Sensor Board Association

Originally published at:

Did I miss the cap sizes in the videos for the Sensor Board? Or just not paying close enough attention?

Hey Mike,

Very possible I didn’t say what size I chose. I’ve gotten into a bad habit of simply putting down 0805 footprints and dealing with sourcing (part numbers) after the fact.

It’s always possible to check against my repos, did you know those were online?

I took a look and they were indeed 0805 :slight_smile:

Sorry about the confusion!

Thanks Chris,

I should have said what is the farad size for the caps. Sorry about that.

Ah…the capacitance!

Another good rule of thumb (especially for decoupling caps) is to assume I am putting in .1 uF or 100 nF caps.

I figured it was a small value but I tend to get in trouble when I assume. :slight_smile:

Usually with caps, it’s only a problem if you go much too small (in the event your device drawing a lot of current) or if you mistakenly replace it with a resistor (GULP) :smiley:

1 Like

How about all the resistors? What are their values? I don’t see them in the repo version or the video.

I actually left them out! That’s something I do later in the assembly, as I promote having a box of various values around and standardizing on 0805s for your assembly. These are the kits I use:

There are versions on eBay and similar. Just get an E12 series or higher (E series chart)

Yup, I have a kit. But I didn’t see an “assembly” video, so wasn’t sure. Is it just covered in the first of the programming videos?

I still can’t find any videos about the assembly. I figured out the led current-limiting resistors, easy enough, but not sure about the rest…

Aha! The R5 and R7 are discussed in the LDR Analog Input video!

Still not sure what R1-R4 should be.

I’ve looked through the DS7505 data sheet, and while it calls for a pull up resistor on the O.S. line, it doesn’t specify one. But it’s just an open source “switch”, so it should work like a plain pull up resistor, as one would with a pushbutton? If so, 10k ought to work. That’d give a current flow of 330uA, which should be safe for the DS7505 - it seems to show max OS current as 4mA, so this would put it a full order of magnitude below that.

But nothing on R2-R4. Are these just protection resistors? The datasheet gives a safe voltage range (-3V to +5.5V), for pins A0-A3, but in Note 6 simply assumes they are either Gnd or VDD.

Hope this is covered in the next video.

10K is fine for R1.
R2-R4 can be 0 Ohm or any small value such as 1K or less.
R5, R7 are not critical, depending on what you are trying to achieve with your measurement. As the LDR has a range of ~25K to 500K you could start with a value of about 22K for both. Alternatively you could use a small value (or even 0 Ohm) for R7 and a large value such as 100K for R5. Any necessary calibration can be done if software.

Testing the photo resistor I got shows a range of 18k when brightly lit (shining phone “flashlight” directly at sensor) down to 300Ohm when completely covered.A 10k resistor in R5 (and 0 in R7) should work - even if the resistance spikes to 20k due to even brighter light, PR_TOP should not exceed 3.3V.

The i2c video does state 0R for R2-R4.

Just wish all this was in one place - I want to finish building the board before starting to program it.

I was going by the LDR listed on the schematic that has a resistance that decreases when illuminated. 10K for R5 sounds okay in your case, this will give a voltage that increases when illuminated when measured at PR_TOP. To get a voltage that decreased with illumination you would swap the values for R5 and R7 and measure at PR_BOT.

This is the one I got:

As the datasheet shows, resistance decreases.

But now I’m concerned - the data sheet says 500k at full dark. That’d push almost the whole 5V across the ADC input, if used with R5 and PR_TOP.

But If I use R7 and PT_BOT, then the risk is on the other side - if it drops to 300Ohm in full bright…

I may need both R5 and R7 to protect the ADC…

So I hooked the cell up to my bench DMM with kelvin probes, because hey, I just got the thing, might as well use it.

Putting the whole assembly into a cardboard box did, in fact, get it to 500k.

Pointing it at a bright LED light dropped the resistance to .05Ohm. That’s right, 50mOhm!

That’s a heck of a range. Really wish the board used 3.3V instead of 5V…

I’m wondering if I can rig some clamping diodes to protect the ADC. Or just bodge in a connection 3.3V…

Okay, so you are using the same one that is specified on the schematic.

I don’t know why 5V was used if the analog inputs are 3.3V.

What I would do is use a value of about 820 Ohm for R5 and 10K for R7 and then connect a 3V3 zener from PR_TOP to ground and measure on PR_BOT.

Actually, there’s an easier solution! The adapter board has a jumper that selects 3v3 or 5v. Can just swap the jumper.

Unless the point of the exercise is to deal with the 5v?

Then yes, a 3v3 zener would be perfect. But I’d have to order one - not something I have on the shelf.