Soeren
If the main controller(PC) shuts down for what ever reason and the DAC voltages go to 0v with the Sabertooth or RobotQ
Drives will go Full reverse and full CCW turn until the batteries drain, drives fail or the robot ends up in Cleveland.
With the P&G drive if the main controller shuts down and the DAC voltages drop to 0v the drive will shut down and show
a fault. also it I power up the drives with the main controller(PC) turned off the Sabertooth and RobotQ will take off, but
not the P&G drive, it would show a fault.

Because I'm not doing random cutting paths(I hope) any deviations from the target path will cause me a problem, with the Sabertooth
and RobotQ drives even if one wheel runs over a toy the kids left laying around or even Dog crap(my mother has a dog) the drive
can make the speed and direction correction very fast(full reverse to full forwards in .5 seconds), because of the Acc/Dec programed
into the P&G drive it takes 6 seconds to go from full reverse to full forwards.

because of the reason listed above(dog crap, toys,gopher holes) I can't use the drive wheel in the velocity or positioning loops

I was thinking about all that friction while turning, I have never used tracks before so my information in very limited.
I willing to give them a try.


Tommy

Hi,

Yes, sunlight is rich in IR and so is a lot of other light sources (energy saving lamps and fluorescent tubes are really bad in that respect) so you need to use a modulated signal (eg. a carrier wave) to get a weak signal out of a lot of noise and sunlight can still swamp the receiver out, so that it gets insensitive to the signals you want through.

Integrated receivers are great. They have built in Automatic Gain Control (AGC) and noise reduction (different types for different problems, so you need to select which type).
The frequency usually won't matter that much, as long as transmitter and receiver is matched.

Vishay have a broad range of receivers with different types of AGC circuits for different noise problems. If you take a look at eg. TSOP381nn or TSOP383nn (Where nn is the carrier frequency; 30, 33, 36, 38, 40 or 56 kHz), you'll see it's spec'd at 45m range typical. This is with zero ambient light, i.e. in a totally dark room.

You'll need to select the carrier frequency for the transmitter (beacon) and you need to select the AGC after both the kind of noise it will see and the type of "telegram" it will see from the beacon - a continuous wave (CW) will need one of the IRRX's made for security curtains, which has a constant gain in their "AGC" circuits (shouldn't really be called an AGC then) and that may be a problem if you have other interference and different angles and distance - as ever so often, the selection will be a compromise between opposite "interests".

hmmmm Webbot I think you took me *waaaay* too seriously on my joke with madsci

I was really just poking fun at him, tin foil hat style and all.

To coworkers of madsci: at some point you just got to trust other people and their code. Webbot has a lot more to gain by building his reputation as an awesome coder, and he isn't just some random guy in China no one has heard of before that works super cheap.


But yea, Webbot is definitely right on the point of optimization.

Hi,


We're good, but we aren't Oracles 

Did you try with a controller installed?
Did you try with more than one controller?
Which OS are you running?
Laptop or stationary?
How did you build the programmer?
Did you use a PCB?
Did you inspect the PCB with a magnifying glass before assembly?
Did you inspect the PCB with a magnifying glass after assembly?
Did you measure for connectivity where there should be and none where there shouldn't?
Did you try it on more than one PC?
Did you try more than one circuit?

Without this info, we'd just be guessing.

The answers you get will not be all that much better than the amount of work you put into your questions.

@Invicta

You don't say whether the answer is 'success' or 'fail'.
So I'll assume 'success' - ie no further action required - unless I hear otherwise

Hi,

Only if made horribly wrong. The high current through the motors will not go through the controller (if made correctly of course).
A bad opto coupler design can still induce noise and transients into the logics.
A transistor is adequate for separation in a properly designed circuit.
(We're not talking isolation from mains voltage for safety here).

Not sure why you two guys got stuck in using Ant to build the .a files from the source code.

Presume you have followed Dunk's tutorial at http://www.societyofrobots.com/member_tutorials/node/382 showing how to do it on Win, Unix and Mac. He doesn't seem to have your problems.

Ant is used to build production releases of about 99.9999999% of ALL java applications - so its not some flakey thing its an industry standard thing - and although it runs under Java it can build non-Java code. Even Americans have managed to use it ! 

"Can I just change WebbotLib to only use .h files" ? - No.
For the same reason that even AVRlib uses C files rather than doing everything via H files.
If your application has multiple C files then it gets even harder to manage.

The benefit of using .a files is that all my code has been built using '-Os' to make it smaller and faster and since I've learned to understand the avr-gcc compiler 'bugs' then its all fine and tested. I think I have only ever had one bug ever reported, long since fixed, caused due to this setting.

But most people choose to use the compiler with optimisation disabled '-O0' - mainly coz they dont want to have to bother with learning the 'avr-gcc' bugs that can be introduced. If WebbotLib could be achieved just using 'h' files then this means that everything would be compiled with larger and slower code - which will have a bad impact on interrupt service routines. Unfortunately avr-gcc has no way to use '#pragma', as done by most C compilers, to change the optimisation settings for any given piece of code - its "all or nothing" - which is bad but I haven't written the compiler!

So if your issue is that the '.a' files haven't been built in America - then learn to use Ant so you can build it yourself.

As for Project Designer then it uses Java jnlp technology. The only 'mysterious transfer' is to compare the version on your pc with the version on the server and downloads the new version if required. This is 'pure Java' ie its all done by Java itself and I do not play any part in that communication - hence it shows my digital certificate. Java is now owned by Oracle - an American company - so ask them what it does as I don't have a clue and don't really care since it works.

Whilst you are at it then why not ask your browser provider (Microsoft, Firefox, etc etc) what information is used in an HTML page load to decide whether the cached version is the same as the latest server version so it can decide whether to download the latest version from the server. That's analogous to what jnlp does. Presumably you are also limited to only being able to access American web pages.

Trust that all the hardware and software drivers installed on your computer are 'made in America'. And that you also only ever run programs at work that are shipped as source code and compiled-by-you.

So in summary: WebbotLib is shipped as precompiled .a files (which you can recompile yourself) and this wont be changing; my Board/Project/Gait Designer tools are pure Java and will not be changed nor ever shipped as source code; and 'yes' I am a 'foreigner' (ie not American) and I have no desire to change that status either.

Live with it - or move on !

Hi,

Watch out for acid traps (you have a couple).

When this one has matured, why not go for a SMD proto module, I'm sure a lot of people would like that, considering the ever shrinking amount of thru-hole components.

Essentially, opto isolators are to physically (in respect to electricity) separate electronic components while still letting them interact.

A common example would be for motor control.  Since motors use a lot of current (and draw a lot during power up), the high currents can damage a micro controller if they belong to the same electrical circuit.  With an optical isolator the controller can still signal the motor circuit without being directly wired to it through light pulses (which interact with photo diodes/resistors)

i built a seial programmer to the atmega32 found in this link :

http://ballwell.blogspot.com/2009/08/serial-programmer-for-avr.html


i am trying to test it using ponyprog , but unfortunately when i open the setup interface i get "Test Failed"

i am searching for any error but i cant find

thnx for any help