CMOS input protection possible troubles ......

[Rykhaard]

I've got at least 1 puzzle here, in working on my 2nd panel for my Lunetta machine.

I finished hooking up 95% of everything for the 1st of my 2 x 4 bit shift registers, within the CD4015.

For looping purposes - I.E. feedback from one of the Q outputs to the Data Input, I've routed Q3, Q4 and the Data Input banana jack, to a 3 position rotary switch, so that I may choose which source I wish to loop from, or take new input from the Data Input.

Almost every single time, when I apply a HIGH to the Data Input as well as the Clock In, Q1, Q2 and Q3 light up steady. ???

Whilst they're still lit, if I switch the rotary switch away from Data, to it's previous selection of Q4, then Q4 BRIEFLY flashes and the other 3 Q LEDs, go out.

I have absolutely no way, to get a bit of data into the 1st Register, for cycling - except for ONE TIME ONLY, when I was able to cycle from Q1, to .... Q3???

Things that I'm wondering about:

- Each data in pin is going through a 1N4148 to the input, which also has a 100k resistor to Ground.

- I checked the voltage of my 9V battery, and it's down to 8V. (In hooking up the panel together with my first panel, I replaced the battery with a new one. No change in any way, using the 40106 oscillators, as the Clock and Data sources ...... though, their minimum frequencies are about 47HZ.)

I have 2 switches set up. 1 for 'BIT' which when switched on, provides a manual HIGH. There is also a status LED for this switch, lighting up when the switch is turned on.
Switching it off, will Ground the input, with the 100k resistor that the pin has normally attached to it.

The 2nd switch is for manual Reset, with an output jack, for it. (Though it's not hooked up; the Reset banana jack IS)

Tomorrow - I'm going to check all of my connections on the miniature circuit board. I'm fairly confident that everything is hooked up properly. Since I've been building with PCBs done from Press N Peel, I've rarely had any failures, almost always having a board work on first try.

1 thought that I had, were that I could have hooked up Q1 to Q3, backwards. I can't logically see that having any troubles.

Another thing that I tried, remembering that the waveforms coming out from the 40106 oscillators peak at about +2.5V from a 9V battery - I wondered if the voltage drop caused by the diodes on the inputs, could have any influence on things?
When I DID get that 1 single operation of being able transfer a bit from 1 Q to another, I had moved the wires from the input jacks, to the OTHER SIDE of the diodes, so that they were both 'out of circuit' having the wires connected directly to the pin of the chip, where the 100k resistor is to Ground.

For now - I'm completely stumped.

Any ideas, anyone? Thanks for any possible help on this one. I'm hoping big time to get this guy up and running as it'll be my first step to minor 'sequencing' from my Deathlehem Machine.

[RF]

About 1/2 volt drop on those diodes, isn't it?
At 2.5, I think you are right AT the threshold between high and low...

Of course - I may be completely wrong

Good luck, Rych

bruce

[Rykhaard]

[RF]

Just a thought - Although I have a bi-polar 12 volt supply for some circuits in my Lunetta - I run the CMOS on 5 volts. I have not had any issues with logic levels.

I think Schmidt triggers are 1/3 and 2/3 supply voltage switching levels - I'm not sure about other CD4xxx CMOS. Anyway - I think it's that 'netherworld' between logic states that's the bugger.


Anyone?

bruce

[Rykhaard]

That's right, Bruce. The 1/3 and 2/3 is for the 4069. Thankee for babbling that bit out at least.

Looking over the circuit, I found a few problems: legs of 2 different transistors (LED drivers) were touching other portions; 3 LED leads were shorted together, causing the mis-lighting across the 2 shift registers.

Now - the 2nd 4 bit shift register is working properly. The 1st one though still has 2 problems:

With the Bit High switch hooked up to the Data In, when the switch is flicked to High, the LED for it will not light.
Logically - the Data Input, through a diode, to a 100k to Ground at the Input pin.

My question now - IS the IC's pin properly connected TO that junction? Nothing else is working on that 1st shift register now, either in that regard.

Oh yeah - ANOTHER problem that I hadn't realized - I'd been using Ken Stone's 22R (or 10R) protection resistor as a 'fuse' on the power input to the PCB. Removing that resistor, took the initial problems away from the PCB! I believe the voltage drop caused by that resistor, may have been part of the voltage level problems that I had mentioned earlier.

As to power for the circuits - I'm going to go full bore with the +/-15V Power One supply, in my Sequencer rack of D.A.M.I.A.N.

Whether I bring it down to 12 or 10 or 6.9V bipolar for each PCB or not, I haven't yet decided. If I drop the voltage down, it'll more than likely be individual Voltage Regulators, per PCB. That might actually be the best way to do it.

Outside of that - I transferred the triple voltage controllable & manual SPDT switches from my old sequencer engine, successfully. It's now up and running and properly functioning, in the Deathlehem Machine.

Panel 3 started today: 8 to 1 multiplexer; 8 bit R-2R DAC and triple 3 input audio mixer with banana and 1/4 outs. Once those are completed, I'll record a new MP3.

Beddy bye time. Zzzzzz.