cmos q's and odd problem

Started by jimmy, September 28, 2003, 03:19:28 AM

Previous topic - Next topic

jimmy

hi all

in the process of building my calavera ive hit a snag. no signal seems to get out of the first inverter. i used the audio probe to sort this much out. i cant figure out why.

only thing i can think of is it might have something to do with the cmos chip. ive heard they are really delicate. now i haveby accident re-soldered a joint in there but havent had the iron on the chip any longer than 5 seconds. i have the chip in a socket, bu i keep forgetting to take it out.
"Who the f*** are the naked chefs?" - Ozzy Osbourne

tubes or bust

Tim Escobedo


gez

If you have a multimeter, hook it up to the output of the first stage and take voltage measurements.  

With the wiper of the pot to ground, the output should be 9V (it does use a 9V supply doesn't it?).  With the wiper to 9V, output should be 0V.  As you turn the wiper 'class A' current will flow at some point, which should bias the output somewhere between the two rails - the output should move smoothly in the opposite direction to the way you're turning the wiper (analogue meter is the best for this).  

If you're getting results different from this, then you may have fried the chip, in which case solder in a socket and slot a new chip in (take care with handling as they're static sensitive).  I do this with all CMOS devices as it saves tears in the long run.

Other than that, check all wires, connections, joints etc.
"They always say there's nothing new under the sun.  I think that that's a big copout..."  Wayne Shorter

jimmy

oh, and another thing, do i have to ground all the rest of the inverters, or can i just leave them and have it noisy for the time being?
"Who the f*** are the naked chefs?" - Ozzy Osbourne

tubes or bust

gez

Whoops!  You're already using a socket...sorry!
"They always say there's nothing new under the sun.  I think that that's a big copout..."  Wayne Shorter

gez

Quote from: jimmyoh, and another thing, do i have to ground all the rest of the inverters, or can i just leave them and have it noisy for the time being?

Ground the inputs, or tie them to Vdd.
"They always say there's nothing new under the sun.  I think that that's a big copout..."  Wayne Shorter

jimmy

hmmmm when i turn the trimpots, nothing happens to the voltage at all. chips cooked i spose. ill get a new one and let you know how everything went.
"Who the f*** are the naked chefs?" - Ozzy Osbourne

tubes or bust

Peter Snowberg

Quote from: jimmyoh, and another thing, do i have to ground all the rest of the inverters, or can i just leave them and have it noisy for the time being?

CMOS chips with floating inputs are prone to a condition called "latch up" where everything gets totally saturated and the chip is effectively frozen until you power it down and back up. With several unused floating inputs, it may latch up as soon as you add V+. 60 cycle hum and AM transmitters can easily provide the straw to break the camels back. The solution.... ground that camel! :)

Another thing to keep in mind with CMOS is that you can blow some of these chips out just by touching them with a 110 VAC soldering iron. Most chips have good protection diodes these days, but if you run across 4000 series parts with an 'A' in the suffix, watch out as these early chips are much more fragile. Quite a few 24 VAC soldering irons are advertised as "CMOS safe" and are grounded.

With a logic chip like this it's really easy to test. Just power the chip up with the inputs all grounded and look for power on the outs using a 1K resistor and a red LED or your voltmeter. Keep in mind that gates like a 4011 don't like to drive LEDs so a meter is better most of the time. CMOS inverters are a special case since that have MUCH more output drive than regular gates.

Take care,
-Peter
Eschew paradigm obfuscation

jimmy

right ive hooked everything up as per schem, and have tested the outputs and the LED lights up at every one of them except the output. i changed the inverter and then that one didnt light up but the one i was using before did. am i ok? should the circuit be fine?
"Who the f*** are the naked chefs?" - Ozzy Osbourne

tubes or bust

Ge_Whiz

The best way to fry a CMOS chip is to leave unused inputs floating.

gez

Quote from: jimmyright ive hooked everything up as per schem, and have tested the outputs and the LED lights up at every one of them except the output.

By 'output' do you mean the output of the inverter you were having problems with?  If so then it sounds suspect.

"i changed the inverter and then that one didnt light up but the one i was using before did. am i ok? should the circuit be fine"

Do you mean you've hooked up the trimpot to one of the other inverters within the chip, and now that one doesn't work, but the original which was 'duff' is now ok?  If so, sounds like the chip might be ok and you've got a wrong connection (perhaps the trimpot isn't grounded?)

Apologies if I've got the wrong end of the stick, but your post was a little unclear.
"They always say there's nothing new under the sun.  I think that that's a big copout..."  Wayne Shorter

Peter Snowberg

If you're checking with an LED, first test from the inverter outputs to ground. Next, flip the LED polarity and check between the outputs and V+. You could have half of a CMOS pair blown. This should tell you.

A voltmeter on the inverter input should tell you if you're 'biased into analog'. You should see a voltage on the inputs somewhere in the middle between V+ and gound. If you don't see any voltage, either your bias resistors are funky or you have a dead chip.

-Peter
Eschew paradigm obfuscation

jimmy

QuoteDo you mean you've hooked up the trimpot to one of the other inverters within the chip, and now that one doesn't work, but the original which was 'duff' is now ok?

yeah thats the one. sorry for the unclear posts im a tad frustrated.

im fairly confident the fault lies with the trimpots, because when i adjust them, and my fingers hit the legs i get signal. can you suggest a fixed value that i could use?
"Who the f*** are the naked chefs?" - Ozzy Osbourne

tubes or bust

gez

A straight forward divider should do ok.  Try a couple of 470k resistors.
You might find that the outputs are slightly (or not so slightly) off-centre, but it should bias somewhere in its 'linear' region.

This is only MY suggestion (it's Tim's circuit and he'd know better)

Edit:  PS it's worth persevering with the trimpots, it's probably just a wrong connection or perhaps a dodgy pot.
"They always say there's nothing new under the sun.  I think that that's a big copout..."  Wayne Shorter

gez

Quote from: jimmyim fairly confident the fault lies with the trimpots, because when i adjust them, and my fingers hit the legs i get signal. can you suggest a fixed value that i could use?

Sounds like a bad joint.  If you have a continuity tester on your multimeter, use it to determine which joint is the offender.  Remove the chip (this time  :D ) and do the joint again.  It's probably the one which connects to ground (this would hold the input at 9V, biasing the output at 0V).
"They always say there's nothing new under the sun.  I think that that's a big copout..."  Wayne Shorter

Tim Escobedo

Tie all un used inputs to ground. That's a given with all CMOS chips. Failing to do so can lead to odd behavior and excessive power consumption, possibly even chip failure. Use sockets, take all static precautions, etc.

That being said, I've never fried a CMOS chip, ever. I used to treat them with kid gloves, but after doing some unscientific testing, I found them to be a bit more durable than I thought.

The voltage at the input of each inverter should be determined by the setting of each trimmer. That it, it should be continuously variable from 0V to V+. Each inverter will "work" with the trimmer somewhere around 1/2 V+. This will vary from chip to chip, thus the need for trimmers. If the input of either inverter is 0V or V+, and does not change with trimmer setting, either there is a faulty trimmer connection somewhere, or a coupling cap is missing somewhere, or you have a seriously fried chip.

jimmy

when i adjust the trimmers, very little change in voltage occurs. im pretty sure ive cooked it.
"Who the f*** are the naked chefs?" - Ozzy Osbourne

tubes or bust

Dan N

Quote from: Tim EscobedoTie all un used inputs to ground. That's a given with all CMOS chips.

Is it O.K. to ground out the unused outputs too, or just leave them alone?

Sorry to briefly hijack the thread. I just need to know what to do with something I am working on. I'm using 2 sections of a 4016, and half of a 4013. I'd like to know the right way to deal with the unused parts...

Thanks,

Dan

gez

quote "Dan N"

"Is it O.K. to ground out the unused outputs too, or just leave them alone?"


With inverters, if you 'ground' the output all you're doing, in effect, is connecting a wire from it to ground, so if it's input is pulled 'low' (ground) it's output is 'high' (positive rail) and a large flow of current results (not a good idea).

You can attach resistors to ground from the outputs of inverters for various applications.
"They always say there's nothing new under the sun.  I think that that's a big copout..."  Wayne Shorter

Dan N

Thanks gez. Can you or anyone tell me what to do with the unused half of a 4013? Set and reset look like inputs. Ground those? Clock and data too?