed out the DLC cable all the way to the OBD connector this morning
 

·
Registered
Joined
·
156 Posts
Discussion Starter #9
This little project to get my TECH2 working again is turning out to be somewhat of a challenge.

The biggest hurdle has been no technical information about the opperation of the tech even just a pinout of the 26 pin connector I have found to be non existant so its been a slow process of trying to work it all out from scratch.

Perfoming the VCI self tests revealed a number of errors one of them been not detecting the loop back adapter.

Thankfully there is a test for this. There appears to be a variety of loop back plugs and their ID is determined by a resistor in the plug.

The ID resistor test will return the Loopback plug ID number in my case it was returning " Cant Ident . Cable ID Resis. "

I have been collecting data sheets for all the major IC's on the board and identified 2 16 channel Analog multiplexer chips who's output connect to the Analog to digital converter. I reasoned that there was a good chance these were going to have something to do with measuring the adapter ID resistor so poking around the input pins with a scope looking at signals didnt show anything useful so I tried injecting 5v into each of the pins. a 1M ohm resistor connected to 5v and touch the other end to each pin in turn and see what changes. (Using a 1M ohm resistor limits the current flow significantly to prevent any damage )

Using this method I did find a pin on one of the MUX chips (U10) which made a signifcant difference. This indicated that that the pin I was connecting to had nothing else of significance connected to it. I got a little excited at this point as this pointed towards an open circuit connection between the MUX and the 26pin connector.
At this point things went sour pretty quickly as I looked at the board more closely It became quiet obvious this is a least a 4 layer board. :eek:

I was through trial and error trace the pin I had identified back to the 26pin connector but it didnt match up with the location of the ID resistor in the connector drawing I had obtained from DIY-TEch2. This is when I discovered the pin number on his drawing was incorrect so fixing that up still didnt point to the right pin.

I did have a bit of luck though after running the TECH 2 for A while I noticed periodically the message would change from cant ident resistor to ID 25 detected. When it was displaying the correct message I was able to see that the address lines for the MUX would change. A3 would change from 1 to 0 V. this meant that when the ID was failing the VCI was actually reading the wrong pin on the plug.

The fault is clearly not in the connection to the car but back in the logic of the VCI. Maybe an intermittant connection between the MUX and what ever controls it.

This 4 layer board has come back to cause me grief again. tracing tracks :(

The only thing I could do was connect a multimeter probe to a pin on the MUX and touch every other pin on the board untill I found a connection. That was found on pin 9 of the Altera PLD chip - its only got 100 pins on it :roll: I was lucky I started at pin 1 ... finding the connection was great but no intermittant connection, the output of the PLD also varied ;oops:

The next step is to either find the inputs to the CPD and continue to work backwards ( which given I have 84 IO pins which any pin can be configured to perform either task with programable logic in between is probably not possible unless I can get extract a copy of its programing)
The other option is to power the tech2 up and wait for it to start working Which it usually does after its been running for a while - I suspect heating of a component is causing a bad solder joint to close up - allowing the tech2 to work. HWen it cools down it opens up and stops working.

So my next step would be to get some freezer spray and wait for the tech2 to start working and cool each component down and see when it stops working to isolate the faulty area of the board.

Or I could just sit down and resolder everything :eek:
 

·
Registered
Joined
·
156 Posts
Discussion Starter #10
Chinese Clones

In looking at this VCI and trying to work my way through it I am blown away by the fact that while the quality of the tech2 could be better and they are susceptible to failure but the very fact that they have taken a device and copied it and made it work takes a lot of skill and possibly even more effort than the original design.

I have a lot of respect for the abbiliites of the people that have made this product available to us.
 

·
Registered
Joined
·
4,197 Posts
Chances are it's more sinister than that. Vetronix likely had the original Tech 2 manufactured in China, and at some point those designs were appropriated. Highly unlikely anyone reverse engineered anything. Clone Tech 2s are regular Tech 2s sans QC. That is a commonly cited problem with manufacturing in China - you have to be extremely vigilant to ensure your product isn't stolen and produced elsewhere without your knowledge. With something as complicated as an iPhone based on a proprietary microprocessor you're a bit safer, but with something as simple as a Tech 2 - especially by today's standards - you're at risk.

That's always been my personal justification for unapologetically using a clone Tech 2 - it's not a $2000 device or really even a $200 device. They charge what they charge because they can, to keep regular people (and businesses) from competing with the dealer. The price does not reflect the actual value of the device. $250 shipped to my door? Close enough.
 

·
Registered
Joined
·
156 Posts
Discussion Starter #12
argh sat down today and hooked up a logic analyzer today to get a better idea of what the Signal to MUX b did when it worked and when it didn’t and of course it has worked perfectly all afternoon. Anyway at least I have more data on what it’s suppose to do so I can compare it to when it doesn’t work again. Unless the very act of soldering pins in to attach the logic analyzer fixed the bad joint
 

·
Registered
Joined
·
156 Posts
Discussion Starter #13
After having it work beautifully for the last couple of months (all in its dissasembled state I had a spare moment this afternoon and figured Id screw it all back together and I found an old laptop case it will all fit into noicely so I screwed it all back toghether and did a self test and back to MUX B not working again - At least I have something to work with again :lol:
 

·
Registered
Joined
·
40 Posts
So I know that this thread might be a bit old but I'm curious, have you found anywhere or discovered the pinout for the VCI serial connector ? I'm at my 3rd one going bad and I reaaally reaaaally am getting very frustrated because I cannot get any info on this.
 

·
Registered
Joined
·
156 Posts
Discussion Starter #15
Sorry No - only because I haven’t had any need. ( your talking about the serial port on the side that connects to the computer ? )

i ended up getting desperate so bought another VCI which incidentally is also not quit right But the functions the SAAB needed worked.
 

·
Registered
Joined
·
156 Posts
Discussion Starter #16
I noticed since the forum update the table in Post #8 is all screwed up. Here it is

Pin out of the DLC cable
D26
Round 19 PinOBD
1​
2​
17​
5​
GND
3​
1​
1​
4​
14​
2​
5​
2​
3​
6​
8​
6​
7​
16​
5 500ohms
8​
18​
4​
9​
18​
4​
10​
7​
7​
K-Line
11​
12​
15​
13​
14​
15​
4​
16​
17​
3​
16​
18​
3​
16​
19​
5​
8​
K-Line SAAB Protocol
20​
6​
9​
21​
11​
10​
22​
13​
11​
23​
10​
12​
24​
19​
13​
25​
9​
14​
26​
12​
15​
 

·
Registered
Joined
·
40 Posts
I noticed since the forum update the table in Post #8 is all screwed up. Here it is

Pin out of the DLC cable
D26
Round 19 PinOBD
1​
2​
17​
5​
GND
3​
1​
1​
4​
14​
2​
5​
2​
3​
6​
8​
6​
7​
16​
5 500ohms
8​
18​
4​
9​
18​
4​
10​
7​
7​
K-Line
11​
12​
15​
13​
14​
15​
4​
16​
17​
3​
16​
18​
3​
16​
19​
5​
8​
K-Line SAAB Protocol
20​
6​
9​
21​
11​
10​
22​
13​
11​
23​
10​
12​
24​
19​
13​
25​
9​
14​
26​
12​
15​
You sir are a blessing !
 

·
Registered
Joined
·
40 Posts
Sorry No - only because I haven’t had any need. ( your talking about the serial port on the side that connects to the computer ? )

i ended up getting desperate so bought another VCI which incidentally is also not quit right But the functions the SAAB needed worked.
I was actually talking about the 26 pin serial port where the DLC connects to the VCI, however you've answered this in the post above :cool::cool:

Right now I am battling with the following situation:

My 3rd VCI was shipped with the errors below, and everything worked just fine up until it didn't . No matter what I did no connection could be made with the vehicle ( CANdi not connected or anything. Candi does boot correctly, flashes are intermitent but are slow flashes opposite to quick flashes when connected)
Tech 2 Wiki says this might be a bad VCI however at this point I don't know what it could be.


269632
269633
269634
 

·
Registered
Joined
·
156 Posts
Discussion Starter #19
Given some of the IO pass and some fail and both MUX’s pass I would be looking for a bad solder connection possibly on the MUX’s
As far as I can tell we don’t need all those data lines to talk to the Saab but candi module might need them?

I ran out energy after we had our 3rd child last year so shelved it all. But I’ll dig out my notes. I spent a lot of time tracing the tracks from the connector but never got a chance to document what I had done.
I try and get some info together for you tomorrow
 

·
Registered
Joined
·
40 Posts
Given some of the IO pass and some fail and both MUX’s pass I would be looking for a bad solder connection possibly on the MUX’s
As far as I can tell we don’t need all those data lines to talk to the Saab but candi module might need them?

I ran out energy after we had our 3rd child last year so shelved it all. But I’ll dig out my notes. I spent a lot of time tracing the tracks from the connector but never got a chance to document what I had done.
I try and get some info together for you tomorrow
Fun fact: it worked with those errors for some time, I was able to connect multiple times with the car, then all of a sudden... not connecting anymore. I am stumped o_O
 
1 - 20 of 43 Posts
Top