Let's Make Robots!

YUMBBBBMH A.K.A. GhostRoachMouseBullRam CrazyBot

Zooms around searching for light

My iteration at building the famous MouseBot, originally (Mousey, the Junkbot?) created by Randy Sargent.

I present to you YUMBBBBMH (Yet Un-named Mouse Bot Being Built By My Hands) previously known as MouseBot Still To Be Named and now A.K.A GhostRoachMouseBullRam CrazyBot

(...or perhaps just GRMBR CB for short? :P)

February, 12th 2012

As per popular request I hereby also add (as an attachment) the proper pronunciation soundfile of this bot's true name (c'mon this was is way easy). 

October, 11th 2012

So after the tip from birdmun regarding what was on the LM386 datasheet application notes regarding gain, I decided to do some tests with capacitors. At first, seemingly I had not much success, but then again my 9V batteries were low on juice and I got lazy and did not recharged them at the time. Last night finally I've put on charging over night, and today decided to fiddle with the mouse bot again. After, making certain everything was connected, the mouse bot lost it's previously typical crazy spin... and instead turned into a crazy shooting arrow. Problem: It crashed so hard against stuff that pop goes the tiny whiskers. So after some brainstorming with ossipee I went looking  for a wire cloth hanger... and a couple of pliers and a few (considerable) twists later I got myself a couple of cross sturdy whiskers and mounted them on some holes on the microswitchers (after so filling so they would fit). For enhanced (animal) security at the end of each whisker I've deposited a blob/ball of hot glue. Turn it on, and success, now it zooms around like an enraged brutish mouse, without the whiskers falling off. I now consider this project, my second robot project, first to be fully completed. So, the MAJOR trick is to connect pins 1 and 8 of the LM386 through a CAPACITOR and not directly (see altered schematic bellow), otherwise you might end up with a [clockwise|counterclockwise] crazy spinner.


September 9th, 2012

So, I've tracked the uneven voltage problem, basically pin 1 and 8 on the LM386 were tied together to provide maximum gain, by cutting that connection I was able to leverage the voltages, however I've lost noticeable gain. Solution (EDIT: or so I thought): Connect pins 1 and 8 through a resistor. My experiments told me that I needed high (>1 kohms) resistor values to even out the voltage differences, however whenever I increase the resistance I also decreased the sensitivity hence this is going to take some fine tuning later on.

So the schematic should be something like (WRONG...):

the value of the resistor (range of potentiometer) will depend on the LEDs used for optical sensors. I found that normal LEDs and LDRs were unsuited for the task, the best results I got were with IR LEDs.


August 26th, 2012

Since my mousebot CCCW (crazy-counter clockwise) spinner I decided to dig it up again and run a bit more of diagnostics.

I've already measured that the voltages where un-evenly divided between both motors (even when the led "eyes" were blinded -- made a couple of little black hoods for them). 

So, I've decided for a "clean" read. Opened the bot up, disconnected the motors and connected the mainboard with the battery. With the switch turned ON I read:

RIGHT motor vcc PIN: ~5V 

LEFT motor vcc PIN:  (-)3.4V

Turning the trimpot either way changes this readout very little. Previous measures directly done on the motor leads while running also shown me that adjusting the trimpot would either raise or lower the voltages but on BOTH motors at the SAME time.

So my guess, is that I inserted the trimpot in the wrong (INDEED WRONG, IGNORE BELLOW) place/way, but I have no idea why or how to fix it. See schema below.

P.S. - See original (from the internet) schema near the end of this post.

July 21th, 2012

It's a few minutes past midnight and finally the still un-named mouse bot made it's first run. Actually it didn't went anywhere given that a wheel popped of after a second or two. And there's low traction. I'm now going to work on securing the wheels better to the motor pinions. Below you can see the final stages of the assembly!

July 20th, 2012


After a busy "night-ful" I got most of the base work done:

- Mouse entrails completely removed and made way for new "guts"

-- base (side holes for motors, padding for motors, front holes and mounting of bumper switches)

-- top (on/off switch mounted)

- Perf-board soldered all the components on a small stretch of perf-board


- Apply proper connectors to dangling leads;

- Connect everything and test; if(notWorkingProperly){curse(HIGH);}

- Mount/attach the motors and respective wheels

- Close everything up, cross fingers & hope for the best

July 19th, 2012


Got a suitable candidate PS/2 mouse to use as a carcass.

July 18th, 2012


Breadboarded it successfully according to the schematic as show at MAKE magazine of the first mouse bot just added another bump switch and a 100k trimpot:

[Colorful] Schematic:

Breadboard look (before trimpot):

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Pretty sweet setup you have there.

that would be a lovely chaser for yer pet cat!

haha, I see you got the gist of my naming conventions :p

As for now it only spins in tight circles either CW or CCW... so I picked it up and chased the cat instead.I proved useful at least as a cat's scare-tease toy. :P 

have you considered attempting to place your pot where the "T" was in your original? I mean where the positive line feeds in from the current limiting resistor to the ground side of your light sensors.

My initial reasoning for adding the pot as I did was that at a "neutral/balanced" state I would be already adding 50Kohm of resistance to each of the receivers, but then again, now that I look at it, I may have created a "path-of-least-resistance" so to speak, is that what you're aiming at? If I shine flashlights in the receiving leds I can still diferentialy speed up the motors, the problem is that is always the aforementioned huge ofset between both, that is only damped by the battery falling voltage, but when that happens is near where the battery provides little moving power...

Well, I've gone and did connect it like so to the same result. Afterwards, I've taken out the trimpot and connected the grounds as in the original. At both points the voltage delta remained the same as before... perhaps there's something wrong with my connections elsewhere. I wonder if the "antenaee" lenghts are too low (and difference) to create such difference in voltage... I suspect that I'll just arquive this bot as-is and just move on.

According to the datasheet, a capacitor across pins 1 and 8 will allow for gain.

Maybe I am confused. :)

I've looked into the datasheet knowing before hand that I usually gleen very little out from them, but yeah I saw the application notes and although what they suggest is related to sound output... I did stick a cap between 1 and 8 (I'm so gansta right now) of the breadboarded version, and out of 8.75V from the battery to the motor leads I got 4.43 V versus 4.12 V. The voltage difference is reducing, but still not negligeable I think. 

good job,pleae upload video of the new version

Will do that once film a I proper video. I've tried to shot a few videos but ran into a lot of troubles, so there's not much to show for at this moment.