General Consensus Regarding OPTO 22?


Thread Starter

Peter Arthur

I am employed by an Automation distributor. I am charged with looking into some different product lines to fill the voids in our product lineup.

One of the biggest holes I've found is in the I/O products-PLC arena. We offer neither. We don't intend to get into the traditional PLC thing which generally means butting heads with AB all day, but I am nonetheless interested.

I am intrigued by OPTO 22. Anyone care to comment pro or con? I am most interested to hear from actual users of these products, but any and all comments are welcome and appreciated.

Thanks in advance-

Peter Arthur
Project Engineer
Applied Motion Products
If you are going to look at I/O products, including Opto-22, you should
also consider Phoenix Contact, National Instruments, and others (NOT an
exhaustive list). NI has probably the largest offering of I/O and PAC
(programmable automation controller) products around. You might also
want to look at MTL.

For a larger list, look at We have product
information posted on many categories, including I/O:

Walt Boyes
Editor in Chief
CONTROL magazine
555 W. Pierce Rd., Suite 301
Itasca, IL 60506
Phone: +1-630-467-1301 x 368
[email protected]

Blogging Sound OFF!! Editors' Blog at
or direct at

Bob Peterson

Here is my take. I work for a systems integrator. One of the things we do is make some specialized equipment. One of the users of this equipment loves Opto22 - specifically SnapIO. They like it because it is completely non-proprietary and works pretty well running off their ethernet network and can be controlled from their Linus and Unix based plant control systems (totally roll your own here).

Having said that, in recent pieces of equipment they have had us add a relay in the equipment cabinets that can remote cycle power to the Snap IO brain boards. The reason - they have had several cases where a brain board has stopped communicating with their network and the only solution was to go cycle power.

These guys know a lot about this product (they have thousands of nodes installed) so I do not think the problem is on their end. Personally, I am, and remain suspicious of using ethernet for IO anyway, and this situation has done nothing to change my opinion.

Opto 22 fills a niche that no one else really does, for those that want to do their own thing. The fact is that there are plenty of people who think they can do it better than AB or the DCS venders, and Opto22 supplies the IO for those people.

They also have some interesting control products that are very cost effective if you can live with a non-RLL solution.

Most of our customers are unwilling to learn a new product. It is just way too expensive to send 20 people to school for two or three weeks each to learn a new product and stock spare parts to save a few hundred dollars on a single project.

Curt Wuollet

Ethernet is great for IO and reliable enough that we all use it extensively every day. The Opto22 Ethernet "Brain" is slow and easily swamped. I expect this might be what was hanging them up. In many PC control apps it would be fine, but high rate polling didn't work for me. Too bad since I was working on a LinuxPLC.


Enough already, Curt. Your characterization of Opto's Ethernet I/O (that you have relentlessly regurgitated on this board countless times) was based on your experiences in 1999, relative to your LinuxPLC project. Believe it or not, much progress has been made in the last 7 years (at least from Opto's side; I can't speak for your LinuxPLC project). To date, tens of thousands of Opto's Ethernet-based I/O products are dutifully serving in thousands of applications all over the world. What can you say about your LinuxPLC installed base?

On the subject of Linux, you should know about the Linux-based SNAP OEM I/O from Opto 22. They've certainly embraced the concept of a Linux-based control system, and have developed a platform for those who wish to focus on the software, and not building the hardware. Support is included free of charge from very abled engineers (one who authored an article for Embedded Linux Journal). In addition, forum-contributor Frank has indicated he's used it successfully in his application.

So we get it. You don't like Opto. And you are entitled to your opinion. But as others have requested, please back up your opinion with data. Otherwise, your actions could be considered a disservice to those who look for guidance here.

With all due respect, -Benson

Michael Griffin

In reply to Benson - By all means, please do tell us the maximum polling rate for Opto-22 Snap I/O. I and no doubt others would be very interested in hearing from you as to what this very important number is so we can judge whether these products can be used in particular applications. I have searched Opto-22's literature but have not been able to find any hard numbers on this matter anywhere.

We had an extensive discussion last year under the topic of "PLCS: Opto-22 vs Rockwell 1756/1769" in which the response time question was raised and Mr. Wuollet offered the results of his experiments. At the moment, Mr. Wuollet's experiments are all we have to go by. If you have any recent detailed numbers which contradict him, I and no doubt others would love to see them.

Curt Wuollet

Hi Benson I have never indicated that I didn't like Opto, I would like very much to have as many non-vertical ethernet IO solutions for automation as possible. I am a steadfast and long time proponent of real Ethernet and Modbus TCP for automation. The reason I put out the admittedly dated information, yet again is because it is likely the only information available, especially from a reasonably objective source. Perhaps the first question a designer would want to know about remote IO is what update rate is possible. Answers are remarkably hard to come by and from what I've seen are not at all what one would expect. I would love to be corrected by hard data from Opto and there are a lot of very expensive proprietary vendors whose real world figures are nothing like the commonly quoted "theoretical" numbers we see in the ad copy. And I was careful to state that in many applications you would be fine with the product.

I would be the first to apologize if I mis characterize the product. But, it was actual real world testing and that's what I got. Is it unreasonable to expect Ethernet IO to perform within an order of magnitude or two of Ethernet speeds? Wouldn't you expect better than 50 Hz sampling on a Profitbus network? Performance might be the skunk on the table, but I didn't bring the skunk, I just noticed the aroma. We need truth in advertising about bus products.

I had used the snap I/O series, the quality of them is not good, a lot of them need to back to the factory.
I would strongly recommend that if you have an issue with OPTO22 that you contact them directly. Benson lurks regularly on this list. Hopefully, he'll see this and respond.

Walt Boyes
Editor in chief
Control magazine

Walt Boyes
BlackBerry service provided by Nextel
Benson here. I am very curious and concerned about any issues you experienced with SNAP I/O. Please feel free to contact me directly at bensonh at opto22 dot com, or contact our product support group at 800-835-6786 in the USA or 951-695-3080 outside the USA.

I look forward to your reply.