Hi there,

I'm trying to repurpose a number of old video cameras that use the old 10 pin connection. One of the pins on the connector is the remote control or trigger pin to start and stop recording. This was used when the operator pressed the record button to make a constant current flow which would start the VTR recording - when pressed again the circuit was cut and the VTR would stop recording. Very simple and universally used across many recording devices (tape recorders, dictatphones, cine cameras) at the time.

I am using a drone FPV recorder to connect with these old cameras as these are tiny devices that capture composite video to micro SD card - a lot lighter than the old VTRs! They come with breakout cable to operate each of the buttons, one of which is of course the record button. This is a momentary start/stop switch and what I want to do is hook it up to the trigger cable on the camera to start and stop recording on the FPV recorder - probably via a reed relay. The issue is that the trigger is a constant current and I just need a momentary pulse. I looked into something using a 555 timer (or two) in monostable mode as per this instruction http://www.circuitbasics.com/555-timer-basics-monostable-mode/. But I hadn't realised that the pulse would be continuous if the current was constant.

I am therefore a bit stuck and looking for a solution, and would welcome any suggestions for how to do this. I just need a single momentary pulse output when the voltage in the trigger circuit goes high and low. I'm sure it's simple enough but I am a newbie when it comes to this kind of thing - simple circuitry is ok, but ICs are a new level - so please forgive the inevitable stupid questions along the way.

Thanks very much.

 

What is the current level you need for the pulse ?

Regards, Dana.

 

I just need a single momentary pulse output when the voltage in the trigger circuit goes high and low.

Do you need a pulse when the trigger circuit goes high and one when the trigger circuit goes low, or only when the voltage goes high?

 

Hi guys,

Thanks for taking the time to respond. Please see my answers below:

1) The pulse needs only to be a few milliamps, enough to activate a reed relay - say 25ma

2) I need a pulse for each, going high and going low - record start / record stop. This may require 2 ICs I realise which is fine.

I have seen similar threads on here which is encouraging but on each the circuit diagram images were no longer showing.

Thanks for your help.

 

2) I need a pulse for each, going high and going low - record start / record stop. This may require 2 ICs I realise which is fine.

How will the circuit receiving the pulse know what the pulse is for (start or stop)?
In other words, how will you sync the camera with the state of the recorder.

To generate a pulse for either direction of the record signal, you can use an exclusive-or gate one-shot circuit.
What is the signal voltage, and what power do you have available?

 

Hi,

Both the camera and the recorder are powered by the same source and will power on and switch off together. As the recorder has the same action to start /stop there is a possibility that they could become de-synced in specific circumstances, but there will be a record button on the recorder that can be used to just on the recorder to put them back in sync.

The signal voltage will be 6-8.5 volts - it depends on the camera, and the voltage of the switch on the recorder is 3.3v - hence I am thinking of using a relay. Can you explain more about the 'exclusive-or gate one-shot circuit'. Newbie here I'm afraid. I don't need anything to complicated.

Thanks again.

 

The LTspice simulation of an example XOR pulse circuit is shown below.
It requires a power source of 3V to 15V to operate (which could be a battery as the circuit draws only a very small leakage current).
What source might you have available at what voltage?

 

Exercise caution using slow risetime signals into CMOS logic -



http://www.ti.com/lit/an/scba004d/scba004d.pdf

http://www.ti.com/lit/wp/slla364a/slla364a.pdf


Regards, Dana.

 

Is your trigger the existence of a constant current or the pulse generated is a constant
current ? If former what is the current ?

Regards, Dana.

 

Exercise caution using slow risetime signals into CMOS logic -

That can be a concern, however the waveforms you posted are for high speed CMOS logic, such as the 74HCxx series, and generally this transition oscillation is less of a problem with the slow speed CD4000 logic.
But if it is, a Schmitt-trigger inverter, such as the CD40106, can be added at each input of the XOR gate.

 

The issue is having an input spending excessive time at input threshold
in a noisy environment. So I would posit CD4000 has same issue. Because
of slower speed on 4000 I would expect oscillations to be of a lower freq
however. Anytime we have a G element with its input sitting at threshold
for a time >> Tpd I would expect problems.

Regards, Dana.

 

Here is a sim of CD4000 with 50 mV of noise input referred. Even 5 mV
of noise a problem.




Regards, Dana.

 

The XOR street, from post #7, can also be implemented by four NAND gates with schmitt trigger, such as CD4093 or HEF4093.

 

a Schmitt-trigger inverter, such as the CD40106, can be added at each input of the XOR gate.

Or, use one of the other 4070 gates as a non-inverting stage and add two resistors for hysteresis. Saves adding another chip and chip type to the circuit.

ak

 

Okay, you convinced me.
Here's the circuit using four Schmitt trigger NAND gates (one IC package) configured as an XOR gate.

 

The CMOS gate approach has to consider the very poor PW inaccuracy due to
C1 tolerance and CD4093 wide threshold inaccuracy and T and V effects.

Might want to have R1 trimmable....?

If accuracy is an issue many UP's have 1 - 5% onboard clocks, that being the prime accuracy
determinator. That's over T and V. And power up/power down glitch free operation.


Regards, Dana.

 

Hi Guys,

Been following the back and forth on this, and trying to follow what I could. So it seems the CD4093B is the key component here. I have looked it up and it seems suitably complicated for my abilities. Could one of you help me out with a more practical circuit diagram. I can follow the diagramatic ones to an extent but in terms of connecting up the IC I'm not sure which pin is which. I assume the output from the NAND gate is one of the outputs on the IC, but not sure what the connections to the NAND gates on the left refer to.

Again, stupid newbie stuff I know but I am trying to treat this as a learning experience.

Thanks again.

 

Here is datasheet with pinout. In this logic circuit crutschow posted it does
not matter what gate in the part/package is in what position in his simulation
schematic. It would matter if you are laying out a PCB and want a "clean"
looking layout, but a proto of 4 gates wiring up by hand not so much.

http://www.ti.com/lit/ds/symlink/cd4093b.pdf

So just draw out his schematic, and pick gates one at a time in part and label
their pin numbers on schematic. Or maybe he can post the file and you add
pin labels in the LTC Spice simulator he used.

Regards, Dana.

 

The 4093 is called a quad gate because it has four identical and interchangeable sections. The post #15 schematic shows them as U1, U2, U3, and U4, but a better designation is U1A, U1B, U1C, and U1D because they are different internal parts on a single IC (U1).

Each section is an identical 2-input NAND gate. Each gate has two inputs (traditionally on the left) and one output. The extended D shape is the standard symbol for an AND gate. The output has a bubble on it to indicate that the signal is inverted with respect to the AND function. The gate is a NOT-AND, or NAND. The 4093 datasheet shows the four gates and the pins associated with each one. For each gate, the two inputs are completely interchangeable. You can assign the gates and pin numbers to the schematic in many different ways, all of them identical in performance. This gives you the freedom to assign gates and pins in a way that reduces the wiring complexity.

Take a swing at it, post a schematic with pins numbers, and we will advise.

ak

 

Thanks @AnalogKid.

I did realise that the 4093 was a multiple gate IC once I looked closer. So attached is a rough diagram of what I think would be right. I hope you can follow it. Any questions please ask, and please do correct me where I'm wrong.

A few of queries:
1) Can you explain, where the VDD and Ground would connect up, if anywhere
2) Does the capacitor just go to ground as I have it here
3) If I run the pulse through the reed relay and then to ground that should to the momentary switch operation right?
4) What would the voltage of the pulse be?

Thanks for your help again.