4026 stuck at zero!

[LFNfan]

Hi All

I have put together a 7 segment display driven by a 4026 with clock from a 555 in astable mode, following various guides online.

After incrementing a few times it is now stuck permanently at zero. The 555 is still working, putting out between 0 and 2.5v at pin 3 of the 555 to the 4026 (the rest of the 5v supply is going to the led I think).

Im also getting that same difference at the 4026 clock pin 1 (unsurprisingly). But the display refuses to shift from showing 0.

I think I have all the other 4026 pins properly high or low so don't think that is what's driving the zero display.  I've also swapped in another 4026 but no change. I have also taken out the led to get a higher voltage at the clock pin. No joy.

If anyone could take a look at the attached pic and point out where I have made a mistake or what to check next that would be fantastic. Once I have this down I'm going to add another 7 seg.

Many thanks!

[grumpydoc]

Could you post your intended schematic please.

[LFNfan]

Sure. Sorry it's pretty rough and ready. Hopefully useful though

[LFNfan]

... and the 555

[retiredfeline]

0 - 2.5V swing may be marginal for the 4026.

[grumpydoc]

0 - 2.5V swing may be marginal for the 4026.

It is (ideally ≥ 3.5V on 5V supplies). But I don't think that's the problem.

My guess is that it is something to do with the way the 4026 is wired, and seeing the (albeit coffee stained) schematic shows it's not just a mistake in transcribing schematic to the breadboard.

Hint (hope that I'm right): What happens to unused CMOS inputs?

[oPossum]

Pin 15 (reset) must be grounded for the counter to run.

[Gyro]

Sorry it's pretty rough and ready. Hopefully useful though

Colourful though. 

Edit: Yes, I think it's sitting in reset due to Pin 15 floating. It's right next to VCC so would probably float high from leakage.

[grumpydoc]

Pin 15 (reset) must be grounded for the counter to run.

Yeah, I was hoping we could get LFNfan there with some hints.

Unused CMOS inputs tend to unreliably float high, in this case I think it took the stray voltages a few seconds to charge the gate capacitance up enough to make it register as a logic '1', hence counting for a few seconds before stopping.

LFNfan - make sure unused high impedance inputs are tied to a supply rail (through a resistor of a couple of k if you want) - not doing so can lead to odd problems such as this one. As oPossum says pin 15 needs to be logic '0' for the counter to work..

[LFNfan]

Hi All
You did get me there with the hint Thank you very much!

I have been reading datasheets and fixing and just come back to post and seen the last few posts.

Very interesting.

Tying down all the pins got me there. Learned something new there

Also the voltage was an issue too - 2.2 at the clock with the led in - no advancing. 3.5 with no led and advances happily.

How could I get the led in the circuit? Maybe a transistor with base to pin 3 of the 555?

Many thanks again!!

[Zero999]

Try connecting the LED and current limiting resistor between the output and +V.

[LFNfan]

Perfect.

Many thanks to you all!

[Ian.M]

The problem with 'classic' bipolar 555 timers like the NE555 is the output cant get more positive than about 1 1/2V below the supply rail.  If you are working with 4000 series CMOS logic at under about 7V, or any lower voltage CMOS logic that doesn't have TTL input threshold voltages,  that really isn't good enough.  Either use a CMOS '555', which will have a rail to rail output or mod the bipolar 555 circuit for more output swing.

There's an alternative 555 oscillator circuit that feeds a single timing resistor from the Output, and optionally you can take the actual output from the Discharge pin.  As the discharge pin is open collector, you need a pullup resistor, but you get an output swing from within a few hundred millivolts of ground, all the way up to the supply rail if the load on it doesn't sink significant current through the pullup resistor.  To get there from your circuit, connect the top end of R2 to pin 3 instead of pin 7, and connect pin 7 to the chip you want to clock.  As 4000 series CMOS draws negligible input current you can also make R1 (which is now the pullup) 10K and save some power.

[Zero999]

Connecting the LED from pin 3 to +V, rather than 0V, as I said before, will help the output swing to near +V into a CMOS input, but it will invert the LED output.

The output pin charging/discharging the capacitor doesn't work quite so well with the bipolar 555 timer and is marginal, with the LED connected between the output and 0V.

Good point about pin 7: R1 = 1k and R2 = 10k, so the 4026 input could be connected to pin 7, with no other modifications to the circuit. When pin 7 is low, it will be close to 0V and when pin 7 is open, and the voltage across the capacitor is ¹∕₃Vcc, it will be:

Vhigh = Vcc-R1*(Vcc*²∕₃)/(R1+R2) = 5-1*(5*0.6667)/(1+10) = 5-3.333*11 = 4.7V

Which is well within the CMOS voltage thresholds.

[Ian.M]

Yes.  In the alternative oscillator circuit, moving the LED to the high side is pretty much essential, (and preferably move it to Discharge as well) otherwise the 555 output may not make it past 2/3 Vcc which will stop it oscillating and leave it stuck high.   A 1K pullup on the output will help if its marginal.

The alternative circuit works much better at significantly higher supply voltages (where the Voh drop becomes small compared to Vcc/3) or with a CMOS 555 which will give you a rather good^* 50% duty cycle squarewave.

* Duty cycle tolerance dependent on internal trigger and threshold divider tolerances, so not as good as the nearly perfect 50% you can get by dividing to f/2 using a toggle flipflop.

[floobydust]

I just noticed the CD4026 has "ungated c-segment output" pin 14 for div/60 and div/12 for timekeeping. Neat if you want to make clocks.

[LFNfan]

Hi
Thanks very much for your really full responses. I feel like I am learning a lot.

I have modded the 555 circuit as suggested by Ian.M and it works but weirdly I seem to get a narrower voltage at the 4026 clock pin. And my led flashes after the 7-segment has changed not when it changes. Undoubtedly I have made a mistake(s) somewhere.  Clues appreciated!!

I have redrawn my schematic without coffee stains, reflecting the new setup and that both setups have been using a LM555 chip (newer than the NE555).

Many thanks for your patience!

[Zero999]

Hi
Thanks very much for your really full responses. I feel like I am learning a lot.

I have modded the 555 circuit as suggested by Ian.M and it works but weirdly I seem to get a narrower voltage at the 4026 clock pin. And my led flashes after the 7-segment has changed not when it changes. Undoubtedly I have made a mistake(s) somewhere.  Clues appreciated!!

I have redrawn my schematic without coffee stains, reflecting the new setup and that both setups have been using a LM555 chip (newer than the NE555).

Many thanks for your patience!

Transpose the 10k and 1k on pins 7, 6 & 8.

You forgot to draw the LED: where is it connected now? Try changing LED part back to how it was in the original schematic: pin 3, resistor, LED, 0V. Now the counter is taking its clock from pin 7, rather than, the same pin as the LED is connected, it should work.

[grumpydoc]

...And my led flashes after the 7-segment has changed not when it changes.

If you followed Zero999's suggestions then the LED will flash when the output is low not high - in fact he pointed out that it would invert the LED output.

I have redrawn my schematic without coffee stains

Thata could be the problem - hand drawn schematics only work with coffee stains!!

[Zero999]

I think there's some confusion between different circuits. If you want to take the clock from pin 3, then the LED has to go from pin 3 to +V, which will invert the operation of the LED.

If the clock is taken from pin 7, R_A needs to be low, compared to R_B and it doesn't matter whether the LED is connected between pin 3 and 0V or +V. This circuit will do what you want. I added the decoupling capacitor.