## Re: [softrock40] 12F683 Question

Expand Messages
• It appears from that table (which I had missed before) that if Vdd is 3.3 volts, then the maximum for a logic low is .15 * 3.3v, and .7 volts will be too high.
Message 1 of 13 , Jul 1, 2008
• 0 Attachment
It appears from that table (which I had missed before) that if Vdd is 3.3
volts, then the maximum for a logic low is .15 * 3.3v, and .7 volts will be
too high. Am I reading the table correctly?

Would it be feasible to use germanium steering diodes instead, with a
forward voltage drop of ~.3 volts, or should I just go to plan B and use a
rotary binary encoder? I'd prefer not to, because mMy objective in this is
to select the 6 contest bands using a single rotary switch (2P6T) to
control the oscillator and bpf selection.

73, Pete

At 07:43 AM 7/1/2008, Jack Smith wrote:
>Table 15-5 in the 12F683 data sheet.
>
>Jack
>
>
>Pete Smith wrote:
> >
> > Does anyone know if a 12F683 running on 3.3 volts in the SoftRock 8.3
> > circuit will operate reliably if, instead of using a DIP switch to ground
> > pins on the chip, a rotary switch through a single silicon diode is
> > used. I think that means that the pin will not be at 0, but rather at .7
> > volts above ground, and I can't find any indication in the data sheet of
> > what the chip considers to be the maximum logic low voltage.
> >
• Yes, that is how the data sheet is to be read. As far as using diode steering or diode logic, a couple points. First is that 0.7 volt drop is the common
Message 2 of 13 , Jul 1, 2008
• 0 Attachment
Yes, that is how the data sheet is to be read.

As far as using diode steering or diode logic, a couple points. First is
that 0.7 volt drop is the common assumption for a diode drop. It's far
less than this for low current. Check the typical 1N4148 data sheet, or,
better yet, set up a quick test with a diode, a 10K resistor, a variable
DC power supply, a pair of digital meters and graph Vf vs If.

For a lower drop, a Schottky diode such as a 1N5711 is preferable over a
Germanium. However, a similar Vf versus If plot will be informative.

I wrote about diode Vf versus If at
http://www.cliftonlaboratories.com/diode_vf_vs_if.htm but it will stick
with you more if you also make the Vf versus If measurements yourself.

The 0.15 Vdd threshold is for TTL-compatible inputs. I have not gone
through the 12F683 data sheet, but some (many) PICs have some ports with
Schmitt type inputs which have a different threshold. Table 15-5 implies
that some 12F683 inputs have Schmitt construction with Vl = 0.2 Vdd.

It's also likely that there is some margin in Table 15-5's values. For a
one-off project, some people will accept operation outside the data
sheet values, if the configuration works. You might try it and see what
happens.

Jack

Pete Smith wrote:
>
> It appears from that table (which I had missed before) that if Vdd is 3.3
> volts, then the maximum for a logic low is .15 * 3.3v, and .7 volts
> will be
> too high. Am I reading the table correctly?
>
> Would it be feasible to use germanium steering diodes instead, with a
> forward voltage drop of ~.3 volts, or should I just go to plan B and
> use a
> rotary binary encoder? I'd prefer not to, because mMy objective in
> this is
> to select the 6 contest bands using a single rotary switch (2P6T) to
> control the oscillator and bpf selection.
>
> 73, Pete
>
> At 07:43 AM 7/1/2008, Jack Smith wrote:
> >Table 15-5 in the 12F683 data sheet.
> >
> >Jack
> >
> >
> >Pete Smith wrote:
> > >
> > > Does anyone know if a 12F683 running on 3.3 volts in the SoftRock 8.3
> > > circuit will operate reliably if, instead of using a DIP switch to
> ground
> > > pins on the chip, a rotary switch through a single silicon diode is
> > > used. I think that means that the pin will not be at 0, but rather
> at .7
> > > volts above ground, and I can't find any indication in the data
> sheet of
> > > what the chip considers to be the maximum logic low voltage.
> > >
>
>
• I should also add that any inexpensive NPN transistor can be used as a switch with a Vsat around a couple tenths of a volt. Hence it is possible to use the
Message 3 of 13 , Jul 1, 2008
• 0 Attachment
I should also add that any inexpensive NPN transistor can be used as a
switch with a Vsat around a couple tenths of a volt. Hence it is
possible to use the switch, a diode and a base current limiting resistor
to turn a 2N2222 or 2N4401 on/off with the PIC's input connected to the
NPN's collector. (Needs a pull-up resistor, of course.) Or, a 2N7000
MOSFET can be used if you don't like bipolars. As long as the pull up
resistor isn't too large, the 2N7000 will also give you a low voltage of
a couple tenths.

Jack

Jack Smith wrote:
>
> Yes, that is how the data sheet is to be read.
>
> As far as using diode steering or diode logic, a couple points. First is
> that 0.7 volt drop is the common assumption for a diode drop. It's far
> less than this for low current. Check the typical 1N4148 data sheet, or,
> better yet, set up a quick test with a diode, a 10K resistor, a variable
> DC power supply, a pair of digital meters and graph Vf vs If.
>
> For a lower drop, a Schottky diode such as a 1N5711 is preferable over a
> Germanium. However, a similar Vf versus If plot will be informative.
>
> I wrote about diode Vf versus If at
> http://www.cliftonlaboratories.com/diode_vf_vs_if.htm
> <http://www.cliftonlaboratories.com/diode_vf_vs_if.htm> but it will stick
> with you more if you also make the Vf versus If measurements yourself.
>
> The 0.15 Vdd threshold is for TTL-compatible inputs. I have not gone
> through the 12F683 data sheet, but some (many) PICs have some ports with
> Schmitt type inputs which have a different threshold. Table 15-5 implies
> that some 12F683 inputs have Schmitt construction with Vl = 0.2 Vdd.
>
> It's also likely that there is some margin in Table 15-5's values. For a
> one-off project, some people will accept operation outside the data
> sheet values, if the configuration works. You might try it and see what
> happens.
>
> Jack
>
> Pete Smith wrote:
> >
> > It appears from that table (which I had missed before) that if Vdd
> is 3.3
> > volts, then the maximum for a logic low is .15 * 3.3v, and .7 volts
> > will be
> > too high. Am I reading the table correctly?
> >
> > Would it be feasible to use germanium steering diodes instead, with a
> > forward voltage drop of ~.3 volts, or should I just go to plan B and
> > use a
> > rotary binary encoder? I'd prefer not to, because mMy objective in
> > this is
> > to select the 6 contest bands using a single rotary switch (2P6T) to
> > control the oscillator and bpf selection.
> >
> > 73, Pete
> >
> > At 07:43 AM 7/1/2008, Jack Smith wrote:
> > >Table 15-5 in the 12F683 data sheet.
> > >
> > >Jack
> > >
> > >
> > >Pete Smith wrote:
> > > >
> > > > Does anyone know if a 12F683 running on 3.3 volts in the
> SoftRock 8.3
> > > > circuit will operate reliably if, instead of using a DIP switch to
> > ground
> > > > pins on the chip, a rotary switch through a single silicon diode is
> > > > used. I think that means that the pin will not be at 0, but rather
> > at .7
> > > > volts above ground, and I can't find any indication in the data
> > sheet of
> > > > what the chip considers to be the maximum logic low voltage.
> > > >
> >
> >
>
>
• Great, Jack, as always. Thanks! 73, Pete
Message 4 of 13 , Jul 1, 2008
• 0 Attachment
Great, Jack, as always. Thanks!

73, Pete

At 08:38 AM 7/1/2008, you wrote:
>I should also add that any inexpensive NPN transistor can be used as a
>switch with a Vsat around a couple tenths of a volt. Hence it is
>possible to use the switch, a diode and a base current limiting resistor
>to turn a 2N2222 or 2N4401 on/off with the PIC's input connected to the
>NPN's collector. (Needs a pull-up resistor, of course.) Or, a 2N7000
>MOSFET can be used if you don't like bipolars. As long as the pull up
>resistor isn't too large, the 2N7000 will also give you a low voltage of
>a couple tenths.
>
>Jack
>
>
>Jack Smith wrote:
> >
> > Yes, that is how the data sheet is to be read.
> >
> > As far as using diode steering or diode logic, a couple points. First is
> > that 0.7 volt drop is the common assumption for a diode drop. It's far
> > less than this for low current. Check the typical 1N4148 data sheet, or,
> > better yet, set up a quick test with a diode, a 10K resistor, a variable
> > DC power supply, a pair of digital meters and graph Vf vs If.
> >
> > For a lower drop, a Schottky diode such as a 1N5711 is preferable over a
> > Germanium. However, a similar Vf versus If plot will be informative.
> >
> > I wrote about diode Vf versus If at
> > http://www.cliftonlaboratories.com/diode_vf_vs_if.htm
> > <http://www.cliftonlaboratories.com/diode_vf_vs_if.htm> but it will stick
> > with you more if you also make the Vf versus If measurements yourself.
> >
> > The 0.15 Vdd threshold is for TTL-compatible inputs. I have not gone
> > through the 12F683 data sheet, but some (many) PICs have some ports with
> > Schmitt type inputs which have a different threshold. Table 15-5 implies
> > that some 12F683 inputs have Schmitt construction with Vl = 0.2 Vdd.
> >
> > It's also likely that there is some margin in Table 15-5's values. For a
> > one-off project, some people will accept operation outside the data
> > sheet values, if the configuration works. You might try it and see what
> > happens.
> >
> > Jack
> >
> > Pete Smith wrote:
> > >
> > > It appears from that table (which I had missed before) that if Vdd
> > is 3.3
> > > volts, then the maximum for a logic low is .15 * 3.3v, and .7 volts
> > > will be
> > > too high. Am I reading the table correctly?
> > >
> > > Would it be feasible to use germanium steering diodes instead, with a
> > > forward voltage drop of ~.3 volts, or should I just go to plan B and
> > > use a
> > > rotary binary encoder? I'd prefer not to, because mMy objective in
> > > this is
> > > to select the 6 contest bands using a single rotary switch (2P6T) to
> > > control the oscillator and bpf selection.
> > >
> > > 73, Pete
> > >
> > > At 07:43 AM 7/1/2008, Jack Smith wrote:
> > > >Table 15-5 in the 12F683 data sheet.
> > > >
> > > >Jack
> > > >
> > > >
> > > >Pete Smith wrote:
> > > > >
> > > > > Does anyone know if a 12F683 running on 3.3 volts in the
> > SoftRock 8.3
> > > > > circuit will operate reliably if, instead of using a DIP switch to
> > > ground
> > > > > pins on the chip, a rotary switch through a single silicon diode is
> > > > > used. I think that means that the pin will not be at 0, but rather
> > > at .7
> > > > > volts above ground, and I can't find any indication in the data
> > > sheet of
> > > > > what the chip considers to be the maximum logic low voltage.
> > > > >
> > >
> > >
> >
> >
>
>------------------------------------
>
>
>
>
• ... Not sure,but use a 1N34A Germanium diodes and you get .3 Volt Drop, a second alternative is a Schotty diode. Cecil K5NWA www.softrockradio.org
Message 5 of 13 , Jul 1, 2008
• 0 Attachment
At 06:05 AM 7/1/2008, you wrote:
>Does anyone know if a 12F683 running on 3.3 volts in the SoftRock 8.3
>circuit will operate reliably if, instead of using a DIP switch to ground
>pins on the chip, a rotary switch through a single silicon diode is
>used. I think that means that the pin will not be at 0, but rather at .7
>volts above ground, and I can't find any indication in the data sheet of
>what the chip considers to be the maximum logic low voltage.
>
>73, Pete N4ZR
>

Not sure,but use a 1N34A Germanium diodes and you get .3 Volt Drop, a
second alternative is a Schotty diode.

Cecil
K5NWA

"Blessed are the cracked, for they shall let in the light."
• ... ground ... at .7 ... sheet of ... Hi Pete, To use a single pole rotary switch, add a TTL 8-to-3 encoder (74xx148) and pull any one input to ground at a
Message 6 of 13 , Jul 1, 2008
• 0 Attachment
--- In softrock40@yahoogroups.com, Pete Smith <n4zr@...> wrote:
>
> Does anyone know if a 12F683 running on 3.3 volts in the SoftRock 8.3
> circuit will operate reliably if, instead of using a DIP switch to
ground
> pins on the chip, a rotary switch through a single silicon diode is
> used. I think that means that the pin will not be at 0, but rather
at .7
> volts above ground, and I can't find any indication in the data
sheet of
> what the chip considers to be the maximum logic low voltage.
>
> 73, Pete N4ZR
>
Hi Pete,

To use a single pole rotary switch, add a TTL 8-to-3 encoder (74xx148)
and pull any one input to ground at a time. The output will be the 3
digit binary number reflecting the highest input that is Low when all
below it are high. Inputs will tend to float high in the absence of
large signals, so the encoding will reflect the position of the
grounded pin. I used this a lot 'back in the day' with the LS family
without pull-up resistors (YMMV with other families). Wire the chip
in 'dead bug' style and bridge a 0.1 uf chip between Vcc and Gnd. and
let us know how this works for you.

73 de Jeff, KB3NAX
• Why the diode? ... From: Pete Smith To: Sent: Tuesday, July 01, 2008 6:05 AM Subject: [softrock40] 12F683
Message 7 of 13 , Jul 1, 2008
• 0 Attachment
Why the diode?

----- Original Message -----
From: "Pete Smith" <n4zr@...>
To: <softrock40@yahoogroups.com>
Sent: Tuesday, July 01, 2008 6:05 AM
Subject: [softrock40] 12F683 Question

> Does anyone know if a 12F683 running on 3.3 volts in the SoftRock 8.3
> circuit will operate reliably if, instead of using a DIP switch to ground
> pins on the chip, a rotary switch through a single silicon diode is
> used. I think that means that the pin will not be at 0, but rather at .7
> volts above ground, and I can't find any indication in the data sheet of
> what the chip considers to be the maximum logic low voltage.
• Good question - I am rethinking that, given that I have a separate switch wafer dedicated to frequency control, and another for switching the BPFs. Maybe I
Message 8 of 13 , Jul 2, 2008
• 0 Attachment
Good question - I am rethinking that, given that I have a separate switch
wafer dedicated to frequency control, and another for switching the
BPFs. Maybe I can dispense with the diodes on both of those functions. I
was thinking by analogy to a band decoder, where the same outputs must be
shared between different sets of choices (like 4 antennas and 6 BPFs).

73, Pete

At 12:15 AM 7/2/2008, you wrote:
>Why the diode?
>
>----- Original Message -----
>From: "Pete Smith" <n4zr@...>
>To: <softrock40@yahoogroups.com>
>Sent: Tuesday, July 01, 2008 6:05 AM
>Subject: [softrock40] 12F683 Question
>
>
> > Does anyone know if a 12F683 running on 3.3 volts in the SoftRock 8.3
> > circuit will operate reliably if, instead of using a DIP switch to ground
> > pins on the chip, a rotary switch through a single silicon diode is
> > used. I think that means that the pin will not be at 0, but rather at .7
> > volts above ground, and I can't find any indication in the data sheet of
> > what the chip considers to be the maximum logic low voltage.
>
>
>------------------------------------
>
>
>
>
• Unfortunately, I need 4 outputs, because the codes for several bands use the 4th digit. Actually, I think I have a super-simple solution - I ll test it and
Message 9 of 13 , Jul 2, 2008
• 0 Attachment
Unfortunately, I need 4 outputs, because the codes for several bands use
the 4th digit. Actually, I think I have a super-simple solution - I'll
test it and post description later.

73, Pete

At 06:13 PM 7/1/2008, jeffthom99 wrote:
>--- In softrock40@yahoogroups.com, Pete Smith <n4zr@...> wrote:
> >
> > Does anyone know if a 12F683 running on 3.3 volts in the SoftRock 8.3
> > circuit will operate reliably if, instead of using a DIP switch to
>ground
> > pins on the chip, a rotary switch through a single silicon diode is
> > used. I think that means that the pin will not be at 0, but rather
>at .7
> > volts above ground, and I can't find any indication in the data
>sheet of
> > what the chip considers to be the maximum logic low voltage.
> >
> > 73, Pete N4ZR
> >
>Hi Pete,
>
>To use a single pole rotary switch, add a TTL 8-to-3 encoder (74xx148)
>and pull any one input to ground at a time. The output will be the 3
>digit binary number reflecting the highest input that is Low when all
>below it are high. Inputs will tend to float high in the absence of
>large signals, so the encoding will reflect the position of the
>grounded pin. I used this a lot 'back in the day' with the LS family
>without pull-up resistors (YMMV with other families). Wire the chip
>in 'dead bug' style and bridge a 0.1 uf chip between Vcc and Gnd. and
> let us know how this works for you.
>
>73 de Jeff, KB3NAX
>
>
>------------------------------------
>
>
>
>
• ... I don t know if the 12F683 has internal pull-ups. If so, then a single wafer switching between ground and open will suffice. Generally with CMOS, an open
Message 10 of 13 , Jul 2, 2008
• 0 Attachment
On 2 Jul 2008 at 7:20, Pete Smith wrote:

> Good question - I am rethinking that, given that I have a separate switch
> wafer dedicated to frequency control, and another for switching the
> BPFs. Maybe I can dispense with the diodes on both of those functions. I
> was thinking by analogy to a band decoder, where the same outputs must be
> shared between different sets of choices (like 4 antennas and 6 BPFs).
>
> 73, Pete
>
I don't know if the 12F683 has internal pull-ups. If so, then a single wafer switching between
ground and open will suffice. Generally with CMOS, an open input is undefined, so it must
either be pulled up to Vcc or down to ground at all times. The only issue in connecting to
ground is making sure the pin cannot be an output when it is connected to ground.

Dave - WB6DHW
<http://wb6dhw.com>
• Well, connecting to ground is exactly what the DIP switch on the 8.3 Xtall board does. It has pull-ups and also series protective resistors against the kind
Message 11 of 13 , Jul 3, 2008
• 0 Attachment
Well, connecting to ground is exactly what the DIP switch on the 8.3 Xtall
board does. It has pull-ups and also series protective resistors against
the kind of unintentional output scenario you describe. Seems like a good
design.

On another subject, after thinking a bit more, I will still need the
steering diodes to keep from unintentionally tying input lines together,
because of the different combinations needed for the 6 center frequencies
of interest.

73, Pete N4ZR

At 11:21 AM 7/2/2008, brainerd@... wrote:
>On 2 Jul 2008 at 7:20, Pete Smith wrote:
>
> > Good question - I am rethinking that, given that I have a separate switch
> > wafer dedicated to frequency control, and another for switching the
> > BPFs. Maybe I can dispense with the diodes on both of those functions. I
> > was thinking by analogy to a band decoder, where the same outputs must be
> > shared between different sets of choices (like 4 antennas and 6 BPFs).
> >
> > 73, Pete
> >
> I don't know if the 12F683 has internal pull-ups. If so, then a single
> wafer switching between
>ground and open will suffice. Generally with CMOS, an open input is
>undefined, so it must
>either be pulled up to Vcc or down to ground at all times. The only issue
>in connecting to
>ground is making sure the pin cannot be an output when it is connected to
>ground.
>
>Dave - WB6DHW
><http://wb6dhw.com>
>
>
>------------------------------------
>