SparkFun Forums

[mpanetta]

Are you sure your xcever is working right and has not been damaged?

Also, 52KHz is very suspicious power supply noise, it may be from an electronic ballast. I would suggest turning off all fluorescent lights in the vicinity (including the compact variety) if possible and then seeing if the noise goes away.

Mike

[angelsix]

I have a bag of 20 tranceivers so I will try another one tonight to see if there is any change but yes I'm pretty sure as I have seen this result for 4 months on 4 different brands of sensors using about 10 attempted circuits. Its getting really frustrating now.

I'll also try battery powering the board and turning off all lighting, TV, speakers and wireless internet router etc... to see if the noise goes away.

With regards receiving a signal, what is the typical amplification needed for a signal to be driven back to its original transmission level from 30cm away reflected back on shiny plastic? I know there is a formula of dB degradation over distance based on frequency etc... but would be nice if there were a site or calculator to punch in a few details and get a rough idea, is it 10x, 1000x, 100,000x 5,000,000x ?

[markaren1]

Just in case it hasn't been suggested already (I haven't read all of the threads)...

I know this is going to sound daft, but can you do these measurers with your PC and display completely unplugged and using a battery power source for the unit under test.

I have seen similar problems from either PC PSU leakage (via a USB/RS232 ground wire), or from the horizontal output magnetic/electrostatic radiation from a CRT monitor.

If you have a PC based scope this may be difficult, and you may need to borrow a good old fashioned analog unit.

Regards,

Mark

[angelsix]

mark, thanks for the suggestion. I am almost certain a lot of the background noise is from the scope internals or the mains power or laptop itself as moving the scope away from the laptop reduces it a lot. I was going to try battery power but managed reduce the noise a lot by moving it around so carried on working. I found out something very interesting but will tell you all tomorrow as internet is down at the moment and writing this my mobile is painfully slow.

[angelsix]

Woooooooo!



All I did here was change the sensor to an indoor can style one


and it works perfectly rail to rail detection. And just to please me more I did this scope from about 5 meters away!




So - with that in mind I wondered if the sensor I was using was faulty. I got my bag of them out and tried another one and that was still the same and far weaker than the indoor style one.

The problem I have is that I have no specs on these sensors that I need to use (they are outdoor ones with grommet attached already) other than that they are 40kHz transducers. I tested this by changing the frequency to 50 and 30kHz and there were no reflections at all, the closer to 40kHz I got the stronger the signal until it was the 1V reflection from 30cm away as shown before.

What I believe here then is that the sensors I need to use need to be driven by a larger voltage, would you say? Or the other option is I add another stage of amplification but I believe that wouldn't be the best idea as it means all other background noise will increase too.

If I want to drive them at 20V or so, what shoulld I use and how would that affect the amplifier having to receive the signal? It would see 20V, I could add an analog switch to stop it while driving, but once the switch is off it is still likely to receive a short burst of 20V as the switch opens.

[waltr]

Wow, that looks a lot better.

Just a crazy thought, could it be that the sensor you were using for a receiver is really a transmitter? How could one tell the difference?

[angelsix]

The sensors I am using are all xceivers so they do both. The can one is off Farnell (the one I got working) so I have all the specs on that. But the one I need to get working I only have specs that it is a 40kHz trasceiver. Its definately a transceiver as I have brought them off an existing product kit as just single sensors without the actual product from a salesman, so I know they are xceivers at least. I think they are probably higher voltage sensors due to the also evident fact that the settling time after driving them is only 2ms, where the others are 3-4ms.

So I'm back to my original issue of having to either drive it harder or amplify more. I'd prefer to drive it harder but am a bit unsure as the best method and how to protect the amp from the higher voltage while still driving it (like possibly using the zener as mentioned). Any suggestions on this?