This entire project has been a learning experience for me. When I look back at the pictures while writing these posts, I cringe a little bit at how I went about doing it. Only because I know better now.
I only started learning about electronics around February of this year. No formal training, just learning how I learn best: by doing (and lots of Google-Fu).
I learned to solder by salvaging components off of scrapped electronics and at the same time, learned about each of those components and their uses. If I needed a certain value resistor, I would first look through the hundreds that I salvaged, then go searching for the specific value on random boards and then de-solder it. That became a nightmare, so eventually I just ordered a bunch from AliExpress.
Last we left off, I had made a successful prototype of a VU Meter. In case you’re wondering what that is, a VU (volume unit) meter simply displays the output audio signal level.
Here we can see that I converted the prototype to a PCB, and just need to wire the LEDs to the board.
I made the decision to use a 12v power supply and create a voltage regulation board that will send 5v to the Raspberry Pi and 12v to the clock LEDs, VU meter and fan (A fan was definitely unnecessary. I tend to add fans to everything I do).
We can see that I have a prototype voltage regulator set up here. Underneath that is the supply I will be using. It just needs to have some wires soldered to the input and output. Fun fact: the 12v supply is from an Apple Airport Extreme!
Converted the prototype to a permanent PCB and tested it out. 4.94v is good enough for me.
Testing out what colour I want to use for the clock back-lighting. Settled on combining a blue and red LED. Also added a potentiometer to the back-lighting to adjust it’s intensity.
We have everything powered up and working in this picture. Mummified the power supply in electricians tape, as you do.
Finally I decided it was time to come up with a better solution for the RPi USB breakout. Unfortunately, here is where I accidentally brick the Raspberry Pi by dropping a tiny bit of solder on the board and bridging a capacitor. I just haven’t realized that yet.
Probably the most difficult part of this project is getting everything to fit into the relatively small space available inside the wooden enclosure. You can see that much hot-snot was used.
Also, this is where I decide to add a fan because reasons.
The component with the white wires with red markings is a rotary selector switch. It has been wired up to switch the audio input source from the 3.5mm jack to the Raspberry Pi’s audio out.
Found a place to mount everything on the metal chassis. Zip-ties and hot-glue for the win.
The front side, everything working. The control potentiometers from left to right are: Tone control, volume control and audio source selector.
The chassis has been stuffed into the enclosure. By this point I have not added a clear-coat to the enclosure. It looks way better after that.
Testing everything out. This picture was taken after I found out that the Raspberry Pi was not working. After that, I put this project aside for quite some time.
After my disappointment wore off I had another attempt at getting this project up and running. Between the last photo and this one I finished the enclosure with 3 or 4 coats of varnish.
I don’t have any photos of putting this revision together, but we can see the chassis in the background.
Here is the more simplified revision I came up with. The changes are as follows:
- A new speaker that actually (almost) fits inside the case.
- A smaller, more contained 12v power supply.
- Re-made the VU meter’s LED bar with larger and more LEDs (also realized I only needed one wire from each LED instead of two).
- Re-positioned the VU meter board.
- Removed the non-functioning Raspberry Pi.
- Removed the piece of board used for mounting the RPi.
Everything set up and looking fine. I was happy with how the clock back-light turned out.
Another setback: It wasn’t until I put the faceplate on that I realized the VU meter’s LED bar was not aligned with the slot on the right. I fixed this by cutting the slot taller so that it now extends to where the FM frequency listing was.
At this point I set the project away, thinking it was finished. Before this revision I had attempted to troubleshoot the seemingly dead Raspberry Pi, but to no avail. Until I made a surprisingly strong magnifying glass with some lenses from an old Polaroid camera and decided to have a look at all the components on the RPi’s board to see if I accidentally unsoldered anything. After a bit of looking I found a tiny capacitor that had been bridged with a speck of solder. Once that was removed, it worked just fine and I went about adding it back to the clock radio.
A sizable amount of time had passed between this picture and the last. Everything is the same except for the addition of the Raspberry Pi, a converted 5v USB charger, a potentiometer on the front for adjusting the VU meter’s sensitivity, a selector switch and the final revision of the RPi’s USB breakout. Plugged into the USB breakout is a WiFi adapter which allows the RPi to access my Plex music library over the local network. I learned that it had to be routed outside of the case to avoid audible RF interference with the speaker.
Overall, I was satisfied with this last revision. I thought I was finished with it, but after some use, I realized that the audio would distort quite easily and I wanted to fix that.
This is where I refer back to the beginning of Part One where I point out the rating on the radio’s original speaker. 0.8 watts at 8 ohms. The amplifier was also only being powered with 6v.
The entire time I was working on this project, I never stopped to think that “maybe the radio’s original amplifier just cannot provide the type of power that I have in mind”.
After this realization, it was back to the drawing board. I wanted to create my own amplifier from scratch to put into the clock radio. So I set out to learn about creating amplifiers.
I had another problem though, all the circuits I was learning about required a split power supply, so one that supplies a positive and negative voltage, like +18v and -18v. Similar to a battery. I had no such supply (the closest being a computer psu with a +12v and -12v rail), so I wanted to create my own. Fast-forward to after I have tested every transformer I could find in the house for one that was centre-tapped, and could provide 18-0-18 VDC. I was out of luck, and this project was yet again set on the back-burner.
A couple days ago, after checking out the datasheet I had on the TDA2050 amplifier IC (ordered these from AliExpress months ago), I realized that this IC would function with a single power supply. I then remembered why I ordered these TDA2050s in the first place! First thing I needed to do was find a beefy power supply. The one I started out with was 19v @ 4.95A which was adequate.
This is what I came up with. I am delighted that I finally have a working amplifier that lives up to my expectations. I wondered how much more power I could get out of the TDA2050, which can theoretically output 50 watts, if I created a higher voltage power supply and added a larger heatsink.
This leads to the next revision of the clock radio which I will go over in Part 3. All that I will say now is that there will not be a clock involved…