Buying EXFO Test Gear: 7 Questions I Asked Before Signing (From the Procurement Side)

What I asked before we bought our first EXFO unit

I handle purchasing for a 240-person network services company spread across three offices. About $180,000 a year runs through eight vendors, and fiber test equipment is the line item that generates the most questions from finance. Test gear looks simple on a purchase order — one number, one box — and then you find out it isn't.

Here's what I actually asked, in the order I asked it. No marketing language, just the stuff that came up in the hallway.

• Is EXFO worth the premium, or am I paying for the logo?
• What does an EXFO DWDM OTDR do that a normal OTDR can't?
• Can I just buy a used FTB-200 and save 70%?
• How do I prove a fiber link is ready for service — and when?
• What costs are missing from the quote I got?
• Why doesn't the model name on the request form match any catalog?
• What happens when a unit dies in the middle of a project?

1. Is EXFO worth the premium, or am I paying for the logo?

It's tempting to think you can line up two OTDRs by spec sheet and pick the cheaper one. But identical specs from different vendors can produce wildly different results in the field, mostly around dynamic range, blind zone behavior, and how the software handles a messy trace at the far end of a long span.

My honest position: I stopped treating price per unit as the deciding number in 2023, after a project where the “value” unit passed every benchmark and still generated three days of re-testing because our crew couldn't get a clean splice reading. The re-test labor cost more than the price difference. That said, EXFO isn't automatically right for every job either. A contractor doing short indoor runs doesn't need the same instrument as a team sweeping 80 km spans.

2. What does an EXFO DWDM OTDR actually do that a regular OTDR can't?

This one took me two conversations with our lead tech to understand. A standard OTDR fires at fixed wavelengths like 1310/1550 nm. On a DWDM link you have a multiplexer sitting in the middle, and a regular OTDR just sees the mux as an enormous loss event — it can't see past it, and it may not be able to test through it at all without disrupting live traffic.

A DWDM OTDR tests at wavelengths tuned to the ITU grid or in an out-of-band window, so it can characterize the span through the mux and per-channel. That's why the price jumps. If your network is all point-to-point dark fiber, you likely don't need it. If you're turning up DWDM routes, you do, and renting one for the duration of a project is a legitimate option.

3. Can I just buy a used FTB-200 and save 70%?

I went back and forth on this for almost a month. A refurbished FTB-200 platform with an OTDR module came in around a fifth of a new unit; new handhelds with current modules were five figures. On paper, used made sense. What tipped me was calibration: a unit without a current calibration certificate is a measurement I can't defend to a customer's auditor.

I still kick myself for the first used unit we bought without asking for the calibration history — the certificate had expired eleven months earlier, and recalibration ate most of the savings. Now the rule is simple: no certificate, no purchase. A used FTB-200 is genuinely fine for internal troubleshooting; I'm just careful about using one for acceptance testing a client will audit.

4. How do I prove a fiber link is ready for service, and when does that clock start?

This is the question nobody asks until an acceptance document gets rejected. The answer isn't “when the test passes.” It's when you can produce a record that shows what was tested, with what instrument, at what settings, on what date.

Practically, that means bidirectional OTDR traces at both wavelengths, with splice losses averaged across the two directions — a splice can read high one way and clean the other, so a single-direction pass is not evidence. You also want the instrument's serial number, calibration expiry, refractive index and pulse width settings, plus total span loss and length. Most vendors, EXFO included, ship software that batch-generates these reports, and that's worth more than it sounds when you're handing 200 spans to a client.

On timing: the link is ready for service when the report is signed and filed, not when the light turns green. I've watched a three-week sign-off slip because the testing was done in March and the paperwork was reconstructed in June.

5. What costs are missing from the quote I got?

Almost always four things. Calibration cycles, because an annual recalibration is a recurring line item, not a one-time cost. Software licenses and firmware support, which sometimes sit outside the hardware price. Accessories — launch cables, adapters, inspection scopes, cleaning kits — which quietly add 10-15% to a “complete” system. And training, which is the one people skip and regret.

I have mixed feelings about paying for vendor training. On one hand, our techs already know fiber. On the other, half the bad test data I've seen came from wrong settings, not bad hardware. We now budget training into every instrument purchase and it has cut repeat testing noticeably.

6. Why doesn't the model name on the request form match any catalog?

Here's the one that surprised me. Request forms arrive with names that don't exist in any manufacturer's catalog. Someone writes down what a colleague called a tool in the field, or what a search result suggested, and it lands on my desk as a model number. “Magic Max,” a “DuraForce Pro 3,” an “EXFO FTB-200” — one of those is a legitimate though aging test platform, one is a ruggedized handset a field crew might carry for photos and tickets, and one I still haven't been able to trace to a real part number.

Take it from someone who ordered the wrong module once: never place an order from a name. Ask for the manufacturer part number, the module configuration, and which wavelength or function the crew actually needs. Fifteen minutes of verification beats a return authorization.

7. What happens when a unit dies in the middle of a project?

Worse than expected, the first time. We had a main unit fail on a Tuesday with two weeks of acceptance testing left. No loaner agreement, no local depot, and the replacement lead time was longer than the project. We rented, at a premium, and I explained the cost to a VP who was not impressed.

Now I ask two things before any purchase over $10,000: what's the repair turnaround, and is there a loaner or rental path while it's out. Not because EXFO gear is unreliable — ours has been fine — but because the cost of downtime isn't the repair invoice, it's the crew standing around. That number never appears in a comparison spreadsheet, and it's usually the biggest one.