/diy/ - Do It Yourself

[Anonymous No.544180]

Is MCP3008 our friend?

[Anonymous No.544195]

I'm a friend of all microcontrollers, except those mounted to arduinos PCBs

[Anonymous No.544198]

Nevermind that's an ADC not a microcontroller

It's ok i guess - Is it cheap at least?, because that resolution and sampling rate certainly isn't good by modern standards, but if you don't need good, cheap is good - if that makes sense.

[Anonymous No.544199]

>>544198
It's like 3 bucks on Digikey.

[Anonymous No.544201]

>2013
>through hole integrated circuit mounting

enjoy your parasitics and more expensive and less efficient PCB assembly

DIP a shit

[Anonymous No.544202]

>>544199
i like digi-key but often their prices seem higher than mouser and other suppliers by 10-15% at least

http://octopart.com/parts/search?q=MCP3008

granted some of those are probably 3k Tape&Reel's, but still

[Anonymous No.544222]

>>544201
Feel better now you have that sand out of your vagina honey?

[Anonymous No.544223]

>>544201
How do you prototype then?

[Anonymous No.544246]

>>544223
With SMDs obviously. He probably tries to get the design right on the first try so that he needs to experiment as little as possible.
He's still a faggot.

[Anonymous No.544250]

>>544246
It seems unreasonable to design and print a PCB for each iteration of experiment.

[Anonymous No.544255]

>>544250
Not really, if you need only one or two iterations to get it right. Or if it's a huge long term project where one PCB round more is nothing.
Well, it's a matter of taste.

[Anonymous No.544262]

>>544223
He doesn't. He gets it wrong because he doesn't into breadboard and perf board, then takes it out on parents/pets/image board inhabitants.

[Anonymous No.544266]

ewwwwwwwwwwwwwwwwwwwwwwwwwwwww

[Anonymous No.544287]

>>544266
Why is Adafruit pushing this garbage?

[Anonymous No.544290]

>>544287
> what is supply and demand
You don't always need a high sample rate ADC.

[Anonymous No.544424]

>>544287
Adafruit is selling crap to people who wouldn't know better. Their prices are over inflated for some of the Chinese crap they RESELL.

Remember kids, it is a corporation run as a business.

[Anonymous No.544490]

>>544424
> they had a good idea and little competition so they make a lot of profit so they suck
Maybe it's got to do something with the fact that they are one of the few who import this shit and sell individually?

[Anonymous No.544569]

>>544222
Twas a joke but hard to convey joking intent online sometimes

[Anonymous No.544570]

>>544490
>Maybe it's got to do something with the fact that they are one of the few who import this shit and sell individually?

>one of the few who import and sell individually

huh?

Mouser? Digikey?

Anyone listed on

>>544202

[Anonymous No.544572]

>>544570
But how will amateurs know what to buy for babby's first DIY?

[Anonymous No.544575]

>>544572

By learning what specifications of a part are important (Resolution in bits of an ADC, sampling rate, input bandwidth, power consumption, etc, etc) and then picking one that meets their needs?

Oh wait

[Anonymous No.544578]

>>544570
Sorry, you're right, was confused with sparkfun and the such which mainly sell solderless components...

[Anonymous No.545407]

>>544490
>Maybe it's got to do something with the fact that they are one of the few who import this shit and sell individually?
Or that they actually have support & docs for everything they sell.

[Anonymous No.546582]

>>544223

a small pcb milling machine built around a dremel, and a soldering iron.

Really, soldering most SMD devices is easier than soldering through hole, since you don't need to flip the board over and keep the part from falling out. Only the REALLY fine pitch stuff and shit like BGAs are hard.

[Anonymous No.546666]

>>544195
my nigga, TTL fucking died in the 80's why do they insist on using such shitty standards for everything, it costs fucking $10 in logic level shifting IC's just to connect them to anything remotely useful.
>>544287
you mean aidsfruit? the one run entirely by feminazis?

btw what does /diy/ use for MCU's? i've found the STM32F4's fucking 10/10

[Anonymous No.546718]

>>546666
I haven't done embedded in a while but ATMega32/MSP430 were common.

PIC seems really popular with hobbyists who know what they're doing, but I’m not sure if it's common in industry

[Anonymous No.546727]

>>546718
PICs are pretty fucking common, if its not a 2 cent chinamen knockoff that are secretly a PIC, its usually a PIC.

Next would probably be 8051's in terms of MCU popularity.

And yeah MSP430's are pretty awesome, i need to get a devkit for one eventually.

[Anonymous No.546788]

One project I worked on used MCP3008s to read user-twiddled potentiometers. Perfectly adequate for the job.

Hobbyists often stare at the numbers without considering the actual requirements.

[Anonymous No.546822]

>>546788
>User twiddled potentiometers
Unless you have like 8 potentiometers, you might as well get a PIC or an AVR since they're almost the same price and do basically the same thing. (But with added gizmos)

[Anonymous No.546837]

>>546822
On-board ADCs usually have notably worse performance (noise, nonlinearity), and are often limited to 8 bits. The MCP3008 also costs about a buck and a half in volume, at least at the time an MCU with the same number of analog channels would have cost far more. Add to that the cost of developing the extra firmware and programming in production. The device had a total of 10 pots.

[Anonymous No.547346]

What's a good alternative to MCP3008?

[Anonymous No.548577]

>>546666
STM32?? I was thinking of switching to them after playing with AVR's the last 2 years.
what software do you use??? Any tutorials?

[Anonymous No.548590]

>>546837
While it is true that putting an ADC into a process that is optimized for digital circuit would make ADC more "noisy", almost no "current" generation microcontrollers with 8-bit ADC. The minimum resolution is 10-bit, but most are 12-bit.

ARM chips comes with one or two fast 12-bit ADC at a price that is comparable with the older 8-bitters. They are going to eat the lunch out of all but the cheapest ones. That's one reason why most of the traditional microcontroller vendors already licensed ARM cores.

[Has the hobby of electronics d(...) No.548639]

>>544180
No, its our mortal enemy

if you ever see it shoot it and burn its corpose

[Anonymous No.548758]

>>546727
$10 shipped worldwide for the Launchpad. $13 for the new 5-series. Hell, the Tiva C-series has an ARM for $13. What's stopping you?

[Anonymous No.548862]

>>548758
Can I use a MSP430 microcontroller in place of a ADC like MCP3008?

[Anonymous No.548876]

>>548862

Sure, you can program it to emulate a dedicated ADC chip, but it's generally cheaper, easier and makes more sense to use an actual ADC instead.

[Anonymous No.548879]

>>544569
Smarmy retort retracted, as you were anon.

[Anonymous No.550192]

>>548590
What does the core have to do with ADCs? The peripherals are all added by the chip vendor. If the vendor has existing 8- or 16-bit lines you'll often see them using the same peripheral blocks. This both saves money and helps existing customers migrate to the new product line.

[Anonymous No.550446]

>>550192

Nothing directly, but the ARM core controllers are made with newer processes than many old 8b controllers and that makes 12b ADCs more practical with them.

[Anonymous No.550549]

>>544201
Do you even prototype?

[Anonymous No.551227]

>>550192
The peripherals don't look the same as their older 8/16 microcontrollers. ADC in ARM are DMA and have much better analog specs (faster, higher res).

Some of the older chips are probably not even done in HDL nor the much smaller geometry and done in different process, so putting in their peripheral is not that smart.

[Anonymous No.551229]

>>550549
Parts that are made in the last 10+ years, their process is so fast that parasitic or the limiting I/O in the packaging would hinder their performance. The physical layout (i.e. the transmission line effect, power plane impedance) in a PCB has became one of the components that you cannot duplicate in prototyping. What's the point if your prototype do not behave the same as the real thing?

Design something right by carefully reading all the datasheets, app notes, eval board schematics / gerber that you can get your hands on. Do a lot of simulation on signal integrity, analog simulation, thermal simulation, mechanical modeling etc up front.

Hold a peer design review for a preproduction board run. Most servers boards (lots of BGA and all SMT parts) at work are done that way and most of the time, they work within a couple of design cycles.

You'll find that you would cut down on the need for prototyping a lot. That is assume you know exactly what you are doing.

If you want to muck around with parts that went into production 20-30 years ago and trial/error to make things, go ahead.

[Anonymous No.551344]

>>551229
But, we are making babby's first DIY electronics, not developing a new product for mass production.

[Anonymous No.551347]

>>551227
Go compare eg. ST's STM32F101 and STM8 chips, you'll see that the control registers for many peripherals are identical. They even re-used the epically bad I2C block up until very recently. NXP have also re-used peripheral blocks both from their older ARM7TDMIs and various 8-bit chips. A tell-tale sign is when control registers are broken up into lots of individual 8- or 16-bit registers even when it doesn't make sense.

You can't make blanket statements like "ADC in ARM are DMA and have much better analog specs" because 1) "ARM" is only the processor core and says exactly fuck all about the peripherals and 2) there's way more varied implementations than you imagine. For example, the Stellaris LM1968 which I've got sitting on my shelf has no DMA controller at all. There are also old 8-bitters from vendors like Analog Devices and SiLabs that have far better analog performance than any commodity 32-bit MCU. High clock frequencies and small feature sizes are not good for signal quality.