cant find a 2ch 12bit ADC simple microcontroller...

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

cant find a 2ch 12bit ADC simple microcontroller...

Jump to solution
2,423 Views
Betadine
Contributor III

Hi,

This is my first post.

When looking for an 8 bits microcontroler with 2 ADC (12bit) I always arrive to the S08LC family:

 

http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=S08LC&fsrch=1&sr=1

 

The problem is that I want a simple microcontroller, not with 64 pins!

 

Please, is there any 16 pins (more or less) microcontroller with 2 ch ADC and 1 UART for transmiting RS232 to a PC?

Also want to use C programming

 

Hope to hear some solution.

Thanks!

Labels (1)
0 Kudos
1 Solution
799 Views
Obetz
Contributor III

Hello Betadine,

 

the JM family has 5V and 12 Bit ADC and is available in QFP32, still not large.

 

Oliver

View solution in original post

0 Kudos
11 Replies
800 Views
Obetz
Contributor III

Hello Betadine,

 

the JM family has 5V and 12 Bit ADC and is available in QFP32, still not large.

 

Oliver

0 Kudos
799 Views
Betadine
Contributor III

  Obetz, the JM family is a nice shot. Instead it has problems for Full speed USB applications, Ive read that for slow communications it works good.

 

I´m considering finally to buy the DEMO9S08JM16 (cause I haven't touch any microcontroller since 2 years) so I need newbie first steps to test the software)

Now comes the big question with this demo board:

Is it really possible to make a RS-232 bridge through USB with this demoboard?

 

What I need is to test a serial RS-232 COM communication through USB, I can do it with a TUSB or something that converts it independently of the micro, I know, but i would like to see if the 'marvelous'  VBA for Excel can really establish a serial RS-232 communication with a USB (RS232 protocol). If it works, the I will design something in Altium for finally release the converter.

 

So, It would be nice if the demo board brings code examples to use RS232 with and without USB.

 

Will I receive code examples for quickly test a 'hello world' or do I have to search through web?

Thanks for all for the suggestions!

 

PD: Mac I will let you know how the 12bit ADC goes.

PD: The packager is not a problem, since the creation of flux you can solder the small pins without problems. One piece of flux, solder and fsssshhhh all pins get sold without shortcircuits, (not always work)

0 Kudos
799 Views
Betadine
Contributor III

Well It´s done, I bought the demo kit, hope to achieve the rs-232 connection through USB. It seems that it brings software examples and applications for that purpose. Thank you for the orientation to all. Let´s see how the JM family works.

 

Cheers.

0 Kudos
799 Views
marbomb
Contributor I

Check out the MC9S08QE8 series.

0 Kudos
799 Views
Betadine
Contributor III

Good, MC9S08QE8 and MC9s08QB8 seems to fit the requirements, both of them has 12 bit-ADC and SCI for RS232 (PC). My main problem is that these are low power consumption (3.6V power supply).

I have two 0-5V analog signals to plug into the microcontroller. And In pages 25 and 20 respectively of:

 

QE8

http://cache.freescale.com/files/microcontrollers/doc/data_sheet/MC9S08QE8.pdf

 

QB8

http://cache.freescale.com/files/microcontrollers/doc/data_sheet/MC9S08QB8.pdf?pspll=1

 

tells that the input signal of the analog source cannot go higher than Vdd...

This 5V signals are read as XXX.X value, (max value: 199.9) -> that is 200*10 = 2000 possible values (0-5)V. The tolerance is around 0.1%. Im afraid to try the conversion 5V to 3.3V cause the increment of the tolerance... I cannot permit any tolerance even in decimal values.

 

PIC18F2423 is knocking at the door, but I once used freescale and codewarrior and thought give it a try in my next project. But it seems 12 bit ADC only applies to ultralow power consumption...

 

I would like to use freescale but I have this ultra low power consumption problem :s

¿better to use PIC for this case?

 

Cheers from Spain

0 Kudos
799 Views
Betadine
Contributor III

Conversion from 5V to 3.3V can be done with a resistive divisor of (0.1%) tolerance resistances, that is true. What I fear is the noise for 2000 values.

With 5 V  ->  1 step in the ADC in milivolts would be 5000mv / 2000values = 2.5 mv = 1 step

with 3.6V -> 1.8mv = 1 step

 

Considering that 'this' is the solution, (dont see another way to do it) I´ve seen the starter kit DEMO9S08QE8 as a quick way to see if the software works.

 

Can I safely use this solution? Just I would like to hear your opinion just in case Im missing something...

0 Kudos
799 Views
bigmac
Specialist III

Hello,

 

There are a number of design issues associated with your analog measurement requirements.

 

You will first need to determine the analog interface circuitry required, which may extend beyond a simple voltage divider.  If the analog signals originate from the "outside world" you will surely need to protect the MCU against high voltage transients.  Additionally you will need to consider the source resistance of these signals - to achieve sufficient accuracy in conjunction with12-bit resolution, the signals will each need to provide a low source resistance to each ADC input. 

 

More than likely additional voltage follower buffers would be necessary, to be interspersed between the voltage divider and the ADC input.  The voltage divider would need to be of sufficiently high resistance for negligible loading of the source, dictating the use of operational amplifiers with very low input bias current.  Without knowing the nature of the signal sources, I cannot be any more specific.

 

The initial accuracy of the voltage divider components can be compensated for if a calibration process occurs during production, so that correction factors can be determined and stored.

 

For the smaller pin count MCUs, the ADC reference will be internally tied to the Vdd and Vss rails.  Any variation of the supply voltage, with temperature or load, will affect the accuracy of the measurements.  It is possible to compensate for long term Vdd variations by also measuring a separate internal bandgap reference within the MCU.  You will need to determine whether the Vdd supply rail is sufficiently stable.

 

Since all MCUs generate internal noise, this can cause problems, especially for 12-bit conversions.  One possible approach is to put the MCU into lower power wait mode immediately after a conversion is started, and then use the ADC interrupt to wakeup and read the result.  Another method might be to take multiple readings of each signal, and use the average result.

 

If you cannot achieve the desired accuracy using the internal ADC module, you might need to use an external ADC device that communicates to the MCU, typically via SPI.  This can be used inconjunction with an accurate external voltage reference.

 

Regards,

Mac

 

0 Kudos
799 Views
Betadine
Contributor III

Well, the source is a current source  passing through resistances. So the output impedance in the 'source box' is very low ... aprox 100 ohms. thats necesary to get 0-5V from the current source.

And Yes, it can be calibrated, but the point is that the voltage is applied to an LCD for measurement... The reference of 5V comes from a regulator. (dont remember the name of the chip right now... TRACO?...yes a traco power.

 

Yes the voltage divisor needs a potentiometer ... -> high tolerance

or some trick with reference voltage and the ADC...

 

Seems like im going to spend more time in the ADC than anything...

Thank you for the answers. I will use PIC :smileytongue:

 

Hope to use freescale next time

 

All the Best

 

Cris

0 Kudos
799 Views
bigmac
Specialist III

Hello Cris,

 

Many of the problem areas that I mentioned will apply generally to 12-bit measurements, whether you are using Freescale or PIC.

 

If the input signals are derived from constant current sources, the simplest approach would be to adjust the load resistances, currently 100 ohms, to a proportionally lower value for 3.3 volt full scale.  It may still be useful to use unity gain buffers, to isolate the inputs of the MCU, and effectively provide transient voltage protection.

 

Regards,

Mac

 

0 Kudos
799 Views
Betadine
Contributor III

Mac I will study that, I was thinking about inserting an insolator, I dont realized of the transients, jojo. About the voltage divider to get the 3.3V is clear, but I will try to avoid a 3.3V for the moment unless there isn´t another way. Much appreciated.

 

MC9S08DZ16ACLC ... I´m going to put an eye in this. Tomorrow more. Thanks!

0 Kudos
799 Views
Mickey
Contributor III

You can choose MC9S08DZ16ACLC or MC9S08MP12VWL, it has a better cost performance than the PIC18F2423.

0 Kudos