Relay timings characteriser project

[matseng]

Usually I make my projects with all parts on a pcb designed by myself, but this time I wanted to try it "new-maker" -style since I'm not at home.  So I mail ordered some of the common cheap modules from local retailers and dug into the resources of the local Makerspace here in Penang and came up with this Franken-contraption.

It actually works despite its ugly looks - and this is not too bad compared to how it looked before I tucked all the modules down to a piece acrylics. :-)

On the board I've got an Arduino Nano, a SD card reader module, a Nokia 5110 LCD module, a 35 volt boost converter module, a LED Voltage display module and the board on the right that contains some linear voltage regulators, a MOSFET to driving the relay and two buttons and a rotary encoder.

All of this makes a relay tester. It measures the energizing and releasing timings and also how long the bounce times are. The info is displayed on the LCD in both numbers (milliseconds) and can be stored onto a SD card if doing longterm measurements of millions of operations to see if the parameters are changing over time.

The output voltage for the relay can be adjusted between 1.2 and 30 volts to measure the changes caused by over- or undervolting it.

The first number on each line is the time it took from applying power until any change in the relay was detected. The second number is when the relay is stable after the bounces.

[MattHollands]

Looks ...interesting While I support "maker-style" electronics I always feel a little ill when I see a bunch of cheap modules wired together. I think I just like the art of a nice PCB.

Have you tried driving a relay at 35V? Just back of the envelope, if you're driving the relay coil at 100mA 35V, then there will be over 700mA being drawn into the boost converter which is more than many PCs will allow on their USB port.

[matseng]

Yea, I really hadn't much of a choice than this "crapfest" since I'm on the road and just temporary visiting this city.  It was basically just a quick hack to have something to do.  But it's kinda cool that it's possible to just cobble together something more or less useful out of cheap Chinese modules, some of which has a horrible design like the placement of the caps for the SD card module. What the f*ck were they thinking when they did this? An also using MLCCs for a 1117 vreg?



I for sure hope that a 35 volt relay doesn't pull 100 mA. The selfheating of a beast like that would destroy it unless mounted on a heatsink... It would be a contactor rather than a relay then. :-)   But any ipad charger can supply you with 10 watts of power it really won't be a problem.

But my target is small signalling relays - like the ones you'd use for building a relay computer with - not monster relays.

Usually my designs looks a bit nicer like this:


or

[Andrey_irk]

Hi!
I'm designing something like that at work. And and these "timings" are just a part of that project.
There are a lot of things I have to measure. For example, did you know, that  by observing a shape of the coil current when relay switches (on and off) you can say whether the relay is good or not?

Are you going to share your progress with us?

[Mechatrommer]

if you put all of them on top of a one big perf board using dip pins socket solder connection, that can be neater, stronger and easily replacable. i only concern about where you plug USB cable to, one day that can be out of place...

[matseng]

Yea, a current sense resistor and an opamp for getting the signal into range for the ADC was in my original plans, but when throwing this together at the makerspace it didn't make it.  For the next revision it will though. As will a logic level MOSFET to get rid of the 7812 and also having the microcontroller set the voltage of the smps so I can drop the 317 as well.

The current block diagram of the unit today looks like this:



Sure thing. I'll keep this thread updated with the progress and upload the available resources to github.

Yes. I guess having a 100x160mm perfboard with solderd down pinheadwrs from the mtidules would have been a gazillion times neater. Will do that the next time I'll use the "messy nu-maker" style again. I'm not really proud of this project, but it works at least. 

I too was a bit afraid that the plugging and unplugging of the USB connector on the arduino nano would sooner or later break some connection's, so I added the red top left pcb which is an old dismantled ftdi board that I glued down to the acrylics and is using that just for power. But during development I still need to use the nano connector.