SparkFun MicroView Review(rheingoldheavy.com)
rheingoldheavy.com
SparkFun MicroView Review
http://rheingoldheavy.com/sparkfun-microview-review/
3 comments
@userbinator I'm Marcus from Geek Ammo (the designers of the MicroView). Our mission, like SparkFun’s is to make electronics accessible to anyone.
I carefully chose the price.
I previously founded Ninja Blocks (ninjablocks.com) and one of the things I learned was that when you price a product too low it is impossible to get distributors. This is what happened with the original Ninja Block.
This time round I wanted the MicroView to be available through SparkFun’s amazing distributor channel, so anyone around the world could get a MicroView locally.
I carefully chose the price.
I previously founded Ninja Blocks (ninjablocks.com) and one of the things I learned was that when you price a product too low it is impossible to get distributors. This is what happened with the original Ninja Block.
This time round I wanted the MicroView to be available through SparkFun’s amazing distributor channel, so anyone around the world could get a MicroView locally.
Hey Marcus... I'm Dan of Rheingold Heavy, the guy who wrote the review. I just wanted to thank you for taking the time to come out here and engage with us.
The problem is there's a gulf between people who don't really know that much, and people who know enough to build the thing they're buying. The people that know think it's a rip-off.
The people who don't know really want something that will plug and play before tinkering with the code. I've taught electronics (AVR/C) at summer camps and trust me - even smart people really struggle, even those with programming experience.
Standard industry pricing for any hardware is manufacturing costs multiplied by 2.5-3. That would peg build costs at around $10-15 for this which is about right. SparkFun wants to make a profit on this and rightly so.
It's an entirely valid criticism that the documentation isn't up to scratch. I thought only machine vision companies made you read through crappy API docs and code examples to figure out how anything works!
The people who don't know really want something that will plug and play before tinkering with the code. I've taught electronics (AVR/C) at summer camps and trust me - even smart people really struggle, even those with programming experience.
Standard industry pricing for any hardware is manufacturing costs multiplied by 2.5-3. That would peg build costs at around $10-15 for this which is about right. SparkFun wants to make a profit on this and rightly so.
It's an entirely valid criticism that the documentation isn't up to scratch. I thought only machine vision companies made you read through crappy API docs and code examples to figure out how anything works!
I'm happy to take the criticism for the overall documentation. However I believe JP did a great job with the library's doco.
Please show me an Arduino library with more documentation than: http://learn.microview.io/doco/html/index.html
Please show me an Arduino library with more documentation than: http://learn.microview.io/doco/html/index.html
http://www.nongnu.org/avr-libc/user-manual/
This is a slightly sarcastic point.
I find the Arduino programming environment is horrible. I had to revert to straight-up C in the end and use the Arduino as a convenient ICSP package.
This is a slightly sarcastic point.
I find the Arduino programming environment is horrible. I had to revert to straight-up C in the end and use the Arduino as a convenient ICSP package.
My preferred platform is a Sparkfun Pro Micro (or clone) with the DFU bootloader on it. Just the right number of pins for most projects and programmable with a common micro-USB cable. Tried to poke around with Arduino, but I like using my own libraries.
That looks like a pretty nice platform. I may have to grab one to play with.
I went with a clone because they're about £4/$6 and that seemed like a no brainer. I haven't had any duds yet! They only come in the 5V option from what I've seen, but it's simple enough to swap out the crystal and the regulator to make it a 3.3V/8MHz.
Only issue I had was to modify the LUFA DFU bootloader to automatically start the micro if there isn't a programming request within 5 seconds. The default code will sit in the bootloader forever until something happens - fine for prototyping, but a problem if you have it in a semi-permanent place and the power goes.
Only issue I had was to modify the LUFA DFU bootloader to automatically start the micro if there isn't a programming request within 5 seconds. The default code will sit in the bootloader forever until something happens - fine for prototyping, but a problem if you have it in a semi-permanent place and the power goes.
Thanks for the info. Where do you get them at £4 a pop?
Not worried about the bootloader problems as these would be prototype only and the voltage is ok as I'm replacing a PLC implementation with some homebrew kit so it's just switching relays and reading opto-isolated inputs, neither which are voltage sensitive.
Not worried about the bootloader problems as these would be prototype only and the voltage is ok as I'm replacing a PLC implementation with some homebrew kit so it's just switching relays and reading opto-isolated inputs, neither which are voltage sensitive.
http://www.ebay.co.uk/itm/Hot-New-Pro-Micro-5V-16MHz-ATMega-...
You'll be waiting a while for shipping, so I'd recommend ordering a few ;)
You'll be waiting a while for shipping, so I'd recommend ordering a few ;)
Perfect - thanks :)
Ordered loads of stuff from HK so know about the wait. Have usually forgotten about it by the time it arrives!
Ordered loads of stuff from HK so know about the wait. Have usually forgotten about it by the time it arrives!
more
Quality is not quantity; I think that's what the parent and the reviewer was trying to say - there's certainly plenty of documentation, but it's not particularly easy to use that documentation effectively. E.g.
I have learned that you need uView.begin() in your setup, but I only figured that out because it’s in every one of the examples, not because I read documentation anywhere that said, “This is how you structure a MicroView compatible sketch.”
(Also brings to mind the absolutely overwhelming amount of documentation available for some of the modern smartphone SoCs, but with no easy way to find out e.g. the relationship between one peripheral module and many others sharing the same resources.)
Quality is not quantity; I think that's what the parent and the reviewer was trying to say - there's certainly plenty of documentation, but it's not particularly easy to use that documentation effectively. E.g.
I have learned that you need uView.begin() in your setup, but I only figured that out because it’s in every one of the examples, not because I read documentation anywhere that said, “This is how you structure a MicroView compatible sketch.”
(Also brings to mind the absolutely overwhelming amount of documentation available for some of the modern smartphone SoCs, but with no easy way to find out e.g. the relationship between one peripheral module and many others sharing the same resources.)
Right, this is my point.
Having lovely hookup diagrams, interactive code editors and Doxygen output for the API is nice, but... What's needed is a mid-level introduction to the library once you've gotten past the basic tutorials and solid examples for the commonly used API calls (most of which are simple, but still).
Having lovely hookup diagrams, interactive code editors and Doxygen output for the API is nice, but... What's needed is a mid-level introduction to the library once you've gotten past the basic tutorials and solid examples for the commonly used API calls (most of which are simple, but still).
Please show me an Arduino library with significantly less documentation for a device of similar price and functionality.
Revision: I didn't intend for that to come off quite as snarky as it reads. You're certainly entitled to price your product however you see fit, but I don't think it's fair to praise the documentation simply because it's no worse than that of existing libraries. Particularly when (anecdotally) the vast majority of such libraries have no financial motivations.
Revision: I didn't intend for that to come off quite as snarky as it reads. You're certainly entitled to price your product however you see fit, but I don't think it's fair to praise the documentation simply because it's no worse than that of existing libraries. Particularly when (anecdotally) the vast majority of such libraries have no financial motivations.
I've bought nearly every single digit dollar cost thing on ebay with "arduino" in the title multiple times over in my short amateur stint teaching local kids beginning robotics classes.
I almost always buy one of the sparkfun or Adafruit units to start with. In my experience they save 2-4 hours of intense googling and Chenglish translation on your first pass at the hello world of the new device.
If you're just buying one, get the sparkfun and be glad you can slip then a 20 to spare you the effort.
I almost always buy one of the sparkfun or Adafruit units to start with. In my experience they save 2-4 hours of intense googling and Chenglish translation on your first pass at the hello world of the new device.
If you're just buying one, get the sparkfun and be glad you can slip then a 20 to spare you the effort.
> priced far beyond what they actually cost.
Very rarely will someone tell you that you're not charging as much for something as you should (perhaps with the exception of patio11 because he wants us all to make more money :) ) but generally there's always someone who will say something costs too much.
As explained elsewhere, as soon as you want a product that will be sold by distributors you're looking at a retail cost of ~4 * the build cost. This is because distributors will be looking at a 100% markup to cover their costs (and give room for discounts etc) and the manufacturer will be looking at about the same to cover their non-direct build costs (research, development, documentation, customer support, staffing, taxes, returns, damages, profit, etc). If they don't price like this then the product is unsustainable.
The name you're looking for is probably "non-volume electronics" or "small-run electronics" where the cost of developing a small-run product is spread over a smaller number of units and thus appears to be more expensive than some apparent mass-manufactured equivalent device (e.g. a low end cell phone).
Let's take your examples of $3 + $4, add $1 for the case, and $2 for the PCB & other components and you're already at a retail price of $40 without considering assembly. It gets expensive fast.
But, for the right market, it's still seen as a fair trade of time/money.
(I've seen a really good spreadsheet somewhere that can be used to make this sort of calculation--but I've not been able to track it down. Any pointers welcome.)
For disclosure: I've met and chatted with Marcus on a number of occasions and contracted with Sparkfun in the past but otherwise have no connection with the product in question. I just know how surprisingly expensive hardware manufacturing can be and how disappointing it can be to hear "it's too expensive" from people who may not be aware of the costs involved.
Very rarely will someone tell you that you're not charging as much for something as you should (perhaps with the exception of patio11 because he wants us all to make more money :) ) but generally there's always someone who will say something costs too much.
As explained elsewhere, as soon as you want a product that will be sold by distributors you're looking at a retail cost of ~4 * the build cost. This is because distributors will be looking at a 100% markup to cover their costs (and give room for discounts etc) and the manufacturer will be looking at about the same to cover their non-direct build costs (research, development, documentation, customer support, staffing, taxes, returns, damages, profit, etc). If they don't price like this then the product is unsustainable.
The name you're looking for is probably "non-volume electronics" or "small-run electronics" where the cost of developing a small-run product is spread over a smaller number of units and thus appears to be more expensive than some apparent mass-manufactured equivalent device (e.g. a low end cell phone).
Let's take your examples of $3 + $4, add $1 for the case, and $2 for the PCB & other components and you're already at a retail price of $40 without considering assembly. It gets expensive fast.
But, for the right market, it's still seen as a fair trade of time/money.
(I've seen a really good spreadsheet somewhere that can be used to make this sort of calculation--but I've not been able to track it down. Any pointers welcome.)
For disclosure: I've met and chatted with Marcus on a number of occasions and contracted with Sparkfun in the past but otherwise have no connection with the product in question. I just know how surprisingly expensive hardware manufacturing can be and how disappointing it can be to hear "it's too expensive" from people who may not be aware of the costs involved.
It takes time to layout the board, time to build libraries, time to do CAD on the case, time to test prototypes, time to negotiate manufacturing rates, etc. Time costs money.
And there's nothing wrong with making a profit.
And there's nothing wrong with making a profit.
I think you are conflating cost with value. Clearly there are folks out there who value the packaging, the work that has gone into programming libraries, etc. They value it enough to put up $40 for one. That has nothing at all to do with the cost, which as you point out in volume is probably less than $6 each although ATMega don't really get cheap until you hit 10K units.
If it isn't "worth" $40 then Sparkfun won't sell any more after they fulfill all their kickstarter rewards, or maybe they will drop the price or something.
If it isn't "worth" $40 then Sparkfun won't sell any more after they fulfill all their kickstarter rewards, or maybe they will drop the price or something.
Yeah, this got me thinking. Pricing at $40 also meant that we were able to recover from our Snafu (https://www.sparkfun.com/news/1575), and provide everybody affected with 2x MicroViews.
I agree but I suspect that they fill the niche of the 200-in-1 electronic kits that were around in the 1970s to 1990s. They were inferior than the parts, overpriced and teach at a way higher level of abstraction than the individual components without providing any real education.
However, they were popular because they were productive and gave people a jump start. I assume the same is true for this.
Even Forest Mimms books were short on the details required to design or build something from scratch and missed vast chunks of fundamental knowledge.
However, as someone who dabbles with electronics occasionally and is a qualified electronics engineer, not that I've ever used it professionally; nothing replicates the joy of sitting there with a pencil and paper for half an hour with a copy of TAOE and building something from scratch without copying a single pattern or module from a kit and not using a single microcontroller or black box abstraction.
However, they were popular because they were productive and gave people a jump start. I assume the same is true for this.
Even Forest Mimms books were short on the details required to design or build something from scratch and missed vast chunks of fundamental knowledge.
However, as someone who dabbles with electronics occasionally and is a qualified electronics engineer, not that I've ever used it professionally; nothing replicates the joy of sitting there with a pencil and paper for half an hour with a copy of TAOE and building something from scratch without copying a single pattern or module from a kit and not using a single microcontroller or black box abstraction.
I suspect that if they sold in GSM phone volumes they could price them cheaper.
Maybe this is the perfect opportunity for you to make something like this at something you consider a more reasonable price and make a fortune?
This reminds of those clients who think "You should be able to knock this out quickly." If they are so knowledgeable that they know how long something should take, why don't they just "knock it out in a couple of days" themselves?
Maybe this is the perfect opportunity for you to make something like this at something you consider a more reasonable price and make a fortune?
This reminds of those clients who think "You should be able to knock this out quickly." If they are so knowledgeable that they know how long something should take, why don't they just "knock it out in a couple of days" themselves?
as soon as you put a western engineer in the loop, the prices just soar.
I am excited to get my hands on one of these, in the video they showed on Kickstarter, they seemed to have some demos on how to program it for kids running on what looked like a Surface Pro.
Great write up! I'll have to get my hands on one in the near future.
0.66" OLEDs are in the <$3 range, the MCU used in there is <$4, and I don't think any of the other components are any more expensive than that. PCBs are extremely cheap too.
When one can buy a fully-functioning GSM phone (complete with OLED display) for a fraction of the price, no wonder it makes electronics look like an expensive hobby...