Emtek Door Hardware, Gas Valves, and MIG Settings: The Spec Mistake You're Most Likely to Make

I've been handling hardware orders for a mid-size commercial GC and a millwork shop for 11 years. In that time I've documented 34 mistakes that cost real money — roughly $41,000 in reorders, restocking fees, and rework labor. I keep a checklist now. What follows is basically that checklist, sorted by project type.

Here's what nobody tells you when you start buying hardware: there isn't one right way to spec it. The mistake you're most likely to make depends almost entirely on what you're installing and who has to live with it afterward. So instead of one generic list of "best practices," here are four scenarios. Find yours.

• Scenario A: Interior door hardware on a residential job
• Scenario B: Exterior and entry sets
• Scenario C: Gas piping and valves
• Scenario D: Fabrication — fasteners and welding

If you're doing more than one of these on the same project (and most contractors are), the traps stack. I'll get to that at the end.

Scenario A: Interior Door Hardware on a Residential Job

This is where people assume the least risk and get burned the most, because interior hardware feels forgiving. It isn't.

Say you're putting together a whole-house package of emtek arts and crafts door hardware — that Craftsman look with square rosettes and rectangular levers. Three things go sideways almost every time:

1. Rosette geometry. Square rosettes are way less forgiving than round ones. A round rose covers a slightly off-center bore or a chipped edge. A square one doesn't. If the door was prepped before the hardware was picked, measure the actual bore and cross-bore centers on site — don't trust the door schedule.
2. Finish matching across product families. Two items can both say "oil-rubbed bronze" and still not match, because one is a lacquered finish and the other is a living finish that keeps darkening. Put them on the same door and it shows up in the first week. I'm not 100% sure every manufacturer documents this clearly on the product page — but I can tell you the sample rings do, and that's why you should order one.
3. Backset and door thickness. 2-3/8" vs 2-3/4" backset, and 1-3/8" vs 1-3/4" door thickness. Older homes are often 1-3/8". Standard stock latches assume 1-3/4". The latch will bolt up. It just won't reach.

In April 2023 I shipped a 47-piece interior package to a jobsite. Every lever was right. Every latch was the wrong backset. That was $1,880 in reorders, plus two days of a carpenter standing around waiting — which cost more than the hardware did.

The fix isn't complicated: pull the spec sheet, then measure three doors. Not one. Three. The first door in a house is almost always the one that was framed correctly.

Scenario B: Exterior and Entry Sets

Emtek exterior door hardware adds two variables that interior sets don't have: handing and exposure.

Handing is the big one. Most entry sets are handed, and the hand is determined by which side the hinges are on when you're standing outside the door, looking at it. Half the returns I've processed over the years were handing errors because somebody stood inside the house and guessed.

Exposure is the second. Coastal and high-humidity environments are hard on finishes, and a lot of finish options that look great in a showroom aren't recommended for salt air. I don't want to overstate this — most finishes hold up fine in most places — but if the house is within a few miles of the ocean, ask the dealer directly and get it in writing. Check the finish warranty terms for the specific product line too. They're not identical across families.

And then there's the prep question. Is this a standard bored prep, a mortise prep, or a multipoint system? Those aren't interchangeable, and the door has usually already been cut for one of them by the time the hardware gets ordered. In November 2022 I ordered six entry sets with the hand reversed on all six. Restocking fee was 25%. That's $780 for a mistake I could've avoided with a photograph taken from the driveway.

If I had 10 minutes with every person ordering exterior hardware, I'd spend nine of them on handing and one on the door thickness. Everything else you can fix on site.

Scenario C: Gas Piping and Valves

This is the scenario with the least margin for error, because a mistake here isn't a callback — it's a safety issue. A 1/2" FNPT bronze gas ball valve looks about as simple as hardware gets. It's not.

FNPT means Female National Pipe Taper. That's it. The trap is what people assume around it:

• FNPT vs MNPT. Female on both ends, or female one end and male the other? Read the product listing twice. I've watched a plumber cut and re-thread four nipples because the valve he grabbed was male-by-female.
• What service the valve is actually rated for. A lot of 1/2-inch bronze ball valves sitting on the shelf are rated for water, oil, and air. That's a different service than fuel gas. Check the valve's listing and rating — not the shelf tag.
• Thread sealant. White PTFE tape is the default for water. For gas, many jurisdictions require a gas-rated sealant or yellow tape. This varies, so check with your local authority having jurisdiction rather than going off what the last guy did.
• Full port vs. standard port. Standard port valves are restricted internally and will affect flow. If the appliance is sensitive to pressure drop, this matters.
• Handle clearance. A lever handle needs room to swing. Behind a water heater in a closet, a tee handle is sometimes the only thing that fits.

Communication failure story: I typed "1/2-inch valve, bronze" on a requisition. The counter guy read "1/2-inch valve, bronze, water." Four of them got installed before the inspector flagged it. Ripping out four valves, re-cutting and re-threading nipples, and re-pressurizing: $1,240 and half a day of a licensed plumber's time. The valve cost $14.

The lesson I took from it is that "1/2-inch valve" is not a specification. It's a description of an object. Write the rating, the service, the port type, and the end connections every single time, even when it feels excessive.

Scenario D: Fabrication — Fasteners and Welding

Two sub-traps here, and they're related: people treat both as commodity decisions, and neither is one.

The hex cap screw trap

Three things bite people on hex cap screws, in order of frequency:

1. Length is measured under the head. Hex cap screws are measured from the bearing surface to the tip. Flat head screws are measured overall, including the head. So a 2-inch hex cap screw and a 2-inch flat head are not the same length in practice. This trips up everybody at least once.
2. Grade markings matter. A head with no grade marking is not a Grade 5. Grades run 2, 5, 8 in inch series and 8.8, 10.9, 12.9 in metric. And don't mix grades inside one connection — the assembly is only as strong as the weakest fastener in it.
3. Thread pitch. 1/2-13 (coarse) and 1/2-20 (fine) are both "1/2-inch" and they're not interchangeable. If the drawing just says "1/2-inch bolt," it's incomplete. Same with coating: hot-dip galvanized adds enough thickness that nuts sometimes won't spin on without chasing the threads.

I once ordered 300 hex cap screws off a drawing that said "1/2 x 2." Wrong pitch. $340 in unusable stock and five days of a fabricator waiting on the right ones. The drawing wasn't wrong, technically. It just wasn't complete, and I didn't push back on it.

How to set a MIG welding machine (the part the door chart gets wrong)

Every MIG machine has a chart inside the door. Everyone uses it. Almost nobody reads the footnotes.

The chart assumes a specific shielding gas mix, a specific wire diameter, and a specific stickout distance. Change any one of those and the chart is wrong. And on most mid-range machines, the dial numbers aren't volts and they aren't inches per minute — they're arbitrary 1-to-10 ranges. So you can't pull settings off a forum and dial them in. You'd be copying numbers that don't mean the same thing on your machine.

What actually matters is understanding the two controls:

• Voltage controls arc length. Too little voltage for your wire feed speed and the wire stubs into the puddle — crackling, popping, spatter everywhere.
• Wire feed speed controls amperage. Too much voltage for the wire speed and you get a wide, flat bead that looks pretty and has poor penetration into the root.

Don't hold me to this, but for .030" solid wire on 1/8-inch mild steel with 75/25 argon/CO2 in short-circuit transfer, I usually land somewhere around 18–19 volts and 200–240 inches per minute. Your machine might need something different. The chart in the door is a starting point, not an answer.

When I compared two identical T-joints side by side — same machine, same wire, same gas, same material — one welded on the door chart and one dialed in by ear — I finally understood what everyone meant by "read the puddle." The chart weld looked fine from above. It had cold lap at the root. The difference wasn't the machine. It was 15 minutes of test coupons on the same thickness and joint type before touching the actual part.

That's the whole MIG answer, honestly: cut two scrap coupons, run them, break them, and adjust. Every machine, every wire spool, every gas mix is slightly different, and no chart can account for that.

Which One Are You?

Three questions will sort you into the right scenario:

1. Will a homeowner or tenant touch this every day? If yes, you're in Scenario A or B. Spend your time on finishes, handing, and geometry. Nobody forgives a lever that's the wrong shade of bronze.
2. Does it carry pressure, fuel, or a structural load? If yes, you're in Scenario C or D. Spend your time on ratings, listings, and grades. Download the spec sheet, don't rely on the shelf tag.
3. Are you replacing like-for-like, or specifying something new? Replacements are low risk — match what's there. New specifications are where every scenario above becomes dangerous, because there's no existing condition to check yourself against.

If you answered yes to more than one of those, you're like most of us, and the risk compounds. Interior hardware and entry sets are usually one order, and getting the finish family consistent across both takes one extra phone call to an authorized dealer who can pull full spec sheets side by side.

If you ask me, the pattern behind all four of these scenarios is the same thing: the expensive mistakes are never about not knowing enough. They're about ordering before measuring. Every single one of my 34 documented errors happened because I placed an order against a drawing, a schedule, or a memory — instead of against the physical thing in front of me.

I'd rather spend 10 minutes on the phone walking a customer through a spec sheet than deal with a $1,900 return and a carpenter standing around. An informed buyer asks better questions and orders once. That's the entire argument for all of this.

Prices and costs in this article are from my own project records and are for general reference only. Ratings, code requirements, and product listings vary by jurisdiction and change over time — verify current requirements with the manufacturer's literature and your local authority having jurisdiction before ordering or installing.