Please comment on schematic

[jbmillard]

This is my first PIC project. I have it bread boarded and everything appears to be working, but I was hoping someone wouldn't mind taking a look at the schematic and making sure I haven't done something stupid.

Since the schematic is pretty big, you can find it here:

**broken link removed**

I'll try and answer a few things up front so people don't have to guess:

1) N1 is a 3.3 V regulator with shutdown capability. B8 turns it on and after the PIC starts, it is held on by the output of RA7. Pin NBP is "noise bypass" -- data sheet just says to put a 0.1 uF cap to ground -- so I did.
2) T1, T2 and voltage divider are for measuring battery voltage. Both T1 and T2 are logic level MOSFETs.
3) U3 is a pressure/temperature sensor
4) U4 is an EEPROM
5) U1 is a 4x20 LCD display from New Haven.
6) I'm not using pull-up resistors on buttons B1-B6 because I have the internal pull-up resistors on CN11 through CN16 on.
7) Capacitor values are uF.

Other than that, I hope it is pretty self explanatory. The PIC reads the pressure and temperature, does some calculations and writes data to the display.

Thanks!

Brad

 

[Noggin]

I only spent a minute or so looking at it so my review is by no means complete.

Your voltage regulator's ground either isn't connected to ground or you have it connected to ground by some attribute in the schematic symbol. If it is connected to ground by an attribute, I think your schematic is misleading and confusing.

C10 appears to be backwards, though it is likely a ceramic cap and it doesn't matter.

 

[jbmillard]

You're right...

...I missed the ground.

The caps are non-polarized.

Thanks for taking the time to look.

Brad

 

[Noggin]

You need a pull down on gate of T1

Your analog reference is 3.3v. I'd use a larger resistor for R7. With a 10k and 1k, the voltage on R7 with a 9v battery is going to be 0.81v. A 9v battery is dead at what? About 6 volts? That is .54v. 0.27v difference is easy enough for the ADC to distinguish, but a higher resistor size would increase the difference and give you finer resolution. I'd use a divider more like 100k and 47.5k. Using these values would let you read the ADC and divide by 100. (Cool trick Nigel taught me )

Example:
At 9v input, the voltage across the bottom resistor is 2.9v. The ADC conversion of 2.9v with a 3.3v vref is 900.
At 7.5v input, the voltage across the bottom resistor is 2.42v. The ADC conversion of 2.42v with a 3.3v vref is 750

 

[jbmillard]

You need a pull down on gate of T1

Good point. I'll do that.

Your analog reference is 3.3v. I'd use a larger resistor for R7. With a 10k and 1k, the voltage on R7 with a 9v battery is going to be 0.81v. A 9v battery is dead at what? About 6 volts? That is .54v. 0.27v difference is easy enough for the ADC to distinguish, but a higher resistor size would increase the difference and give you finer resolution. I'd use a divider more like 100k and 47.5k. Using these values would let you read the ADC and divide by 100. (Cool trick Nigel taught me )

Example:
At 9v input, the voltage across the bottom resistor is 2.9v. The ADC conversion of 2.9v with a 3.3v vref is 900.
At 7.5v input, the voltage across the bottom resistor is 2.42v. The ADC conversion of 2.42v with a 3.3v vref is 750

I ran into problems with a higher impedance bridge. Conversion time was much longer than I wanted, so I went to smaller resistors. Resolution isn't much of a worry since it's just a battery indicator (nearest 5%).

Thanks!

Brad

 

[Oznog]

Did you check the dissipation on that package? 24F can draw a lot of current when the MIPs are turned up, and that's a 9v batt so it's dropping 5.7v there. Max pkg dissipation for the NJM2870 in the spec sheet I googled \was an MTP5) is 200mW at 25C and that's bringing the junction to Absolute Max temp, so you don't wanna be anywhere near there.

Two AA or AAA is enough to run it a good long time. The voltage isn't 3.3v... it may be down to 2v by the time the batts are really dropping off- so you should take that into account.

 

[jbmillard]

Did you check the dissipation on that package? 24F can draw a lot of current when the MIPs are turned up, and that's a 9v batt so it's dropping 5.7v there. Max pkg dissipation for the NJM2870 in the spec sheet I googled \was an MTP5) is 200mW at 25C and that's bringing the junction to Absolute Max temp, so you don't wanna be anywhere near there.

Two AA or AAA is enough to run it a good long time. The voltage isn't 3.3v... it may be down to 2v by the time the batts are really dropping off- so you should take that into account.

Dissipation is 350 mW when on a glass epoxy board, 200 mW when not, according to the data sheet. The circuit only draws 15 mA when powered up. When I need to do calculations I clock switch to 80 MHz/40 MIPS and then switch back to the 7.87 MHz internal clock when done -- this takes less than 1/10 of a second.

By my calculation, it's currently dissipating 86 mW (0.015*5.7) on the breadboard. Is this enough of a design factor?

I'll have to try a couple of AA batteries for the fun of it. I'm not sure how I'd turn it off though.

Thanks!

Brad

 

[Oznog]

Yup. Your safety factor sounds dandy. It'll be physically warm, but you've got plenty of safety factor.