Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Is this concept worth presuing as commercialy viable

Status
Not open for further replies.

Dialtone

New Member
I am looking for advice as to whether an idea I have prototyped is really worth further development as a small scale commercialy viable project. Please be exceedingly blunt and give me the good and the bad as you see it.

In essence, the project takes a PC parallel port and allows it to be used to access up to 256x8 bit addressable data ports, in a modular physical design, that can be written or read in either a latched or buffered I/O mode. Testing on the original prototype went quite well and from my tests, the data thru-put capability was around 4 mega-bits/second. I am now into a second gen design that should allow the speeds to increase by a factor of 1/3 (assuming the PC can handle the I/O that fast).

I have also been able to easily interface a wide variety of currently available devices that would allow a PC to act as a medium/large scale automation control point. (That was the whole purpose of designing it in the first place)

The real question now is not so much CAN it be done, but SHOULD it be done and is it of any real value, and would ANY sector really need that much I/O. I know DI/O PC boards are available, but at the price/port they seem very expensive if one needs a large amount of I/O. Even my 1-off prototype boards came in much cheaper that what I have seen, so I am sure production quantity runs would compete even better from a price perspective. I also know that parallel ports seem to be destined for the scrap heap.

So give it to me straight as you see it.

Thanks
Dialtone
 
Unless you can do it very cheap, as in made in China cheap, including the cost of supporting the product/software, I would unfortunately say no.

One place I worked at did something very similar, using a PC-AT ( it was the 80's ) to control a mess of relays on a factory production line test station. Originally the station was controlled by a very large drum switch driven by a motor - it was a mechanical nightmare - Reuben Goldberg would have been proud. We ended up by a 24 bit I/O card ( 3 x 8 ) for very reasonable money.

What we got was very similar to this device by Accees IO Products, except ours had a row of the small weidmuller type isolated terminals.

https://www.accesio.com/go.cgi?p=../pci/pci_dio_24d.html

One of the nice things about your idea is using the parallel port in that a laptop could be quickly connected to the device for testing or troubleshooting. The biggest problem though is making it cheap enough to compete with those using China.[/url]
 
I agree price is definately a consideration.
From your example, the price per port is $US 6.25 and some of the larger I/O boards are not much better on a price/port basis. My one-of-a kind prototype came in at under $3.00 per port, so I know a production quantity run of 1000 units will net some real cost reductions.
Excluding case & power supply, I really think I could get the cost down to the $1.00- $1.50 per port range.
If one only needed 24 digital /O channels then I guess $6.25 would be an acceptable cost.
I am however, looking for the market where 24 is too few and 25-2048 gets expensive in a really big way.

Software is not much of a problem.. Since this device, is nothing more than an interface expansion device, the only real software it needs is a simple modified printer type driver. In reality, the end user would need to supply their own software depending on the application.

Dialtone
 
Providing a "hardware-only" solution would be a wise move. Seriously cuts the true cost of the product. Expanding to large numbers of I/O is very advantageous, often during design of equipment you end up in a debate of culling the herd of data to a limited number of ports.

Your price does sound good. Access I/O does have lower priced cards, stuff like this: **broken link removed**

the product sounds interesting, there is definitly a market for these things, it all comes down to price and functionality.

One thing that comes to mind is the connection interface. Providing space for more than 24 inputs gets demanding, and expensive for cases / PCB area, and terminals. I would suggest a high-density connector, allowing the customer to have one small connection to your device. AMP make numerous high density connectors, like their Multilok series. Another option would be to use ATA style connectors, to ribbon cable. You could offer "break-out" boxes - More product to sell!
 
I have pondered the interface over many times. Originally I wanted to bring the 8/group I/O leads out to individual 10 pin IDC headers that then could be interfaced thru a ribbon/DB-9 but I soon realized the cost would be prohibitive.
I think I have finally settled on bringing all the I/O out to a 60 pin IDC and a 34 pin IDC on one edge. By removing 1 row of pins every 6th row on the header(s) you can access each group of 10 (8 data bits, ground, & signal validity) with individual 10 pin IDC ribbon plugs (non-keyed of course), for a simple mix-and-match interface solution, without being locked into some propriatary connector scheme.

Dialtone
 
Abandon hope. The trend in industrial control is against you. I/O done dirt cheap uses low end field buses like ASI and DeviceNet. Higher end products, like robot controllers, use Profibus or Ethernet. Everybody uses a PLC instead of a PC because control engineers still speak ladder logic.

You have almost no chance of selling a high IO point count device controlled by a PC. Just the cost of the wire to and from the PC for 256 points will blow you out of the water.

If you are still determined to make a product then make it 256 I/O points with an Ethernet/IP interface, and get set to compete with Allen-Bradley.
 
Status
Not open for further replies.

Latest threads

Back
Top