Nice video. Vibration analysis is a science and an art. If your device under test is vibrating at its resonant frequency, you will see a phase shift, meaning, the needed counterweight will not be opposite the apparent imbalance, but shifted 90 degrees lagging.
Man, this is good stuff. Back in 2007 when i went to ford asset school, technology wasn't this available yet. Sure they had chassis ears, but for cost, it wasnt justified usually. I just got my 4ch pico recently and im looking to add NVH capability, so this video is perfect. I also dabled in arduino for a couple things. I used my arduino for 5v for a pressure sensor to do waveform analysis yesterday. I also have a 3d printer, so if you would like, come up with an enclosure design for housing the parts and i can CAD & print it for you, no charge.
Could you do a detailed assembly video with close ups and a diagram? I am not an electronics guy but I would love making one of these. I’m rebuilding my vintage car engine and want to balance everything. Excellent video, your students are lucky to have you! Thanks!
This is great. If you wanted to improve the accuracy of the accelerometer at picking up high frequencies, you want to use a stiffer bonding material. Hard Resin is better than glue. Using the usb cable is a good idea because of the shielding. Also, for the frequency, I like to run the frequency spectrum to see all frequencies. Sometimes you have a shaft issue, for example, but also have a bearing worn that is really hard to see unless you look at the spectrum.
If you're drawing power from a power bank or stand alone battery the voltage reg is smooth but if you are going to power it from a running car battery you need the capacitors to smooth out the alternator noise
Hello from Down Under in Sydney Australia. I tried a similar Idea using the same 3 axis Arduino accelerometer connected to three separate LM3914 ten LED voltage indicators. I was trying to identify the X, Y, or Z vibrations in my 1965 Mini Cooper S engine. No matter what I did I could not get them to "Zero" at the same output voltage when they were stationary. FINALLY I realized that ZERO for the Z axis really means that it is in freefall. When it is stationary it is actually accelerating upwards at 1g. You have combined the three axis outputs for the single CRO trace. If you had kept them separate you would have have had the different output for the Z axis. Maybe that was the cause of your wide "zero" trace? Maybe gravity is -1g down here......... ????? Thanks Dave D.
Pretty straightforward. Biggest thing I'd change is making one box that holds the batteries, voltage regulator, and all the connections. That way the loose wires can connect to some terminal blocks, and you can put in some test-point style wire loops to connect the scope probes to(or binding posts/banana jacks if preferred). I'd probably 3d print an enclosure, it's not much work if you use the same form factor prototyping PCB for multiple different projects. Edit: that accelerometer module already has a voltage regulator on it, so you probably don't need to add one, just might need to tweak the resistor values for the led/phototransistor.
Great presentation Sir! I was looking for a way to develop a system to balance numerous items; truck drive shaft (custom), grinders which I use to make tooling for my lathe and CNC mill, props for RC aircraft, etc. I cobbled up a setup using a piezo connected to my Pico 4425, it worked great for concept. I wanted to be able to include angular measurement to facilitate accurate placement of balancing weight. Your system provides a great platform upon which to build; thinking of adding a 3 axis accelerometer into it also. Bottom line, you did great: lead-in, demo, and production! Thank you very much for the production of the video, and BTW, I did "like" and subscribe. Best wishes!
Can you show how you hooked both your ir sensor and accelorometer to your scope or lap top and what software are you using? Thank you for the great video!
The little IR sensor requires it to be very close to the rotating surface to work. No good for balancing wheels. I switched to a EK1254x5C IR sensor device from Amazon. It works up to 30 cm away and has a trim pot that allows you to adjust the sensitivity and trigger distance. It takes a 5 vdc power supply, so no resistors to add. 3 terminals to connect, ground, 5v in, and digital on/off out. Has an LED for power, and one that lights up when target is NEAR, all on the little board. I'm using the accelerometer and the EK1254x5C on a Y shaped wire running to a block of wood with terminals for the oscilloscope leads. I bought a HANTEK 2 channel PC based oscilloscope from Amazon, and connect it to my tablet that runs the Open HANTEK app. Entire rig incl the oscilloscope but minus the tablet cost a bit over $100 Cdn.
Can you please make a follow up video? I made the vibration analyzer, how to you use it to monitor the condition of motors? What parameters to look for and how to perform predictive maintenance. Thank you
What tape are you using for the position sensor? I want to use this to sync a cylinder on older diesels for example, i just diagnosed a stump grinder with a 4bt cummins that had mechanical injection using a pulse sensor on exhaust and a pressure transducer on the dipstick tube but could id the cylinder on the scope.
1. How does the accelerometer output in mV gets converted to displacement/velocity/acceleration. Does picoscope software does it automatically? 2. Can we retrieve the discreet data points in x and y coordinates from picoscope software to MS excel for any further signal processing ?
Very COoL. Really GR8T explanation(s). The ruler that was 'flipping out' is worth the price of admission right there. Thank you for the video. You RoCk p.s. My GF says she loves the Z axis. Just saying.
i wonder if you can balace a shaft / multy cilinder crank with this , if you use two of those xyz thingies and mount them on the front and rear holder (each holder made from two main bearingcaps bearings hung from treaded rod hung from a solid object (might be a concrete floor or wall ) and use each to drive an oxiloscope or switch between the sensors and have a way to overlay them , get the timing / degree mark on the crank of where the imbalance tops out , now inbalance in a shaft should cause the front and back to move independently so you need to remove weight at each end and in different spots per side to get it to run in a straight line without swinging side to side one option i was thinking on was using two disks for degreeing the camshaft and a pointer on each end (though that might mess up the balance on its own ) or a numbered piece of sticker pasted on a round part and point the strobe to that in such a way and at the right angle so that a change in direction will light up and freese the strip / disk and pointer and shows the degree number on the disk / strip think you can use differen weight neomedium magnets and move them around a little untill it balanced to find out the weight needed to be added or removed across from where the weight is along with using bobweights you should get a pretty acurate way to get the imbalance though the amount of wieght i would need to remove would take some trail and error as a last option i wonder if you in some way can mount a rotary encoder to the shaft with a flexible holder (like a piece of glasfiber fishing rod mounted to the same frame as the treaded rods of the crank holder )or hang it from the axle and put a weight on the body to alow it to spin in the air without the body moving to prevent the encoder again from messing up the measurments ((Single-phase 600 pulses /R,Two phase 4 frequency doubling to 2400 pulses per rotation ) but it handle the rpm ?) , that way you should be able to find the exact location of the inbalance back without the use of the infrared sensor and shiny surface , still need to ponder about finding out the exact weight you need to remove or add but thats probably like you said getting to know youre setup , a larger stroke crank should take less material removal than a shorter trow crank , spinning it faster or slower allso should affect vibration / amplification
That was a really interesting question! I don’t have any “bare” accelerometers at hand right now. But I tried with an app on my phone and used my loudest acoustic guitar. Unfortunately the vibrations were too weak. But the thought is very interesting and I think it would be possible to make it work. As a sound engineer I’m always looking for odd things to try, and I’ve never thought about this. So thanks for the idea! 👍
@@VarionJimmy glad too read this🙂my idea was to use it as additional controller for other devices or something like this - that is why I was finding such videos like yours. Tell me please if you do something with this idea 🙂
seem that the accelerometer and the optical position sensor get a phase shift (before or after) depending on how much I press the sensor against the bearing where I measure the vibration 🤷♂️
Great Video, I am currently working on performing the modal analysis of a wheel rim but I don't have analyzer. Can I use the same procedure to perform the modal analysis? Any suggestion will be helpful. Thanks
I have an annoying harmonic vibration buzz going on in my HotRod and I want to analyze different locations under the car when it is running down the road. I know what you put together will do the job but I was hoping you could direct me to an app I can use on my Windows tablet that I can connect your sensor package into. I'm hella rusty on my electronics prowis so if you could maybe make a schmetic on how to build the USB connection wiring to the devise to communicate w/ the Oscilloscope Software that would be a huge help..... Thanks so much.
Hello sir, If you see this you might be able to help me, on the infrared pickup, I just finished building both these items. Is there in a way to look at the rpm easily. Seems everything is hertz or cycle time, was wondering if there is a way where it figures pulses per minute so I can move the rulers in and measure the rpm in different spots of the recorded measurement. Thanks so much
would it be possible to wire all 3 axis together so as to see on 1 single channel? Ideally so you could have 4 sensors on a vehicle and watch each corner and not miss any detail. Thanks for sharing this!
That wouldn’t give you a good signal. It would likely blend them all together and that would not represent the vibration accurately. You could use four sensors and just measure one axis on each.
That's great. I will make one of these. Would I be able to balance a crankshaft with this? Would I need to do anything special to make a crankshaft balancer?
@@siuautomotive I have made a start. It's for a ducati crank so I'm using a pair of old cases to spin the crank in with a 12v motor in one if the gearbox output locations. For Bob weights I'm cutting down some old rods and adding extra mass. Might make a good educational project, nice an cheap too. I'm attaching the accelerometers and rotory encoder to an arduino, so lots to learn. Thanks again.
Adafruit makes all kinds of electronics modules etc. They are not only intended for one single purpose. You simply choose what you need for your project. Check out their website and you’ll see. 👍
Thanks for the video explanation. This is great information. I have some Piezo sensors from Spark Fun and I was wondering if I could use them in place of the accelerometer you used. I searched "vibration analysis piezo" and came across your other video about the pulse sensor but it was a little different concept I think. Do you think a Piezo could be used in place of the ADXL335 in this video?
@@eddysmit i can give half of an answer since I went ahead and gave it a try with my piezo sensors. I was able to do some balancing with the piezos. The signal from the piezos is pretty noisy. My shaft was pretty bad so the piezos still gave me a waveform I was able to use to make corrections. I made a dramatic improvement, so much that now I can't distinguish any meaningful waveform in the noisy piezo signal. But my shaft still have some vibration. I think in order to do this you ideally need the accelerometer. I have got a couple on order now.
Thanks for watching. I agree with your comments. For vibration, I think you'll need an accelerometer. Those piezos work well for noise, but not lower frequency vibrations. At least that's what I found.
@@siuautomotive while impatiently waiting for my accelerometers to arrive, I discovered an alternative. I mounted a desktop computer speaker tangentially to both end bearings of the drum/shaft I wanted to balance, and affixed a tube to each bearing, making contact with the speaker cone. A speaker is usually used to turn an electrical signal into a mechanical vibration, but it can also do the opposite. This gave me signals I was able to use with my oscilloscope to balance my drum.
i get the part where you made the sensors ... i also understood the PC is on the other hand ... what i dont understand/can't see is what is in the middle of those both. can you please explain a bit on that part
@@siuautomotive i mean is it a data logger which is getting signals from the sensors before sending it to PC?. Where the wires are going from the sensors?. I think that part is not in the video.
It does a good job with vibrations that can be felt. Probably 1hz to 300-400hz. A piezo sensor would probably work better for vibrations that can be heard.
With this tool, the output goes to a scope, not a computer. There are three analog output, x,y, and x and they vary voltage up and down from a static 2.5 volts to indicate vibration.
I'd have to say spot on accurate. The amplitude of the vibration might take a little figuring out. That's where production tools shine is they can relate the voltage amplitude to a unit of gravity (g). I do have the production units, like the pico NVH, EVA, and MTS4100. I need to do some experimenting and put all four on the same components and make a scale for my homemade device.
