On The Inside

On The Inside

Monday, April 27, 2015

LC Meter

Oh look, it's my LC meter made from a kit. I find this useful, it's fast and accurate enough. It certainly gets a lot of use.  (Looks like I need to tap that cal button again)






IIRC, it was built from this design:





SmartCans Power, Doh!

Power in the edison system is something of an issue for any project that isn't designed from the ground up to deal with the various voltages involved. This includes, of course, most hacker style projects.

For this project I'm going to need several different voltages and one of the major concerns that I have is in powering the OTG port so that I can use a USB sound card. Any end product would forgo the USB interface here and tie a codec directly to the I2S bus, but, that's more than I want to do right now considering that it's not clear whether or not it will actually work.

Even if I could get it to work, most boards are a bit too large  for my needs and I don't want to make a custom surface mount board at this point only to find out that there is insufficient software support.

So, for the time being, I'll stick to trying to make the USB audio work. If I can power the OTG port via the console port on the edison then wer're home free. Coming up with a charging circuit/inverter to manage the 3.7 v battery and generate 5v for USB power is something that I can deal with on short notice. Fry's had just what I needed, a micro USB to OTG cable.



 Doh! I managed to pick up the cable that blocks the console port when the OTG port is plugged in.

Wednesday, April 22, 2015

Etched@Home PCBoard for the SmartCans Headphone Amplifier

The SmartCans are going to need a headphone amplifier that is of sufficient audio quality, and sufficiently powerful to satisfy typical critical listening by a musician or sane audiophile. In addition to decent sound, however, building an amp into headphones comes with other constraints. The two primary concerns are power consumption and size. For this post, I'm going to discuss size, I'll talk about power consumption in a later post.  If this were going to be a consumer product and I had a larger prototyping budget, I'd just layout everything with surface mount, and send the boards out to be done professionally.

Back to reality.

For many of my DIY projects I tend to use one of several rapid prototyping methods such as ugly, or manhattan construction.  Here's a really nice example of the latter from Jim Kortge, K8IQY, of a class AB audio amplifier. 


Parts are soldered to pads which are glued to the double sided pcboard which acts as a ground plane. Ugly construction uses the same idea, but just glues the parts to the board upside down instead of using fancy pads. Manhattan construction is fantastic for working out a project as you go yet still having something sturdy when you're done. If you've only ever breadboarded with solderless breadboards, you owe it to yourself to give manhattan construction a try. You can prototype very quickly building projects almost as fast as you can layout the schematic.

The limitation of this type of construction, however, is size.  It's really challenging to use this method for compact construction, it can be done, but you're going to have to work for it.

Point to point methods can be really compact, but they are usually time consuming and very difficult to correct mistakes. For compact construction, there's really no substitute for a printed circuit board. In some cases, a commercial pc board is easy to justify, if it's reasonably complex, yet reasonably small, I'd almost always rather send it off and pay the relatively high onsey-twosy price. ExpressPCB charges about $50 for their quick return three small board deal.  They also provide free software that you can use to layout boards and schematics. The intention is that you order boards from them, but, their software is useful even if you want to make your own boards.

Ok, back to the amp, here's the first amp I'm trying for this project. Sijosae even provides a nice layout for a simple single sided PCB.



This is one of those cases where it's just not worth it to get it professionally made, yet, the circuit is complex enough that I don't want to do it point to point. So for this, I chose to do it at home. I use the toner transfer method. Basically, you layout the board, print it out, iron it on to your pc board blank, clean off the paper, and etch. There are lots of great tutorials about this on the web.

So here's what you get when you do it at home with a few shortcuts. The pseudo-groundplane around the outside was just a quick afterthought since I'm still prototyping. It was created by filling in the region with a sharpie before etching. That's why it's rough in places. Had I used toner, it would be much cleaner.

The toner didn't transfer perfectly, so I touched up a couple of the traces with a sharpie (one incorrectly, as it turns out) and, unfortunately, the lettering didn't come out clean. I was experimenting with paper and my color laser, which isn't supposed to work well. For paper, I just used a page out of a magazine, it doesn't have to be white, the toner transfers, but the ink on the page does not. None of the traces are bridged, the one you see at the top is intentional. For such a quick turnaround, and not very much effort, I'm pretty happy with the results;  it will get the job done and it saved me a bunch of time. You can't beat the price either, all told, I think that's about a twenty five cent circuit board!


Fully populated the compactness of the circuitry is more readily apparent. The upright resistors have to have their body facing outwards in order to fit and the 8 pin dip socket has been modified to allow two additional resistors to live underneath the chip. Putting something like this together on a pcboard isn't too bad, trying to do it point to point is bound to be a solder bridge headache.


She's loaded up with an inexpensive TL072 for the initial tests, I'll replace that with something more expensive after she passes the smoke test.