T&L Publications

See attachment circuit.
Is there any way this circuit can be modified to a 3 digit frequency counter? I realized the software will have to be changed. But if someone out there can do this it will be much appreaced. I also realized that the D1, Q3, RLY1 can be omited.

Greetings Craig,

Craig Kendrick Sellen wrote: See attachment circuit.
Is there any way this circuit can be modified to a 3 digit frequency counter? I realized the software will have to be changed. But if someone out there can do this it will be much appreaced. I also realized that the D1, Q3, RLY1 can be omited. :smile:

The circuit is not a frequency counter, but it does have
the bare structure for one (uC, displays, user switches).

To expand it to a third digit would require one more
display and a dedicated low-side driver transistor for
it. Unfortunately, there are no spare pins on the uC.
A work around would be to encode the two existing pins
and add a decoder IC (two bit binary decoder).

By using a larger uC (with more IO pins) all the
external ICs can be removed (only one transistor
per digit would remain, plus some Rs and Cs and
switches). A crystal timebase would be superior.

Take a look at the cover story in Nov 2007 Nuts
and Volts for a similar uC and display solution.

What frequency range do you need?
What precision do you need?

Comments Welcome!

Hello,

Apparently the OP didnt come back.

I would use one of the outputs to make the third digit, but i would
use NPN transistors, not PNP. The programming goes accordingly.

my first thought was to take one of the LED outputs for the 3rd digit. My second thought was "why are they using PNPs for low side drivers?".

Of course, if they used a similar sized PIC instead of a PICAXE, they would have plenty of outputs.

Hi again,

Phliba:

Yeah, i was wondering about those PNP's too that's why i mentioned
the NPN's instead.

I'd also double check the value of the seven resistors (now 47 ohms
each) to make sure the current works with the type of seven segment
display being used. The average current through a segment is:

I=DC*(Vdd-(VIC+VLED))/R

where

I is the average current through a segment
DC is the duty cycle (determined by programming)
Vdd is the supply to the segment driver chip
VIC is the output voltage drop of the segment driver chip
VLED is the voltage of the LED segment
R is the resistor value

With appropriate programming the duty cycle can be made as
high as 33 percent when using three digits.

There are more Freq Counter projects on the web than you can count. Take a look at This Page.

TO ALL
I can tell you this.
There are common cathodes and common anodes on the displays. Thats why there PNP's or NPN's.

To the one on the web, whats the PIC# and what software does it use?

Greetings Craig,

Craig Kendrick Sellen wrote:TO ALL
I can tell you this.
There are common cathodes and common anodes on the displays. Thats why there PNP's or NPN's.

Good to know. The OP was discussing this circuit.
I only see PNP transistors (BC557), and the LED displays
are common cathode. I'm not aware of any LED displays
with both common anode and common cathode at the
same time.

The other poster mentioned PNP and NPNs because they
would have used NPN types for the common cathode
role (which has slight advantages but requires the uC
digit data to be inverted, it's a popular arrangement).

Comments Welcome!

Greetings Craig,

Craig Kendrick Sellen wrote: To the one on the web, whats the PIC# and what software does it use?

Found this following the previous link (here)
"very low component count: a PIC 16F628"

The PIC firmware for the frequency counter can be downloaded
from this link (includes hex file for all display variants and the
assembly source code).

Comments Welcome!