Switch branches/tags
Nothing to show
Find file History
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
..
Failed to load latest commit information.
assets
.catalog_metadata
README.md
tap.fzz

README.md

#285 The Tap

Harking back to a 7400-series logic circuit published in a 1974 Elektor article, this retro Boldport Club project is a classic.

Build

▶️ return to the LEAP Catalog

Notes

The Touch Activated Programmer (TAP) Sensor from Elektor magazine of December 1974 was a very early example of a capacitive touch circuit.

And Boldport Club members were lucky enough to get a "modern" rendering of the same circuit. The PCB design is as psychadelic as you would expect from a 70's vintage design. What is amazing is that the parts are still quite readily available.

The PCB is well up to Boldport standards. Altough this can create some assembly confusion - you must pay close attention to which side of the board the parts are inserted. It looks just as good both ways.

Electrically, touch pads A, B and C are connected to an NPN transistor pair in a Darlington configuration. I've covered the operation of this kind of switch in LEAP#130 DarlingtonTouchSwitch.

All manner of outputs are provided. For each touchpad there are four:

  • Q/Q1¯ - the logic-level output and its complement
  • S/ which are logically identical, but an open-collector configuration able to sink greater current

Behaviour

Action Q1 Q1¯ Q2 Q2¯ Q3 Q3¯
At power-on 1 0 1 0 1 0
Touch A 0 1 1 0 1 0
Touch B 1 0 0 1 1 0
Touch C 1 0 1 0 0 1
Touch RESET 1 0 1 0 1 0

S/S¯ and Q/Q¯ Outputs

S/S¯ and Q/Q¯ outputs follow the same logic behaviour.

It appears the only difference is that Q/Q¯ are mostly suited to digital integration as they are limited in the current they can sink (16mA).

  • Q and Q¯ source current, logic 1 : 0.4mA
  • Q and Q¯ sink current, logic 0 : -16mA

The S/S¯ outputs are controlled with low-side NPN switches and can sink up to 1A - the absolute maximum collector current for the PN2222ATA. Higher currents would need closer attention to heatsinking and wire/connector ratings.

RB (Reset Bar) and CB (Contact Bar)

The RB and CB pins are intended to allow chaining of mulitple sensor units with a common reset/contact bus.

Parts and Unboxing

Ref Item Qty
T1-7, 9-14 bipolar (BJT) single transistor NPN, Fairchild PN2222ATA x13
T8 bipolar (BJT) single transistor PNP Fairchild BC556BTA x1
IC1,2 NAND gate 4 gate, 2 input, TI SN7400 x2
D1-3 Small signal diode 1V Multicomp 1N4148 x3
R1-3 100KΩ resistor, Multicomp MCF 0.25W 100K x3
R4-7 10MΩ resistor, Multicomp MCF 0.25W 10M x4
R8 1KΩ resistor, Multicomp MCF 0.25W 1K x1
R9-15 27KΩ resistor, Multicomp MCF 0.25W 27K x7
C1,2 220pF capacitor, Vishay K221J15C0GF53L2 x2
C3 47nF capacitor, Multicomp MC0805B472K500A5.08MM x1
14-pin IC DIP socket , TE Connectivity 1-2199298-3 x2
20-contacts 2.54 mm header, Multicomp MC34739 x1

kit_unboxing1 kit_unboxing2 kit_parts kit_pcb_rear kit_pcb_front

The transistor markings on the silkscreen threw me at first. They are special! I was confused because they don't replicate the standard NPN/PNP schematic exactly. Just focus on the arrow flow: NPN points out from the emitter; PNP points in to the emitter.

transistor_symbols

Construction

The Boldport and Elektor resources already provide thorough details of the circuit. But as is my way, I just had to redraw the circuit. This time with Fritzing:

Breadboard

Schematic

kit_complete_rear

kit_complete

Running some initial current tests. It never peaks beyond 23mA.

kit_initial_test

Hooked up for monitoring under test (analog scopea and logic analyzer)

kit_scope_test

The final build:

Build

Performance

The first trace I took was from a transition to "A" on (from RESET state). CH1 shows Q1 output, CH2 shows Q1¯ output. As we can see, there's some pre-emptive voltage reduction prior to the switch and some jitter and one significant bounce. From this we can see that using the TAP as an input for a sensitive instrument would still need some debouncing.

scope_A_analog

Here is the same switch with alogic analyser. There's quite a delay (2-3µs) between the analog transition and the digital - but I think that is more the scope than the circuit. The important point is that we still see one bounce at logic levels, and the other outputs are as expected.

  • CH1 - Q1 output (trigger)
  • D0 - Q1 output
  • D1 - Q1¯ output
  • D2 - Q2 output
  • D3 - Q2¯ output
  • D4 - Q3 output
  • D5 - Q3¯ output
  • D6 - unused
  • D7 - unused

scope_A_la

Credits and References