PIC sinking voltages higher that Vdd???

[paul borgmeier]

But it also says that PORTA.4 has protection diodes to VDD and VSS.

Which one are we supposed to believe?

Figure 5.4 in the datasheet for the 16F628 shows no clamp diode to VDD.

[Darrel Taylor]

This figure 5-4 ?


<br>

[sougata]

Hi,

This one is from a PIC18F452 datasheet page 90. Note that I/O pin has protection diodes to VSS only.

[Ioannis]

Darrel, take a look at Mid-Range Reference Manual for 16 series. At page 145, Figure 9-3 it clearly states that clamp diode is only to Vss. It could not be otherwise for higer voltage connection. So I suppose the 5-4 you posted might be wrong.

Ioannis

EDIT: Damn, Sougata just got me for a few msec...!

[Darrel Taylor]

Hmmm, well I guess it depends on which PIC you are looking at.
manumenzella never specified. And the original post regarding RA4 as much as assumed they are ALL like that.

Paul was referencing the 16F628, which according to the most recent datasheet, Has the diode.

If you look at a 16F628A, then it doesn't.

If you look at the 18F452 datasheet from a couple years ago, it says it does, but the most recent one doesn't.
Was it a change in the datasheet, or the silicon?

Like my original comment,
"Which one are we supposed to believe?"

Personally, I'd go with the transistor.
<br>

[Acetronics2]

Hi, Picmen

Pic Inputs clamp Diodes are rated @ 500µA ...

So, applications that limit the input current @, say 50µA ( good safety margin !!! ) can be done.

That includes sensing Mains through a 4.7 MOhm 1/2 W or 2 x 2.2 MOhms 1/4W in series ( care to the max voltage across the 1/4W resistors : 250 v PEAK )

I have built lots of 8 Pïns Pics working on mains as zero voltage switches or dimmers ...

BUT Remember TOCKI or MCLR pins DO NOT HAVE the VCC Clamp diode : an external diode has to be added then ...
and it's no pain if a clamp diode is already built in the package !!!

Alain

[Ioannis]

Personally, I'd go with the transistor.
<br>

I double that Darrel. That's what I always do and keep my headaches far away...!

As for the datasheets, they may have errors. Some are found, some not. What about the errata? I have not time to look right now.

Ioannis

[sougata]

Hi,

Spending on a single NPN transistor and a couple of resistor do give you the flexibility to use any pin without the need for painkillers.

[paul borgmeier]

From an earlier post the OP had

>>Hi. My name is Manuel, and I am building a project around the PIC16F628A and a piezo

http://www.picbasic.co.uk/forum/showthread.php?t=5773

I, in a sloppy manner, left the A off of my reply about the 16F628 datasheet. I had checked the 16F628A sheet and saw no clamp and was just reporting that observation.

However with an actual look at the 16F628 datasheet and the others noted, I see why Darrel asked

>> Which one are we supposed to believe

Great question by Darrel ... here is my version for an answer - if it were a one-off by a school boy intending to rudely disrupt the class room I would go with the 9V on RA4. If it were for a medical device or other lawyer-attacting application, I would go the NPN route and would want a written response from Microchip as an effort to help keep by butt out of court should there ever be a problem before going the non NPN route.

[keithdoxey]

Hmmm...interesting this one.

Just looked at the datasheets on my PC for 16F84, 16F628 and 16F628A

The 84 doesnt show diodes but mentions them and specifically says that RA4 is open drain and only has a diode to VSS. It also says that the maximum voltage on any pin is VDD +0.6v except MCLR wich is 14V

The datasheets I have for 628 and 628A both SHOW a diode on RA4 to VDD but also show it as an open drain pin. The spec says max voltage on any other pin is VDD + 0.3V EXCEPT MCLR AND RA4 which are 14V

Obviously something is not right in the datasheet and I would suspect that the diod ISNT present. My understanding of having an open drain pin is specifically to allow it to sink a load driven from a higher voltage hence the electrical spec stating 14V maximum. If the diode IS present then you wont be able to do that.

[Ioannis]

So, if one is in doubt, take the multimeter and check for a diode on pin RA4 to Vdd!

Ioannis

[keithdoxey]

So, if one is in doubt, take the multimeter and check for a diode on pin RA4 to Vdd!

Ioannis

Or put a 10k resistor between RA4 and +12v and measure the voltage.

If its 12v then no diode, if its 5.3-5.6 then there is a diode.

[Acetronics2]

So, if one is in doubt, take the multimeter and check for a diode on pin RA4 to Vdd!

Ioannis

How much current does your multimeter use for the test ??? more or less than 500µA ???

Good question ...

Alain