DIY Hysteresis Test: Build Your Own!

Hello,

I've done some research on hysteresis testers and I've come to the conclusion that it should be possible to build one yourself, provided that the necessary data is available.
Yes, my TDI has the gentle jerking .
If someone can provide me with this data, I will build the circuit, test it, and share the results with all of you . To do this, I have made the following considerations. Please correct any mistakes I may have made!

1. To control (setpoint) the MSW, the control unit generates a pulse-width modulated rectangular signal to achieve efficient current regulation. Optionally, control can be achieved via DC, and a simple current sink would be sufficient for this. What is the range of the control current? Otherwise, a rectangular wave generator with variable pulse width is needed. The data required for this are: frequency, pulse width (min. - max.), and the maximum pulse current .

2. The feedback is provided either via a potentiometer or an inductive sensor. In the ESP system with a potentiometer, a voltage is directly fed back to the control unit (ECU). The inductive sensor consists of a measuring coil and a reference coil. To eliminate temperature dependencies, the difference between the two inductances is measured and converted into a voltage, which then represents the actual value and is reported to the ECU.

Okay, here's the translation:

'Now for the tricky part. To understand how the ECU measures inductances, it's important to know how it processes the data. What information does the ECU receive? Does it receive the individual inductance values, their difference, or a DC voltage? In the most complex scenario, the ECU generates an alternating voltage with a constant frequency, which is applied to both coils. The resulting current then depends only on the inductances. The ECU calculates the current difference and converts the result into a voltage – and that's it. The question is, which 'inductance difference' corresponds to which voltage.'

4. If it had worked, we would have finished the electrical part. The setpoint and actual values still need to be compared. But are these values proportional, and if so, what is their ratio? If they are proportional, both values could be compared using an operational amplifier (op-amp), and the control difference would be present at the output of the circuit with proper design and adjustment. I know – an op-amp can also work with current.

That's all for my theory. Do you find the idea interesting, and more importantly, feasible? If so, I kindly request your support .


Igor.

Igor wrote:

1. To control (setpoint) the motor-driven valve (MSW), the control unit (STG) generates a pulse-width modulated rectangular signal to achieve efficient current regulation. Alternatively, control via direct current (DC) is possible, and a simple current sink would be sufficient for this. What is the range of the control current? Otherwise, a rectangular wave generator with variable pulse width is required. The data required for this are: frequency, pulse width (min. - max.), and the maximum pulse current .

Hi Igor,

The active range of my DC power supply is approximately between 2 (lifting from the zero position) and 4 amps (maximum load).
However, I believe a pulsed actuation would be better because it would introduce "realistic" small vibrations into the pump, and "micro-stalls" could be better avoided.

Quote:

2. The feedback is provided either via a potentiometer or an inductive sensor. In the ESP system with a potentiometer, a voltage is directly fed back to the control unit (ECU). The inductive sensor consists of a measuring coil and a reference coil. To eliminate temperature dependencies, the difference between the two inductances is measured and converted into a voltage, which then transmits the actual value to the ECU.

Okay, I'm ready. Please provide the text you want me to translate from German to English.

Quote:

In the most complex case, the STG (Steuergerät - control unit) generates an alternating voltage with a constant frequency, which is applied to both coils. The resulting current depends only on the inductances. The STG calculates the current difference and converts the result into a voltage - that's it. The question here is which "inductance difference" corresponds to which voltage.

That's how it's done, AFAIK. It's measured, IMO, using a 10 kHz signal.

Quote:

That's all for my theory. Do you find the idea interesting, and more importantly, feasible? Then I kindly request your support .

An interesting project .
I think you should develop a measurement circuit that is as simple as possible and that as many people as possible can reproduce.
They use this to measure the performance of their fuel pump and report the results along with the mileage, whether or not the engine is misfiring, and whether or not a "lean correction" has already been attempted to address the misfiring.

Based on the collected data, it might be possible to estimate which measurement values a "healthy" pump should produce.

@ Igor

It would be great if you could create something like that. I'm very interested in it. I also have that 'slightly' annoying stuttering... It's getting worse... But I still haven't been able to bring myself to spend 1000€ on it.

Deleted upon request.

Hello Igor,

. Now, the tricky part. To be able to measure the inductances, it's interesting to see how the ECU measures the inductances and further processes that data.

In the journal 'Motortechnische Zeitschrift' (MTZ) 53 (1992), issue 5, it is explained how the position feedback works via the half-differential short-circuit ring sensor (HDK). As Gustav Diesel already mentioned, this is evaluated in the control unit using a CMOS circuit, specifically the CC212. The article briefly describes the functionality of this circuit. You will probably need to find a datasheet for the CC212 somewhere.

Hello.
Wolfgang.

Hello fellow DIY enthusiasts,

Thank you for the suggestions. I received the parts for the OBD adapter yesterday. I will start working on this project first, and then I will continue with the hysteresis test. So, it might take a little longer.

I've been searching the web for quite some time now to find the MTZ article about the HDK and the datasheet for the CC212. Unfortunately, I haven't been able to find them yet. Can anyone tell me where I can find these documents and who manufactures the CC212?

See you soon.

Igor.

Hello Igor,

You can find the MTZ publications in university libraries or, for example, at the Munich City Library. If you don't have access to them there, I can also send you a copy, or someone could scan it for you. There are also other interesting articles available.

Hello.
Wolfgang.

Hi,
CCxxx indicates the manufacturer Bosch (at least, the Intel CAN controllers produced by Bosch are also named CC followed by some numbers).
The sensor IC in the MSA15 is from National Semiconductor and has the Bosch part numbers B58328 or B58328TS.
I once disassembled an MSA15 that had clear text printed on it (starting with 'LM'), but a search on the National Server only yielded a reference to a discontinued OEM product that was being replaced by another OEM product. No datasheets were available for either. An inquiry to National only resulted in a polite response stating that it was a product manufactured for a specific customer, and that they unfortunately could not provide any documentation about it.

Best regards, Rudi.

Deleted upon request.

diesel.gustav wrote:

A self-built solution for "personal use" is therefore not really practical for me. Why don't you ask some Bosch service centers how much they charge for a hysteresis test on a VP36 or VP37 injection pump?

Hi Gustav,

This is, after all, a "DIY forum," where it's not just about cost-effectiveness... one's own time, as is well known, has no price there .
. . . unlike workshops, where you usually get a standard measurement , without any chance to chat about details .

Regarding the measuring device: It is not necessary to use or meticulously copy the CC212.

A 10 kHz generator is relatively easy to build (using an astable multivibrator), and an evaluation circuit for measuring differences (using a rectifier/integrator) is probably feasible as well. Furthermore, the control system for the signaling system is, as far as I know, still largely unresolved.
So, we should come up with something to design a way to determine the hysteresis in some way with sufficient accuracy.
If this measurement is consistently applied here, then, in my opinion, we should definitely be able to } obtain comparable data.
However, it is highly likely that these measurements are not comparable to those taken with professional equipment from brands like Bosch or other professional manufacturers.

I will primarily focus on VAGCOM adapters, and (for the time being) will not be involved in the development of the hysteresis project.