RF and Sharps not playing together?

Community General Discussions
1

Osiloscope_mixdown.mp3 (2211447Bytes)
No_Sharps_Plugged_in__mixdown.wav (3381652Bytes)
Two_Sharps_Plugged_in__mixdown_3.wav (3875848Bytes)
four_Sharps_Plugged_in__mixdown_3.wav (2990320Bytes)
Audio_Ossilo_TX_mixdown.wav (3719616Bytes)
TX_RX_Stereo_NOTWorking_mixdown_2.wav (2326768Bytes)
TX_RX_Stereo_Working_mixdown.wav (3580144Bytes)
TX_RX_One_Sharp_plugged_mixdown.wav (3046068Bytes)

 

***Update 3.21.09***
With my new found osilloscope skills and a little tinkering, I think I have found the noise. As far as I can tell, it is all on the ground. Here's what I did:

I wired the RX unit to a different picaxe on a test board using an isolated (seperate from walter's batteries) battery pack. Running a test transmit/receive program I brought this test board over to Walter to check for RF noise. I even went as far as phyically wrapping as many of sensor and power wires around the antenna. I got 100% clean data. Second, again with the same battery pack on the test board, I connected the ground of the test board to Walter's ground and the signal wire from the RX to an input pin on walter's brain. Boom went the dynamite. Zero, no data. Again, unplug the sharps with this "shared ground, different batteries" set-up and I get 100% data. This would lead me to believe all the crap is on the ground wire. This leads me to a question: I remember (vaguely) back from my car stereo days something called ground loops. Also, I went to the Texas Instraments website and searched for isolators where I found some sorta chips that mentioned isolating grounds --I really didn't understand the data sheets but I did like the phrase "isolated ground". The question is, is there any way to isolate a ground? Can the RX unit continue to share the same batteries as are powering the sharps but somehow isolate the ground going to the RX unit from them? It seems this sharp noise is not affecting any other systems on Walter... Yet.

 

I am hooking up my new RF RX unit to walter and have found a wierd thing...

RX unit: http://www.rentron.com/remote_control/RWS-434.htm

Sharps: Just the regular old Sharp IR distance sensors

Now, each sharp has a .1uf cap across + and - to deal with some noise I had earlier -they work perfectly now. With all 4 plugged in to the adc inputs of my 40x1, the receiver simply does not work. I have tested it's output with a speaker and indeed it is spitting out data. However, the picaxe does not recognize anything going into it's input pin. I am using debug and it simply sits there waiting and waiting. Now, when I unplug the sharps, all is well and I am getting 100% of data coming through. Even better -as I plug the sharps back in, one-by-one, I can watch the debug window count slower and slower until it stops entirely. I unplugged the sharps and replaced them with just a cap (to see if it was the caps causing the problem) and it seems that is not the issue. It is not a code problem as I am just using:

main:

serin...

debug...

goto main

Any thoughts, you electronics genius-type guys got any ideas?

Oh, there is plenty of power, all is getting a clean 5v from a regulator. --Checked it with a measure-thingie and the sharps are not stealng anything.

 

Info for Frits and the Poor-man's ossiloscope...

Included, as you will see is the MP3, here are the screen shots:

No Sharps plugged in -Serin working great, no data lost

No_Sharps_plugged_in_0.jpg

 

One Sharp Plugged in: Debug working, serin working but very slow

Two_Sharps_plugged_in.jpg

 

Four sharps: Debug stuck waiting -serin not serin'ing

 

Four_Sharps_plugged_in.jpg

Coming out of TX:

Coming_out_of_TX.jpg

 


 

 

2

-=see=-
https://www.robotshop.com/letsmakerobots/node/5916 -> Ground to grounds & other wire to “data” on both!

3

What about the 100uf?
What type of cap should the 100uf be? Ceramic?

4

A 100uF electrolytic
A 100uF electrolytic polarized capacitor.

5

Hey, brooks!!
Isn’t is just my luck… The RF unit works flawlessly, it’s the old parts I am having problems with!! --Off to Radio Shack tomorrow for the caps -thanks, my friend.

6

What’s going on here?

Chris, I thought we made a deal, you should have a look at it.

All that happens now is more “I think”!

Can you at least as rik suggested send me an MP3 of input as you plug in one Sharp after each other, so I can prove my point?

I thought you where a man of he’s word!

7

I’d be curious to see if you

I’d be curious to see if you get this setup to work. I’ve tried every schematic and trick that I know with mine but for the life of me, can’t get it to work. I’m thinking mine is just a bad unit at this point…just haven’t had any chance to replace it.

8

wasn’t me

For the record: I didn’t put him up to all this!

And the mp3 remark was actually a joke.

Hmmm, come to think of it; make it lossless compression Chris.

9

We want a third alligator in this picture
When RX and TX were sampled simultaneously, we could compare input to output in real time.

10

First analyze

Take a look at these 2 images, things are not really how they should be:

Billede_10.png

Billede_11_0.png

11

(And I have not yet gotten
(And I have not yet gotten the "real" samples from CTC, soon I hope I will, and then…!)

12

The fix!

OK, after a loong struggle, conclusion:

Have a look at this little slideshow :slight_smile:

The Sharps are sending bad vibes into the power of the RX unit. That may be obvious, but now we know for sure, and can head on for a targeted fix, no more guessing!

Possible fixes:

  • If you knew that you whould not use RX when using Sharp’s, and you have a spare output pin, you could have power feed to Sharps by seperate output pin, only turning it on when not recieving serial.
  • If you need to have sharps going all time, you would have to either use a seperate power for Sharps - which may and may not work, depending on where they return the buzz. (if it’s on G, it would not help)… Or, use a diode to make sure the Sharps only get power, not return it.

There may be other electronic wiz-tricks, but now I have helped you SEE what’s going on, and it will be very easy to SEE if the fix that you test (Cap, diode, external power or whatever) is working 100% - Your signal should LOOK like this :slight_smile:

I have spoken!

13

Gotta give you credit…

I will tell you actually what we learned here:

Sure there was noise, we all knew this going into this. However, now that we have gone through these steps and I have now got a very solid grasp on what I am looking at in terms of the Audio visuals, as I play with different "fixes" I can repeat our tests and know for sure. I appreciate all your help and work, here frits but there is one problem… Once again, one of your silly MacGyvered test-systems worked flawlessly… Remember "check it with a ADC input"…?

At any rate, I guess I will get to soldering some of those 100uf caps and give that a shot. If it does not, I guess i will have to just turn off the sharps during RC drive mode. I will have to figure out how to introduce an inturrupt during Auto drive to be able to get back into menu/opitons mode

In the end,I think this is what I actually like about these robots… If i didnt run into these problems, what would I do every night after the kids go down?

Thanks again, frits --Can I start listening to rik now?

14

PIVOTS!
just saying

15

silly?!

"one of your silly MacGyvered test-systems"…?!? SILLY?! man!!

Here where the guessing before my silly stuff:

“I suspect that the short burst of power used by the sharp sensors to pulse the IR LED within them is spiking the power supply enough to upset the receiver.”

"I’m just wondering if that noise is “leaking” into your RF"

My silly stuff shows you excactly what is going on. And, not to mention; When you test caps and shielded cables and what they come up with, you can see if they work - tune in with the right uf’s etc, not guessing, but seeing!

:smiley:

(A blind carpenter was trying to navigate in a troublesome terrain. I made him see so he can navigate. And he tells me that it was silly because people had already told him what the world most likely looks like, how silly!)

16

- And may I add…

In all my self righteousness, allow me to proclaim that you (calling this silly method "poor mans oscilloscope") could never get a clear debugging vision like this on a oscilloscope:

vision.png

17

Crappy outputs/Inputs

CTC,

there are some worrying things in your traces, but first a few questions!

How many Sharps are being used? (?) Fritls stated above it is the IR LED spiking the power supply, however looking at the datasheet the IR LED’s are not switched, so I don’t think it is the IR LED’s however the Phototransistor could be to blame.

What power regulator are you using? what is it’s current out put capacity?

On the trace if you have a look at the pulse chain you will notice a couple of things, if you look at the top trace you will notice you are not getting a nice square wave, however I am not sure if that is an artifact of the PC CRO, notice how the pulse train starts low and creeps up, the other thing to not is notice how the top of the individual pulse degrades with time.

Looking at the trace below that you notice that there another wave form over laid, it looks to be a 50/60Hz hum, this could becomeing from inductive pick up in your Robot or in the laptop it’s self. The second trace is problematic two as the pulses seem to be wider (i.e. lower baud rate).

Can you do a trace of the power and earth rails with the Sharps in and out please?

My guess is at the moment you are pushing the power req a bit hard and it is not keeping up the the transient power requirments, do you use the one battery for both drive and logic?

If you can do those two traces I will have another look at it tonight.

The PC CRO is a good Idea but I will stick to my CRO I think, instrument grade stablity means I can trust the desplay.

best regards Cliff

18

And, your point is?
And, your point is?

19

Frits

"Silly"…

This is not in any way a poke at you, my friend… This is what one would say if they didn’t listen and then was shown how well something works!

–The “one” would be me…

20

Cliff

Hey Cliff,

I don’t know how much weight you should give to the “oscilloscope” information… I am not using PC CRO, instead i am using Frits’s method of the audio-in jack. The screen shots you see are of Adobe Audition, audio recording software and in addition, after the screen shot had been taken, they were cropped and blown-up to see them a bit better. You may have something in terms of the square-wave thing, but then again, one image was taken directly from the transmitter’s pic -wired directly from the output pin bypassing the RF TX unit. I would assume the lack of a clean square wave is the recording software.

I am using a regular ol’ 7805 (with the standard caps on the board) for power on the brain side. Drive is a seperate 14.4v supply, sharing a ground, but was not connected during the tests. I think the 7805 can spit out an amp, but I can tell by the fact that it is stone cold during use, it is not even getting close to that. I’ll see about the new tests but really, I am simply looking at switching the sharps off during the auto-drive mode. The one problem with this is I need to be able to send an RF “inturrupt” command to be able to get out of the auto drive and into a menu/option mode…