I have used Proteus design suite for simulating sensors (such as temperature sensor) interfaced to LPC2138. The sensors which were not available in the software, for them I have modelled them as voltage divider circuit with potentiometer. If I were to show mapping/compare the voltage divider results in software simulation and in hardware design, will the results have any dependency on the type of controller as all I am doing is simulating a controlled environment (by changing the R values)? For example if I use Arduino for hardware testing and LPC2138 in Proteus. Would this be correct? Please confirm.
Add a comment
|
1 Answer
\$\begingroup\$
\$\endgroup\$
6
Proteus is highly unreliable in circuit simulation. Sure, you can run your code to check if its working or not but that is the extent of it. Never depend on any simulation results from Proteus. You said that you have modeled sensors as Voltage dividers which is exactly what I am talking about: You can check if changing the voltage has any effect or not and that maybe fine but do not make the mistake of comparing it to your hardware results. There will always be a big error in Proteus results. Trust me, I learned it the hard way.
-
\$\begingroup\$ Ok thanks. One more question, I have modelled a 2k and 1k (pot) as voltage divide circuit for LPC2138 which works on 3.3 V. For the voltage divider I have kept Vin as 5V. Since you said results can vary a lot, now if I use Arduino I should try to bring the output voltage to vary from 0 to 5V right? Can you suggest the R value and Pot value ? \$\endgroup\$– MeghCommented May 13, 2019 at 8:44
-
\$\begingroup\$ Or in this case I don't require a divider circuit, I should simply use potentiometer, read analog values, convert them to digital, software calibrate (by using some multiplicative or divisive factor)them to simply show them in some sort of format example percentage or degree Celsius? \$\endgroup\$– MeghCommented May 13, 2019 at 9:01
-
\$\begingroup\$ Simulators are merely mathematical models, the biggest errors you'll come across are from inherent RLC values in materials where their significance depends on the application. A simple DC circuit should work just fine, RF applications and RLC oscillators not so much. \$\endgroup\$– lucasgcbCommented May 13, 2019 at 9:08
-
\$\begingroup\$ Yes, the span should be 0-5V. You can use 1k Res and 10K Potentiometer. This will give you a swing of ~0V to 4.56V while measured across Potentiometer. And the change will not be smooth i.e. you will get 0V-2.5V from 0R-1K value of Pot and 2.5V-4.56V from 1k-10K value of Pot. Of course, you can design a better divider if you really want stable sensitivity throughout. \$\endgroup\$– AhmedCommented May 13, 2019 at 9:22
-
\$\begingroup\$ If you want to simulate voltage change, you will need a divider, read its output as analog value, convert to digital and do whatever you want. \$\endgroup\$– AhmedCommented May 13, 2019 at 9:24