Most people assume the hardest part of an automation project is the programming. After eight years of ordering Omron PLCs, relays, contactors, and variable frequency drives, I can tell you from experience: the most expensive mistakes happen before the programmer ever opens the software.
In May 2017, I submitted my first solo order for a replacement PLC. The part number looked correct. The price looked reasonable. The distributor confirmed stock and gave me a delivery date. I approved the order without reading the full specification sheet one more time. The mistake cost us $890 in restocking fees, expedited freight on the replacement, and a week of avoidable downtime.
Since then, I have made enough similar mistakes to fill a notebook. I am not proud of that. But I documented each one, and in Q1 2024 I turned the notes into a pre-order checklist. That checklist has caught 47 potential errors in the last 18 months. This article is the story behind it.
Here is my position, stated plainly: spec verification before purchase is the cheapest insurance an automation team can buy. A better distributor relationship will not save you from a wrong component. A more expensive PLC model will not save you either. The only thing that stopped my recurring mistakes was a disciplined check before the purchase order went out.
The PLC model trap: close enough is not close enough
Omron's PLC lineup is genuinely broad, from the modular CJ series to the compact CP series, the NJ and NX machine automation controllers, and dedicated safety PLCs. That breadth is an advantage until it becomes a trap, because the model numbers look deceptively similar.
One suffix can mean the difference between relay outputs and transistor outputs. One letter can mean AC power instead of 24 VDC power. One digit can mean a CPU with enough memory for your application, or a CPU that forces you to rewrite your program to fit.
My 2017 mistake was an output type error. I ordered a compact CPU because the I/O count matched our panel. I did not confirm whether the output circuits were relay or transistor. The application needed high-speed pulse output, which requires transistor outputs. The controller powered up, the program loaded, and the machine jogged. But the high-speed indexing move never executed correctly. Relay contacts cannot switch that fast. Mechanical reality caught up with my checkbox.
What most people don't realize is that a distributor's ordering system will happily confirm a compatible part number and ship the order without ever knowing your application. To be fair, they cannot know your application. I was expecting them to carry a responsibility that belonged to me.
Relays are where projects go to die
If you think the PLC is the riskiest part of a control panel, you have not lost a week of production to an $8 relay.
In September 2022, I ordered a batch of output relays from a new relay distributor. The electrical list was correct: coil voltage, contact rating, number of poles. What I left out was the physical configuration. The existing sockets in the panel took the 11-pin style. I ordered the 8-pin style. When the electrician went to seat them, nothing clicked. Not the pins, not my confidence.
Eleven brand-new relays, $450, straight back into boxes. Plus a three-day delay while the correct parts arrived.
We were using the same words and meaning different things. I said 'standard relay.' They heard 'standard relay.' Neither of us looked at the socket drawing until it was too late.
The most expensive component in your panel is not always the one with the highest price. It is the one that stops the whole panel from working while the PLC sits there doing exactly what it was told.
The VFD mistake that forced me to write a spec guide
The last big mistake happened in November 2023.
A conveyor drive had failed, and the plant wanted a replacement fast. The old nameplate was still readable. I ordered based on the motor power rating, '11 kW', without comparing the full-load amps on the motor to the output current rating of the drive. On paper, the kW matched. Under load, the drive tripped on overload during startup, especially when the conveyor was loaded.
The drive itself was fine. The match was wrong. I had purchased a component that was technically correct and functionally unsuitable. That distinction cost about $2,700 in restocking, reordering, and two more weeks of temporary bypass operation.
After that, I sat down and wrote the VFD specification guide I should have used all along. The short version looks like this:
- Copy the motor nameplate: voltage, full-load amps, frequency, insulation class, and duty rating.
- Check the drive's rated output current, not just the motor power in kW.
- Confirm the input supply voltage and phase at the installation point.
- Define the overload requirement: constant torque for conveyors and extruders, variable torque for fans and pumps.
- Decide whether you need a braking chopper or resistor.
- Confirm the enclosure type: IP20 for panel mounting, higher protection for standalone installation.
That list is not groundbreaking. That is exactly the point. It is boring, obvious, and it would have saved me thousands of dollars.
What I learned about Omron PLC distributors and component suppliers
This brings me to a topic that does not get enough attention: how you choose the people you buy from.
I have worked with Omron PLC distributors of every size, from global stockists to local specialists. The best ones are not the ones with the fastest checkout. The best ones ask questions. When a quote comes back suspiciously fast with no questions attached, I have learned to be suspicious of my own specification, not of the vendor.
A good relay distributor asks whether you need AC or DC coils, socket style, and contact material. A good VFD distributor asks to see the motor nameplate before quoting. A good PLC distributor asks which CPU, which firmware version, and which software package you are using. Those questions are not an inconvenience. They are a free design review.
Here is something vendors will not tell you: a fast quote is not always a quote from their own stock. Sometimes it is a quote from a supplier list, and you only discover that when the promised delivery date slips. I am not accusing anyone of dishonesty. I am saying that speed on the quote is not the same thing as reliability on delivery.
The 'we can fix it later' objection
Let me address the main objection I hear from other maintenance and engineering people: 'We can just return it if we got it wrong.'
Returns on automation products are not like returns on consumer goods. Try returning a PLC after firmware has been loaded, or a VFD after power has been applied, and you will meet restocking fees, freight costs, and a project delay that nobody budgeted for.
I totaled the direct damage from my documented mistakes. It came to roughly $8,500 in wasted spend, not counting the expedite fees, the overtime, and the damage to my credibility when I had to explain to a plant manager why the spare part I had personally ordered was the wrong one.
Five minutes of verification beats five days of correction. My notebook proves it.
The 12-point checklist I use before every order
In Q1 2024, I consolidated everything into a single checklist. We now use it for every order over $100, and for every component that goes into a live panel.
- Omron PLC: confirm the exact model number, including CPU power supply type.
- PLC outputs: relay or transistor, and does the application need high-speed pulse output?
- I/O count: verify against the current panel drawing, not the original bill of materials.
- Input voltage: AC or DC, and does the control transformer match?
- Relays and contactors: coil voltage, contact rating, pole count, socket style, and pinout.
- VFD: motor nameplate voltage and full-load amps.
- VFD: input supply voltage and phase at the actual installation point.
- VFD: constant torque or variable torque duty.
- Enclosure rating: IP20, IP54, or whatever the environment demands.
- Software and firmware: confirm compatible versions for the controller and configuration tools.
- Lead time: get written confirmation, not a verbal estimate.
- Return policy: know the restocking terms before you need them.
This checklist has caught 47 potential errors in 18 months. Some were small. A few, like a 230 VAC relay coil specified for a 24 VDC circuit, would have caused another embarrassing delay. The estimated cost of those avoided mistakes is somewhere in the $8,000 to $12,000 range, but the real saving is simpler: nobody on our team has had to make the 'I ordered the wrong part' phone call since we started using it.
I am not saying a checklist will catch everything. Commissioning always has surprises, and good engineering judgment still matters. But in my experience, the ratio is clear: for every issue that only appears during commissioning, there were three that a fifteen-minute spec check could have prevented. And I know which three cost more.
So here is the view I keep defending, even when it slows an order down: check before you buy. Not after.
