Let's Make Robots!

No need for oscilloscope, nut cracked, bright debugging ahead!


I do not have an oscilloscope, and I "always" wanted one. "So I could debug & understand & see what was going on".

Yesterday I got so frustrated that my cheap RF still was not 100% rock solid - even though I had made all sorts of initializing strings.

i got so desperate that I decided to buy a used oscilloscope. While checking out what was on sale in my area, I got the idea that perhaps I could make one myself..

That got me to look into .. "perhaps someone out there made a SW oscilloscope - how hard could that be?

And there ARE SW oscilloscopes, and they have demos!

Great, I thought. They work like this: You have minimum 2 channals / rays. Each is simply represented bya stereo channel on your soundcard's input.

.. "great" ..hmm.. yes.. well.. I realized that an oscilloscope is really not that great for my kind of debugging. They only show the present pulses, but what if your problem is little fast glimses?

Really annoyed, i got to think of an old HW music sampler on my desk.

"Heck", I thought, let's sample the outputs then..


This way I could let the robots do their buggy business, record it, and rewind, zoom, watch what's going on.. and the best of it all at the very moment: IT CRACKED THE CHEAP RF-NUT!

Now my only problem is to find time to make a real functional walkthrough on the cheap RF's, and similar stuff..


But just for now, check this blurry crap image, taken with my phone, at the heureka-moment, when the nut was cracked, and I made cheap RF rock solid with minimum of wasted bits etc:


The interesting part is the middle section, and this is what it shows:

Top: What is recieved from the RF reciever
Bottom: What is send by the RF sender

- And from left to right:

First the R is creating Noise (what makes it do this, I also know now, but more about that in another post)

Then, The S is sending a High and  Low, just the right length not to be recognized as serial input. This stops the noise, and serial TR/RX is going on, without problems, all bits are recieved.

Finally, after 2 bits are send and recieved without problems, the R-side responds with a sound, resulting in a wave on that part.

Year, baby!

A soundcard can easily handle the logic 5V as in, and you have 2 channels on your soundcard. It is that easy - you can compare analoge current with alternating current, see what PWM does to your motor, see how a magnet creates a spike somewhere else etc!!

Cool :D 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
THis is brilliant, Frits - howcome you never thought of it before?

I know!

And I don't know why I never thought of it before - with all that playing around with pulsating lights, LDR's, phototransistors, magnetic pulses and what have you!

Wow you need help! You are talking to yourself about yourself? I count at least 3 different personalities!
Fritsl aka Sybil!

The discovery / concept is brilliant, and nobody else sees it, so I had to comment it myself.

All geniuses talk to themselves!

Boy, you are so right!
Where's the picture?

huh? Wtf happened to my blurry picture? I re-inserted.. strange?!?

Anyway, more pictures to come, crappy documentation to thankless nerds takes time, have only like 15 minutes of robo-time on an average day :P


Walkthrough on it's way.

Image: Top is what is recieved on RF, buttom is what's send. Note how I now manage to cancel out noise before sending, giving rock solid transmission :)