Help! CMOS 74C373 Switching Problems

[Frost]

Hey there,

I'm trying to build a channel switching rig for my guitar amp using the design from R.G. Keen: http://www.geofex.com/Article_Folders/fxswitchr/fxswitchr.htm but I'm running into some major headaches right out of the starting gate.

I'm just testing out the logic right now but cannot get it to work for the life of me. Right now, I'm using momentary switches to apply voltage to the D1-4 inputs and having the unit drive 4 LEDs. However, all that happens when I turn it on is that all 4 LEDs light up and pressing any of the switches does nothing. Here is a schematic of what I'm running right now: http://f2.pg.photos.yahoo.com/ph/heiko_frost/detail?.dir=/b407&.dnm=5baa.jpg

If anyone has any insight on what's going wrong or if I hooked up a pin incorrectly, I sure would appreciate it. Thanks,

-Heiko  :?:

[Nasse]

Does anything happen at pin 11 where the 4007´s go?

[Frost]

Hey there,

I'm not sure what you mean regarding pin 11. I'm quite a newbie when it comes to solid state circuits. I'm not sure what pin 11 does exactly, for the clock perhaps? I could not get too much info that helps from the data sheet unfortunately.

-Heiko

[Maneco]

first,you need to put pull down resistors on all cmos inputs,pins 3,4,7,8,11.100k is ok

[Frost]

Hey Maneco,

Is this what you mean?

http://f2.pg.photos.yahoo.com/ph/heiko_frost/detail?.dir=/b407&.dnm=f0a5.jpg

-Heiko

[Peter Snowberg]

Try adding pull down resistors to pins 3, 4, 7, 8, and 11 as Manceo suggests. Depending on the logic family you use you might want to go as low as 10K but for C and HC families 100K will probably be fine. That may solve the problem.

The 373 will have the outputs follow the inputs when pin 11 is high; it's almost like it's not even there. For this reason it's called a "transparent latch". When pin 11 goes low, the previous output is latched and doesn't change regardless of what you do to the input.

The 374 is similar, but it uses more of a clock style input that takes a snapshot of the inputs when the clock input falls from high to low. The rest of the time the 374 outputs don't change at all.

For easy wiring, check out the 74HC573 & 74HC574 devices too.

Good luck,
-Peter

[R.G.]

Your momentary switches to V+ bounce when you press them. They may "make" to V+ and then bounce off half a dozen times every time you press them. That gives the logic circuit - which is quite fast enough to follow the bounces - a lot of false inputs with every press.  Debouncing is what those schmitt trigger inverters and resistor/capacitor pairs are for in the original schematic.  All mechanical switches connected to logic need debounced in general because of this false triggering problem.

Some logic circuits are tolerant of this, some are not.

You might be able to get this to work if you put a capacitor across each switch and a resistor to ground from the switch (this is the input pulldowns mentioned earlier) but not a direct connection to V+. That lets the R-C network debounce the switch and the 74C373's schmitt trigger inputs cleal up the slow, sloppy rise and fall times. Try 10K and 0.01uF.

Pin 11 may need pulled down with 1M on some brands of 'C373.

Cleaning up switch bounce and the sloppy rise times from it is the reason I put those inverters in front of the logic chip. Quoting Albert - everything should be as simple as possible - but no simpler!

[/list][/list]

[Frost]

Hey Everyone,

Thanks for all the input. I'm going to give it another go tonight after work. Thanks R.G. for both the schematic and the tip regarding the OP amps. I ommited them with my first schematic to keep things simple only while getting the logic to work. I had no idea that bouncing could be a problem. Take care,

-Heiko

[R.G.]

....and the tip regarding the OP amps

ACK!! Those are not opamps!

Those are logic chips - the CD4049 is a hex (that means six-way) logic inverter chip. Opamps can work there, but it takes a lot more components around them. Use the CMOS logic chip.

[Frost]

Thanks for the clarification RG. Like I meantioned before, I am quite new to solid state equipment. I've got a numbers of years of working with tube amps under my belt but all of this is something new...and quite interesting too!

-Heiko

[Frost]

Hi there,

Thanks again to everyone who helped me get this switching rig off of the ground. It works quite well but it has one glitch:

When the unit is first turned on, all of the outputs are usually activated until I press one switch which cancels the other three. I have played with different values for the resitor to ground from pin 11 but to no avail. Any suggestions? Here again is the schematic I'm using:

http://us.f2.yahoofs.com/users/7c0ba17c_m40b53511/bc/b407/__sr_/f0a5.jpg?ph5YZABBfHWy2JOB

Thanks,

-Frost

[Nasse]

I have seen it done in some similar circuits just with a small cap from pos voltage connected to one input, so that the cap gives a pulse to the desired input, but there were diodes at inputs so the inputs don´t interfere (could not see your circuit, it was not available for me) but don´t know if that method works with 373 in that context

There are circuits around that give a pulse at power on, search for "reset swith" or something

EDITED

Try connecting a 10 nF cap from V+ to one input, maybe it´ll do the trick

[Frost]

Hey Nasse,

Thanks so much for the suggestion...it worked! I had to go up from a 10nF cap to a 1uF cap from V+ to an input but now it always turns that input on initially as long as the power supply has been discharged enough.

Also, thanks to everyone who helped me with this, I am truly grateful.

- Heiko Frost

[R.G.]

I think that if you tie that same cap to the latch enable pin (LE) that it will clock in an "all off" condition, not one input on.