Adapter Optimization: Tips & Best Practices

Hello everyone.

Although I believe that my VAGCOM adapter version 1.1 will generally perform more reliably than the initial design, I cannot rule out the possibility that problems may still arise.

To minimize potential issues, I need data about the interfaces that will be connected – specifically, from the diagnostic connector on the car and from the COM port on the PC.

It is especially important to consider the signal levels, impedances (internal resistances), and detection thresholds for both high and low states of all signal pins involved.

Apparently, not all cars or PCs/laptops have the same specifications, because otherwise, the same adapter wouldn't work in some cars/computers and not in others.

Therefore, it wouldn't be very helpful for me to continue measuring the conditions within my own system, as those would only represent one example within the possible range of interface data.

Instead, I need something like normal ranges for levels, impedances, and high/low thresholds, within which (theoretically) all pairings of cars and PCs should operate (let the mischievous think what they want ).

Who knows data, links, etc.?

Just a question... is it possible to measure these voltages and everything around them using a standard multimeter, or to determine them in some other way?

I would like to help you... but unfortunately, I'm not quite sure how..

best regards, BB

BlauBaer wrote:

just a question...is it possible to measure these swellings and everything around them with a regular multimeter, or to determine them in some other way...? I would like to help you... but unfortunately, I'm not quite sure how..

Hi.

No, that probably won't help.

First, you would need to bring the COM port on the PC to specific states and maintain them for the measurement duration (I'm not sure how to do that).
On the other hand, your PC would "only" provide a 2nd example that falls somewhere within the range of possible (and compliant, if such a standard exists) values.

The same applies to cars.

I, on the other hand, need, as I said, the possible overall bandwidth of the interface parameters.

Thank you anyway .

look:

http://www.carbussystems.com/pdfs/ECT2002Workshop.pdf

http://techtrain.microchip.com/masters2002/Classes/647/647%20LIN.pdf

CU Gremlin.

Hi.

Great, I now have what I need for the car section.

Does anyone else know of anything similar for the PC's COM port?

Hello,

auf der PC-Seite ( meist eher Laptop ) sieht es eigentlich ganz gut spezifiziert aus. But!!!
The RS232 standard specification allows for voltage levels between +/-2.5V and +/-15V with a 3kOhm load. Older desktop PCs had drivers that were connected to the +12V and -12V of the PC power supply.
"For example, in the years 1488/1489, these circuits produced voltage levels of +/-10V. Laptops, however, don't have the luxury of a +/-12V power supply. Therefore, RS232 drivers generate these voltages from a charge pump, starting with +5V to create +9V, and then inverting it to -9V. Typically, only 100nF capacitors are used for the charge pump, and good charge pump drivers require silicon, which costs money. Modern laptops use multipliers based on the 3.3V logic voltage."
A better example here http://www.linear.com/pdf/1348fa.pdf

I estimate that most laptops can handle a load of 3 kΩ with a voltage level of +/-5V, while better ones can handle +/-6V. The extent to which the voltage levels sag with smaller load resistors (i.e., how high the internal resistance of the individual driver is, and how high the resistance of the entire charge pump is when multiple outputs are simultaneously high) is unknown. The specifications are usually given for a load of 3 kΩ. However, they typically drive 10mA without any problems. It should be clear that exceeding the RS232 load specification can lead to significant differences due to this.

The situation is slightly better with the RS232 inputs, which typically operate in the TTL voltage range, i.e., between +0.8V and 1.5V, with some hysteresis (approximately 0.3V).

Are all questions answered?

Best regards.
Christian

now I'm going to completely confuse things

Okay, so the side facing the car is now clear. Now you just need to ensure proper voltage levels on the RS232 connection. You can use a MAX232 or similar to make sure everything works correctly with the laptop as well.

However, I must reiterate: Jeff's wiring diagram, with a 1-meter cable to the car and a 1.5-meter cable to the laptop, has worked without issues on A4 (1997), A4 (1999), Octavia (2002), Passat (2001), Golf (1994), and Ibiza (1995)...more to come.



CU Gremlin.

Gremlin wrote:

So, the side facing the car is now clear. Now you just need to ensure proper voltage levels on the RS232. You can use a MAX232 or similar to ensure everything works correctly with the laptop as well.

Hi.

As someone who's not very tech-savvy when it comes to PCs, I obviously don't know what a MAX232 is...

...but I've devised a circuit with an "afterburner" that essentially replicates the Jeff plan and includes the transistor stage that I initially removed.

While this makes the whole thing a bit more complicated, I think, in a grandiose way, "it can't get much better for a universal solution."

Quote:

However, I must also say this again: Jeff's wiring diagram, 1 meter cable to the car, 1.5 meters cable to the laptop, and no problems with A4 (1997), A4 (1999), Octavia (2002), Passat (2001), Golf (1994), Ibiza (1995)...more to come

What always surprises me is how this thing can even function properly, considering how primitively the driver transistor is controlled towards the PC (only 140 k ohms between the K-line and the base), and how only levels below approximately 0.6 volts are recognized as "really low".

. . . but practice is when it works and nobody knows why .

chris11 wrote:

The RS232 standard specification allows for load levels between +/-2.5V and +/-15V at 3 kOhms. Older desktop PCs had drivers that were connected to the +12V and -12V of the PC power supply. "For example, in the years 1488/1489, these circuits produced voltage levels of +/-10V. Laptops, however, don't have the luxury of a +/-12V power supply. Therefore, RS232 drivers in laptops generate these voltages using a charge pump, starting from +5V to create +9V, and then inverting it to -9V. Typically, only 100nF capacitors are used for the charge pump, and good charge pump drivers require silicon, which costs money. Modern laptops use multipliers based on the 3.3V logic voltage." I estimate that for most laptops, you can calculate with a load of 3 kOhms and a voltage level of +/-5V, and for good ones, +/-6V.

Hi Chris,

...and that's the situation we're in.

In my opinion, an LED in an optocoupler should be able to be safely powered with 10 mA.

However, 5 volts across a 3 kOhm resistor only results in a current of 1.7 mA, which is only about 1/6 of the desired current.

15 volts at 3 kOhms is already 5 mA.

To effectively manage the relatively low impedances on the device side with such a wide range, it appears necessary to use an impedance converter, as Noxon and Stautz already planned, but I initially didn't understand the purpose – I wasn't aware that some computers have such weak output signals.

So, I'm going to expand the adapter plan again and send Rainer version 2.0.
(hopefully he doesn't feel completely screwed over yet ).

Quote:

Things are slightly better with the RS232 inputs, which usually operate in the TTL voltage range, i.e., between +0.8V and 1.5V, with some (0.3V) hysteresis.

What do you mean by "mostly"?
Where are the actual low-high decision thresholds located?
And what are the input impedances?

During my oscilloscope measurements, for example, I observed both positive and negative voltage levels of varying heights on the RXD pin.

Quote:

Are all ambiguities cleared up?

Yes, quite so. Thank you very much for getting this far!

Hello Ulf,

ulf wrote:

(hopefully he doesn't feel completely screwed yet )

No, don't worry, besides, I had already suspected something like that.
Tomorrow, I'll take a look at the PDFs. Maybe we can incorporate a MAX232 chip into the circuit. The MAX232 is a level converter that converts TTL signal levels (i.e., 0V / 5V) into RS232 levels, and vice versa.

Best regards, Rainer.

Rainer K. wrote:

Perhaps it would be useful to integrate a MAX232 into the circuit.

The MAX232 is a level converter that converts TTL signal levels (i.e., 0V / 5V) into RS232 levels and vice versa.}

Hi Rainer,

What the farmer doesn't know, he won't eat.

I have some experience building things with discrete transistors (unlike with a MAX232), so I'm using them as an add-on.

I'll email you the plan right away.

Perhaps you could make both versions available for download: version 1.1 for PCs with a "strong" COM port output, and version 2.0 for devices with weak driver levels or unknown data.

Hello Ulf, okay, upload it .

Best regards, Rainer.

Oh, guys...

It doesn't force me to design an interface either .
'We don't have time.'

@ulf

RS232 defines 12V as the LOW level and -12V as the HIGH level. Therefore, ICs like the MAX232 generate ±12V from 5V using a charge pump, which is then used to drive the RS232 interface. On the other side, a TTL signal is output. It's extremely simple to use such a component.

CU Gremlin.

Gremlin wrote:

RS232 defines 12V as the LOW level and -12V as the HIGH level. Therefore, ICs like the MAX232 generate ±12V from 5V using a charge pump, which is then used to drive the RS232 interface. On the other side, TTL signals are output. It is extremely simple to use such a component.

Hi.

It might be possible, if you have the application documentation... but where exactly did you want to build something like that into the adapter ?

1. The PC does not provide sufficient power via the RS-232 interface.

2. The fundamental galvanic isolation between the car and PC sides should be maintained.



According to my understanding, the best way to install the Max in a PC/laptop would be to bring the RS232 up to the required power level... or am I completely wrong?

It seems to me that an adapter with output transistors acting as afterburners (powered by the vehicle's electrical system) is the simplest overall solution.

According to my understanding, the best way to install the Max into a PC/laptop would be to bring the RS232 up to the required power level... or am I completely wrong?


Actually, yes, because the laptop's RS232 port should, in theory, comply with the specifications.

Please send me your address via private message, and I'll send you something.

CU Gremlin.

ulf wrote:

chris11 wrote: The situation is slightly better with the RS232 inputs, which usually operate in the TTL voltage range, i.e., between +0.8V and 1.5V, with some (0.3V) hysteresis. What do you mean by "mostly"? Where are the actual low-high decision thresholds located? And what are the input impedances? During my oscilloscope measurements, for example, I observed both positive and negative voltage levels of different magnitudes at the RXD pin. . .

Hello,

The above datasheet specifies that all voltages above +2.4V are recognized as electrically high (not to be confused with logically 0), and all voltages below +0.8V are recognized as electrically low. In normal operation, with a short cable, the output voltages of an RS232 driver are +/-5 to +/-10V at the receiver. Therefore, it is normal to measure negative voltages during operation.
The input impedance of the receiver is specified in the same datasheet as 3kΩ to 7kΩ relative to ground. This should also apply to most laptops and PCs.

To maintain isolation, you will either need to build discrete circuits using transistors or use an RS232 driver chip like the MAX232 or LT1181, along with an external 5V power supply (e.g., a 9V battery with a 78L05 regulator). Based on my experience, batteries often run out of power when you need them on a Saturday evening, and it's important to figure out why you can't get the vehicle to the gas station (where you can get a replacement battery) before resorting to a solution that requires an external power source. Therefore, I prefer a solution that doesn't require any external power.

Best regards.
Christian.