Creating a schematics marketplace!

[Sckem]

Hi all,

I have been in the electronics industry for more than 20 years.

Now I have an idea to create a schematics marketplace, where freelancers can upload their schematics (following a set of schematic drawing guidelines), and users can download a proven schematic to "get started" quickly. My idea is to have 3 main categories:
* FPGA
* MCU
* CPU

A user would browse existing solutions and pick one that is close enough to what he needs, then he edits that schematic to add the features/functions he needs. The idea is to get a project started quickly while reducing the risk of schematic errors.

A central authority would review schematics, ensure its authentic, and allow a "rating" system from users.

This is - i concede - still a very vague idea, and there are many many other aspects i would have to deal with to get such a project up and running.

So my question for this brilliant community is: How useful this would be for you? (assuming i can get all the parts right, and build a community of users around this)

I'm eager to learn your honest point of view, either as a user or as a freelancer who would be willing to put his work on sale.

Thanks

 

[ronsimpson]

There are "free" circuit sights on the internet. From what I can see they steal from each other. (and will steal form you)
A large amount of circuits do not work, have not been tried, and are just silly. (how will you verify you circuits)
Ideas spread like weeds. I should name a circuit some thing like "simple audio amplifier 555 using 3 parts, under $0.12" and watch the not working circuit spread around all the forms over night. And if I really want it to go viral "free energy using a 555"

I want to see a "schematics marketplace" but I do not see how.
I make a living making working projects, complete with schematics. So in some ways this is against what I do.

 

[Sckem]

Hi Ron,

Thanks for your feedback!

There are "free" circuit sights on the internet. From what I can see they steal from each other. (and will steal form you)
A large amount of circuits do not work, have not been tried, and are just silly. (how will you verify you circuits)

Indeed. there are many free schematics, and as you pointed they are full of rubbish.

First, what i intent to do is not an "all kind of schematics" sort of thing, but rather a focus on complex scheamtics that are build on an FPGA, MCU or CPU.

There are many ways to test circuits. Rating from users one. Hieing experts to analyze posted schematics and add a "Validated-by-us" flag to it is yet another solution.

As for the theft, they'll have to buy the schematics in the first place before putting it online. If that's the only problem, i am sure we'll find solutions (aren't copyright meant to protect us? sorry, bad joke!)

I make a living making working projects, complete with schematics. So in some ways this is against what I do.

Could you explain why you think this is against what you do? My feeling is that it will only help you get things done faster. Instead of spending hours to draw all the wires of a DDR3 memory connected to an FPGA, you would be better off using your expertise to quickly check the provided schematic and build upon it, no?

 

[alec_t]

I doubt this would fly.
I don't see any way you could realistically and economically test more than a handful of schematics to see if the circuits actually work as it says on the tin. Who would build and verify them? Who would pay for the very expensive test equipment needed for the "complex schematics that are built on an FPGA, MCU or CPU"?
As a potential customer I wouldn't want to rely on user reviews/ratings: they're too easily spoofed.
What warranty would you be able to give? I can foresee endless arguments along the lines of : "I built the circuit exactly as shown but it doesn't work. I want a refund.". "Well the circuit is fine but you obviously built it wrongly or used it out of spec. No refund".
Would the site be aimed at hobbyists or professionals? Hobbyists would be reluctant to pay for something they think should be free. Businesses are likely to adopt a "not-invented-here" stance and want to do their own thing, especially for anything proprietary or security-related.

 

[Sckem]

I doubt this would fly.
I don't see any way you could realistically and economically test more than a handful of schematics to see if the circuits actually work as it says on the tin. Who would build and verify them? Who would pay for the very expensive test equipment needed for the "complex schematics that are built on an FPGA, MCU or CPU"?
As a potential customer I wouldn't want to rely on user reviews/ratings: they're too easily spoofed.
What warranty would you be able to give? I can foresee endless arguments along the lines of : "I built the circuit exactly as shown but it doesn't work. I want a refund.". "Well the circuit is fine but you obviously built it wrongly or used it out of spec. No refund".
Would the site be aimed at hobbyists or professionals? Hobbyists would be reluctant to pay for something they think should be free. Businesses are likely to adopt a "not-invented-here" stance and want to do their own thing, especially for anything proprietary or security-related.

Hmmm..

I was kind of hoping to create a community where "trust" can be built, day after day, schematic after schematic.

A user would still be required to verify a schematic (i.e. verify the wiring of an FPGA), but at least we would be starting from a white sheet.

Also, a support channel between the schematic author and the user would help to remove doubts about specific parts of a schematic.

Again, my aim is not the hobbyists who "just want to take a ready schematic and apply it", my aim is to prevent engineers from wasting precious hours drawing hundreds of lines and buses. My idea is that they would be wasting less time verifying a schematic then modify it to their need instead of starting from zero.

I for example, if i'm going to start a project with a spartan 6 FPGA, i would absolutely never start from 0, but duplicate my last schematic that contained a spartan 6 and start modifying it.

Ain't this how we all get along?

 

[Sckem]

Businesses are likely to adopt a "not-invented-here" stance

I don't think that the wiring of a DDR3 to an FPGA along with the configuration memory and JTAG interface is where the invention lies... (for example)

 

[granddad]

I for example, if i'm going to start a project with a spartan 6 FPGA, i would absolutely never start from 0, but duplicate my last schematic that contained a spartan 6 and start modifying it.

I would imagine 'engineers ' already have a library of schematics they know and trust , buying off a web site would be like doing themselves out of a job ? I would aim at the hobbyist must be a big market for 'trusted' designs but not to buy perhaps post one download one ...

 

[DerStrom8]

Unfortunately I also don't think this would go over well. Why pay for a schematic, even if it's 100% tested and proven, when you can find hundreds of similar schematics online for free, even if they might be a little less reliable? I'm willing to bet that this is the mindset of most people searching the internet for example circuits, so even if you were able to get such a marketplace up and running with reliable and tested circuits, etc. I doubt it would go anywhere. Besides, enforcing copyright on these schematics would be impossible. Anybody can redraw it and change the way it looks and call it their own.

 

[OBW0549]

A user would browse existing solutions and pick one that is close enough to what he needs, then he edits that schematic to add the features/functions he needs.

That would entail just as much work, if not more, than creating the whole schematic myself from scratch.

A central authority would review schematics, ensure its authentic, and allow a "rating" system from users.

As a design engineer, there is only one "central authority" I trust: myself. Everyone else is suspect, unless I'm very familiar with their work.

So my question for this brilliant community is: How useful this would be for you?

This would have ZERO usefulness to me. I wouldn't use it, and I wouldn't contribute to it.

And I suspect it would have little usefulness to anyone else, either: anyone who would need such a service is going to lack the experience and/or talent to use it without screwing something up; and anyone with experience and talent simply isn't going to need the service. Engineers will neither need nor trust it, and hobbyists will not be willing to pay for it.

 

[JonSea]

An existing schematic with its "hundred of lines" isn't going to save somebody much time if it's available only in print form. If the source file is available, it might save someone some time, if it's available for whatever schematic capture program they use - 6 or 8 or more common incompatable options.

An who will want to sell a schematic for a few bucks with "a line of support and communication" expected.

 

[Sckem]

An existing schematic with its "hundred of lines" isn't going to save somebody much time if it's available only in print form. If the source file is available, it might save someone some time, if it's available for whatever schematic capture program they use - 6 or 8 or more common incompatable options.

of course, the idea is to offer the native format (Eagle, Altium, kicad, etc..)

 

[JonSea]

of course, the idea is to offer the native format (Eagle, Altium, kicad, etc..)

All, or what the oeiginator used to draw it? I don't believe tbere are any conversion tools between formats, so it will be a big rask to manually convert and verify each schematic.

 

[Sckem]

All, or what the oeiginator used to draw it? I don't believe tbere are any conversion tools between formats, so it will be a big rask to manually convert and verify each schematic.

Well, there are some conversions tools between some of the formats. If that's the only problem, i am sure we can work out a solution.

 

[Sckem]

And I suspect it would have little usefulness to anyone else, either: anyone who would need such a service is going to lack the experience and/or talent to use it without screwing something up

I still don't see where is the talent in drawing 32 lines for an address bus on a schematic?

 

[gophert]

Well, there are some conversions tools between some of the formats. If that's the only problem, i am sure we can work out a solution.

Are you an engineer or a marketing guy? You must keep uttering that same phrase "I'm sure we can work that out".

Eachprojec not only has a function (do this when that happens), they also ave power requrements/limitations and space requirements must fit this footprint) and even orientation requirements (the i/o be here..., power tap must be here...). And a recent project had, connector must be of type xyz.

A library of custom boards is a waste of time because they would have never been designed as a cust9m project if there was any chance that many people could use them. And, at the same time, if many people could use a circuit, I think the designer (or his company) would be smart enough to sell the completed device themselves rather than selling plans in an online catalog.

Your plan would be better suited for architects selling log cabin blueprints. Everyone needs a house - not everyone needs a blinky-thingy-with-a-2x16-LCD-and-temp-sensor-with-less-than-0.25%-drift-and-0.1°F-readout-and-user-input-for-when-they-want-a-TacoBell-taco-delivered

 

[gophert]

On the other hand, don't let me stop you from investing time and money, hiring web developers and getting this thing rolling.

3% of GDP is spent setting up failed businesses. The economy will welcome your spending.

 

[Sckem]

Thanks all for your answers so far. It has been very helpful in understanding how you view such a project and how you react to it.

I totally understand the reasons why you doubt this, and I think you're right. That being said, I still think there is room for improvement: With today's digital schematics, there is no art nor talent in drawing repetitive lines for wide data and address buses. I know there are some kind of solutions for that, but nothing quite convincing to me.

I'll keep that problem in the back of my head for now until a bulb lights some day

gophert I don't want to waste money (even if it's not mine), that's why i am doing some very preliminary research here (literally!) and there.

 

[AnalogKid]

of course, the idea is to offer the native format (Eagle, Altium, kicad, etc..)

There are way over a dozen "common" or "popular" schematic CAD packages, dozen*s* when you include things that have come and gone over the last 10 years. I have current hands-on skills with 4 systems, legacy skills for 6 more, and some experience with native file conflicts. Native file format management will be a huge problem.

Well, there are some conversions tools between some of the formats. If that's the only problem, i am sure we can work out a solution.

I'm sure you can't. Most major packages do not import their own previous formats without significant issues if more than a few years old, and importing someone else's files never works as advertised. Separate from that is the leap of faith it takes to believe that there were no conversion errors because the software said so.

Consider that if this kind of thing were practical and saved real companies real dollars, either TI or ADI would have worked it out long ago as a marketing tool against the other.

ak

 

[OBW0549]

I still don't see where is the talent in drawing 32 lines for an address bus on a schematic?

That's exactly my point: drawing those lines-- or any number of other lines, for that matter-- is a trivial exercise. So why in the world would I want to pay your service to do it for me, especially since I'm almost certain to have to make extensive modifications/additions to any of your generic schematics anyway to end up with the design I need?

It's been my experience from 40+ years as a designer that the actual drafting of a schematic is by far the easiest part of any development effort. The much harder part is all the design work that has to be done beforehand, as well as the circuit board layout, and the packaging and the design testing and test results analysis that has to be done afterward. And you will not be offering any of that.

What are the qualifications of the "freelancers" who will be contributing schematics to this marketplace of yours? How can I get assurance that any particular freelancer is a competent designer and not some slack-jawed, drooling idiot who just dabbles in electronics? Are you going to offer any warranty on these designs? Will I get my money back if a design doesn't work? Will I get my money back after I've made the necessary modifications/additions to one of these designs, and it doesn't work? Will I get technical support from your freelance contributor if I use one of his designs and encounter difficulties with it? I'm not willing to pay you so much as one thin dime unless I get all that, and more.

And you can forget about saving me any time with your canned schematics: whatever small amount of drafting time will be saved by using them, will be spent checking and double-checking your freelancer's work for errors and design shortcomings-- because unless I personally know the freelancer and his work, I'm not going to trust him to be able to even walk and chew gum at the same time.

No, I'm not interested.

 

[KeepItSimpleStupid]

Its hard enough for PCB footprints to be compatible. Both 2 and 3D.

The only way I see what I proposed happening is if the manufacturers supplied the info in a universal importable format so you get package, silk screen, 3D signals and spice from a trusted source.