Using Vddcore as Vref

[Diver300]

I'm using a the A/D converter on a pic18F25J10. The supply voltage is not well regulated so I need a separate Vref. I want to know if there is any downside in using Vddcore for Vref.

For those of you who don't know, the pic18F25J10 runs on 2.7 to 3.6 V, but the core runs on 2.5V, provided by an internal regulator. There has to be a capacitor on the Vddcore/cap point to stabilise the regulator.

The core takes between 4 and 12 mA typical, depending on frequency. I will be running at 10 MHz so the current will be nearer to 4 mA.

Vref takes a maximum of 150 :mu:A. Is it OK to connect Vddcore to Vref and use the regulated 2.5 V as the A/D reference?

 

[Nigel Goodwin]

Never looked at one, but I would imagine the datasheet would mention doing so if it was a good idea?, doesn't it say anything about it? - like you, I think it sounds an obvious thing to do, and I would expect the datasheet to be telling you how clever it is.

 

[eng1]

I'm using a the A/D converter on a pic18F25J10. The supply voltage is not well regulated so I need a separate Vref. I want to know if there is any downside in using Vddcore for Vref.

I've noticed that Vddcore is not software selectable as the reference of the A/D converter... I have a PIC18F45J10 but I've used a voltage derived from an external reference on RA3... what about opening a ticket and get an answer from Microchip? I would be interested in knowing the answer!

 

[Diver300]

The data sheet doesn't mention where to find Vref from. The A/D converter is specified with Vref > 3V so I suppose it won't be quite as accurate if run off 2.5 V.

However, Vref is specified as > 1.8 V if the supply is less than 3V, so it can't be that critical.

I suspect that Microchip don't want to encourage the regulator to be used for other things, but 150 :mu:A peak from a regulator that has to be rated to 20 mA doesn't seem to bad to me.