-
Price Tag
01/09/2020 at 16:29 • 0 commentsI've managed a quote from PCBway for this board: including parts and assembly, it would cost $138 for five, which comes to $27.60 each. That's not terrible; there are FeatherWings on the market that cost more. I wish it could be lower, though.
The total component price is $88.20 -- $17.64 apiece, not counting the board itself.
(For comparison, the BOM at Octowing tells me I can get all the components for $7.75 going with the lowest overall price, or $14.43 if I get them all from Digi-Key -- in both cases not including shipping.)
The real killer is the DAC8571, which runs $5.25 per unit (online prices run $3-6 apiece). That's almost one-third of the component price, and almost one-fifth of the overall price!
At $3.88 on their BOM, the MCP3422 isn't much better (though prices I've found have been in the $2-3 range).
I'd love to find something less spendy for the DAC and ADC; and, like I say in the Description, one chip to perform both functions would really be ideal. Any recommendations?
-
Don't Panic!
12/24/2019 at 15:53 • 0 commentsI fixed a couple of minor things this morning: a typo in the description, and the solder header on the PCB (which had been represented by a row of vias). With a week to go before the contest deadline, I'm pretty sure this will be the final revision.
Looking at the other entries in the Take Flight With Feather contest, I think this one has a decent shot at taking the top prize, though I'm far from convinced -- there are a lot of turkeys in there, but some really cool ideas too, including a couple that I wouldn't mind having myself.
If it doesn't win, I'll probably see about starting up some real-world experimentation in the spring.
-
"But I'm feeling much better now!"
12/07/2019 at 18:52 • 0 commentsI decided to rework the I2C version of the board, as though I was putting it together from scratch. I started with the ADC and DAC sitting right next to the I2C ports, then I set up the mike and speaker connectors to the other end of the board with their respective chips nearby. Then I put in the passive components one at a time, using the shortest routes possible. I think it came out fairly well. Certainly there are fewer twists and turns around the board (though about twice as many vias),
I'm calling it the "semifinal" because it has yet to be tested; passive components may need to be added, removed, or changed in value, and other changes may be needed.
Heck, if I was confident that the CoolAudio V1000 was I2C-compatible, I would've added it. I'm pretty sure there'd be room. (Not 100%, but fairly sure.)
We'll just see what develops from here.
-
Going with the I2C version (for the contest, at least)
12/06/2019 at 19:36 • 0 commentsI've made a new version of the board, which I think I'll treat as my "official" contest entry version. It replaces all of the resistors with 0602 SMD versions, removes all of the wire jumper connectors, and redirects the input and output to a couple of chips for ADC and DAC (respectively) with I2C connections. I probably need to add a couple of pull-up resistors for the I2C wires, but there's room for that; I made sure of it. I just need to work out the best values.
Once I make that modification (and possibly others based on viewer feedback), I'll post that update and start editing the Details section to reflect what I've done.
PS: Based on a commenter's recommendation, I've downloaded and tried out KiCad. I'm having a bit of a learning curve with it, but I'll get used to it in time. For now, for anything I want to complete I'm going to continue with Fritzing, and then try to duplicate with KiCad until I get the hang of it.
(I'd actually love to see what any of my boards -- but especially the two connector boards -- would look like in 3D!)
-
SMD vs TFT: I'm asking you...
12/04/2019 at 15:47 • 1 commentHere's a question for all of you folks following this project:
How much of an improvement would it be to convert all of the resistors to SMD (instead of TFT)? It would save a good bit of space, but I'm not sure what I'd be trading off for that.
(I'm re-learning electronics after some 40 years away, so bear with me here! I'm not sure there even were SMD components back then!) -
Darn you, bias!
11/29/2019 at 16:08 • 0 commentsWell, it turns out that I wasn't quite as done as I'd thought. That hanging bias pin was bothering me, so I took a look at what would be needed on it. I didn't really find anything specific for this circumstance, so I took a cue from the data sheet and added a 100K resistor and 100pf capacitor there, connecting the series to Ground. I think only experience and/or experimentation will tell for sure whether that's the right approach, and whether those are the correct values.
I do hope I don't have to add anything else. Board space has become premium. I'd have to seriously juggle some things around for that. (EDIT: I probably will need to do that anyway, so what's just been done will probably not be the final version of things.)
-
Design phase finished! (I think.)
11/28/2019 at 16:19 • 0 commentsI took some time last night to complete the board's design. After a last-ditch effort to see if there was some way to include I2C on this board (short answer: no flippin' way), I moved things around so the mike output could go to any of A1-A5 with a wire jumper.
In a satisfying twist, removing the trace from the mike output to A5 allowed me to neaten up the other traces in that part of the board considerably.
I've also checked the measurements for the headers and mounting holes to make sure they match up with the Feather standard. They didn't, but they do now.
Overall, I feel pretty satisfied with how this has turned out so far. I may tweak the design a little, mostly depending on any feedback I get, but I think it looks pretty good from here.
It'll be at least January before I can start actually getting the hardware to test this out, which I'll do depending on (among other things) what happens with the Take Flight With Feather contest.
-
Not as "done" as I'd thought
11/27/2019 at 18:01 • 0 commentsFrom the Department of Always Double-Check Your Stuff: The A1 pin on the Feather is not usable for analog output, so A0 is the only place to hook up the NCS2211. The A1 has thus been completely cut out of the equation.
While I was at it, I migrated the components as much toward the left and bottom ("southwest") as I could, leaving lots of room around the I2C pins. Not surprisingly, there's still not room for an I2C chip, especially considering that two of the ports would have to be analog. I'd still love to make this I2C compatible, but unless I'm missing something here, it's just not meant to be. (There might be room there for something else, though. I just don't know what, yet.)
-
A different approach to the mike
11/25/2019 at 16:42 • 0 commentsOkay, I was wrong. I just did a big update to the circuitry around the MAX4468.
That had been bothering me a bit, and I wanted to see if what I had could work with a piezoelectric mike, so I did a little digging and found that my approach was (as far as I could figure) altogether wrong, and would probably not work with an electret either. The electret was the main point (the piezo is only secondary), so I needed to fix things.
The recommendation I found was basically to rely on the schematic for the MAX4466. I'd switch to that chip altogether, but now that I have a grasp on what the Shutdown does I'd like to keep that functionality. This does leave the Bias pin open, though; I'm not sure whether I should do that (which is my impression at present) or hook it up to Ground, 3.3V, or IN+.
I also moved all of the electronics for the MAX4468 as far west as I could (though there's probably room for even more), just to leave more room for new electronics for the NCS2211 (or something altogether new).
Anyway, I've updated the parts list, and I've also kept the previous version of the board here in case anyone wants to compare them (and tell me that I had the whole thing right to begin with -- even though I'm pretty sure I didn't).
I still wouldn't want to try this with a piezoelectric mike without breadboarding first, but I think I at least have a model that would be worth trying out on an electret now.
-
Mostly Done (The Design Stage, Anyway)
11/23/2019 at 21:10 • 0 commentsI think I have a handle on the Shutdown pin; it looks like a nice, power-saving Mute button. Thus, I've put in a Jumper to allow the user to take its signal from D12, or just sit on the Ground.
Other than that, I've more or less left things as they were.
I'm contemplating replacing the TFT components (the resistors, anyway) with SMD versions to reduce space, but I'm leaning against it. I'm not sure how much benefit it would bring; and, from what I can tell, hand-soldering would be a pain in the keester.
That, and the aligning of the Feather-related headers and mounting holes, are basically just "design tweaks." Unless someone posts something about it, I'm done with adding, changing, or significantly rearranging the actual components.