PLC input circuit

This is a work project (again). Sorry I can't share with you a lot of circuit details but I can describe the global working.
We had an issue with an PLC input circuit. A piece of equipment with a mains power output (230V ~) gives a signal to a PLC in another machine. The two machines are connected by means of a relay. Most of the time this works reasonably well. However, sometimes the PLC sees multiple triggers because of relay bounce.
Five years ago, I developed this circuit it is supposed to function as a electronic signal relay. An issue was that there was no additional power available to power the electronics. The PLC input circuit is standard circuit with a 24V output in series with an unknown resistor. So to test the power available from input of the PLC I checked with several resistors whether they would trigger the input. The result:

Resistance	Trigger (J/N)
470	J
1k	J
4k7	J
10k	N

Conclusion the output of my circuit should look like a 10K resistor (or larger) when not triggered.  This means that the total power consumption should be 24V/10K < 2.4mA....
Unloaded the input of the PLC produces 23.8V, loaded with a 10K resistor the output voltage drops to 23.1V. Using the formula for a voltage divider (V_out = V_in * R₁/ (R₁ + R₂) ), this indicates an output resistance of circa 300 ohms. Also the short circuit current can then be calculated to be around 23.8V /300 ohm is 79mA.


First setup

The circuit is build around the following modules:

1. 230V impulsgenerator and opto-isolator. The purpose of this module is to generate 50Hz pulses as long as 230V mains is applied.
2. Retriggerable Monostale multivribator with a period > 20msec. The input of this block are 50Hz (= 20 msec) pulses. So as long as those pulses are available at the input a steady low pulse is generated at the output. 
3. Differentiator and inverter. This block generates, at the start of the the long output pulse of the MMV, a short positive pulse. The inverter then inverts this, to a short negative pulse.
4. A Non-retriggerable Monostale multivribator with an adjustable period. Triggered by the inverter a pulse is given with a fixed  but adjustable time period. The period is independent of the length of the input pulse and can be configured using a multi turn potmeter.
5. The output stage is switched on by the 2nd MMV en simulates the switching action of the relay by creating a short circuit for the 24V input.
6. The power supply block converts the 24V to a stable and consistent 12V suitable for all components. A Graetz diode bridge is used for easy connection of the input.

The circuit was so successful it ran for over six years without problems. Now the time has come for multiple circuits so a PCB was designed and ordered. Below, you'll see the populated PCB:

]\

Telemecanique beacons

Sometime ago I bought on Marktplaats (a dutch auction site like eBay) six Telemecanique beacons. They're smartly engineered. You can stack up to six different lights on top of each other. They automatically connect with the beacon below by using a build-in connector system. Because every light needs to be turned a little compared to the one below they're all individually addressable. The base has a built in screw terminal with six positions and one common. By applying power between the selected position and the common the corresponding light is turned on.

No drawbacks? Yes a major one! The light bulb in the beacons consumes a whopping 6.5 Watts! At 24V this implies over 300mA! Just for an indicator light...

 This is more than the whole application, for which I bought them, together. Yes, there is a LED replacement.... but the price tag is ridiculous. They cost €51.03 a piece !

DIY? Yes! But I wanted to keep the beacon unchanged. So I needed to upgrade the light bulb. I decided to remove the innards of the bulb and to replace it with a custom made LED bulb. To make it flexibele I wanted to use a current source and not a series resistor.

I've choosen the following current source for two reasons:

1. it is floating

2. the two transistors stabilizes each other

3. and it uses all low costs parts whivh I had readily available

D1-D4 are ordinary 1N4148, NPN=BC547B, PNP=BC557B. For R1 and R2 I've choosen 100 Ohm. gives a current of Ube/R = 0.7V/100 Ohm is 7mA per transistor. In total thus 4mA.

RL in this application is a string of 4 LEDs.

The beauty of this little circuit is the inherent self stabilising of this circuit. The NPN transistor sinks constant current hrough the reference diodes of the PNP transistor, which sources a constant current through the reference diodes of the NPN transistor which sinks a constant current through the ....etc...etc.

The final result:

SMD prototyping PCB

  Frequently I need to do some SMT prototyping. Partly because some components are only available in a SMT enclosure ( like the MCP4725, a nice small 12 bits DAC), partly because it needs to be small and sometimes it is Saturday 5 past 5 and the only type of a critical part available is SMT.
Two weeks ago I saw on SMD shop a SMT prototyping PCB.
It features pads in 0.05"x0.05" square pattern and every fourth pad is connected to the groundplane on the bottomside.
Looks quit nice, with the only drawback a price tag of 8.22€ for 100 X 80 mm board. (yes I am a skeap skate dutchie)

HP 456A Current probe - restoration

This current probe is a real beauty. Released as a new product by the Hewlett-Packard Company in 1960, the 456A was HP’s first solid-state, stand-alone, clip-on current probe. Its elegantly designed amplifier uses two— then “state-of-the art”—PNP germanium transistors. The Original Probe In 1960, The Hewlett-Packard Journal (July-August, Vol. 11) proudly announced: “This new probe measures current over the full range of the frequencies most commonly used in typical work—25~ to 20 megacycles—and over an amplitude range from below 0.5 mA to 1 A rms.

HP456A current probe

Below you'll find the circuit diagram of the HP-456A probe.

Maybe the concept of the circuit is unclear at first but actually the idea behind it is really simple and effective. (Especially considering the fact that this performance was achieved, with what would be considered nowadays, mediocre components at best.)
Q1 functions as the input transistor in common base configuration. As you will know in a CB setup, the base is hold at a fixed voltage (0V in this case), the emitter is used as the input and the collector is used as the output. Main characteristics (and in this advantages) are the low input resistance and the high frequency. The latter is caused by the fact that the base is at a fixed voltage effect and the Miller effect is eliminated. Also because the emitter current equals the collector current, there is no current amplification.
Q2 is the output amplifier and is set up in a common emitter configuration (emitter hold at a fixed voltage).
The combination of the two transistors works as opamp in current to voltage topology (eq. transimpedance configuration.) See below

Rf is implemented as R9 parallel tot R10. Beautiful isn't it?

The reason the setup didn't work in the one I bought is that the cable actual current probe was broken on were it connected to the PCB in the amplifier case. (in the circuit near C2 and C3). After fixing this it works like a charm. Not bad for a 25+ years old device!

I you have any questions let me know!

73 PA3COR