VAGCOM Adapter: Never Ending Story?

Hello everyone.

If anyone is wondering why the optocoupler adapter construction guide appears again in a new version within the diagnostic technical article:
The main difference between version 2.3 and 2.4 lies in the use of LEDs.
Number 3 only illuminates during the dialogue when data is received from the vehicle, and a sufficiently accurate comparison for VR 1 and 2 can be achieved without measuring devices by performing the VAGCOM LED test based on the brightness of LEDs 1 and 2.

The achievable transmission quality of 2.4 is only slightly better than that of 2.3, provided that the latter is configured on the diagnostic PC used for the VAGCOM LED test as follows:
Adjust VR1 and VR2 so that the voltage at pin 3 of OK 1 and 2 does not exceed approximately 8-9 volts (this is a more reliable criterion than the collector voltage of transistors T1 and T2).
Adjust VR3 so that LED 3 no longer emits light at its maximum brightness.
If you encounter (unlikely) dialogue issues, fine-tune them again gradually.

If you want to bring the performance of the 2.3 up to the electronic level of the 2.4, you can replace D4 with a wire jumper and connect it in an antiparallel configuration to the output of OK3 (as D3 is connected on the 2.4). Then, you should also replace R3 and R4 with 2.2 kOhm resistors each, to reduce the load on the COM port and potentially make LED 3 shine a little brighter, which in turn would facilitate the adjustment of VR3.

But previously, there were reports that the adapter only worked with quite low values for R 3 .
That's correct, but the cause of this phenomenon – the (over)saturation of the OK 3 output – can be eliminated with VR 3, which was not yet included in the original circuits.

Adjustable adapters such as 2.3 or 2.4 can circumvent the inherent drawback of optocouplers, namely error-prone timing distortions in the transmitted bits, the extent of which is also heavily dependent on the inconsistent voltages and impedances at the respective COM port.

I suspect that this problem with purchase adapters, which often leads to complaints, is the main reason why there's a trend away from optocouplers in those devices. However, this comes at the cost of an increased risk of damage to the PC and onboard electronics under worst-case conditions.
However, this doesn't count when it comes to sales. What's more important is the certainty that a well-designed adapter, without optocouplers, will work on virtually any PC, and therefore, at least, no compatibility issues are expected.

"If you are experiencing communication problems with an older version of the optocoupler adapter and do not want to upgrade it to versions 2.3 or 2.4, you could try replacing the 1 kΩ resistors in series with the optocoupler inputs with 1.5 kΩ or 2.2 kΩ resistors. For optocoupler input 3, you might also try values up to approximately 6.8 kΩ (as long as the adapter is still recognized)."
If the length of the transmitted bits, as altered by the adapter, was previously (partially) within the error range, it may, with a bit of luck, be shifted into the "green" zone.

A carefully calibrated 2.4 should, in my opinion, be reliable enough for all current needs of the VAGCOM system, even if fluctuating operating voltages (engine running or stopped, battery charge level) may still cause slight distortions in the transmitted bit length.
I could provide a corresponding "repair" with additional components if experience shows that the trimmers need to be constantly readjusted depending on whether the motor is stationary or running, especially during demanding = particularly fast communication processes.

Sorry for the frustration of VAGCOM users who had or are having dialogue problems with my previous build suggestions - and my envious congratulations to those whose Noxon minimal adapter (by chance) works so well with their diagnostic PC that no problems have occurred so far .

ulf wrote:

With adjustable adapters such as 2.3 or 2.4, it is possible to bypass the actual "Achilles' heel" of optocouplers, namely error-prone timing distortions of the transmitted bits, the extent of which also depends heavily on the inconsistent(!) voltages and impedances at the respective COM port.

Hello,

I've been reading here for quite some time, and I'm particularly interested in the VAGCOM adapter.
I also have a suggestion to make the adapter more stable: why not combine your solution and Maik's solution?
With a MAX232, one could potentially solve the main problem, which is the different voltage swings and driver strengths of the COM ports used. For power supply, I would use the PS/2 port, which is often found in older laptops. At the output of the MAX232, there would be optocouplers, whose configuration should now be easier to calculate. For newer laptops with a USB port, one could replace the MAX232 and PS/2 port with a microcontroller from FTDI.
What do you think of the idea?

Best regards,
Andreas

ulf wrote:

I apologize for the frustration of VAGCOM users who have experienced or are experiencing dialogue problems with my previous construction suggestions.

Hi Ulf,

That's going too far; you have to apologize for your efforts.


My/Your 2.2 adapter works perfectly for me (with 3 TDIs and 1 gasoline car). I even tried it with my old notebook running Windows 95 (potentiometers set to the middle position) -> no problems at all!

Best regards, Christian.

AndreasK wrote:

Why not combine your solution and Maik's solution? With a MAX232, one could potentially solve the main problem, which is the different voltage swings and driver strengths of the COM ports used. For power supply, I would use the PS/2 port, which is often found in older laptops. At the output of the MAX232, there would be optocouplers, whose configuration should now be easier to calculate. For newer laptops with a USB port, one could replace the MAX232 and PS/2 port with a microcontroller from FTDI. What do you think of the idea?

Hi Andreas,

That doesn't sound bad for a first attempt .

If this is feasible in practice, and especially if the microcontroller doesn't interfere with the timing protocol of VAGCOM, it would likely be a worthwhile project.

Right now, I'm just wondering how to get VAGCOM to communicate via a USB port instead of a COM port.

ulf wrote:

At the moment, I'm just wondering how to get VAGCOM to communicate via a USB port instead of a COM port...

Not at all.
"VAG-COM won't detect it. A miniport driver is provided for the microcontroller, which makes it appear as a COM port. That's what makes this whole thing so interesting."

Best regards,
Andreas

AndreasK wrote:

A miniport driver is provided for the microcontroller, which makes it appear as a COM port. That's what makes things so interesting.

Hi Andreas,

Okay, let's build a prototype and let us know how it works .

I believe that there will be an increasing demand for such solutions in the future, especially for laptops without a COM port.

Can you provide a rough estimate of the material costs here?

@ulf:

I believe VAG-COM version 2.082 was the first one to support USB connectivity. You can select 'USB' as the interface option.

joergs wrote:

I believe the VAG-COM version 2.082 was the one that already supported USB connectivity. You can select "interface USB" when configuring it!

Hi Jörg,

Hmm? So, my 2082 only offers the options COM 1 to 4 .

ulf wrote:

A miniport driver is provided for the microcontroller, which makes it appear as a COM port. AndreasK wrote: That's what makes things so interesting. Hi Andreas, Okay, let's build a prototype and let us know how it works ?

What else should be added?

As soon as optocouplers are involved, their outputs exhibit relatively low slew rates, which delays the bits and, depending on the switching point of the following stage from high to low, also distorts the bit lengths to a greater or lesser extent.

In practice, not only the data from the COM port but also variations in individual components can cause errors. Therefore, even in a precisely defined electrical environment, it is necessary to include trimmers, such as those found in Adapter 2.4, if one wants to optimize the system for minimal errors within the capabilities of the OK system.

@ulf

Sorry, I'm completely confused after trying and downloading so many different versions.
It wasn't the 2082! It was the 212.2! You can download it from the Ross-Tech website as a beta version, both with and without USB support.

ulf wrote:

Then go ahead, build a prototype and let us know how it works . I believe that there will be an increasing demand for such solutions in the future, especially for laptops without a COM port. Can you provide a rough estimate of the material costs based on a "Pi x Daumen" approach here?

Hello, hello.

I'm planning to start working on the prototype. However, it will likely take at least another 14 days before I can begin. Firstly, I am not currently in the country, and secondly, I still need to receive the evaluation board from FTDI. The microcontroller costs approximately €12 as a single chip, which means surface-mount soldering is definitely required. Let's see how much the evaluation board costs; it will certainly be more convenient for "manual soldering."
Regarding the optocouplers, I'm going to consult with a few more acquaintances; perhaps they have some ideas on how to improve the slew rates and thus reduce the delay times.

Best regards,
Andreas

Hey guys,

wenn ihr den PC mittels Max232 oder USB-COM Adapter auf TTL Pegel bringen wollt, würde ich zur Potentialtrennung den OK 6N136 von Sharp empfehlen (Reichelt 0.77 Euro). He is for it.
The system is designed for 5V applications and should introduce a delay of approximately 1 microsecond without any calibration.
You can find the datasheet using Google or a similar search engine.

Okay, then.

BERT.

'Version 2.4 has brought about quite a few changes. I've now tested version 2.3 with the T3=BC879 setting and acceleration capacitors on both my '95 Polo and my '02 A3. Even with a 9600 baud rate, the communication works smoothly. As for the laptop, I'm using an old Toshiba 4800 (with a 486DX40 processor) that has 16 MB of RAM and runs Windows 95.'
I'm going to test the modified return process with OK3. It looks promising.
I would be grateful if someone could share their experience with the runtime performance of version 2.4.

Hi Andy,

You have mail.

Hi there,

I have a tool that allows you to address RTS (L-line) and TXD (K-line) connections both statically and dynamically.
If there is interest, I could email that document to a central point of contact.


Okay, then.

Yes, then I could measure whether the voltages are being reduced sufficiently, without having to solder anything or rely on guesswork.
The LED test in VAG-Com is running too fast.