It speaks! (And I broke something else, maybe)

A project log for K-9

Raspberry Pi powered K-9 GUV with a NES controller style remote

Andrew BillsAndrew Bills 02/13/2017 at 02:500 Comments

It's been stinking hot here so I haven't really had a lot of motivation to be out messing with electronics over trying to stay cool but I have done some stuff so thought I should do a small update.

Connected up the head sensors etc as mentioned previously however was getting no sound. I did think this was my wiring but after much much effort figured out that it was my hookup (followed the instructions which were wrong telling me to use the L/R channel instead of just a single channel + ground to the mono amp). Sadly my wiring was the only issue, with a little bit of effort to run it through an arduino to test I worked out my 2.5w amp was dead.

Not a problem as Little Bird Electronics who I got it from were super helpful in replacing it providing I covered the postage. Which was good because I did buy the amp months and months ago. When the new one arrived I was able to hook it up and test without an issue.

I previously did look into some K9 audio files for reuse which is common and popular to do but wasn't pleased with the quality/consistency of the files that are used in prop builders. I also wasn't intending to go and sample my own so I instead elected to go the voice modulation path with espeak. While it will not be 100% perfect I can use the voice to provide pre-checks for when I am not looking/using the touchscreen panel. The video below is just a simple test playing a wav then a phrase via espeak. I may tweak the voice settings some more later.

That being done I hooked everything back up and started to run through retesting things I had already tested to be sure. My current state is I have some problem with the GPIO in use/Ultrasonics. Which is strange due to it being fine prior.

The only explanation I can think of is I bumped my wiring with all the moving about for the amp. When I hit another cool patch in the weather I will plug in one of the other ultrasonics (the side ones that I haven't mounted yet) and test with them. I want to believe it is a software issue at this stage as I got one return result from the ultrasonics while messing about but will have to figure it out.

Separately I have put together the missing side panel for mounting so it can flip out to access the motors/batteries. See the comment about weather for progressing that as it still needs some putty touch-up & paint before it can be mounted in a final sense.

Hopefully cool weather isn't too far away as it is a lot easier to figure stuff out when it's not too hot to think.