Moving from mikroBasic to OshonSoft Basic

[bcz]

A/D coversion and the entire code simulation is ok, except when values for AN0 are set small.
Then the LCD does not show "Input is too low".
P.S.: The logarythmic signal detector (ahaed of the MCU circuit) output is ca. 0.45 VDC (broadband noise signal) without any input signal applied. In this case I want to display the above message.

again:
Lcdcmdout LcdClear
ADC_Read 0, ADCRead 'Result of A/D conversion from AN0
ADCResult = ADCRead*5/1024
Lcd_Out(1,1, "ADC OUT = " + #ADCResult +" V")
If ADCResult<0.5 Then
Lcd_Out(2,1, "Input is too low")
Else
dBmResult = (ADCResult-ZEROXING)*SLOPE
Lcd_Out(2,1, "PWR = " + #dBmResult + " dBm")
Endif
WaitMs 500
Goto again 'Make a new A/D conversion

Goto main
End

 

[DogFlu66]

Well, I have tested the program in more detail and it seems that in these new versions of the compiler ADC_Read configures the pin as analog and as input. I thought that the problem was that the pin was not output as analog in the simulator. Anyway, if you wanted to do this, it does work.

I didn't see this configure the AN0 pin: Sub GlobInit.

 

[bcz]

Gentlemen,
My project works (simulation and hardware), mission accomplished.
Thank you all who contributed with advice, and help.
I have learnt a lot and will continue exploring OshonBasic.
Best regards,
Robert

 

[bcz]

I want to eliminate the 0.01 dBm variation on the LCD display that happens for a constant input RF signal level to the ADC of the circuit. The ADC conversion slope is 19.61 dBm/Volt.
Therefore 0.1 dBm level change is caused by 0.1 dBm/(19.61 dBm/V) = 5.1 mV change in the analog input voltage.
My goal is to keep the display value unchaged for 2*5.1 mV = 10.2 mV, i.e. <= 2 bit change in ADC output.
Please find attached the source code. Would the error message mean that the syntax "abs" is not known in OshonBasic? Any cure for that?

 

[Ian Rogers]

The only way to "eleminate" this tiny variation is to over sample.

If you are running at 5V a bit change is 4.88Mv. Then you need to decide how to get the best information to screen.. You set your decimal places to two.

I do not use floating point numbers on small devices as I know how they work and its not how you think.

lets take your calculation 2.5V should be 512 bits.. 512 * 5 = 2560 / 1024 = 2.5v BUT!!!!! ADCRead is an interger only the result is floating so it may need casting.. Vlad does allow casting

So
adcread = adcin(0) * 5 'integer multiply
adcresult = adcread ' basic cast to float
adcresult = adcresult / 1024 ' now it works..

But!! you want MORE precision

do this
for idx = 0 to 10
adcresult = adcresult + adcin(0)
' you can place a tiny delay here to slug readings
next idx
adcresult * 5 / 10240 ' gives 2.5V

You can play around ie read adc 5 times then times by 100 then / 10240

The point is the output will be more stable.

 

[bcz]

Sorry, I meant to say "I want to eliminate the 0.1 dBm variation" in my previous message.