Audio Sequencer Circuit

Thread Starter

DanRilley

Joined Jan 13, 2008
107
Hi all, I've been asking questions here on and off, but I'm finally undergoing a larger project combining a few things and figured I could use some help in the execution.

The circuit is a 10 step sequencer, on each step the circuit closes the switch to an audio play button on a small lo-fi digital recorder and the sound is played. It is a basic concept but I will also be adding in functionality to record to each recorder through one input and then have all signals mixed out through one output. The user can select from 3 timing options, 1) the internal 555 / tempo pot, 2) an external clock pulse, 3) a manual pulse button.

Right now I am working on the schematic that deals with the sequencing and playing of the sounds (the mixing of the output I will tackle once this is complete). It is almost a copy of the LED chaser circuit, but instead of just turning on the LED it uses an NPN transistor to close the switch of the audio circuit as well. There is also a manual switch so one can play each sample step by pressing a button. Furthermore the ten steps can be cut short to any number of beats via a switch at each step which sends a signal to the 4017 counter to reset, allowing for different timing bars.

I am posting a .pdf here I made out of LTSpice. The switch diagrams exported kind of weird with the text going vertical so look out for that and let me know if you see any problems that might crop up. Each aspect of the circuit I have tested independently, however I'm just not sure if they're all going to hook together as I intend.

The one thing I know is kind of janky is the Manual Pulse Generator, it's basically just a capacitor that charges up and when you hit a switch it sends a low-high charge over to the 4017, it seems to kind of work, but I know there must be a better way to do that.

I'll keep posting updates as I proceed. In the physical world, I'm still looking for a way to mount these tiny PC boards from the recorders onto a larger prototyping board. They have tiny screwholes but I need some kind of tiny L-bracket that I can then solder into the proto board. I just haven't found anywhere that sells miniature L-brackets oddly enough. Too small for home depot!


Oh yeah on the schematic, for the audio play circuits U1-10 are manual spst switches. D1-D10 are LEds U11-U20 are the reset switches.
 

Attachments

SgtWookie

Joined Jul 17, 2007
22,230
you're going to have pretty limited current available from the 4017 outputs, even in a b series - perhaps 5 ma if you're running them really hard. your leds will cause the output voltage to be clamped at their level, since you have no current limit resistors for them.

you also have u11-u21 - what do those represent.
 

Thread Starter

DanRilley

Joined Jan 13, 2008
107
Yeah I'm not sure exactly how much current it takes to close that NPN switch, but when I tested it worked, so it should work across all hopefully.

U11-U22 are spst switches all hooked to the 4017's reset pin. So that, when closed, the counter will reset at that particular time-step. The only thing is that if they're all open then the last switch, U21, must be closed to keep it grounded (it's kind of janky but I'm keeping it simple so it has a higher chance of working). This way I can have the counter reset every 4,5,6,etc. beats depending where the switch is closed.

Any suggestions about making a more professional Manual Pulse Generator?
 

nomurphy

Joined Aug 8, 2005
567
The house brand for Sears electronics was named Silvertone (Sears Silvertone), and their schematics would use a single line for ground that wound everywhere around the schematic. It was very difficult to keep track of that, and your schematic falls into that category. So, with apologies to Frank Zappa, "...is that a real schematic, or is that a Sears schematic?"

I suggest you use a ground symbol at the negative 9V battery terminal, and then at any component that connects accordingly.

I don't think R2 is a very good idea, it means that all the xstrs become voltage followers instead of open-collector switches. Whenever a "-PLAYx" line is supposed to be at GND potential, it will actually be at something else, based on the emitter current times R2 (plus the LED drop across the same resistor). This would be more evident if you used a ground symbol.

Sarge is right about the LEDs -- I suggest you use low power LEDs designed for 2mA forward current, place a 3.3K resistor between each LED cathode and "GND".
 

SgtWookie

Joined Jul 17, 2007
22,230
nomurphy is right about the 'run-around ground' - it makes your schematic very hard to follow. i see that you have used three grounds, but the battery isn't connected to ground anywere, except via their vcc points.

you should always try to make your ground symbols 'point' downwards. if they're pointing up, they can be confused with v+/vcc/vdd symbols. use them liberally, it will eliminate a lot of visual clutter.

i missed the r2 in the battery negative supply. as nomurphy mentioned, that's not good.
 

Thread Starter

DanRilley

Joined Jan 13, 2008
107
Great thanks for the tips. I'll update the ground as you mentioned, I think I was thinking of how I would actually wire the thing up, rather than good schematic practices, so I'll fix that asap. I pretty much grabbed R2 straight from an LED chaser circuit. Should I completely remove it and then place 3.3K resistors at each LED cathode?
 

SgtWookie

Joined Jul 17, 2007
22,230
If you already have some super-bright red LEDs and you don't really want to order the low current variety, you could go ahead and use them - just use limiting resistors so that they don't get more than about 3ma each. That would mean using resistors of around 2.3k. You'll be able to see them; they just won't be very bright.
 

nomurphy

Joined Aug 8, 2005
567
Some schematic tips:

1) Inputs on left, outputs on right. Preferably, schematic should read left/right as when reading a book.

2) Consistantly use connection dots on nets that interconnect, do not create four-way ties/connections, nets that cross without dot are not connected (no need for humps or hoops).

3) Use ground symbols liberally, do not thread ground net around schematic, this can also apply to power signals such as Vcc or Vdd.

4) Terminate open or off-page nets with appropriate input/output arrow.

5) Label any/all "no-connect" pins/nets as such (I often use the refdes an pin #, such as U7NC1).

6a) Use net names, and ones that are explanatory that help understand functions.
6b) Net names should not be ambiguous, do not use simply ~SEL (what are you selecting when signal is low?), use ~SELA or SELB, or ~SELVIDOUT.
6c) Always show voltage/power (Vcc, Vdd, Vss) on each component, either explicitely by net connection, or implied by turning on symbol attribute(s).

7) If you create or have control over symbols:
7a) Symbols do not need to look like the IC, place inputs on left and outputs on right (bi-directional pins, such as a data bus, can be either side and help balance symbol).
7b) Preferably, place clocks & control lines (CLK, RESET, EN, CS, etc.) at upper portion of symbol.
7c) Indicate active low pins with bubbles on pins, and overbars on pin name.
7d) Place pins only on sides of symbol, try not to place pins coming from top or bottom of symbol.
 

Thread Starter

DanRilley

Joined Jan 13, 2008
107
Hey didn't get to see your post til after I already redid the layout, so everything's going right to left! I think I just end up doing that because I'm lefty but who knows. Again, I really appreciate all the tips, having never taken a class. I'm not even fully sure what net names are, but what I gather from google, they are just labels for each part of the diagram e.g GND, POWER? Attached is a schematic with the last changes you mentioned. I think I'll start just putting it together on a breadboard to see if it's gonna work out.
 

Attachments

SgtWookie

Joined Jul 17, 2007
22,230
Oops!

Move all of the 3.3K resistors so that they are only in series with the LEDs.

Example: below the 555, you have D6. A ways below is R31. However, R31 is also the ground path for R14, Q6, and U7!

R31 should ONLY be limiting current through D6. You can place R31 either immediately above D6, or immediately below it. My personal preference is immediately above, but you can do it either way.
 

nomurphy

Joined Aug 8, 2005
567
See redlines in attached, change all "PLAY" outputs like this.

Also, I think your manual switch should be rearranged, otherwise the input to the 4017 is floating. Even so, this circuit may not work as you intend, but try it first before going on to something more involved.
 

Attachments

Thread Starter

DanRilley

Joined Jan 13, 2008
107
I see what you mean by 'floating' where the 4017 might pick up noise and fire by accident? The reason I put it like this was because I thought the capacitor would charge up since theres a full circuit from Vcc to the capacitor then to GND, then when the switch of the capacitor gets changed from GND to the 4017 it would fire off what it had charged up. I don't think that method really works perfectly (but its all i could easily come up with), but will it still have that effect if I rewire as you had diagramed?
 

nomurphy

Joined Aug 8, 2005
567
There are a number of issues with this switch circuit that I wasn't going to get bogged down in, because of the lack of design information.

Wire it the way I've shown it, but place a 10K resistor from the cap/switch to ground. Also, place a 10K resistor from the 4017 pin-14 to ground.

See if that works for you.
 

JoeJester

Joined Apr 26, 2005
4,390
What frequency is your 555 timer operating? I noticed a missing wire between threshhold and trigger, or is the intention to have the 555 to act as a retriggerable monostable?

What is the maximum load connected to the +play and -play connections?
 

Thread Starter

DanRilley

Joined Jan 13, 2008
107
Here is an updated schematic that I cleaned up a bit and added your changes in nomurphy. Yeah that manual pulse switch may not end up working, however, just to follow what you were saying, did I place R5 right to ground? It seems like the flow of electrons would all just get pulled to ground and none would go through the switch, but then again I have almost no experience so just let me know if I drew that right.

As for JoeJester, I took the 555 wiring straight from an LED chaser circuit once I test it I'll let you know if it fails or not, so I couldn't really tell yo, but I'll keep an eye out. Also the load from the audio circuit I'm not sure about either, however the other circuit operates from it's own power supply, this is just closing a switch on it. I'm not sure if I need to use diodes to keep voltages from passing back and forth between the two circuits but when I tested one it didn't seem to affect it since they're very small circuits.
 

Attachments

nomurphy

Joined Aug 8, 2005
567
Wow, that's a lot better drawing.

Sorry, but remove the 10K from the switch. The normally closed position should be at ground. When you push the switch to the cap, you will get a pulse. But, you may have to hold the button.

You might be better off just getting rid of the cap, but you can always try variations and see if things work for you or don't.

And, you may find switch bounce to be an issue.
 

Thread Starter

DanRilley

Joined Jan 13, 2008
107
Hey guys, so I started building the circuit. It worked great on the proto board going into just one LED and switch circuit. However, I have now started wiring it onto a perf board, connecting the leads by wire and a problem has cropped up.

I have connected 5 of the 10 LED/switch circuits as well as the 4017 and the 555 connections. I'm skipping the other clock options for now obviously til I get this working. The problem is that it works fine for about 2 or 3 revolutions of the decade counter but that it slows down/freaks out and get's caught on the first step and won't move. It's as if the reset pin was getting a voltage, but I have zoomed in on the reset pin with the oscilloscope and it's grounded very well.

Does anyone have any suggestions? I soldered everything pretty messy, but I double checked that no leads were touching. It's weird I didn't have the problem at all with just one switch circuit hooked up, could it be that the 5 circuits are drawing too much? Note: the 555 continues to work flawlessly and delivers the pulse after the 4017 has crapped out. Also I never connected R5 (the 10K resistor from Clock to ground (since I hadn't done the manual pulse generator and it was related to that), do you think that would help?
 
Top