Find PWM for DE-ACCM3D2 LCD Display

[noagname]

I am having a hard time finding a proper pwm to go with DE-ACCM3D2. Then outputting to a LCD display.

 

[noagname]

Can anyone give me any kind of advice for me to find a PWM that goes with the accelerometer?

 

[MATLABdude]

I just Googled for the chip and glanced over the datasheet:
http://www.dimensionengineering.com/datasheets/DE-ACCM3D2.pdf

Are you aware that there is no PWM whatsoever involved in the operation of this chip? Even if there were (for instance, in the Analog Devices ADXL-202) the PWM would be the output of the chip (duty cycle corresponding to acceleration, frequency is designer-adjustable with a resistor and capacitor or two).

Take a look in the referenced datasheet; they give you examples based off of analog-to-digital conversion (which microcontroller are you using for this project?)

 

[noagname]

I am not sure which microcontroller I am using and thanks for fixing my confusion over pwm. The moment I find a microcontroller I should be able to do this

 

[Bob S]

There are two ways to interface the accelerometers; pwm or voltage to frequency converter. Both provide a noise immune asynchronous signal suitable for distance communication. Do you intend to process these signals at the receiving end, or put them into a LED display?

 

[noagname]

To put them to a display, in the end I want it so I can place this device on the dashboard of my plane, car and other vehicles to see my xyz axis g's. And to also have a button that zero's(puts xyz to 1g) so when its at a different angle x doesn't start out at 1.5g's or something like that.

 

[Bob S]

I think your 3-axis accelerometer uses the STMicroelectronis chip LIS344ALH described at
http://www.stmicro.fr/stonline/products/literature/ds/14337/lis344alh.pdf .
I believe it is dc coupled so does not need resetting. I suggest you consider using straight analog coupling to your display. You might consider using 3 off-board buffer amplifiers that provide a current signal (e.g., 0 to 10 mA). You could wire it directly to your display meters.
I have used one of these meters (see below) that I have used for more than a year: (eBay product)
http://cgi.ebay.com/3-1-2-Red-LED-Digital-Volt-Panel-Meter-AC-200mV_W0QQitemZ370208376639QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item563223873f&_trksid=p4634.c0.m14.l1262&_trkparms=%7C301%3A1%7C293%3A8%7C294%3A30
You should decide on what display you want: LED or LCD, (red,green, blue) and panel height, etc. This same company has a frequency meter that would work nicely with a voltage to frequency converter.

 

[noagname]

I don't really want a display meter, I just want a chip that will take an analogue input then convert that based on my programing to output on a 3x16 lcd display. Using those display meters will give me the voltage or hz(using a v to hz converter) and I have to using a math problem get a proper g force reading.

Edit: I think a serial LCD display will be the easiest

 

[noagname]

Unless anyone can give me any advice I think I will be going towards the pic 18f452.

 

[MATLABdude]

Unless anyone can give me any advice I think I will be going towards the pic 18f452.

Do you have a programmer for that?

 

[vk6kro]

It seems like you would need some way of displaying the peak value of your accelleration.

Either when just straight line accellerating (like from the traffic lights) or when you are turning a corner, you would usually be too busy to watch the display, but you might like to see the data later.

So, you could try for a display that holds the highest value reached.
Or you could take readings every 100 mSec, maybe, and store them for later transfer to a computer or stepping through with a push button.
Then you could see how the accelleration changed as the gears were changed.

Connecting it to a display with road speed and engine RPM as well would be pretty useful.
Race cars have this sort of technology.

 

[noagname]

I don't have a programmer yet and connecting it to my computer so I can see everything is a cool idea but I can't even get past the ADC.

 

[MATLABdude]

I don't have a programmer yet and connecting it to my computer so I can see everything is a cool idea but I can't even get past the ADC.

It would be cool, but you need to get a programmer before you can pass go and collect your $200. Or even get the ADC functional to the point where it can start to give you headaches (figuratively).

I was trying to steer the conversation to the point that, if you're starting out in the field of microcontrollers, you need to also find a programmer suitable for programming said microcontroller, which is not always a straightforward task (especially not for Microchip ones):
http://en.wikipedia.org/wiki/PIC_microcontroller#Device_Programmers

Microchip Programmers generally cost a fair bit (couple hundred bucks new), but you can find homebrew ones (that don't cost as much, but also aren't always going to work). There is a homebrew ICD2 (Microchip programmer + debugger) that I heard about, but I don't know how well it works:
http://www.mp3car.com/vbulletin/hardware-development/43981-homebrew-microchip-icd2-clone.html

The ATMEGA line by Atmel has a very cheap ($30-$50) in-circuit programmer, the AVRisp mkII--works with all their chips, even if it lacks the more advanced debug features (and significantly higher price tag) of its debugger brothers. It's also got a (free) C compiler based on GCC, if that factors into your decision at all (Microchip products are designed to be programmed in assembly, though not-quite-standard C compilers are available, for a price).

 

[vk6kro]

I don't have a programmer yet and connecting it to my computer so I can see everything is a cool idea but I can't even get past the ADC.

If the input is a problem, I really suggest you have a look at the Picaxe chips.

I did a test on a Picaxe 14, which is a $6 chip.
It has 48 bytes of internal memory where you can store data quickly. Not much, but maybe enough for what you want. There are other chips in the same range with more memory.

I could read the ADC port 48 times and fill this memory with the data in about 0.2 seconds.
So, in the real world, if you wanted to, you could take a reading every 0.5 second and continue for 24 seconds. Quite enough for the average burn-rubber takeoff.
The data would then be there for you to examine on a LCD display or feed to a computer for graphing.

This assumes you can get the accellerometer output into a 0 to 5 volt range and that 8 bit bytes are adequate. The chip can do 10 bits but they take more room to store, so you get fewer readings. 10 bits gives you 1024 steps of 5 milivolts per step.

This chip is programmed using free software and a cable you can make from an old serial mouse cable.