Raspberry Pi Circuits

Forum: LinuxTotal Replies: 12
Author Content

Nov 19, 2014
9:04 PM EDT
Over the next few weeks [maybe months] or so, I'll post a few circuits that go along with or can be used with the Raspberry Pi. Make kindly phrased comments, suggestions, additions, or build a few yourself as you please.

I use gEDA for circuit drawing and can find a place to distribute the drawing files if requested.


Nov 26, 2014
11:37 AM EDT
Very Cool, can't wait to see!

Nov 29, 2014
5:51 PM EDT

Gonna try this one tonight.

Nov 29, 2014
7:28 PM EDT
It's not a good idea to drive an antenna directly from the Raspberry Pi GPIO. There will likely be a non-optimal impedance match... which will probably result in extra loading and/or reflection of transmitted power back into the transmitter pin. It's a much better idea to create a simple RF front end using a high frequency FET and then attach the antenna onto that...

I believe the maximum output is 16mA per pin. So, if you do drive an antenna, the maximum output is 80mW at 5V... or about 80 times the typical limit for unlicensed devices.


So be careful, you may get a knock on the door and a fine, like this guy:


Nov 29, 2014
8:06 PM EDT
Here's a partial and slightly inaccurate schematic for the dusklight circuit I promised... just so it is known that I haven't forgotten...

I'll post more later ---

Nov 30, 2014
1:15 PM EDT
It isn't "unlicensed." I have a Technician license. Besides, if I leave off the "antenna" (wire), the actual output is likely less than the FCC limits, given that the effective range is reduced to about 1 meter.

Nov 30, 2014
1:37 PM EDT

It's good that you have a license. I let mine expire... However, it is important to note the need for a license or limited power without a license. I don't remember exactly what transmit power is allowed within which bands as a Technician. I didn't take the Morse code test -- so I held a Technician license as well.

As far as antenna use goes... The reflected power may be significant, which may fry the BCM2835. It's good practice to isolate RF circuits from logic circuits. Without a isolation circuit, the output GPIO pin will be radiating it's maximum power directly at the other GPIO pins. So, besides the possibility of overloading that particular pin, there is a strong possibility of rendering the remaining GPIO pins unusable during RF transmission due to excessive noise. OR... it may just work with no problems... I don't know, because I haven't tested this particular circuit, just indicating best practices...

Nov 30, 2014
2:25 PM EDT
Not having internal BCM2835 schematics, I can't speak with authority. However, given that the RPi's pin header is a direct, unamplified connection to the BCM2835, I'd say that, if the proximity of the signal inside the chip doesn't fry things, then odds are they're safe outside the chip as well.

I worry more about the square-wave noise. Once the signal leaves the room, how much harmonic interference is it causing?

Nov 30, 2014
3:34 PM EDT
If you are using the header pin as an antenna, the signal should propagate as a omnidirectional plane wave... meaning circular wavefronts with very little signal being propagated in the 'z' directions...

Like this:


Transmitted power adheres to the inverse square law... so power decreases with the square of the distance. So the power at 2 meters will be 1/4 of the power at 1 meter. Since the antenna is not tuned to the transmitted frequency, it may be possible that harmonics could be quite powerful. For best performance the antenna should be 1/4 wavelength, 1/2 wavelength, or full wave length... at 100MHz full wave length is about 3 meters... I think the header pin is about 11 mm or 0.00367 wavelengths.

I haven't played around too much with automated smith charts, but you may be able to plot the RF circuit using xsmc available in most GNU/Linux repos:


Of course, the owner of the website seems to have a Technician license -- maybe it's even you... however, here's a bit of a tutorial on smith charts if you are not familiar with them...


An RF transmitter may seem like an open circuit, but it's not.

Please check and correct my numbers.

Nov 30, 2014
7:35 PM EDT
Heh. I only learned about Smith charts this afternoon.

My frequency of choice (for now, anyway) is 52.0MHz. I couldn't pick up any discernible signal on my hand-held, so hopefully I won't be interfering with anybody's transmission when I do "go live."

Nov 30, 2014
7:52 PM EDT
If you are learning about smith charts, then you must have learned about oscilloscopes... or I imagine...

If you are a hobbyist or a student, I'd highly recommend picking up an older oscilloscope on ebay. I've purchased lots of o-scopes there... most are old government or university owned. An o-scope may be the most versatile piece of test equipment, and ebay usually has several for around $150-$250 USD. I'd recommend a 1980s HP digital storage with GPIB or a Tektronics analog scope. I have the HP 54501A DSO that I use for most of my circuit tests.

Anyway, the point being -- if you have or have access to a high frequency o-scope you can easily check for interference on other GPIO pins and/or configure a relatively simple horn antenna with chamber resonator in order to check antenna patterns or RF power in at a given frequency and distance.

Nov 30, 2014
9:11 PM EDT
I had an o-scope once upon a time. It disappeared when my unicorn ran away. :(

Dec 01, 2014
9:30 PM EDT
I noticed I did not have any high power SPDT relays left in my parts bin... I've placed them on order, and will update this thread with schematics and circuit test results when I receive the relay order.

You cannot post until you login.