Maker Pro
Maker Pro

Voodoo 74HC161 counter

Hi,

I'm not that knowledgeable in TTL but know enough to get myself in
trouble. ;-) I built a very simply TTL circuit that is not behaving as
it should IMHO.

I have one 54LS00J NAND chip, one 74HC161 hi-speed CMOS counter, and
some phototransistors. I am simply trying to tie the phototransistor to
the counters clock input. I am doing this the common way by placing a
resistor (47K) from Vcc to collector and grounding the emitter and then
using the phototransistors collector as the output, which goes to the
counters clock input (pin 2). I understand 47K is not common, but it
works best with my phototransistor since the phototransistor goes from
over 2 mega ohms in darkness to several kilo ohms in light. So as a
test, I simply move the phototransistors from light to dark, then back
to light and so forth. It is a gentle transition and I see no noise. Of
course it is not an instant transition either. The counters clock input
looks fine as it goes from nearly zero to 5.22 volts. For some reason
the counter won't respond. Below you'll see that I've tried adding
various size capacitors to eliminate noise.

So then I use two of the four NAND gates to create a J/K flip-flop just
to see if the counter is working. So now I have two phototransistors
tied to a J/K flip-flop. Both phototransistors are in the dark and I
move one of them into the light and then back into the dark then do the
same to the other phototransistors. This makes the flip-flops R input
go from H to L to H and then the S input goes from H to L to H. That
causes the flip-flop to change. I tie either one of the flip-flops
output to the counters clock and it works! So fine, I figure the
counter just don't like my phototransistor. I have no idea why because
the voltages look fine. I don't want the phototransistor to go to a
flip-flop, that was just a test. So instead of feeding two
phototransistors to a flip-flop, I feed one single phototransistor to
one NAND gate and then feed the NAND gate output to the counters clock.
So then I verified that the NAND gate is working fine; i.e., it's happy
with the phototransistor, BUT the counter is not happy! So then I feed
the NAND output to another NAND (on the same NAND chip) and then that
NAND output to the counters clock input. Still doesn't work! Also I
tried placing various size capacitors from the phototransistor's
collector to ground just incase their was some noise, but it made no
difference.

BTW, I am using a simple digital multimeter to measure the voltage.
Here's where some weird voodoo stuff comes in. If I reset the counter
by making the master reset (pin 1) go low then all four output bits of
the counter goes low. Then if I move the phototransistor in light and
back to dark the least significant bit goes high. After that point,
regardless if I move the phototransistor in light and back to dark,
etc. it will not change. Although, if I remove the multimeter and then
move the phototransistor in light and back to dark and then put the
multimeters clip lead back on counters least significant bit (pin 14)
then it is low! If I remove the multimeter again and move the
phototransistor back and forth in light / dark and put multimeter back
on then nothing changes, but if I leave the multimeter on pin 14 and
then move phototransistor back and forth in light / dark then the
counter bit goes high. I can't understand how my multimeter is
affecting the counter. I have no reason to believe there's anything
wrong with my multimeter. Regardless, the counter will not work if the
multimeter is connected or not.

I always thought a voltage was a voltage. The counters clock input (pin
2) goes from 0.11 to 4.11 volts coming from the NAND gate, but it does
not work. Yet, when I make two NAND's into j/k flip-flop then the
counter works even though the clock input voltages are not any
difference as they still change from 0.11 to 4.11 volts. Why would the
counter care? I refuse to believe there's any noise in my NAND output.

I would appreciate any help or suggestions!

Thanks,
Paul
 
J

John Fields

Jan 1, 1970
0
Hi,

I'm not that knowledgeable in TTL but know enough to get myself in
trouble. ;-) I built a very simply TTL circuit that is not behaving as
it should IMHO.

I have one 54LS00J NAND chip, one 74HC161 hi-speed CMOS counter, and
some phototransistors. I am simply trying to tie the phototransistor to
the counters clock input. I am doing this the common way by placing a
resistor (47K) from Vcc to collector and grounding the emitter and then
using the phototransistors collector as the output, which goes to the
counters clock input (pin 2). I understand 47K is not common, but it
works best with my phototransistor since the phototransistor goes from
over 2 mega ohms in darkness to several kilo ohms in light. So as a
test, I simply move the phototransistors from light to dark, then back
to light and so forth. It is a gentle transition and I see no noise. Of
course it is not an instant transition either. The counters clock input
looks fine as it goes from nearly zero to 5.22 volts. For some reason
the counter won't respond. Below you'll see that I've tried adding
various size capacitors to eliminate noise.

So then I use two of the four NAND gates to create a J/K flip-flop just
to see if the counter is working. So now I have two phototransistors
tied to a J/K flip-flop. Both phototransistors are in the dark and I
move one of them into the light and then back into the dark then do the
same to the other phototransistors. This makes the flip-flops R input
go from H to L to H and then the S input goes from H to L to H. That
causes the flip-flop to change. I tie either one of the flip-flops
output to the counters clock and it works! So fine, I figure the
counter just don't like my phototransistor. I have no idea why because
the voltages look fine. I don't want the phototransistor to go to a
flip-flop, that was just a test. So instead of feeding two
phototransistors to a flip-flop, I feed one single phototransistor to
one NAND gate and then feed the NAND gate output to the counters clock.
So then I verified that the NAND gate is working fine; i.e., it's happy
with the phototransistor, BUT the counter is not happy! So then I feed
the NAND output to another NAND (on the same NAND chip) and then that
NAND output to the counters clock input. Still doesn't work! Also I
tried placing various size capacitors from the phototransistor's
collector to ground just incase their was some noise, but it made no
difference.

BTW, I am using a simple digital multimeter to measure the voltage.
Here's where some weird voodoo stuff comes in. If I reset the counter
by making the master reset (pin 1) go low then all four output bits of
the counter goes low. Then if I move the phototransistor in light and
back to dark the least significant bit goes high. After that point,
regardless if I move the phototransistor in light and back to dark,
etc. it will not change. Although, if I remove the multimeter and then
move the phototransistor in light and back to dark and then put the
multimeters clip lead back on counters least significant bit (pin 14)
then it is low! If I remove the multimeter again and move the
phototransistor back and forth in light / dark and put multimeter back
on then nothing changes, but if I leave the multimeter on pin 14 and
then move phototransistor back and forth in light / dark then the
counter bit goes high. I can't understand how my multimeter is
affecting the counter. I have no reason to believe there's anything
wrong with my multimeter. Regardless, the counter will not work if the
multimeter is connected or not.

I always thought a voltage was a voltage. The counters clock input (pin
2) goes from 0.11 to 4.11 volts coming from the NAND gate, but it does
not work. Yet, when I make two NAND's into j/k flip-flop then the
counter works even though the clock input voltages are not any
difference as they still change from 0.11 to 4.11 volts. Why would the
counter care? I refuse to believe there's any noise in my NAND output.

I would appreciate any help or suggestions!

---
The first think I'd do would be to check the wiring (especially Vcc
and GND) for high resistance or intermittents.

The next thing is that since your counter is CMOS, it's _mandatory_
that _all_ the inputs be terminated either to Vcc or GND, as
appropriate, for the counter to work. That means, for example, that
the MR- terminal needs to be pulled up to Vcc when it's not being
used, as do CET, CEP, PE-, and D0 through D4.

After that I'd suspect the problem might be the speed of the edge of
the "clock" from the phototransistor. The slowest thing non-Schmitt
trigger CMOS inputs want to see is about 500ns, and that
phototransistor output is going to be way, way slower than that.

The next problem is that HC isn't designed to be driven by TTL, so
there may be something iffy going on with that, depending on your
parts. The way to fix that is to pull up the CMOS clock input and
drive it with the output of one of the TTL gates (to get the speed
you need) or, preferably, get yourself a comparator to do the
dark-light switching and use its output as the clock input to the
counter.
 
Top