Budget ConnectR / Rovio

[Fraggboy]

Nice!! It seems this transformation is quite a chore!!! But, I'm sure you are enjoying this.

[bhylak]

We need a like button.

[mfortuna]

I may ask our rework tech at work to solder the 3 tough connections (to SMT resistors), the rest should be fairly easy.

[mfortuna]

Another update. Ready to start soldering. The pictures show the steps in preparing the 25 pin serial extender cable. The first step is easy, just snip the cable in half. The next step is to remove most of the jacket, exposing the 25 signal wires and the chassis ground wire which attaches to the foil shield that wraps around the other 25 wires. Finally, I identified which conductor goes to which pin. I wrote down all the colors and also labeled each wire with a label maker.

[mfortuna]

Made some more progress but have an issue. I have all the wires connected that should allow the BAM to connect to the dirt dog. I confirmed one suspicion I had, the BAM uses switched 5V from the Create. The BAM can be cheaper than a Rootooth because it doesn't need a voltage regulator to turn raw battery voltage into 5V.

I can plug the battery connector into the controller board and the BAM pairs with my computer. The issue I have now is I am not getting any data back from the DD. I checked and double checked I had RX and TX correct but I have a gut feeling I may have swapped them. It could also be the fact both the SCI port and BAM are connected. The extra length of cable to the SCI may be causing signal integrity issues.Time to break out a scope, but not this weekend.

Here is a picture of the wiring and of the BAM with no cover. The BAM uses a BT module and has some assortment of transistors and other parts on it.

[bhylak]

Nice! Thats really interesting.

[mfortuna]

Minor update. I swear I tested this before but now I'm not so sure. It looks like roomba status and sci tester don't toggle the device detect pin which is supposed to wake up the roomba. What I figured out is my wiring job was OK. I need to turn the dawg on first before trying to connect via roomba status. I can connect to the dawg using either the SCI port or my 25 pin connector with BAM providing I turn on the dawg first using the front panel.

Another mystery is when I have the sci port or BAM connected pressing the single button on the dawg just results in a beep and green LED. I guess that is the way it is supposed to work.

The next step will be the three remaining wires to control the camera and then some driving tests with the security dawg all back together.

It's coming together, just not all that quickly!

[mfortuna]

Another update. The last three wires are soldered and tested. Next up will be transferring the camera, servo, power supply, and servo board from the create.

I did run into an interesting issue. The dawg seems to have a stuck wheel drop switch. All three switches test OK with a meter but the diags fail. It could be a circuit board issue. If I use roomba status to get into safe and then full mode I have no control over driving or lighting LEDs. Because of the wheel drop issue when the robot enters safe mode and detects a wheel drop it goes back to passive mode. I found that if I use the send command to issue a 130,131,132 sequence I get control of the robot.

As I suspected from before my left wheel motor seems to be slow and may need to be replaced. Here are a few pics and a screen shot. Sorry for the fuzzy picture.

The robot looks put together but one screw is holding the front bumper and the dust bin is holding the top plate out.

[mfortuna]

The dirt dog actually displays independent status of the drop sensors. If I flip it upside down all three come on and when I put it upright the left stays on. I will retest the switch.

The left motor has a lot of resistance compared to the right. It may be a mechanical issue.

[mfortuna]

Wheel motor issue solved. The binding was caused by a major build up of crud under the inner bushing. Once I cleaned it up the wheel works fine. Funny, with no brush deck or debris bin vacuum motor the roomba lifts its rear end up in the air when going backwards. Hopefully the camera and circuit boards will make it drive more normal.

Next up will be a system test of the equipment currently in the create. Then I will have to figure out where to mount the camera.

[sageman]

ive found this even with no bin tho it comes down again a second later

[Fraggboy]

That is SO cool Mike!!

[mfortuna]

Thanks Chris.

Here is another update. At this point the Security Dawg Project is at the chassis complete stage! I put it all back together and I mounted the BAM to the top panel via the existing screws and a piece of velcro strap. I tucked the extra wires under the strap. Not sure if they will ever get used.

The only other mod I'm considering is to cut the lip off the dirt bin so the circuit boards can be mounted flat.

I'm ready to pull the equipment from my create.

[bhylak]

Wow so not including the dirt dog, how much did it cost just to get to the stage you are now. (With the BAM hooked up)

[mfortuna]

The BAM is $59.99, a 3 ft (the smallest length I could find) 25 pin serial extension cable is about $2.50. So less than $63.00 If you have a soldering iron and tools.

$36 cheaper than a rootooth with the advantage of bringing out additional signals if you want.

There is some cost savings because the BAM relies on 5V from the roomba while the Rootooth requires a regulator to get the battery voltage down to 5V. That is not $36 worth of parts with markup. I think the BAM is priced with a lower profit margin than the rootooth.

[bhylak]

Nice... Another question, could you hook up a command module?

[mfortuna]

That's a good question Ben. I don't see any reason why not. If a BAM works a command module should work. I guess I will have to give that a try.

[bhylak]

Yes, then you could have a true budget Create.

[bhylak]

Also, consider using RoboRealm for the project. They may give it to you for free if you contact them and tell them about your project. (worked for me)

[mfortuna]

I will contact them Ben.

The one issue with using a non-create with the command module or BAM is you don't get full functionality. I managed to get the three DO (data out) signals by using the motor controls no longer needed in the robot. The DI, PWM, and analog pins the create supports are a lot more difficult, if not impossible. I guess you could replace the bump or wheel drop sensors since in full mode they don't stop the robot anyway. That could provide a DI function. The other functions are not really supported at all by non create firmware so they would be tough.