Electrical Engineering V3

[paindoc]

Anyone know of a decently cheap breakout board for an audio decoder that could manage playing a few sounds at once, or at least some way to get something like a Teensy to do that job?

I'm refurbishing my old RC tanks, starting with my partially-repainted and kinda still ugly King Tiger. The internal electronics are utter shit, so I'm replacing those with an arduino and adding a 9ch receiver to work with my fancy plane transmitter. I'm not yet diving into replacing the speakers and audio stuff and such, since I'll just be using a mega 2560 to control the damn thing, but it is something I want to do once I get the thing moving again. I'd like to just [URL="http://openpanzer.org/"]buy one of these boards[/URL] for controlling things, but they're dealing with manufacturing issues and have no window for availability. For audio, [URL="http://openpanzer.org/wiki/doku.php?id=wiki:tcb:tcbinstall:sound_op_v1"]I'd like to try this[/URL] by ordering the bare board (maybe), but I need to total up the materials on the BOM and see what it'd cost me.

If I do order bare boards from OSH park, the bit about having to order 3 works since I have another tank to refurb, and I'm definitely looking at getting a modern vehicle so I can play with getting some kind of vertical stabilizer working. Really, I just want to justify buying another of the damn things :V

I'll have to post pics shortly, once I get internet at my new place. To get away from the software-only work I do at work now and to take a break from my game projects I decided it was time to return to my roots (of where I started programming, at least)

[editline]edited[/editline]

seems like this might do, and a teensy would be easy to fit behind the transmission/gearbox [url]https://www.pjrc.com/store/teensy3_audio.html[/url]

[LoneWolf_Recon]

[QUOTE=paindoc;52576258]Anyone know of a decently cheap breakout board for an audio decoder that could manage playing a few sounds at once, or at least some way to get something like a Teensy to do that job?

I'm refurbishing my old RC tanks, starting with my partially-repainted and kinda still ugly King Tiger. The internal electronics are utter shit, so I'm replacing those with an arduino and adding a 9ch receiver to work with my fancy plane transmitter. I'm not yet diving into replacing the speakers and audio stuff and such, since I'll just be using a mega 2560 to control the damn thing, but it is something I want to do once I get the thing moving again. I'd like to just [URL="http://openpanzer.org/"]buy one of these boards[/URL] for controlling things, but they're dealing with manufacturing issues and have no window for availability. For audio, [URL="http://openpanzer.org/wiki/doku.php?id=wiki:tcb:tcbinstall:sound_op_v1"]I'd like to try this[/URL] by ordering the bare board (maybe), but I need to total up the materials on the BOM and see what it'd cost me.

If I do order bare boards from OSH park, the bit about having to order 3 works since I have another tank to refurb, and I'm definitely looking at getting a modern vehicle so I can play with getting some kind of vertical stabilizer working. Really, I just want to justify buying another of the damn things :V

I'll have to post pics shortly, once I get internet at my new place. To get away from the software-only work I do at work now and to take a break from my game projects I decided it was time to return to my roots (of where I started programming, at least)

[editline]edited[/editline]

seems like this might do, and a teensy would be easy to fit behind the transmission/gearbox [url]https://www.pjrc.com/store/teensy3_audio.html[/url][/QUOTE]

Looking at the BOM for the audio board there's a few components you might have trouble soldering if you're not used to leadless SMT. But other than that you'll probably be paying ~$35-40 for PCB and parts. So it looks pretty good.

The only other polyphonic sound effect breakout board that I could find is somewhat expensive but has all the bells and whistles that you want. [URL="https://www.sparkfun.com/products/13660"]~$50 on SparkFun.[/URL] and it'll still require a separate Class D amp if you want to drive it loud enough compared to just a headphone jack on the wav trigger.

[URL="https://www.sparkfun.com/products/12767"]This $15 one seems to atleast have up to four sounds polyphony[/URL] but also requires a Class D amp. And this [URL="https://www.adafruit.com/product/1552"]$10 Class D breakout[/URL] seems to be on par with Open Panzer's board.

[paindoc]

[QUOTE=LoneWolf_Recon;52576424]Looking at the BOM for the audio board there's a few components you might have trouble soldering if you're not used to leadless SMT. But other than that you'll probably be paying ~$35-40 for PCB and parts. So it looks pretty good.

The only other polyphonic sound effect breakout board that I could find is somewhat expensive but has all the bells and whistles that you want. [URL="https://www.sparkfun.com/products/13660"]~$50 on SparkFun.[/URL] and it'll still require a separate Class D amp if you want to drive it loud enough compared to just a headphone jack on the wav trigger.

[URL="https://www.sparkfun.com/products/12767"]This $15 one seems to atleast have up to four sounds polyphony[/URL] but also requires a Class D amp. And this [URL="https://www.adafruit.com/product/1552"]$10 Class D breakout[/URL] seems to be on par with Open Panzer's board.[/QUOTE]

For the leadless SMT stuff, I can probably solder those here at work, since I just got my ESD recert and can go into our production area. I've done [I]some[/I] SMT stuff before, just not very well. Having done a more bit more looking around and peeking through your links, it looks like the open panzer board is pretty competitive tbh.

If its not too much to ask and you've got the time, what would you estimate the price of the BOM ([URL="http://www.openpanzer.org/downloads/tcbmk1/bom/OpenPanzer_TCB_Mk1_BOM.pdf"]link[/URL], [URL="http://openpanzer.org/forum/index.php?PHPSESSID=64f7ecfa5c6c4407221a863e50eaab93&page=download"]main page link[/URL]) for the control board to be, at a glance?

[LoneWolf_Recon]

[QUOTE=paindoc;52576472]For the leadless SMT stuff, I can probably solder those here at work, since I just got my ESD recert and can go into our production area. I've done [I]some[/I] SMT stuff before, just not very well. Having done a more bit more looking around and peeking through your links, it looks like the open panzer board is pretty competitive tbh.

If its not too much to ask and you've got the time, what would you estimate the price of the BOM ([URL="http://www.openpanzer.org/downloads/tcbmk1/bom/OpenPanzer_TCB_Mk1_BOM.pdf"]link[/URL], [URL="http://openpanzer.org/forum/index.php?PHPSESSID=64f7ecfa5c6c4407221a863e50eaab93&page=download"]main page link[/URL]) for the control board to be, at a glance?[/QUOTE]

I'd say about ~$35 to $40 in parts.

[paindoc]

[QUOTE=LoneWolf_Recon;52576518]I'd say about ~$35 to $40 in parts.[/QUOTE]

Damn, thanks for giving that a look then. Going to have to see about ordering all the bits I need to get two of those operating

[WontLurk]

I was doing a homework today and one of the questions was to write what is the current passing through a generic LED when it is polarized reversely. I quickly google the datasheet of a red LED and found the "reverse voltage" characteristic, which had the measurement of the current passing through it. It had a weird symbol in it, it was specified as 100 lambda-ampere (an actual lambda symbol). What does this mean?

[Chipset]

[QUOTE=RockyTV;52595398]I was doing a homework today and one of the questions was to write what is the current passing through a generic LED when it is polarized reversely. I quickly google the datasheet of a red LED and found the "reverse voltage" characteristic, which had the measurement of the current passing through it. It had a weird symbol in it, it was specified as 100 lambda-ampere (an actual lambda symbol). What does this mean?[/QUOTE]

link the datasheet? Lowercase lambda usually refers to wavelength.

[Chryseus]

LEDs do work as photodiodes with sensitivity at equal and lower wavelengths, although I'm not aware of any reliable equation to calculate such, it would depend on the LED used I would think.

[WontLurk]

[QUOTE=chipset;52595474]link the datasheet? Lowercase lambda usually refers to wavelength.[/QUOTE]

[url]http://www.alliedelec.com/m/d/6355b8aba0b01578df0bb7b871ceefd7.pdf[/url]

It's literally Lambda-ampere.

[pentium]

[b]NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO[/b]

[quote]*Dear Valued Customers*

We would like to inform you that due in part to the passing of both Phil and Daniel,
the founders of our family business, Main Electronic Supplies is now "Closed."

We are currently looking to sell the whole lot or large sections of our inventory.
If you or anyone you know (including businesses) are interested,
please email us at: [email][e-mail][/email]

Once our inventory has found a new home,
we will gladly provide an update below as to where it has moved.

We sincerely thank you for your understanding and continued support since 1971!
[/quote]
-[url]http://www.mainelectronics.com/[/url]

That was the last place I knew of where you could ask for help at the desk and they knew what you needed help with. Danny helped me extensively when I rebuilt that Westinghouse television.
I'm going to miss you. :cry:

[DrLuke]

Sorry to hear that :(

[r0b0tsquid]

[QUOTE=RockyTV;52595922][URL]http://www.alliedelec.com/m/d/6355b8aba0b01578df0bb7b871ceefd7.pdf[/URL]

It's literally Lambda-ampere.[/QUOTE]

The units are number of wavelengths times number of amps. Dimensionally, it's still a current, because "number of wavelengths" is dimensionless.

So if the LED's wavelength is 660nm, and they specify the current as "100 λA", then I would read that as the current being 66 uA.

This is a complete guess, and you should buy an LED and measure it if you want to confirm it, rather than taking me at my word.

The only reason I could see you wanting to do that, is that there is a (kind of) inverse relationship between λ and the reverse leakage current due to a smaller bandgap at longer wavelengths, and this is a convenient way of specifying a maximum leakage to be expected across [I]all[/I] of their LEDs of different wavelengths.

This is supported by the suspiciously round number they give. I would guess this is a very conservative maximum.

[editline]blah[/editline]

In this case the notation is kind of dodgy, but parameterising things in terms of wavelength is common. [URL="https://en.wikipedia.org/wiki/Planck_constant"]Lots of things scale with it.[/URL]

[LoneWolf_Recon]

[QUOTE=RockyTV;52595922][url]http://www.alliedelec.com/m/d/6355b8aba0b01578df0bb7b871ceefd7.pdf[/url]

It's literally Lambda-ampere.[/QUOTE]

[QUOTE=r0b0tsquid;52601188]The units are number of wavelengths times number of amps. Dimensionally, it's still a current, because "number of wavelengths" is dimensionless.

So if the LED's wavelength is 660nm, and they specify the current as "100 λA", then I would read that as the current being 66 uA.

This is a complete guess, and you should buy an LED and measure it if you want to confirm it, rather than taking me at my word.

The only reason I could see you wanting to do that, is that there is a (kind of) inverse relationship between λ and the reverse leakage current due to a smaller bandgap at longer wavelengths, and this is a convenient way of specifying a maximum leakage to be expected across [I]all[/I] of their LEDs of different wavelengths.

This is supported by the suspiciously round number they give. I would guess this is a very conservative maximum.

[editline]blah[/editline]

In this case the notation is kind of dodgy, but parameterising things in terms of wavelength is common. [URL="https://en.wikipedia.org/wiki/Planck_constant"]Lots of things scale with it.[/URL][/QUOTE]

This is why there are dedicated photodiodes that typically provide their units in A/W or V/uW (if using a built-in Op-Amp). I agree those units are dodgy and non-standard.

I honestly suggest getting a dedicated photodiode like the OPT101 and use that as a reference for calibrating against LEDs to use as colored photodiodes (Ofc the A/W will vary from LED to LED even within the same lot/batch). You might also want to buy some colored filters to use with the photodiode as well.

[r0b0tsquid]

[img]http://i.imgur.com/Oqw1NMy.png[/img]

eurocircuits.com can go die in a fire

[pentium]

I love it.

I've always avoided circuit bending as a cardinal sin considering how hideous and butchered you end up making shit just so it beeps differently.
Then along comes a nasty 70's electronic organ (Lowrey "Teenie Genie". Who the fuck thought that was a great name? :v: ). The cabinet was moldy but the actual organ came out in one piece, plus the pedal and the speaker. Now I have a purely analog organ with no cabinet but totally open for some fucking about with the oscillators (the Flute is literally just the sine wave generator). I feel like actually circuit bending it.

Lmao, nope. Googling about seems to find that circuit benders are lazy fucks. There seems to be several models of toys and digital keyboards they like to carve up (and drive the price up in the process) but when it comes to that tacky as fuck electronic organ your grandma had at her house when she died and now its at the end of the driveway because you couldn't get rid of it on craigslist? Nothing. I can literally find digital keyboards made by Lowrey they circuit bend but no love for the analog units.

[Zero-Point]

So these MPU6050 boards I got some time ago stated that they could be set to one I2C address or another by tying a certain pin to either 3.3V or ground. Normally when tinkering with mine, I set it to ground, but when building a "proper" (read "ugly as fuck") perf-board that plugs into my BeagleBone Black, I tried tying it to 3.3V.
The internal temperature sensor was reading 135 degrees Fahrenheit when I did this. :wideeye:

[r0b0tsquid]

[cpp]module spi_master(
input wire clk,
input wire rst_n,

input wire [7:0] din,
input wire din_valid,
output reg din_ready,
output wire [7:0] dout,

output reg sck,
output reg mosi,
input wire miso
);

reg [2:0] bitcount;
reg [7:0] shift;

assign dout = shift; // valid when din_ready is high

always @ (posedge clk) begin
if (rst_n) begin
if (din_ready) begin
if (din_valid) begin
shift <= din;
din_ready <= 1'b0;
end
end else begin
if (sck) begin
bitcount <= bitcount + 1;
shift <= {miso, shift[7:1]};
mosi <= shift[0];
sck <= 0;
end else begin
sck <= 1;
end
if (bitcount == 7) begin
din_ready <= 1'b1;
end
end
end else begin
mosi <= 1'b0;
sck <= 1'b1;
din_ready <= 1'b1;
end
end

endmodule // spi_master[/cpp]

[cpp]>>> REPORT
Number of flops: 14
555s for flops: 84
555s for logic: 289
Total 555s: 373
luke@thinkpad:~/proj/fivesynth$
[/cpp]

Getting there! The logic is pretty huge at the moment because I haven't implemented any of the post-mapping operations.

[r0b0tsquid]

Baby steps. It can now dump out a verilog netlist for the 555s, which I can simulate with a behavioural 555 model.

flop.v:

[cpp]module flop( input wire clk,
input wire i,
output reg o
);

always @ (posedge clk)
o <= i;

endmodule
[/cpp]

synthesis result:
[cpp]module synth_top ( input wire clk,
output wire o,
input wire i
);

wire __n_0;
wire __n_1;
wire __n_2;
wire __n_3;
wire __n_4;

triple5 u0000 (__n_2, __n_0, 1'b1, o);
triple5 u0001 (__n_1, clk, 1'b1, __n_0);
triple5 u0002 (1'b1, __n_2, i, __n_1);
triple5 u0003 (1'b1, clk, 1'b1, __n_3);
triple5 u0004 (1'b1, __n_3, __n_0, __n_4);
triple5 u0005 (1'b1, __n_1, __n_4, __n_2);

endmodule
[/cpp]

Xilinx ISIM:

[IMG]http://i.imgur.com/Y77DR3w.png[/IMG]

[editline]5th September 2017[/editline]

DrDevil what's so funny? You don't take 555s seriously?

[editline]5th September 2017[/editline]

Simulating a counter crashes ISIM :goodjob:

[r0b0tsquid]

[IMG]http://i.imgur.com/vBp6JMz.png[/IMG]

That's 402 555s in action right there! I also synthed a small CPU at 1410 555s

[r0b0tsquid]

Went from 402 555s to only 305, for the SPI master. 84 of those are for flops. Added 2 optimisation passes:

- Simple rules-table based pass which does things like turning muxes into and/or gates where possible. This gets us from 402 to 372 555s.
- Simulated annealing pass, applying De Morgan's and moving inverter bubbles around, using number of mapped 555s as a cost function. This gets us the rest of the way down to 305.

This is after the standard logic minimisation algorithms in the synthesis frontend (I'm using YoSys). 555s are a pain to map logic onto, because there are some strange functions they implement very well (like !A & B) but they aren't very good at NAND/NOR etc.

[hakkoru]

man i need to pick up verilog again. i really enjoyed it in my introductory class to it because the prof was awesome, but it was immediately soured by the advanced class because the prof sucked.

what resources/books do you recommend?

[r0b0tsquid]

[QUOTE=elitehakor;52656290]man i need to pick up verilog again. i really enjoyed it in my introductory class to it because the prof was awesome, but it was immediately soured by the advanced class because the prof sucked.

what resources/books do you recommend?[/QUOTE]

It's amazing fun!

I'm not really following any books. I use [URL="https://inst.eecs.berkeley.edu/~cs150/fa06/Labs/verilog-ieee.pdf"]the standard[/URL] when I need definitive answers (you can safely ignore large parts of it unless you're doing crazy modelling), and the usual StackExchange + google combo otherwise. I think the main thing is to have projects to get stuck into, that's one way of finding out what you need to learn about next.

SPI went back up to 320 555s when I told the annealing pass it wasn't allowed to invert any signals entering/leaving the module :badzing:

[hakkoru]

thanks! i'll definitely give it a skim-through

and also yeah i agree, doing projects that you really enjoy is my favorite way to learn. my biggest personal project right now is in a language that i have limited proficiency in and i'm learning a lot of new things on the way.

[pentium]

I'm working on installing a relatively simple burglar alarm but it uses a momentary circuit to arm and disarm the system. Radio Shack used to sell for about $35 a self-contained digital keypad for this purpose (item 49-535) but clearly they do not anymore and everyone wants $75-$90 per keypad which is nuts.
Thankfully however because Radio Shack was awesome that way the complete schematic for the keypad as well as the full theory of operation is printed in their instruction manuals. ([url=http://schematicsforfree.com/archive/file/Alarms/Products/Alarm%20-%20Radio%20Shack%2049-535%20Digital%20Key%20Switch.pdf]>clicky clicky<[/url])

I found a great source for nice looking keypads (Moose MPI-275, specs can be found [url=http://www.pdf-zoo.com/moose-products/7305/mpi-275-digital-remote-control-station/preview.png]here[/url]) with the two LED's for $3 from [url=https://www.surplussales.com/switches/SWKeypads-1.html]this surplus website[/url]. My idea is that because the Radio Shack keypad operation is so simple I could put an arduino in the Moose keypad and make a near perfect clone. Does this seem like a good idea? I don't really like the current designs being passed around for Arduino controlled security keypads, nor do I like those absolutely shitty 4x4 membrane keypads.

[F.X Clampazzo]

Could also use a Teensy if you want to get really in depth here and then your keypad could be quite literally just a controller board, you'd then be able to write drivers and have the ability to make a pretty much modular security system on a linux or even windows environment in the event you're wanting to upgrade from basic to cameras and shit. (they also have an addon for working with Arduino IDE) Otherwise I don't see a problem, no.

[url]https://www.pjrc.com/teensy/index.html[/url]

[pentium]

Not looking for that though, which is partially my initial problem because a lot of the cheap keypads on the market add those kinds of bells and whistles which I don't need and create potential problems. There is absolutely nothing on the cheap side from china that is nothing short of a clone of the Radio Shack units. They all have [I]some[/I] gimmick.
In order for me to find again the kind of stuff I am looking for in a keypad and only those features I gotta go above $150 per keypad with products like the IEI 212 series which has remained relatively unchanged since it was announced about 30 years ago and you can still buy brand new.

An Arduino Pro Mini should be more than enough.

[F.X Clampazzo]

Well with a Teensy you'd be able to build the keypad for much cheaper, which was my entire point, the 8bit one is only $16 and you'd be able to build a keypad for rather cheap on top of that. idk how else to help you because as it seems you've done far more than enough research to say that it's unlikely you'll find the exact product you're wanting without building one.

[Van-man]

[QUOTE=F.X Clampazzo;52665145]Well with a Teensy you'd be able to build the keypad for much cheaper, which was my entire point, the 8bit one is only $16 and you'd be able to build a keypad for rather cheap on top of that. idk how else to help you because as it seems you've done far more than enough research to say that it's unlikely you'll find the exact product you're wanting without building one.[/QUOTE]

Teensy 2.0 is essentially a Arduino Pro Micro or Leonardo, which there is plenty of Chinese clones of.

So it depends on the amount of IO pins he needs whenever or not he's better off with the 2.0++ [sp]or add some shift registers[/sp]

[F.X Clampazzo]

That's fair. I'm not one for Chinese clones but if that's what he's after then yeah that'd work too.

[thefreemann]

Uni closed for a week during Harvey and now we're playing catchup :v:

Learning somewhat advanced circuit theory and the physics behind material properties at 2x speed is not fun :cry:

Also started learning Verilog today! It's actually kinda fun to use