Do 556 multi-vibrators mess with your circuit's power?

[dentaku]

This is definitely a beginner question and I don't really know how to describe it.

This week I built an up/down counter with a 4029 counter and a 4511 BCD to 7segment decoder/driver. I needed a clock so plugged the output of a 556 astable I had already built on a tiny breadboard and ran it off the same 9V battery as the counter circuit.
I found this example and it's almost exactly what I built http://www.electro-tech-online.com/attachments/up-down-counter-gif.26692/ except I have a 556 and not a 555.

THEN I connected another breadboard with a VCO on it to the same 9V battery as the counter circuit and used the BCD outputs of the 4029 counter to modify the frequency of the simple little saw wave VCO http://electro-music.com/forum/topic-49901.html
You get little two note sequences at whatever tempo your clock is running at and they're different depending which of the 4 outputs you connect it to, like I was expecting. That's all fine, it's like a little step sequencer synth people use all the time.

QUESTION: Why is it that even when there's nothing connected to the control voltage input of the VCO I STILL get the frequency of the VCO going up and down a little bit at the same speed as the 556 of course just sitting there connected to the same power source (9V battery) as the counter circuit?

I even tried it with just the little 556 breadboard running (not connected to the counter at all or anything for that matter) and it still happens as long as the VCO and the 556 circuit are sharing the same 9V battery so obviously it has something to do with the 556 multi-vibrator messing with the power of the VCO.
How would I avoid this situation so the saw wave VCO stays at a stable frequency even when the 556 is running?

This IS and ancient LM556CN from what looks like 1984. Would using a 74c14 Schmitt Trigger or a modern 7555 to generate the clock signal for the counter help?
I guess that's what I should try next

[T3sl4co1l]

The VCO is pretty awful so it will be sensitive to power supply ripple, average voltage, variation in the control inputs (which, if they in turn depend on the supply..), temperature, phase of the Moon, and so on.

Why 556?  Just what you had on hand?  A 9V battery won't last all that long powering a crusty old brute like that -- some, what is it, 5 or 10mA just to run the chip?  If you're planning on working with a lot of battery powered circuits, consider buying 7556 or LMC556 (CMOS versions), which will be much more kind to the battery, also disturbing the supply less.

Tim

[dentaku]

It just happens to be that I had a 556 set up on a little breadboard so I used it.
The fact that it did something unexpected is almost more interesting than if it all worked perfectly on the first try.
Like Dave says, the best way to learn is to fail and have to figure out how to fix it.
I'm gonna try a 74c14 Schmitt Trigger or a 7555 and see how it turns out.

As for the VCO.... It runs off a single supply and it's fun to experiment with. I might build a VCO with a LM13700 OTA sometime soon.

[Bored@Work]

Ok, ancient NE556, battery supply. Yep, makes sense. The original 555 and, if I am not mistaken, his brother the 556 are tough on the power supply. They create large current spikes when switching. Depending on the internal resistance of the battery and the wires connecting the battery to the circuits the supply voltage is going down a bit whenever the 556 switches. So what might happen is the switching is coupled into your VCO via the power supply.

Breadboard you said? So lots of long wires? It could also be the switching couples in wireless (EMI) in to the VCO.

[SeanB]

Most 555/556 applications have a 470uF capacitor right at the supply pins of the chip to try to decouple that switching spike. In some cases I have used up to 2200uF to keep it manageable.

[StryderL]

I had a project similar to this a while back, I'm not entirely sure how your project is set up but the problem might have to do with your LED display.  In my other project I had two 7-segment displays counting and an ADC picking up values to feed to them through a microcontroller.  Although the displays functioned fine, the ADC was always rather off.  It turned out that every time the displays changed, they were taking a lot more current than I expected, and what the power supply could handle.  A probe with the oscilloscope showed a very noisy power bus, with 1v hiccups.  I managed to fix the issue by adding a capacitor of around 1mF to the power bus, as well as another right next to my ADC as well as my display driver.

You might try disconnecting the power to your LED display to see if that current is disrupting your circuit, I might be mistaken but if it's a counter then the driver is probably switching with your clock.  An oscilloscope probe on all your power lines might also be useful for finding issues, if you have access to one.  I also don't believe 9V batteries are all that great at providing a lot of "burst" current, so adding a larger capacitor on your power bus might help considerably.

[dentaku]

The first thing I tried was to disconnect the 556 circuit completely from the counter circuit and it's the thing that was messing with my power.
This morning I built square wave generators with a 74c14 Schmitt Trigger and another one with a 7555 all by themselves, not connected to the counter at all and tested out what the power rails looked like on my old $40 analog scope.
The 74c14 was almost as noisy as the 556 and the 7555 was surprisingly worse than the other two.
The scope shows me a square-ish waveform with a rising edge that's curved at about 0.1V Peak-to-Peak.
The 74c14 creates a very fuzzy square-ish wave on the power rails that's only 0.02V peak-to-peak so it's definitely better than the 7555. I don't have the 556 circuit anymore to test it again.
It's worse when it's connected to the counter circuit of course, switching on and off leds with the 4511.

I guess I'm going to have to experiment with different types and sizes of decoupling capacitors and see if I can learn more about quieting down these spikes.
I'm not building anything useful here anyway, just trying to learn new things.

Of course, like T3sl4co1l said, the Saw VCO I'm using that's picking up this noise is "pretty awful so it will be sensitive to power supply ripple, average voltage, variation in the control inputs (which, if they in turn depend on the supply..), temperature, phase of the Moon, and so on."

I had a project similar to this a while back, I'm not entirely sure how your project is set up but the problem might have to do with your LED display.  In my other project I had two 7-segment displays counting and an ADC picking up values to feed to them through a microcontroller.  Although the displays functioned fine, the ADC was always rather off.  It turned out that every time the displays changed, they were taking a lot more current than I expected, and what the power supply could handle.  A probe with the oscilloscope showed a very noisy power bus, with 1v hiccups.  I managed to fix the issue by adding a capacitor of around 1mF to the power bus, as well as another right next to my ADC as well as my display driver.

You might try disconnecting the power to your LED display to see if that current is disrupting your circuit, I might be mistaken but if it's a counter then the driver is probably switching with your clock.  An oscilloscope probe on all your power lines might also be useful for finding issues, if you have access to one.  I also don't believe 9V batteries are all that great at providing a lot of "burst" current, so adding a larger capacitor on your power bus might help considerably.

[dentaku]

I did some experiments yesterday with a 9V, 1.5A, 13.5W Max "regulated and filtered" AC Adapter and it IS definitely better than a 9V battery.
I wish it had screws so I could open it up to see how it's designed