Atmel AVR + Motorola MPX Sensors approach - Notes and Discussions to the project

[MrHijet]

Today I completed my order at Digikey.com:


FreeScale MPX5010DP ( diff pressure sensors with aprox. 0-40 inch water rise)
MSP 1210A-002D-3L ( diff pressure sensor with aprox 0-60 inch water rise)
(There is another MSP 1210A-001D-3L, which is 0-25 inch water rise, but its out of stock at the moment)
FreeScale MPX5050GP ( Absolute sensor to play around for the development kit and my mity vac).
Analog Devices AD7705 ( 2 channel 16 bit A/D converter with serial (SPI) interface)

The AD converter is a little bit overkill, but the very promissing MSP 1210 series is very precise (but costs factor 2,5 more than the MPX from Motorola/Freescale). So I can start learning to connect it via a nice A/D converter.

My Idea is to make a very cheap and basic set: Atmel AVR + cheap LCD Display + a very precise, but cheap A/D converter.
Then build around the periphicals like the MPX 5010 or Honeywell 26PCB/PBA or MSP 1210 series, so that every can choose between several sensors.

Lets see when the stuff arrives.

Cheers,

Daniel

PS: The specs for the MSP 1210 series:

[Tony]

This sounds like an excellent project Daniel.

I doubt if you are going to get 16 bits of useful resolution because of electrical and particularly acoustic noise. All these pressure transducers respond like microphones that extend well above and well below the range of human hearing for frequency response.

Air movement in your bench will cause air turbulence, rumble, motor noise, and motor vibrations, that will be directly picked up by the pressure transducer which is really a giant ear. Every reading by the a/d will be a snapshot of instantaneous pressure plus instantaneous noise. I would expect many of the least significant bits out of sixteen to end up being totally random garbage.

My own past experience with this type of measurement is that a very great number of readings will need to be averaged in software to get a steady pressure figure of sufficient resolution.

[MrHijet]

Hi Tony,

I thought about the same. Ofcourse I will never get 16 bit "clean" signals, however the AD7705 ( or the pin compatible maxim MX7705) do also include some filter mechanism. I like the idea of utilising the low cost atmel mcu with a low cost 16 bit AD-converter as a basement for all following periphicals. The processor plus ad convertier is cost below 10US$ (together!), so it fits into the cheap display idea.

With an SPI connected AD converter, I have more IO-ports left open at the atmel mcu for additional toys, displays or connections and we still have easy to solder DIP chips.

Cheers,

Daniel

[Tony]

I agree this is a really great project, I like the way it is heading.

In the past I have had some interesting problems with data logging pressure transducers, that turned out to be acoustic noise in the air lines. Both acoustic muffling, and analog low pass filtering help, but there is a definite limit to the effective noise reduction that can be achieved that way.

I have had much more success with statistical averaging in software, and it is really easy to do.

[Moriniman]

[MrHijet]

Hi Paul,

Thanks for the hint. Where have you seen the MAX1416 available to buy ? I had no chance to get one.

The AD7705/MX7705 cost around 5 euros and the oscillator + 2 condenser below 1 euro. However the MAX1416 sounds great.

Cheers,

Daniel

[Moriniman]

I obtained samples, but you can buy direct from Maxim; cost in singles US$9.70.

[MrHijet]

There was a little forum board issue with my account, so my apologies for the posts which are now "Guest" instead of my username.

Some news:

I received my sensors yesterday. So Time to get my LCD and LED output running to step on to the sensor readout work.

Cheers,

Daniel

[MrHijet]

Today was the big parcel day: I received 6 parcels including some with electronics inside.

As you can see I received some Sensors:

There are two MPX5010DP and one MPX5050GP. I don't really need the MPX5050GP for the bench, but that one is easier for the first development (measuring from the AVR).
Then there are the 16Bit A/D Converter and one MSP sensor with the metal tubes with 2 psi measuring range. This MSP sensor can be very interesting because its very precise and is very sturdy again to high pressures. There is a 1 psi version available in general, however that one was out of stock when I ordered everything.

Then I had a bit luck on ebay and catched some extra sensors to play around with them.

I never know if I really need them, but better have them at home already to play with them if there is a need.

The same seller had some LCD Displays. I catched them too to make the shipping more profitable.


Now I have some parts to get in touch with and can try out which display suites best of this job. Additionally I will start with the sensor read out (but I still wait for some ceramic caps).

Cheers,

Daniel

[Tony]

Wow Daniel,
Plenty of new toys to play with. It can be pretty exciting sometimes when the mail arrives.

[MrHijet]

Sorry for the long delay, but I'm a bit ill, so the project had to wait.

In the meantime, I received a second batch of sensors from the same ebay seller. Lucky me



Hope I will be a bit fitter again after next week.

Daniel

[MrHijet]

Due to my problems with my back, not so much updates, but at least this photo:



Thats my development board with the 16 bit analog/digital converter, an Motorola/Freescale 5050GP and a little filter
setup. A 2x20 character LCD interface is connected and working too.

For the first tests I play around with a Non-Differential sensor, because thats a bit easier when trying out in theory at home. Additional I have connected a little test circuit to simulate voltages between 0-5 volts, but thats just for learning purposes.

Cheers,

Daniel

[bruce]

It's good to see these projects coming together!

Opens up a whole new aspect for the DIY'er not being stuck with buying a commercial product!

[Tony]

Great work Daniel.
Getting started off initially with something like this, is always the most difficult part. What you have looks like a really great development setup.

What some here may not realize, is that the big development circuit board is only required to develop, test, and debug the original software. It will not be required as part of the final system.

Look after yourself Daniel, back problems can be tricky.

[larrycavan]

You guys have completely lost it!........Looks like you're well on the way to a flow computer....Nice Job!