Display running engine vibrations.

[john1]

Hi,

I'm thinking of making a unit to display my car engine vibrations.
Maybe an oscilloscope type of display.

Just fishing for any ideas or suggestions.
It's a diesel engine, so i will have to sort out some way to synchronise
the display with the engine revolutions.

John

 

[Pommie]

Save your energy, it's a diesel, it vibrates, what more do you need to know.
If it vibrates lots then get the injectors cleaned out.

Seriously, why do you want to see the vibrations?

Mike.

 

[crutschow]

For a display, the easiest and probably cheapest would be a small digital oscilloscope such as the Velleman (**broken link removed**.

Analog Devices makes a number of accelerometer chips that could be used to sense the vibration (**broken link removed**)

I don't think you particularly need to synchronize to the engine revolutions. The display will likely have a prominent fundamental vibration from the engine that the oscilloscope can synchronize to.

 

[RadioRon]

I strongly recommend that you get a "slip-synch stroboscope". I have used these in the past during vibration testing of electronic products to find resonances, loose components and all sorts of other problems. This is a great tool.

It is a strobe light that you synchronize to the source of vibration and then do some magic. In our case, we synchronized to the shaker table which was fairly easy. In your case, you would probably want to synchronize to a tachometer or something related to the engine revolutions. Now, the magic part. This special kind of strobe light has a trigger in it that has its own oscillator. You can choose to synchronize the oscillator to the external input (the tach) or you can adjust the frequency of oscillation to vary some number of Hz away from the frequency of the external synch signal. Its not just a phase shift, it is a true frequency shift. You can also phase shift if you like. When you "slip" the synch like this, you start to see relative motion of things and this relative motion becomes very easy to see. Thats the magic part.

It might be easier and cheaper to buy a triggerable strobe lamp and build the slip synch trigger function yourself. I had a look around and there aren't very many ready-made used ones available.

 

[The Mad Professor]

Hi,

I'm thinking of making a unit to display my car engine vibrations.
Maybe an oscilloscope type of display.

Just fishing for any ideas or suggestions.
It's a diesel engine, so i will have to sort out some way to synchronise
the display with the engine revolutions.

John

Is this display for show value or a more serious application?

If it's just a whistle & bells gimmick to light the dashboard then
something like an LM3914 with a multi-plexed display will make nice patterns from the input off a "knock" sensor or microphone.

Of course if you want to get real serious , then there is allways the in car computer option. There is some free oscilloscope software lurking about on the web that uses the microphone input on the motherboard.

Don't worry about spending a fortune on a Mini-ITX or Pico size motherboard that will fit in the dashboard, scour the skips for an old pentium III ( or II at a push) computer as these should fit nicely under one of the front seats of boot when stripped down.

Alas no easy solution for monitor, but seven and 5 inch models are to be found specifically made for automotive use and look the business. Plus the other small joy of having a computer aboard is entertainment value.

 

[john1]

Hi POMMIE, Hi CRUTSCHOW, Hi RADIO RON, Hi MAD PROFESSOR,

Thank you all for your responses,
I will try to answer properly,

John

 

[john1]

Hi Pommie,

Seriously, why do you want to see the vibrations?

By way of engine diagnosis, like early warning of problems.

John

 

[john1]

Hi Crutschow,

I'm afraid that 'Parts Express' url just brought up a 404 for me,
but i did find some 'Velleman Oscilloscopes' on the Maplin site
**broken link removed**

However, i am not likely to spend anything like that sort of money on this
project, as it will be only for displaying the engine vibration pattern
for me to examine from time to time.
And probably compare with previous displays, to look for any bits that
might be getting worse.

The 'analog.com' page looked much more interesting.
I had not considered accelerometers, i was thinking of using small microphones,
like the ones in mobile phones and such.
However, microphones like that might be very susceptible to oil, maybe these
accelerometer chips are tiny, like transistors, and maybe they wouldn't be
susceptible to the oil and diesel on the engine.
I will try to check that out.

I have been giving quite a bit of thought to the synchronising.
As want to compare the display at intervals, probably over long periods, then
i want to make sure that it isn't going to synch up to something else at a
later date.
Also just to make it interesting, i want it to sychronise nicely over a wide
range of engine revs.

So i was thinking of making a device to drive the horizontal positioning
directly driven from the cam-shaft. Or some other part that runs at half
engine revs.

Unless theres an easy way to get the horizontal sweep to follow the engine
revs, and also stay the same length.

I'm kind of assuming a CRT is about the easiest option here.

John

 

[john1]

Hi RadioRon,

Yes, those units are very helpful tools for finding all sorts of loose bits and parts
that resonate under odd or awkward conditions.
We used to use very large motor driven shaker rigs to check out all sorts of assemblies.

However, i am hoping to display the vibrations with a view to gauging the wear on the
bearing parts, as a sort of indicator of the general state of the wear and tear.
In particular, the progress of the wear and tear.

John

 

[john1]

Hi Mad Professor,

Well, i must admit i had not considered that there might be some entertainment or
amusement value in making a display of the vibration patterns.
However this is not the purpose i had in mind, but it may indeed prove to have a
degree of gimmicky attraction.
I was not thinking of an in-car display, but it might be interesting to try that
later, but for the time being i am intending this to be more of a diagnostic tool
for external use, probably run from the electric mains, probably built using a CRT
unless i happen to find a ready made program to run such a virtual oscilloscope
especially for this purpose.

Such a possibility is unlikely, but the amount of work it would save me if such
a program were to be available as a download is such, that it is worth looking
for just in case such a program were to be found by searching for a few hours.

I have looked at a few of the free "PC-Oscilloscope" programs, and some of them
are interesting, but none of them seem to meet my desires.
I feel that my requirements are not unreasonable, nor impossible.
I am probably going to try and throw together my own unit, probably based around
a CRT.
Unless something crops up which will do it using a PC (well, it might)

John

 

[blueroomelectronics]

Why not just use the diagnostic port on your car?

 

[crutschow]

I am probably going to try and throw together my own unit, probably based around a CRT.

So are you trying to generate an oscilloscope type display? Or did you have some other ideas?

Actually, when they analyze machinery for vibration they often use low frequency spectrum analyzers (signal analyzers) that show a display of frequency versus amplitude. This is generally easier to interpret and compare with previous readings to detect changes then using the time domain. I believe there are some free PC programs that do spectrum analysis (some designed for audio room and frequency response analysis) that may work for you. An example of such audio software is at https://www.hometheatershack.com/roomeq/

 

[stevez]

It does sound like an interesting project. It may not be a surprise to you that real time monitoring and alarming of machinery has been done for years. What you propose makes a lot of sense. The challenge will be to sort out normal from abnormal. About 10 years ago I was involved in a project where we fitted a commercially available system to our machine ( fills a large building) at select points. The "select points" were simply bearings. There's some fairly well understood parameters that can be measured to indicate the health of the bearing. Periodically they'd collect real time data, look at on spectrum analyzer. The continuous monitoring part involved looking at amplitude at only one or two frequencies - probably selected based on rotational speed of shaft. If amplitude at select frequency (selected with filters) exceeded limits an alarm would sound.

Sorry for all the babble - the point I wanted to get to - take a look at the work that's been done to see if there are any clues that will help you. I'd expect that lots of info is available via the internet - some from people who make these systems.

 

[Nigel Goodwin]

I had a tour round a coal fired power station a number of years ago - they had acoustic sensors at various parts of the plant, and monitored the vibrations which they compared to 'normal'. In this way they were able to detect any problems before they occurred.

In a similar way they took temperature readings with an expensive Minolta lensed IR thermometer - and continually checked for temerature changes or sudden increases.

 

[john1]

Hi blueroomelectronics,

I don't know if my old Peugeot 205 has a diagnostic port. I will check, but i
doubt it would include the sort of signals i would be looking for.

John

 

[john1]

Hi Crutschow,
I had not thought of using a spectograph for this. I should have really, cos now
i think thats one of the diagnostic pieces that were used at the research place
where i used to work.
And yes, there are some free programs for these out there. So yes, that would be
a good start, to get one of those rigged up and working.

That is an excellent suggestion, and i will start looking at ways to do this first.
I will check out that link you posted.

John

 

[john1]

Hi Stevez,
your post confirms the value of this type of display, but i suspect that it will
not be easy to interpret the information, as i wont be able to fit sensors directly
to the bearings.
There are free downloads for this type of program available, so i will be sorting
stuff out to get a PC running out in my little garage, and looking at various
transducers, accelerometers, maybe microphones.
And maybe a small pre-amp to drive the 'Mic' input.

John

 

[john1]

Hi Nigel,
That sounds very similar to what i have in mind.

I had not considered monitoring temperatures around the engine, and if i did then
it would be with small fixed sensors. Those IR types are way out of my price
range.

John

 

[The Mad Professor]

Regarding sensors you might want to look at Piezo transducers.

Normally used as sounders / loudspeakers in everthing from greetings cards to cell phones,

however they do work the other way as well and make for very effective vibration sensors.

Small, cheap and impervious to dirt and oil contamination and tough enough to survive inside the harsh enviroment of an engine bay.