Vanishing Point debugging...

[Arno van der Heijden]

I'm currently building a Vanishing Point (2.1), but I'm having trouble getting the random mode to work. 
I haven't built the filter part yet. Instead I'm first trying to get the sequencer part to work, to make debugging easier.

The sequence mode works fine right now, but when I switch to the random mode, it does something similiar to the sequence mode, only the switching time is not constant for each LED. It goes something like this:


X---X-X---X-X---X-X---X  (etc.)
1   2 3   4 5   6 7   8

Here X is a LED flashing and - is a certain period of time.

What could be wrong here?

[RonaldB]

Hi Arno,
Find your answer here:
http://www.diystompboxes.com/smfforum/index.php?topic=64348.0

Greetz RonaldB

[Arno van der Heijden]

I don't think this is the same problem.

My unit does some sequencing in random mode, it's just not random 

I hope Andrew will chime in...

[The Tone God]

Ok so sequence mode works fine. I'm wondering if the inhibit pin is being driven correctly. Could you hook in the random clock to the clock pin by disconnecting the sequencer clock from pin 14, moving pin 13 to pin 14, then grounding pin 13 ?

Report back.

Also is this a CD4017 ?

Andrew

[Arno van der Heijden]

Hi Andrew,

Thanks for chiming in...

Ok so sequence mode works fine. I'm wondering if the inhibit pin is being driven correctly. Could you hook in the random clock to the clock pin by disconnecting the sequencer clock from pin 14, moving pin 13 to pin 14, then grounding pin 13 ?

Report back.

Hmmm... I hate to say this, but this is almost impossible, given the fact that I built it on pcb.

Also is this a CD4017 ?

yep...

Any other things I could check? I'm checking for layout and soldering issues as we speak...

[Arno van der Heijden]

?

[The Tone God]

Sorry. I've been very busy. Are your IC's socketed ? If not can you desolder one end of parts, say a resistor, to hack in test connections ?

Andrew

[Arno van der Heijden]

Ok so sequence mode works fine. I'm wondering if the inhibit pin is being driven correctly. Could you hook in the random clock to the clock pin by disconnecting the sequencer clock from pin 14, moving pin 13 to pin 14, then grounding pin 13 ?

Ok, with a lot of messing around I was able to pull it off.

The result is that it behaves like it was in sequence mode. The sequence speed changes when the speed control is turned.
If I understand it right, this is the result of the transistor switching the HF signal on and off.

BTW, with the help of an improvised audio probe I was able to verify that the first (modulating) HF clock works fine. I have doubts about the second one. There's certainly some oscillation going on there, but it doesn't sound as loud as the first one (ripple?)... needs some more investigation...

[The Tone God]

Is there a slight "dead" space between steps in random mode compared to sequence mode ?

Andrew

[Arno van der Heijden]

Is there a slight "dead" space between steps in random mode compared to sequence mode ?

Hi Andrew,

Are we talking about the 'test configuration' here, or just normal operation?

BTW, I don't have the opportunity to test this before the weekend, so I'll report back next Saturday or Sunday.

[The Tone God]

Are we talking about the 'test configuration' here, or just normal operation?

Wired as per the schematic. You can control the length of the beat time with the 0.1uF/1M resistor combo that comes off the sequence clock. Increasing either part value will increase the beat length. If it doesn't change then there maybe a problem stage beat feature which would screw up the random mode.

Andrew

[Arno van der Heijden]

Is there a slight "dead" space between steps in random mode compared to sequence mode ?

Hi Andrew,

I just checked. There doesn't seem to be a "dead" space between the steps. Is that good or bad?

Like I said in my first post, every other LED is skipped very quickly (it lights for a very short time) in random mode. When I increase the rate, the difference between the long and short lit LEDS seems to become less.

You can control the length of the beat time with the 0.1uF/1M resistor combo that comes off the sequence clock.

Any ideas which values I should try? It's really a pain to desolder and solder these parts on my tight pcb layout, so I don't wanna have to swap back and forth too often.

[The Tone God]

I just checked. There doesn't seem to be a "dead" space between the steps. Is that good or bad?

Yes this is bad. The stage LEDs are driven by the same signal that should be controlling the random cycle. No dead space means no random "spin" cycle. If you have another LED connect it's anode to the gate output that drives the two transistors (output LED transistor and the inhibit pin transistor). Connect the cathode to ground through a limiting resistor. You should see it blink very shortly after each clock pulse. If you don't then the pulse detector is not working which means the random cycle is not be given a chance to run  the LED quickly.

Any ideas which values I should try? It's really a pain to desolder and solder these parts on my tight pcb layout, so I don't wanna have to swap back and forth too often.

Probably the resistor is easier to chance. You will want to increase the value to increase the time so disconnect one side leaving the other end hanging connected to the board. You can then add resistance in series with the 1M. I usually hook up a large value pot with a couple of alligator clips to tune in the value.

Andrew

[Arno van der Heijden]

Yes this is bad. The stage LEDs are driven by the same signal that should be controlling the random cycle. No dead space means no random "spin" cycle. If you have another LED connect it's anode to the gate output that drives the two transistors (output LED transistor and the inhibit pin transistor). Connect the cathode to ground through a limiting resistor. You should see it blink very shortly after each clock pulse. If you don't then the pulse detector is not working which means the random cycle is not be given a chance to run  the LED quickly.

Hi Andrew,

Thanks again for your help. Now that I understand the reason for your question, I think i've unintentionally misled you here, since I ditched output LED transistor (I have the pots, LEDs and filter parts still on the breadboard) for now. No output LED transistor means that there can't be a dead space.

I did check the output of the mentioned gate with a LED. It doesn't light up on the LEDs that are jumped over. Also, this isn't just a 'short blink', but looks just like one of the stage LEDs that is fully lit. I guess this means that the pulse detector isn't working correctly?

[The Tone God]

Thanks again for your help. Now that I understand the reason for your question, I think i've unintentionally misled you here, since I ditched output LED transistor (I have the pots, LEDs and filter parts still on the breadboard) for now. No output LED transistor means that there can't be a dead space.

Ah hence why I asked the follow up question.

I did check the output of the mentioned gate with a LED. It doesn't light up on the LEDs that are jumped over. Also, this isn't just a 'short blink', but looks just like one of the stage LEDs that is fully lit. I guess this means that the pulse detector isn't working correctly?

Yes so that means both the random cycle is not happening and of course the LED dead space if you had it. Disconnect the transistor that comes from the reset switch so that all that is connection the pulse detector gate is the RC network from the sequencer clock. See if that helps. Next would be to make adjustments to the RC network to increase the pulse time. The LED on the output of the pulse detector should blink quickly and for a short period at the beginning of each clock cycle.

Andrew

[Arno van der Heijden]

Yes so that means both the random cycle is not happening and of course the LED dead space if you had it. Disconnect the transistor that comes from the reset switch so that all that is connection the pulse detector gate is the RC network from the sequencer clock. See if that helps. Next would be to make adjustments to the RC network to increase the pulse time. The LED on the output of the pulse detector should blink quickly and for a short period at the beginning of each clock cycle.

I removed the reset transistor and increased the pulse detector resistor to 2.2M. Both didn't help in any way.

FWIW, the LEDs that were 'jumped' over are now lit somewhat longer than they were before, but still shorter than the others. The output of the gate that follows the pulse detector still doesn't show the expected behaviour when probed with an LED.

[The Tone God]

I removed the reset transistor and increased the pulse detector resistor to 2.2M. Both didn't help in any way.

FWIW, the LEDs that were 'jumped' over are now lit somewhat longer than they were before, but still shorter than the others. The output of the gate that follows the pulse detector still doesn't show the expected behaviour when probed with an LED.

Try connecting the clock directly to the pulse detector and see what the output of the pulse detector is.

Andrew

[Arno van der Heijden]

Hi Andrew,

I just had some time to work on this again.

Try connecting the clock directly to the pulse detector and see what the output of the pulse detector is.

Since it's very hard to do this on my pcb, I set up a CD4093 on my breadboard with the LF clock and pulse detector sections of the schematic. It behaves the same as the pcb version: when I probe the output of the actual pulse detector (cap+resistor) the LED blinks very shortly after each clock pulse. When I probe the output of the gate that follows the wave shaper, the LED is lit much longer.

Isn't this the correct behaviour? The signal coming from the pulse detector (short blink) is inverted and as a result, the LED is off very shortly and lit much longer.

[The Tone God]

Yes that is correct behavour so that is good. Since you have the circuit on the breadboard can you now build the circuit on the board minus the HF clock and modulating clock instead connecting pin 13 through a 10K to posistive ? It should step normally like this.

Andrew