Thesis - ulepszanie zawieszenia (skyhook)

[belial666]

Filtr dolnoprzepustowy prawdopodobnie jest po to, żeby wyeliminować szumy o wysokiej częstotliwości które nie wynikają z "odczytu drogi" tylko z zakłóceń.

Odnośnie czujników dlaczego 5. Gdzieś tam poniżej Vladimir napisał że czujnik przechyłów bocznych w bagażniku prawie nigdy nie siada i raczej jest sprawny. Padają najpierw przy kołach, później te w komorze silnika, główna przyczyna uszkodzenia to temperatura.

No i 5 za zderzakiem przednim pod prawą lampą. Liczę na to, że ktoś to przetestuje i zda relację. Ja mam "kwadratne" czujniki.

[s0krates]

Zawsze możesz wymienić wszystkie z centralką na używki od Heliosa albo brygadzisty1, a potem ogień. Tylko dobrze by było skoro już V jest taki otrzaskany w temacie jakby się wysypał jak te czujniki sobie posprawdzać.

[belial666]

Zawsze możesz wymienić wszystkie z centralką na używki od Heliosa albo brygadzisty1, a potem ogień. Tylko dobrze by było skoro już V jest taki otrzaskany w temacie jakby się wysypał jak te czujniki sobie posprawdzać.

Miekkość mojego auta mi się podoba, jak na Thesis. Zawieszenie w Thesis wymaga poprawek by to chodziło inaczej. Skyhooki też. No czekam na chętnego, by to przetestował.

Poprawek, które bym chciał w zawieszeniu Thesis, te czujniki nie ogarną... Bedzie jak po fabryce, tyle.

[vlad47]

As mentioned the low pass filter is there only to eliminate the noise. Original sensor has nearly endless frequency. Skyhook shockabsorbers work on 2KHz and modified sensors transmit at more than 4KHz, so also Shannon-Kotelnik theorem is applied. At 130km/h the sensor can still measure each cca 0,8cm.
The best is to change all 6 sensors but for those who want to save money it's advised to change at least 5 of them (except the one in the trunk). The sensors are getting older and after some time they got broken. The problem is that none of us has an oscilloscope that can measure such a frequency as mentioned higher, so there's no way how to check the sensor only change and try if there's some difference. You can measure only if the sensor is in very bad condition.

[s0krates]

First - welcome to our forum, great to have You posting here
My scope measures up to 20 MHz - it's quite bigger range than 2kHz, but for someone having good connection to university it should be no problem to access scopes for even 600Mhz.
But as you know, the bigger the frequency is, the less power it is giving at the same amplitude, so HF noise should count less than LF.
Are You sure that Sachs didn't make filter inside the control unit? It's unbelievable that they ommit such thing..

I would be very grateful if You could post here pins description of a sensor. First step would be to check static noise and then compare to Your sensor. I guess somebody here (Bartek?) could also send me sensor that is 100% defect..

[vlad47]

That's only a good info you can measure it . Sachs used a filter but so small that compared to the frequency of shockabsorbers, it can't be even considered. The noise is not only caused by external conditions, but also by internal especially the G measuring chip itself. It has implemented error calculation, but you can see that after some time there are more and more errors that can't be handled by chip itself and the amplitude of the noise is big. And this is the problem, the bigger the amp, the bigger the problem. So we are not talking about noise from magnetic field etc, but caused by the chip itself.
That scope measuring seems to be quite easy, but it's a bit complicated. You have to realize that there are more than 4000 samples per second (my version, original has more than a few millions) and if there's only one bad, it's a problem. Sometimes the chip behaviour is so strange that it's working perfectly a few minutes and only than starts sending wrong G measurements for a few seconds, sometimes longer. So with the scope you have to record a log (as your eyes are not able to catch each sample in that rate) for a few minutes, with original sensor one minute would have a few hundreds of million samples and you have to check each one that there's no bad sample in a few minutes. Than if you check that, it's still not for sure the sensor is ok, because of that strange behaviour mentioned before. Some digital scopes have good software for analyzing the logs, maybe that should be the way.

[AUFER]

Welcome

It's great that you're with us

[bronx]

Hi vlad, is nice to meet you.

... The problem is that none of us has an oscilloscope that can measure such a frequency as mentioned higher, so there's no way how to check the sensor only ....

Nothing could be more wrong,
My oscy :
http://teledynelecroy.com/oscilloscope/ ... series=339

and some realy very very good textronics oscy


Yet I didn`t have time for the sensors ... yet, I have hope.
But I'm open to suggestions if you want to try something or measure

Try to convince me to measure

[vlad47]

Many thanks for your effort, it's nice to see that open space for help, but this is luckily behind me and now I've solved most of the issues.
To add, the point of my sensor production is not to produce better sensor than original, but that the sensors are getting old and working worth and worth so they need to be changed, the modification is just a bonus.

[s0krates]

OK, so then we should measure in "peak hold" mode to catch maximum of noise.

Can You give us info about which pin is which function? Is the output signal 100% analog?

Im seriously interested, but first i need to know that my sensors are really faulty - easiest would be to drive Your Thesis, but we're living far away unfortunately..

[vlad47]

Oh I'm really sorry, I've forgotten.
White -> ground
Red -> power +
Yellow -> output signal

[s0krates]

thanks, power is +12V?

ps. how's Your project of daylights for Thesis?

[Bartek-Lodz]

Zadam może strasznie głupie pytanie ale...
co to za historia z różnymi czujnikami to znaczy, które są inne w zależności od rocznika - bo jeżeli chodzi o czujniki na nadkolach i belce oraz w bagażniku to w moich Thesisach z 2002, 2003, 2004, 2006, 2007 i 2008 są identyczne. Fakt, centralka z 2008 ma trochę inne napisy ale numer katalogowy ten sam, zresztą działa z powodzeniem w aucie z 2004 roku bo sobie przełożyłem w ramach walki z zawieszeniem.

W takim razie pozostały czujniki te na amortyzatorach, może to o nie chodzi. Jeżeli ktoś wie to proszę o informację.

[bronx]

Kształtem sie roznily, nie dam ci 100% odpowiedzi bo dam nie wiem. Jedne byly kwadratowe drugie chyba prostokątne. Zaznaczam chyba!

[vlad47]

Power supply is 5V.
There are three different versions in two different cases. One case is cube and another one rectangle. In rectangle there are two different versions inside but you have to disassemble it to see.