Sharp Is471f

[russbow]

Has anyone had any experience of this little gizmo? Struggling to get my homebrew IR stable on the TX side, so looking at alternatives. This is cheap at a little over £2.00

Thought it a bit nifty the way it syncs tx/rx. Suspect it is a bit insensitive though. Any comments?

 

[slurp]

Yea, I've had a few of these for IR object detection the basic circuit at the bottom of page 4 works well.

I found the range to be about 50-60mm when the supply voltage was 4.7v and 60-70mm with 6.3v (un-regulated battery supplies). It's worth noting that detection is a low output... seems to trip people up from time to time.

There was a website that showed an application with LEDs driven via a transistor on the GLout, unfortunately the site has gone but 160mm detection range was reported by someone on the minisumo forum. Multiple LEDs in this was can work around narrow beam angles as well as brightness of output.

The IS471F is very susceptible to IR coming from the side of your LEDs, make sure it's well shielded. I used either heat shrink or mounted LED on the opposite side of the board to this device. When using multiple devices the LED on one easily triggered the other.

I'm intending to salvage a few of these from old circuits to try the transistor drive circuit to gain more range... it'll make it more effective for opponent detection on minisumo!

regards,
Colin

 

[russbow]

Thanks Colin. Certainly worth a play. I suspect some of my problems are not matching the led with the detector. Something to do with wavelength?
I think the spec sheet leads me in that direction.

I would be very interested in your findings using a transistor driver. Think I'll get a couple of these and have a go. Watch this scace !?*

Russ

 

[slurp]

What wave length was the LEd you tried?

While 940nm is the peek I would have expected anywhere between 840 and 1000nm to give good results... 90% of peek performance. I can't remember just how close I was, I think it was within about 10-20nm.

regards,
colin

 

[Andres Rodriguez]

Yea, I've had a few of these for IR object detection the basic circuit at the bottom of page 4 works well.

Colin

Colin,
From the circuit in page 4 they connect the cathode of the IR LED to pin 4. Did you leave that pin floating?

Regards,
Andrés

 

[russbow]

Colin,

I am using the TSOP 2438 rx. Data sheet suggests 38khz. I cetainly get a level shift indicated on the scope when it's on the workbench, but eratic.

The tx is a TSAL 6400 ( or 5400 - my supplier stocks both, but I am not sure which I bought !). They both use a wavelength of 940nM so all should be compatible. I think the osc is iffy. The preset shifts it through the band, but too broad, so a bit of restistor juggling. It is a dual op amp. Don't like it so I think I'll try a NE 555. Mused the idea of using a picaxe output to a transistor driver, but dumped it as I need a continuous output.

Why is a carrier freq used? Is it so you can hit the tx harder for a short period of time?

Had a look at your link Colin. Very interesting. Must have a deeper look into the site. he links to the german sites don't work by the way. Well, guess the problems are self inflicted, so it's back to the soldering iron.

 

[hippy]

Why is a carrier freq used?

It makes it easier to detect against any background light levels. Regardless of what the ambient light level is that the light level is going up and down is noticeable.

 

[slurp]

Marmitas,

No pin was left unconnected. The IS471F device modules the IR transmission for you, hence the LED connected between Vcc and GLout.

Point the LED away from the device and shield for side (stray) IR transmission, then the device will pick up reflected IR from the object that's being detected.

Russbow,

TSAL 6400 looks spot on, I thought the TSxx 5400 was 950 nm but it shouldn't make much difference. We're jumping to a new problem when you start to talk about the TSOP devices.

Un-modulated IR is ok at close quarters where there is little excess from surrounding ambient lighting... 1-2mm at best 5(ish) depending on the actual device and set-up. They're often used for line following or minisumo edge detection but you're mounting close to the target.

As Hippy says, modulation helps you pick a signal out against back ground light levels. With object detection you firing your signal out into open space and hoping to pick out a reflection. How do you know what is reflected and what's not if there' no distinctive signal?

Considering your NE555 and detection using the TSOP xxxx, there are a number of things to consider. First tuning your components to give 38Khz output from the oscillator... breadboard will be a pain, there's so much capacitance before you start but it's not impossible.

Second is the off time, the TSOP will try to attenuate a continuous signal. If you're trying to keep your target in "sight" (or expect it to be there for more than say 600us) then you'll need to turn the signal off for a short period. The an 08M would make a good candidate for output as you can get a good match to the 38kHz and control the periods. I can't say I've had great success with the TSOP devices for object detection, I should try again but then they're not designed for this purpose.... I might try again. I've got a neat board made up for dual sensors but need to spend a little more time on setting up the period of modulation.

Purpose built object sensors may be a little bit more expensive but they are sooooo much easier to use!

The (missing) German site is the same site as I said was gone in my first post the concept is still good

best regards,
colin

 

[slurp]

This is fairly similar to the broken link on the minisumo forum...

http://www.kreatives-chaos.com/artikel/abstandssensor-mit-is471

Google language tools seem to do a reasonable job of translation, it gives a little explanation on delivering higher currents to increase the range. I've not tried this, or checked the maths on their time periods vs current on the LED (yet).

regards,
Colin