I am changing email addresses to avoid the constant influx of emails about this project. I hope everyone who finds this page does well on their projects, but I can no longer be a source of information.

———————————————————————————————————————————-

I’ve gotten lots of emails about this project. Sadly, I don’t have the time to answer each one as they come in, and I’ve already forgotten some of the info I figured out for this project. Before I forget it all I decided to write this FAQ to help out anyone who is having problems.

First, make sure you have read every post and page on this website. If you haven’t, then get to it. Most of the very important information is in the “pages” of this site that does not show up on the blog. This sitemap shows all the pages. The “Quick Reference” page has links to other sites using the HM2007. Please read all of those sites as well. I have updated this page to reflect even more information.

Que: “Where Do I get the HM2007?” (or other parts)

Soln. All parts can be acquired from Images Co.

Que: “My HM2007 project doesn’t work”or “I am getting no reaction…”

Soln. Make sure you have not just blindly built what is in the schematic. Triple check your pins on the datasheets for your chips and make changes if you need to. In my case, The RAM chip I used had a non-inverted enable on pin26, but the schematic assumed the enable was inverted. This caused the circuit not to work until I figured it out.

Also, feel free to move the address lines around. These lines are mixed up on the schematic only to facilitate the PCB designed from Images co. Aligning SA0-A0 from the HM2007 to RAM and D0-D0 in the same fashion caused no problems for me. It works great and is easier to build on a breadboard.

Que: “I am using CPU mode …”

Soln. I am sorry that I cannot help in these cases at all as I did not choose to use this mode on the chip. However, there is some information in the Images Co datasheet Addendum and some more info on the talking toaster page as well as some info for interfacing the Arduino. (<– that last link is the HM2007 wiki, great resource.)

In General:

This project is very reliable. If you build the HM2007 correctly, it will work exactly as described. If it does not, then do the following:

1.) Double check the schematic connections

2.) Double check wiring connections

3.) If possible, test your chips to confirm they work.

4.) Redo the schematic from scratch (not as hard as you might think!!!) Use address lines SA0-A0 and data lines D0-D0 like in this schematic. Draw your own schematic. You may need to enlist the help of a professor or teacher to help you with this.

5.) If nothing you do is fixing your project, it is very likely that your RAM chip or HM2007 is bad. This is not the most likely problem, however, I did encounter a bad HM2007 chip and a bad RAM chip. The best bet here is to order new chips. I made several orders from Images Co and was charges overall about $60 for shipping on everything.

With that said, this project is closed. I am sorry, but I will no longer be answering emails or comments about anything on this project.

- Mik

Try number 1:
we disconnected the speaker inside one of the radios and attached it directly to the microphone input on the voice module. That shorted out the circuit or gave us an FF on your BCD Display. Failed

Try number 2:
We reconnected the speaker to the radio and put the microphone inputs to the voice module in parallel. We figured that maybe the resistance of the 16 Ohm speaker would be sense instead of a high/open circuit resistance. We kept getting a “word too long” error or a “word not found” error. We found this had to do with the “transmission over” signal that happens after the push button is released on the radio after a transmission. Failed

Try number 3:
To cut the microphone off after a voice command so the extra noise after a transmission wasn’t heard by the voice module we shorted the push button leads on the radio and wired the push button in series with the microphone as a force stop. That didn’t work either. Failed

On top of all the failures we managed to burn off one of the surface mount capacitors. We were able to patch it with another, larger capacitor but since the capacitor is to big to fit in the radio housing with out shorting we are back to the drawing board.
Maybe we’ll have better luck tomorrow.

–Mike

I only used 3 commands: “Forward”, “Reverse”, and “Stop”. These commands worked great. However, I noticed that eventually I’ll need a microphone preamp on the input so I won’t have to yell at the robot.

Also, notice how it takes a bit for the robot to react to what I say. This is because I have a 1 second delay to make the servos turn. Literally, when I say “Forward” it will turn the servos on for 1 second, then check to see if the input has changed. This length of time can and will be changed on future tests.

–Adam

In the video, when you see the code “66″ that’s actually an error code, meaning “word not recognized” if I am remembering correctly.

Click the pic for full size.

–Adam

Hardware:

We have all the servos mounted, and tested them out. You can see more information on this on the testing page.

Somehow Mike touched the IR rangefinder sensor and it started working correctly. Don’t ask me how, but it just started giving the correct singals after Mike held it for a few seconds. I guess he’s magic or something.

Since the wrong glue was used for the polyethylene bodies, it did not stick well. The bodies will be reinforced with L-brackets and self-tapping screws. Check out more pics of the bodies at the bottom of the Robot Base page.

Finally, I’ve fixed the voice controller circuit. It turns out that the schematic we were using is wrong. Pin 26 on the RAM (CE2) was tied to ground for some reason in the schematic we were using from the datasheet. Comparing that schematic to the schematics of other people who had gotten the circuit working correctly(from the talking toaster example), I noticed they tied this pin to Vcc. Taking another look at the datasheet for the CY6264 chip, I noticed the CE2 pin was not an inverted input. I tied it to Vcc and it suddenly worked perfectly. I’ll post a video of this soon.

–Adam

The robot base build started today. You can see its progression so far in the pictures after the jump.
Some Complications

After opening the package containing the HDPE I found a piece of mangled very floppy plastic….. to say the least, we couldn’t use it. I called up a local plastics company in Dallas, TX and got a new piece of plastic that was 1/4″ in stead of 1/8″. We feel that the added thickness will be more ridged but in turn add weight. The weight will not be that big of a deal since the mounting hardware used to mount the wheel motors to the base is very sturdy.

Pan and Tilt System

The pan and tilt system is really neat and it adds much more fluid motion. As you can see in the pictures below the robot head is mounted to the pan and tilt mounting plate. There is a second servo that allows the head to move in a semi circular direction forward giving the head a tilting motion. The first servo is will be mounted to the tower and will provide the pan motion for the tilt mechanism.

Drive Motors and Wheels

The drive motors are housed in a servo mounting bracket. The servo also has a machined hub that bolts directly to the wheels using beveled machine screws. You as see the servo, mounting bracket, wheels, hubs, and assembly in the pictures below.

Base

The base of the robot was made of the same 1/4″ HDPE as described above. Five main pieces were cut and glued to make the tower and base. The motor mounting brackets where assembled with machine screws and locking nuts. In the pictures below you can see the build process and the both wheels attached to the base.

–Mike

Hit the jump for more and a video.

The startup cycle (as I understand it ) for the HM2007 chip is as such:

Startup and check the RAM. If RAM checks out, then initialize the RAM. Then light the “Ready” LED and output two zeros on the D-bus (00) to signal recognition mode.

The circuit I built does something a bit different. It outputs FF on the D-bus at startup. Oddly this is the same reaction I get from the circuit when I completely remove the RAM chip. If I hit any of the keypad buttons, it cycles through 4 states that aren’t documented in the datasheet nor anywhere else.

All components have been switched for other chips to test them, and the wiring has been done several times using several wiring diagrams (all ended up being the same) and the results are the same. I can only assume I fried the RAM while trying to solder them to the SOIC to DIP PCBs. This would not check out when the HM2007 starts up. I can only assume the FF output as well as the other states are all error codes that aren’t documented.

Mishaps are expected and I know I will figure this out sooner or later.

Here’s the example of what the circuit does so far:

- Mike

high-density polyethylene (HDPE)
A snazzy robot head

More details for the hardware configuration to come.

The Plan

The base for this robot was derived from a previous forgotten project you may have seen on Mike’s personal website known as DDS. We chose a configuration that utilized a caster in order to maximize the maneuverability of our robot. A rough outline of what the robot may look like can be found in the pictures below:

And, YES, these plans were drawn, not computer generated…..can you tell?

Mobility Hardware

Servos
Currently the robot’s configuration will utilize four servos. Two servos will be used as motors for the wheels. These servos have been hacked for continuous rotation. The other two servos will be combined into a pan and tilt system for the head of the robot. There are also four additional servos that will control two arms . These four servos may be added later to the project given the amount of time left before our deadline. In the side view you can see two of the four possible servos for the arms.

Wheels
The wheels that will be used for our robot will be 6″ ABS black plastic wheels with a rubber tread around the circumference. You may be wondering about the head of our robot. at the back end the robot will be able to pivot using a 1″ caster. You can see that by using a small caster there will be a three degree tilt in order to allow for better weight distribution while the robot is in motion.

Base Construction Material
The material we have decided to use is a 1/8″ HDPE plastic. The HDPE is extremely strong and light weight. Since we are using servos, a microcontroller development board, the voice recognition module, and a hefty battery, weight was a major concern.

Robot Head
You may be curious about our robot head. The truth is both of us were walking through one of a particular chain of stores and stumbled on an alarm clock that looked like a robot head. We both through it would be neat to you in this project. If you want to see what the alarm clock looks like visit Mike’s website and look for the AVR project.

- Mike