4017 - Baby 8 Sequencer: Clocking Problems, Please Help.

[J2600]

Hi there,

I don't know if this is the right forum to post in but I hope you can help.

I made a 8 step sequencer based on the 4017 decade counter. It's basically a Baby 10 sequencer from this schematic:

http://www.midiwall.com/gear/babyseq/

Everything was working well when I was running it off a battery, it was clocking from the internal clock (a 401016) I built and clocking externally from my Doepfer LFOs.

Now, I decided to hook up a DC socket to run the sequencer off a power supply (9V) and now I'm getting some strange behaviour from the external clock. It kind of is clocking, but it's very sporadic and not following the Doepfer LFOs properly.

As a test I dropped the voltage on the power supply to 7.5V and it works perfectly (this would explain why it was working on a weaker battery).

Shall I just divider the incoming voltage down to 7.5V or shall I do something in the circuit itself/incoming external voltage? Strange it doesn't mention this is the original schematic and suggests such a high voltage?

Any help will be much appreciated.

Cheers!

[Inventor]

Is the input clock varying from 0V to Vcc? Does it go negative? If so, that's not a good thing. I'd imagine that your Doepfer LFO's are providing full scale square waves that go both below ground and above Vcc. Both are not recommended. You need to determine signal levels when interfacing circuits. If I am off-base here, my apologies, but that is at least one thing to consider.

Another possibility is that the signal level from the LFO module is not high enough in amplitude. The clock threshold of a CMOS chip is typically half of the supply voltage but it can vary. If lowering the supply voltage causes correct clocking and raising it causes intermittent clocking, then perhaps your input signal level is not high enough. For example you might be putting a 0V-to-5V signal into a circuit that has a 5V threshold, causing intermittent clocking.

Hope that helps.

Les

[ezekiel]

Inventor's suggestion are probably correct.

The Doepfer A-145 LFO output varies between -2.5 and +2.5 volts according to its spec sheet. You can't feed those negative voltages (below the "ground" voltage) into these CMOS chips--they don't really handle that.

And, +2.5 volts is certainly not a very high max voltage input for a CMOS Schmitt chip running at 9 volts. At 9 volts, you really need the input signal to go below 0.5 volts and above 8 volts to be sure it works. At 7.5 volts, you need the input signal to go below 0.7 volts and above 6.5 volts. These are rough numbers; see the 40106 spec sheet for the specific voltage ranges for the low threshold and high threshold; see the graphs in there. If you don't meet these criteria, you will be operating in the Land of Gambling where you cannot be sure what the chip will do.

Very important also: Be sure that the Doepfer module and your Lunetta have a common ground. The Doepfer ground is on the ring of the 3.5mm jack and plug.

If you really want to feed waveforms from the Doepfer to your CMOS lunetta, you should wire up the standard comparator-based circuit so that you keep the CMOS happy with signals only between 0 and Vcc--and make sure that square waves fed in actually swing all the way from 0 to full Vcc.

[J2600]

Thanks so much, guys.

I think both things that Inventor mentioned are the problem here.

I tried boosting the Doepfer signal and (BANG!)... it works. So I was gonna build in a little amp circuit for the input as well as the comparator mentioned to make sure only positive voltages get in.

As you can probably tell I don't fully know what I'm doing and just working it out as I go along... so much appreciated for the advice.

Do you guys have the simple comparator schematic you mentioned? Just thought I'd ask...

Thanks again.

[ezekiel]

To use a comparator to square off inputs at the correct voltage, here is a good page:
http://www.musicfromouterspace.com/analogsynth/oddsandends.html
See "Gate Buffer With Comparator Input" for a full-blown circuit. But "12V to 5V Gate Reducer" might be close to what you need.

As an example, this page shows a schematic using this approach and explains it with oscilloscope diagrams:
http://yusynth.net/Modular/EN/COMPARATORS/index.html

For more DIY synth connections to manufactured modular synths, see also:
http://www.doepfer.de/DIY/a100_diy.htm

[J2600]

Thanks again for all the help.

I was looking at the "12V to 5V Gate Reducer" circuit, I'm gonna try and add this the sequencer... But obviously I have to boost the signal as opposed to reduce.

One more amatuerish question... the schematic says the TL082 (pin 8 and 4) are powered by +12V and -12V. Can I power the TL082 with just the 9V (same power supply for the sequencer chips)?

Also, I'm waiting on some TL082s, could I achieve a similar thing with a LM386 amplifier chip combined with the 1N914/final section of the "12V to 5V Gate Reducer"?

Again, I really appreciate the help!

[ezekiel]

[quote]Can I power the TL082 with just the 9V[/quote]

No. Substitute an LF353 (or LM324 or 741) op amp to run with just 9v and ground. You might have to experiment with some components values.

Someone more knowledgeable needs to jump in and point to a really appropriate circuit that is exactly what you need. Anyone?

[Inventor]

Well I don't want to oppose what has already been said about clocking levels, but for a simple easy circuit, just use a cap and two resistors. Use a series capacitor of about 10uF with - connected to the signal input and + connected to the middle of a voltage divider formed by two resistors. Connect the top resistor to Vdd and the bottom to ground. Use 10k resistors or similar.

How it works is the cap floats the input value and the divider repositions it's level to Vdd/2. Now, you won't be fully driving the input but it should be sufficient for CMOS. Give it a try, it's only three components, should be ok.

Les

[DGTom]

[J2600]

Sorry for the late reply. Been away...

Thanks again for the help... I'll give all these suggestions a try today and report back.

You've all been really helpful.

Cheers