-
When You Think You Know What You're Doing (But Don't)
-
The 5-Step Pre-Purchase Checklist
-
Step 1: Define the Actual Load Profile (Not Just the Nameplate)
-
Step 3: Yes, Step 3—Mind the Feedback Loop (Encoder vs. Resolver vs. Sensorless)
-
Step 4 (Wait, Actually Step 3): Calculate the Real Peak Power (And Check the Bus)
-
Step 4: Verify the Cable and Connector Compatibility (Yes, This Matters)
-
Step 5: The 'What If' Test—Expect an Error
-
Step 1: Define the Actual Load Profile (Not Just the Nameplate)
-
Common Cost-Cutting Traps That Increase TCO
When You Think You Know What You're Doing (But Don't)
In my first year handling orders for industrial motion components (2017), I assumed that the cheapest VFD with the right kW rating was always the smart choice. I was wrong. Two weeks after installing a bargain-bin drive, we had a full line shutdown. The production loss and the rush-delivery of a replacement unit cost us over $3,200—more than the original 'savings' of $400.
That was my first expensive lesson in Total Cost of Ownership (TCO). Since then, I've made another 18 significant errors (documented them all), and I swear by a checklist now. This guide is for the engineer or buyer who is about to spec a servo, stepper, or VFD for a new machine or retrofit, and doesn't want to learn the hard way. Here are the five critical steps I never skip.
The 5-Step Pre-Purchase Checklist
Step 1: Define the Actual Load Profile (Not Just the Nameplate)
You'll read lots of specs: continuous torque, peak torque, speed range. But the real trick is understanding duty cycle. I once ordered a Hydac servo motor (great unit) based on a 'continuous current' rating, but the application needed a high dwell time at near-zero rpm with constant torque. The motor overheated.
Action: Before even looking at the catalog, sketch out the velocity vs. time profile for 10 minutes of operation. What are the peak torque demands? Is there any regenerative braking? Stick this profile in the email when you ask for a quote.
Step 3: Yes, Step 3—Mind the Feedback Loop (Encoder vs. Resolver vs. Sensorless)
I know the structure says Step 2 comes after Step 1, but this is so often overlooked I'm skipping ahead. Most people assume that a servo motor needs an encoder. But what kind? Absolute or incremental? Single-turn or multi-turn?
I skipped the feedback specification on a stepper motor controller order for a pick-and-place robot. The controller assumed an incremental encoder with index pulse. I ordered a multi-turn absolute encoder. oops. Should mention: the $150 unit difference turned into a $700 rework because the cable was pinned differently.
Checklist:
— For servo: specify resolution (e.g., 23-bit absolute), interface (EnDat, BiSS, SSI), and whether you need multi-turn or single-turn.
— For stepper: confirm if you need a closed-loop controller with an encoder, or if open-loop is acceptable. Open-loop is great unless you lose a step under load.
— For VFD: do you need a feedback card for closed-loop vector control, or is sensorless vector enough? For general pump/fan applications, sensorless may save you $200.
Step 4 (Wait, Actually Step 3): Calculate the Real Peak Power (And Check the Bus)
This is the one most people miss. A motor's peak torque spec is nice, but the drive (VFD or servo drive) needs to deliver that peak current from its DC bus. I ignored the fact that my Hydac VFD was rated for 7.5 kW continuous, but the peak current for a short 500ms could be 3x that. The bus capacitors weren't sized for the surge.
Action: Calculate the peak power required. If using a VFD for a high-inertia load (like a centrifuge), make sure to oversize the VFD or add a braking resistor. For servo drives, check the peak current rating (for 1 second, 3 seconds, etc.). The drive should cost a bit more upfront than the motor TCO calculation will save you.
Step 4: Verify the Cable and Connector Compatibility (Yes, This Matters)
You have a Hydac servo motor. You have a third-party servo drive. Do the connectors match? I ordered 5 motors and 5 drives for a multi-axis project. The motor had a standard M23 connector. The drive had a terminal block. $890 wasted on custom cable assemblies and a 1-week delay.
Checklist:
— Is the motor feedback connector compatible with the drive's feedback card?
— What about the motor power cable? What gauge braid for shielding?
— Are you ordering pre-terminated cables from the catalog (e.g., Hydac's own cable assemblies) or planning to field-terminate? Field termination saves cost but risks errors (I've done it).
Pro tip: if you're ordering a VFD, check the cable length limit for the motor. Long runs can cause voltage reflection issues. Most VFD manuals state max cable lengths without a dV/dT filter. (Based on manufacturer specs, check your specific model.)
Step 5: The 'What If' Test—Expect an Error
After three rejections in Q1 2024, I created a final sanity check. Imagine you've wired the system, powered it up, and the motor doesn't move. What's the most likely mistake? The encoder mismatch? The wrong direction of rotation? A missing brake release? When you spec the parts, plan for that failure mode.
For example: I now always order a safety relay and a regen resistor for VFDs—even if the application doesn't 'need' it now. It's a $40 resistor vs. a $3,000 burned-up motor drive when a web breaks and the inertia spins back. (Note to self: add this to the budget checklist for every order.)
Common Cost-Cutting Traps That Increase TCO
It's tempting to think you can just compare kW ratings. But identical specs from different drive manufacturers can result in wildly different outcomes. The $500 VFD quote turned into $800 after adding the EMC filter, external braking resistor, and the software configuration license. The 'all-in' solution from a single supplier (maybe a Hydac integrated servo+drive) at $650 was actually cheaper.
Hidden costs I've seen:
— Setup fees for drives that require a licensing dongle.
— EMC filter boxes that aren't included and cost $150.
— Programming cable that's proprietary ($80 for a USB-stick).
My rule of thumb: once you have a final bill of materials, add 15% for 'integration surprises'—cables, filters, brackets, and software. If the cheapest quote still beats the others by 20% or more, triple-check the feedback compatibility. (Based on 25+ fails over 5 years, I'm now at a 75% success rate of predicting the hidden cost category.)
One last warning: don't forget the logo. When you're ordering a branded system (like Hydac), you're paying for the catalog, the online configurator, and the tech support. That's worth something. I learned this after ordering a 'generic' servo controller that had zero documentation in English. Time is a hidden cost too.