1
\$\begingroup\$

Maybe I'm biting off more than I can chew, but let's see what my options are.

I'm making a pachinko type game. It will use bottle caps for the dropped object (can switch to 1.25" washer for weight if that helps). Every time it hits a peg, I want an LED on the end of the peg to flash. Pegs will be #8 machine bolts, and any wiring will have to go under the bolts or behind the backboard.

My background in this is that I aced a DC electronics class 30 years ago, but can learn new stuff pretty quickly.

What kind of sensor am I looking for?

Any help on how to do this or to warn me that this is too difficult would be appreciated.

\$\endgroup\$
7
  • \$\begingroup\$ How many pegs?? \$\endgroup\$ – Leon Heller Jun 16 '18 at 19:03
  • \$\begingroup\$ What's your budget limit in time and cost vs desired duration of indicator? Judging by choice of budget caps, this may be more than you can chew. \$\endgroup\$ – Tony Stewart EE75 Jun 16 '18 at 20:35
  • \$\begingroup\$ maybe an optical sensor placed on either side of each of the pegs may work \$\endgroup\$ – jsotola Jun 16 '18 at 22:58
  • \$\begingroup\$ another choice is to use a metal detector on either side of each peg ..... instructables.com/id/Simple-Arduino-Metal-Detector \$\endgroup\$ – jsotola Jun 16 '18 at 22:59
  • \$\begingroup\$ Sorry. The 1.25" refers to washers instead of bottle caps. Updated the original question. \$\endgroup\$ – Tom Collins Jun 16 '18 at 23:03
2
\$\begingroup\$

Can you make the backplate conductive and isolate the bolts? That would make it a simple continuity test.

enter image description here

Figure 1. A sensitive touch switch could be made as shown. Electrical contact between the washer and the screw would indicate contact.

A metal back-board would eliminate a considerable amount of wiring and assembly as only one wire would connect all the "washers".

\$\endgroup\$
6
  • \$\begingroup\$ It could also improve rebound with the right matched spring response time and spring time constant of cap. But it could also lead to oxididation from insufficient wetting current where only gold plated contacts are reliable. Ability to adjust the spring constant , k and contact gap, x is a mechanical challenge for impact for F>=k/x. Ultimately alignment of the peg to the back-board would be a constant problem. The smaller the neoprene ring, the better this aligment can be. But an entire rubber sheet would be impossible to adjust. \$\endgroup\$ – Tony Stewart EE75 Jun 16 '18 at 20:15
  • \$\begingroup\$ Of course wire tension cannot interfere with moving pin force on the attached wire. \$\endgroup\$ – Tony Stewart EE75 Jun 16 '18 at 20:22
  • \$\begingroup\$ The LED pulse would be so short that it would be hard to visualize the entire path better than just watching the path, unless the time is stretched for the duration of the fall. Thus an electronic solution only can make that worthwhile. \$\endgroup\$ – Tony Stewart EE75 Jun 16 '18 at 20:32
  • \$\begingroup\$ The question is only for a sensor. \$\endgroup\$ – Transistor Jun 16 '18 at 20:33
  • \$\begingroup\$ The sensor also must be reliable and effective for a visible length pulse (TBD) Since there are no specs, any answer can be wrong. \$\endgroup\$ – Tony Stewart EE75 Jun 16 '18 at 20:38
2
\$\begingroup\$

What kind of sensor am I looking for?

I think your best bet is piezo sensor to which a peg is attached. Unlike mechanical switch it would not change "bouncing" property of the peg much. And it should not have much cross-talk with other pegs if amplification gain is right. You can drill a hole in non-sensing part of the sensor and mount a bolt with a nut. Or, even better, solder a piece of brass tube to it, which would give you a conduit for LED wire (the second is the tube itself).

Interfacing the piezo sensor is simple and it will give you a pulse than can be stretched to light up the LED.

UPDATE

Here is alternative idea that might be somewhat cheaper to implement. For this you need long bolts mounted on a back panel and coming forward through slightly larger holes allowing for a tiny bit of bending. You glue strong magnets to the bolts right behind front panel and put coils around them.

From here the idea is clear - you amplify the coil current and set a threshold for strike detection.

In either case I must agree with @tony-stewart that giving the amount of pegs in the typical game it won't be cheap no matter which way you go.

UPDATE 2

Here's another one :)

There are digital accelerometer modules available under a buck. You'd need the chips with lowest possible range (something like 0.5g). Use same setup as above but glue these modules to the pegs instead of magnets. You don't need to read them, just program "interrupt on tap" feature. This will cover all the analog circuitry required in other solutions and you'll get clean digital signal right away.

\$\endgroup\$
1
  • \$\begingroup\$ Re, cross-talk, my guess is that a high-pass filter would be helpful there. Depending on what the backboard is made of, it may attenuate higher frequencies more than lower frequencies. \$\endgroup\$ – Solomon Slow Jun 17 '18 at 15:05
1
\$\begingroup\$

Can you define how long you want to stretch pulse for LED? A mic on end of bolt can work with an ~? kHz high pass filter to reject backplane rumble and some crosstalk of adjacent bolts.

Extending the back side bolt length will amplify the sound depending on board stiffness. Since the backplane conducts all sounds , discrimination of the strongest may become critical.

Then a pulse stretcher from diode peak detected signals into CMOS inverter with slow RC decay to DC bias level drives each cathode low from the Schmitt inverter using 74HC series 5V with 10mA and ultra Brightled series R drop V.

Tuning the threshold could be select on test R divider’s on mic pullup to get the optimum DC bias and AC gain the trip the Schmitt trigger inputs.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.