1. The four numbers that define a screw seat
Every countersunk screw seat is fully described by four values. Three of them are visible in the finished work. The fourth is the one that decides whether the other three mean anything.
| Number | What it is | Who decides it | Visible afterwards? |
|---|---|---|---|
| Pilot diameter | The hole that accepts the threaded core of the screw, sized to the root of the thread and to the density of the wood | The screw gauge and the species — hardwood takes a larger pilot than softwood | Yes, once the screw is withdrawn |
| Cone top diameter | The mouth of the cone at the surface, which must equal the head diameter of the fastener | The fastener, not you | Yes — it is the visible circle around the head |
| Cone depth | The distance from the surface to the apex of the cone | Fixed by geometry once the top diameter and the angle are known | Yes |
| Cone angle | The included angle of the cone, measured across the full cone rather than as a half angle from the centre line | The fastener standard — and usually not printed on the bit | Not without cutting the work apart |
The trap is in the last row. Every other drilling tool on this site is bought by its own dimensions: you choose a diameter and everything else follows. A countersink is bought by somebody else’s dimension. You cannot look at a cone and read its angle, most catalogues do not state it, many sets do not mark it, and by the time the error becomes visible the screw is in the work.
2. Where 82°, 90°, 100° and the rest actually come from
The angle is not a manufacturer preference. It is written into fastener standards, and the standard your customer’s screws follow decides the tool they need. The commonly cited mapping is:
| Included angle | What it belongs to | Governing documents commonly cited | Relevance to woodwork |
|---|---|---|---|
| 60° | Lathe centres, spotting, deburring small holes | Matches the standard included angle of lathe centres (ANSI B5.4) | Rare in joinery. Useful as a deburring cone and for starting small holes true |
| 82° | Inch-series flat head screws: wood, machine and tapping | ASME B18.6.1 (wood screws), ASME B18.6.2 (machine screws), ASME B18.6.3 (tapping screws) | The North American default. If a workshop there owns one angle, this is it |
| 90° | ISO metric countersunk heads | ISO 2009 (slotted), ISO 7046 (cross recessed), ISO 10642 (socket cap); DIN 963, DIN 965, DIN 7991. The countersink itself is dimensioned in DIN 335 | The metric default. Flat-pack hardware, European cabinet fittings and most imported machinery use it |
| 100° | Aerospace flush fasteners and aircraft structural rivets | NAS 523, along with the AN and MS series; ASME B18.6.5 covers 100° flat heads | Outside general woodwork. Occasionally appears with imported low-profile hardware |
| 110° | Specialist aerospace and some European automotive flush fasteners | Specified on the drawing rather than assumed | Rare, and visually too close to 100° to identify by eye |
| 120° | Shallow wide chamfers, deburring, lead-in chamfers | General practice rather than a fastener standard | Useful as a chamfer cone for edge break and for starting hardware |
The practical consequence for anyone buying these tools for resale is blunt: there is no such thing as a universal countersink, and a set of one angle is only right in a market where every fastener follows one standard. Almost no market today is that clean. Flat-pack furniture is metric, construction screws are inch, European cabinet hardware is metric, and a shop may use all three in a week.
3. What a mismatched angle does to the joint
Think of it as two cones: the one you cut, and the one on the underside of the head. If the angles are identical the two surfaces touch everywhere, the whole cone carries the clamping load, and the head finishes where you planned. If they differ, the surfaces can only touch along a single circle. All of the force the screw can deliver travels through that circle, and where that circle lands depends on which of the two cones is steeper.
| Bit wider than screw — the everyday version is a 90° bit under an 82° inch screw | Bit narrower than screw — the everyday version is an 82° bit under a 90° metric screw | |
|---|---|---|
| Where contact lands | At the small end of the cone, near the throat, because the head cone nests inside the wider cone and bottoms out before its rim reaches the surface | At the rim of the head and therefore at the mouth edge, because the blunter head cannot enter the steeper cone past its own rim |
| Immediate visual symptom | Head stands proud. The rim is above the surface and the joint does not close | Head will not seat. It rocks on the rim; if forced, the rim digs a groove |
| Size of the error on a #8 flat head | About 0.6 mm proud (see the depth table below) | Contact reduced to a knife edge; the head cannot be made to seat at any depth |
| Behaviour in softwood | Driving harder crushes the cone until it looks flush. Six months later the crushed fibres have moved and the ring around the head telegraphs through paint or clear finish | The rim crushes a narrow ring of surface fibre, then the head sinks a little more every time the joint is re-tightened until the screw no longer pulls |
| Behaviour in hardwood | The proud rim holds the joint off its mating surface, leaving a hairline gap at the joint line that invites over-driving and a stripped thread | The rim chips out rather than crushing, leaving a permanent halo around the head, and thin stock can split from the concentrated radial force |
| Behaviour on veneered board | Crushed ring plus a proud head: the finish bridges it and later sinks into it | The cone cuts past the face veneer into a core of a different colour, giving a dark ring no amount of sanding removes |
| Structural consequence | Clamp load concentrated on a small ring near the shank, so pull-through resistance falls well below what a full conical seat delivers | Clamp load concentrated on a knife edge at the surface, which both loses holding power and works as a wedge driving capillary splits along the grain |
Two real-world versions of this account for most of it. The first is an all-metric hardware package — hinges, drawer slides, connector bolts — being installed with inch-series 82° bits bought locally. The second is exactly its mirror image: a North American cabinet shop working with imported metric flat heads and a set of cones bought years earlier for local screws. In both cases the shop has a drawer full of good, sharp, correctly sharpened tools that cannot produce a correct joint.
4. Cone depth arithmetic: how deep is flush
Depth is not a judgement call. Once you know the head diameter and the cone angle, the flush depth is single-valued, and it comes straight out of right-angle trigonometry:
Since the divisor is a constant for any given angle, that collapses into a multiplier you can memorise and put on a card above the bench:
| Included angle | Depth multiplier k | Example: an 8.4 mm (#8) head | Where you will meet it |
|---|---|---|---|
| 60° | 0.866 | 7.3 mm | Centres, spotting, deburring cones |
| 82° | 0.575 | 4.8 mm | Inch flat head screws |
| 90° | 0.500 | 4.2 mm | ISO metric countersunk heads |
| 100° | 0.420 | 3.5 mm | Aerospace and some imported low-profile hardware |
| 110° | 0.350 | 2.9 mm | Specialist flush fasteners |
| 120° | 0.289 | 2.4 mm | Chamfering and deburring |
Applied to the flat head sizes most of the world actually screws together, using head and shank diameters from ASME B18.6.1, the table below is the one worth printing. Note the final column, which is the proud-head error from the previous section expressed millimetre for millimetre:
| Flat head size | Head dia. max | Shank dia. nominal | Flush depth at 82° | Flush depth at 90° | Proud error, 90° bit on this screw |
|---|---|---|---|---|---|
| #4 | 0.225″ / 5.72 mm | 0.112″ / 2.84 mm | 3.3 mm | 2.9 mm | 0.43 mm |
| #6 | 0.279″ / 7.09 mm | 0.138″ / 3.51 mm | 4.1 mm | 3.5 mm | 0.53 mm |
| #8 | 0.332″ / 8.43 mm | 0.164″ / 4.17 mm | 4.8 mm | 4.2 mm | 0.63 mm |
| #10 | 0.385″ / 9.78 mm | 0.190″ / 4.83 mm | 5.6 mm | 4.9 mm | 0.73 mm |
| #12 | 0.438″ / 11.13 mm | 0.216″ / 5.49 mm | 6.4 mm | 5.6 mm | 0.84 mm |
| #14 | 0.507″ / 12.88 mm | 0.242″ / 6.15 mm | 7.4 mm | 6.4 mm | 0.97 mm |
Half a millimetre does not sound like much. It is enough to catch a fingernail, hold a door stop off its seat, keep a drawer front from closing flush, or lift the edge of a carpet gripper. It is also enough that a shop will “fix” it by driving harder, which is how a crushed ring ends up permanently in the work.
5. Why over-sinking costs more than the depth you added
The instinct after a proud head is to go deeper. That instinct is expensive, and the reason is that deepening a cone does not just move the head down — it widens the visible circle at the same time. The relationship is:
At 82° that multiplier is 1.74. At 90° it is 2.00. At 100° it is 2.38. In words: every millimetre you sink too deep widens the visible cone by nearly two millimetres, which is why a slightly over-driven countersink looks not slightly wrong but obviously too big.
| Over-sink past flush | Visible mouth dia., 82° cone on a #8 head | Visible mouth dia., 90° cone | Visible mouth dia., 100° cone | What it means on a real panel |
|---|---|---|---|---|
| 0 (flush) | 8.4 mm | 8.4 mm | 8.4 mm | Head flush, cone invisible under it |
| 0.5 mm | 9.3 mm | 9.4 mm | 9.6 mm | A visible ring outside the head. Noticeable under a gloss finish |
| 1.0 mm | 10.2 mm | 10.4 mm | 10.8 mm | 20% wider than the head. Reads as sloppy at arm’s length |
| 1.5 mm | 11.0 mm | 11.4 mm | 12.0 mm | Cone well beyond the head; plug or cap no longer covers it |
| 2.0 mm | 11.9 mm | 12.4 mm | 13.2 mm | 42% wider than the head. Structurally meaningfully less material left above the cone |
| 3.0 mm | 13.7 mm | 14.4 mm | 15.6 mm | On 18 mm board you have removed most of the remaining section locally |
Three things follow from that table, and they are the reason depth control is worth paying for:
- On veneered sheet your entire error budget is the veneer. A cabinet-grade face veneer is nominally 0.6 mm and less after sanding. Sink past roughly that and the cone has cut through into a core that is usually a different colour and always a different density. There is no repair: the core ring accepts finish differently for the life of the piece.
- On MDF and particleboard the cone walls are already the worst surface in the workpiece. Widening the cone exposes more of that porous edge, which drinks primer and swells with waterborne finish. The halo around an over-sunk head on MDF is a finishing problem before it is a structural one.
- Painting over it does not work, because the ring that shows through is not a colour difference but a change in surface texture and absorbency. The only fix is filler, which then means the fastener can no longer be removed without destroying the surface.
The reverse rule is equally useful when you choose deliberately rather than by accident: if you intend to hide the screw under a plug or cap, deliberately sink for the plug thickness plus a little, and size the cutter to suit. Matching a Forstner bit or plug cutter to the cone is its own subject, and the fit tolerances there are far tighter than most people expect.
6. Measuring an unmarked cone angle without a gauge
A cone and its cut is fully described by mouth diameter and depth, so the same formula that sets the depth also lets you recover the angle. Invert it:
The method, which needs nothing more than a rule and a piece of scrap:
- Drill a pilot first, so the cone runs true and you are measuring the cone rather than a wobble.
- Cut the cone into scrap of the material you actually use, then stop the machine.
- Measure the mouth across its widest point, and the depth from the surface to the apex.
- Divide half the mouth by the depth, take the arctangent, double it.
A cone measuring 10.0 mm across and 5.0 mm deep gives half the mouth (5.0) divided by depth (5.0), which is 1.000, and the arctangent of 1.000 is 45°, so the cone is 90°. A cone 9.0 mm across and 5.2 mm deep gives 4.5 divided by 5.2, which is 0.865, and double the arctangent comes to roughly 82°. If you would rather not reach for a calculator, cut the scrap lengthwise through the middle of the cone and lay a screw of the target standard into the half-cone: any light gap between head and cone is the mismatch you were looking for.
7. The rim does the work: why cones burn at the mouth
There is one piece of physics behind almost every countersink wear pattern, and it is not obvious until someone points it out. A cone rotating about its axis has zero cutting speed at its apex and maximum cutting speed at its widest point. Along a single cutting edge, the surface speed rises from nothing to whatever it is at the rim.
Three consequences follow, and all three show up on the bench within the first hour of use:
- The rim wears first and the tool does not fail gradually. Everything else still looks sharp long after the rim has stopped cutting cleanly, which is why a cone can feel fine while producing a scorched ring and fuzzy walls.
- Burning appears at the mouth before anywhere else. Any combination of a dull rim, a dry hardwood or too much speed shows up as a dark ring exactly at the widest point of the cone, and no amount of extra feed pressure improves it.
- Always drill the pilot first. The apex of a cone is barely cutting at all; without a pilot it rubs and polishes rather than shears, generating heat at the bottom of the cone and pushing the tool off centre. The pilot is not just a guide for the screw, it removes the worst part of the cone’s workload.
Over-sinking makes all three worse, because it plunges the fastest part of the cone deeper into material than the joint requires. This is also where the over-sink arithmetic above and the wear question meet: the deepest usable cone is also the shortest-lived one.
On material selection, we keep this page short on purpose. In clean timber an ordinary high speed steel cone is the right answer and stays the right answer for a long time. Carbide earns its premium only on abrasive board — MDF, particleboard and laminates, where glue and resin wear a steel rim quickly. The full argument about tool steel grades sits on our HSS twist drill set page and about coatings on our Ti-coated set page; neither is specific enough to this tool to be worth repeating here.
8. Set composition is a matrix, not a ladder
Every other set of drills is a ladder: a list of diameters, usually in even steps, and the buyer’s only question is how far up it goes. A countersink set cannot work that way, because it has two independent variables — angle and body diameter — and a six-piece set can legitimately mean six diameters of one angle or three diameters of two. Those are different products aimed at different markets, and they look almost identical in a catalogue photograph.
| Market the set is aimed at | Angle(s) | Body diameters that earn their place | Why this list |
|---|---|---|---|
| North American carpentry and casework | 82° only | Roughly 6, 8, 9.5, 12 and 14 mm — sized to cover #4 through #14 flat heads | Every common fastener follows ASME B18.6.1. A second angle would sit unused |
| European, Asian and general metric assembly | 90° only | The DIN 335 form C ladder: 6.3, 8.3, 10.4, 12.4, 16.5 and 20.5 mm | The metric standards and DIN 335 are written around these steps, so catalogue pairings to metric thread sizes land on them cleanly |
| Mixed hardware — the realistic case for most importers | 82° and 90° | The three body sizes that cover the middle of the gauge range, in both angles — about 6, 9 and 13 mm | The middle of the range is where volume sits, and having both angles there prevents the common failure entirely |
| Furniture and cabinet production with visible plugs | Matching the fastener, usually 90° | One body diameter per fastener used, plus a matching plug arrangement | Repeatability matters more than coverage; a small matched set plus spare individual cones beats a wide set |
Two cautions on building that list.
First, specify the set by the fastener it must serve, not by piece count. A twelve-piece ladder of one angle is worse value to your customer than three diameters in each of two angles, even though it looks twice the product.
Second, this is a different argument from the one we make about holding spares. On our single auger page we discuss how many of a hot diameter to keep in stock, which is a consumption question driven by replacement rates. Here the question is narrower and comes first: which combinations must exist in the box at all, because a missing angle cannot be worked around and a missing diameter frequently can, with the next size up and a little extra settling.
9. Cutter forms: fluted, cross-hole, and combination units
| Form | How it cuts | In clean wood | Where it wins | Where it loses |
|---|---|---|---|---|
| Multi-flute, even spacing (3 to 6 edges, symmetrical) | Each edge takes a small successive bite; several edges share the load | Fast removal, good wall under a rigid setup, can produce faceted cones hand-held | Drill press work, production runs, hardwoods where rigidity is available | Hand-held use; each edge re-enters the cut at the same phase, which is what lets a chatter rhythm build |
| Multi-flute, irregular spacing (“chatterless” geometry) | Same principle, but no two adjacent gaps are equal so the impacts never synchronise | The most forgiving multi-edge form. Noticeably rounder cones at hand-held feed | General purpose; the default recommendation for most wood going through a portable drill | Costs more to grind; slightly less aggressive than an evenly spaced cone of the same size |
| Single-flute | One cutting edge with a large open gullet opposite it | Very free cutting, excellent chip clearance, will not self-excite | Softwoods, plastics, and deliberately heavy cuts where chips would pack a narrow gullet | Slower to finish a cone; a long stringy chip can clog if you let pressure off at the wrong moment |
| Cross-hole / zero-flute | A solid cone with one hole drilled through it; the edge of that hole is the cutter, and chips leave through the hole | The cleanest cone wall available with a hand-held tool, and effectively free of chatter | Hardwood and veneered work where the wall is visible; also excellent for hand deburring | Slower removal; the single port can pack in gummy material; usually the most expensive form per piece |
| Pilot and countersink combination (drill plus cone plus stop collar) | Drills the pilot, cuts the cone, and stops — in one plunge | Identical seat every time, automatic concentricity between hole and cone | High repetition: decks, carcass assembly, trim, anything with hundreds of the same screw | The pilot diameter is fixed to the unit, and the correct pilot changes with wood density |
| Adjustable pilot systems | Cone body accepts interchangeable pilots of several diameters | Fixes the rigidity problem of the fixed combination while keeping one-plunge speed | Shops that move between species with the same fastener gauge | More parts to lose, more joints to keep tight, and concentricity depends on the pilot seating properly every time |
The honest recommendation for a wood-only buyer is unglamorous: own a separate twist drill for the pilot, an irregularly spaced multi-flute cone for the seat, and a stop collar, unless your work is genuinely repetitive on one gauge and one species, in which case buy combination units sized to that job and accept that they are a consumable set-up rather than a general tool.
That matters more than it sounds, because hardwoods need a larger pilot than softwoods for the same gauge — the dense material cannot compress out of the thread’s way, so it splits instead. A combination unit set for pine is wrong the first time it meets oak, and correcting it means replacing the whole unit rather than swapping a twist drill out of an index.
10. Chatter in wood: ranking the real causes
The received wisdom is that multi-flute cones chatter and single-flute cones do not. That is roughly right for metal and incomplete for wood, and buying on it alone leads people to the wrong tool for the wrong reason. Chatter is self-excited vibration: one edge bites, deflects the tool, the next meeting occurs at a slightly different depth, and if the rhythm reinforces itself you get the faceted cone everyone recognises. A cone with several equally spaced edges gives that rhythm a clean beat to lock onto; uneven spacing deliberately destroys it, and a single edge has no partner to synchronise with at all. Hence the existence of irregular “chatterless” geometry and hence the truth in the old advice.
But in wood, those are third-order effects next to these three:
- No pilot hole. Without one, the cone has nothing to keep it on axis and spends the first half of the cut wandering until the walls pick it up, which produces oval entry mouths and poor concentricity with the screw. This is the most common single cause.
- Feed too light. Operators ease off, expecting a better finish, and get the opposite: the edges rub instead of shearing, friction rises, heat appears, and rub-to-grab is exactly how chatter starts. Feed steadily enough that the tool is always producing a chip.
- Runout in the drive line. A worn chuck, a bent shank, or a quick-change extension all move the cone off its own axis. This is the specific cost of convenience formats — we go into the trade-off in more detail on our hex shank drill bit page — and it shows up badly on a tool whose whole purpose is to produce a surface.
Fix those three before you spend money on a different cutter form. If you still have trouble, the order of escalation is a cross-hole cone, then a hand-held cross-hole used slowly by hand for the final pass on visible work, then a drill press.
11. Depth control, and why wood punishes sloppiness harder than metal
In metal, an over-deep countersink is mostly cosmetic. In wood it is progressive, because wood is cellular, absorbs moisture, and crushes before it recovers. Three mechanisms are specific to our material:
- Crush then rebound. A proud head can always be driven home by crushing the cone under it. It looks perfect on the day. The crushed fibres have been densified and their moisture response has changed, so weeks later — or after the first coat of waterborne finish — the ring around the head swells or shrinks at a different rate and telegraphs to the surface. This is the mechanism behind the majority of “it was fine when I fitted it” callbacks.
- Non-recovering crush in board. MDF and particleboard do not rebound at all. Once the cone is crushed it stays crushed, and any waterborne product applied afterwards swells that ring further. The result is a permanent shallow dish around every fastener on a painted panel.
- Veneer breakthrough. Covered above, and worth repeating because it is irreversible: roughly 0.6 mm of facing is the whole budget, and breaking through exposes a core with different colour and different absorbency.
| Depth control method | Repeatability | Cost | Drawbacks | Best used for |
|---|---|---|---|---|
| Setscrew stop collar | Reasonable, but slips under load | Low | A grubscrew bears on a round hardened shank and can creep, especially with vibration. Fit one that bears on a flat, or accept drift | Site work, short runs, and as a first line of control |
| Cage / roller depth stop | Best available hand-held | Medium | Adds length to the assembly and marks sensitive surfaces unless the bearing face is protected | Cabinet work, veneered sheet, anything where every seat must match |
| Drill press depth stop | Excellent | Already paid for | Requires the press; awkward for large assemblies that will not fit under the spindle | Shop work, repeat components, jigged production |
| Tape flag on the shank | Poor | Free | Tears, moves, and is routinely ignored. Better than nothing for a one-off | Emergencies only |
| Integral stop on a combination unit | Excellent within its set-up | Medium | Fixed to one pilot diameter and one depth until reset | High repetition on one gauge and one species |
The single rule: set the stop on scrap of the same material and the same thickness, with the same screw, and test-fit that screw. Setting a depth off a photograph or off last week’s panel is how the cones get too big.
12. Twelve countersink failures and their causes
| Symptom | Most likely cause | Material where it shows worst | What to change |
|---|---|---|---|
| Faceted or polygonal cone wall | Chatter from no pilot, light feed, or runout | Hardwood, or any board under a hand-held drill | Drill pilot first, feed to a chip, check runout, then move to irregular-spacing geometry |
| Fuzzy or torn cone wall | Dull rim, or a too-aggressive form in a soft species | Softwood, MDF, veneered board face | Sharpen or replace; use a cross-hole cone for the final light pass |
| Scorched dark ring at the mouth | Worn rim doing all the work, or surface speed too high at the widest point | Hardwood, resinous softwood | Replace the cone; do not compensate with more pressure |
| Screw head proud after a week | Angle mismatch, then crushed-to-fit | Softwood and MDF especially | Correct angle; set the stop for flush and leave it there |
| Dark ring around the head on veneered sheet | Cone cut through the face veneer into the core | Veneered plywood and laminated board | Reduce depth to within the veneer; use a cage stop |
| Cone off-centre from the pilot | No pilot, or runout, or the operator angled the tool | All | Pilot first with a sharp twist drill; keep the tool square |
| Visible cone far larger than the head | Over-sink, made worse by a wide angle | Painted panels and visible faces | Use the depth table; remember each millimetre costs nearly two millimetres of width |
| Head sits crooked in the seat | Angle mismatch in the narrow direction, or the pilot itself is angled | Hardwood, where nothing compresses to hide it | Correct angle; check the pilot was drilled square |
| Polished ring around the cone from the collar | Stop collar spinning against the surface | Pre-finished board and lacquered surfaces | Use a bearing cage stop, or protect the face |
| Combination unit drifts deeper through the run | Stop collar creeping under load | All, worsening as the day goes on | Reset against a hard stop regularly; prefer a cage unit for long runs |
| Chips packed in the cone, then burning | Gullet too small or the port blocked; easing off at the wrong moment | Damp softwood, plastics | Withdraw to clear; prefer single-flute or cross-hole forms in stringy material |
| Split along the grain near the edge | Pilot too small for the species, or load concentrated at the rim from an angle mismatch | Hardwood, and any near-edge fixing | Larger pilot for the species; correct angle; move the fastener away from the edge |
13. Full specifications
| Product name | Wood Countersink Drill Bits |
|---|---|
| Series / SKU | GUSHI-CSK-SET |
| Included angles offered | 82° (ASME B18.6.1 / B18.6.2 / B18.6.3 inch flat heads); 90° (ISO 2009, ISO 7046, ISO 10642; DIN 963, DIN 965, DIN 7991); 100° aerospace, 60° and 120° to order |
| Body diameter range | 6–31 mm (1/4″ – 1-1/4″) |
| Metric ladder available | To the DIN 335 form C series: 6.3, 8.3, 10.4, 12.4, 16.5, 20.5, 25.0, 31.0 mm |
| Standard set contents, metric | 6 pieces: 6.3, 8.3, 10.4, 12.4, 16.5, 20.5 mm at a single included angle |
| Standard set contents, inch | 5 pieces covering #4 through #14 flat heads, approximately 6, 8, 9.5, 12 and 14 mm body diameter |
| Mixed-standard set | 6 pieces: approx. 6, 9 and 13 mm body diameter in each of 82° and 90° |
| Cutter forms | Three-flute with even spacing; five-flute; irregular-spacing chatterless geometry; single-flute; cross-hole zero-flute; pilot-and-countersink combination |
| Number of cutting edges | 1, 3, 5 or 6 depending on form; cross-hole forms have a single cutting port |
| Cutting edge material | High speed steel as standard; carbon steel for economy lines; carbide-tipped for abrasive board |
| Body material | Steel body, hardened and tempered to a stated range on request |
| Shank forms | Round, three-flat, or 1/4″ quick-change hex |
| Shank diameters | To DIN 335 convention by size — typically 5, 6, 8, 10 and 12 mm across the metric ladder |
| Overall length | 45–71 mm across the metric ladder; longer overall lengths to order |
| Cutting direction | Right hand |
| Pilot arrangements | Plain cone (pilot drilled separately); fixed pilot combination; interchangeable pilot system |
| Depth control options | Setscrew stop collar; bearing cage stop; integral stop on combination units |
| Angle marking | Included angle stamped or etched on the shank of every piece, plus on the case |
| Suitable materials | Softwood, hardwood, plywood, veneered board, MDF, particleboard, laminated board, plastics |
| Restricted materials | Thin sheet metal and thin acrylic — a high-rake wood cone grabs and can tear the work |
| Unsuitable materials | Masonry, ceramic tile, glass, hardened steel |
| Machine guidance | Hand-held drill or drill press; see our Forstner set page for the general speed and feed method for large wood cutters |
| Surface treatments | Uncoated (bright), black oxide, or titanium nitride. See our coating page for what a coating does and does not change — it never changes the edge material |
| Packaging options | Individual hanging card; PVC pouch; blow-moulded case; wooden presentation case; bulk polybag for trade packs |
| Optional extras | Spare stop collars; matched plug cutters; spare pilots; branded depth gauge card |
| Inspection | 100% visual inspection before shipment; cone angle verified on a sample basis by the two-measurement method in section 6 |
| Country of manufacture | Made in China, or to your own origin requirement where specified |
14. How to specify a countersink set on an export order
Everything above reduces to a short checklist. Miss the first item and nothing else on the list matters.
- State the included angle explicitly, in degrees, on the order line. Not “metric”, not “standard”, and certainly not nothing. 82° or 90° written on the line.
- Name the fastener standard your market actually buys — ASME B18.6.1 for inch wood screws, ISO 2009 / 7046 / 10642 for metric — so a substitution cannot be argued.
- List the body diameters, or the gauge range they must cover. Do not ask for a piece count and leave the diameters to the supplier.
- Choose the cutter form against how your customers work: irregular-spacing multi-flute for general hand-held joinery, cross-hole for visible hardwood and veneer, combination units for high repetition.
- Specify the shank form by the chucks common in your market — round, three-flat, or 1/4″ quick-change hex — and remember every quick-change joint in the drive line is runout you will pay for at the cone.
- Decide the depth control and whether every unit ships with it. A stop collar in the same box converts a tool into a system and costs very little.
- Require angle marking on every piece. This is the cheapest quality upgrade available anywhere in this category.
- Attach a drawing or reference photograph and request a counter-sample. Two manufacturers can both deliver a compliant “6-piece 90° set” that share nothing but the count.
Two questions worth asking a supplier before the first container, because the answers sort out quality faster than any certificate:
- “What steel grade and hardness range is the cone, and will you put it in writing?” — because “industrial grade high speed steel” is a marketing phrase, not a specification.
- “Can you show me how the included angle is verified in your process?” — because a shop that can describe its check generally has one, and a shop that cannot usually goes by eye.
15. Incoming inspection
A countersink arrives either perfect or useless, and nothing in between can be seen from across a warehouse. Seven checks:
- Confirm the angle. Cut scrap, measure mouth and depth, compute. Check one piece per size per shipment, not one piece per shipment.
- Check concentricity. Spin the bit in the intended chuck or holder, not in a perfect test fixture, and look at the rim. A rim that moves relative to the shank will produce an oval mouth.
- Read the marking. Angle and diameter must both be legible on the piece. Unmarked stock cannot be re-ordered correctly and cannot be issued correctly.
- Inspect the cutting edges under modest magnification. Look for chipping, burnt grinding, and inconsistent relief between one edge and the next.
- Check the cone for true form. A cone ground off-axis is visible as an uneven land width along one flute.
- Test the depth control. Fit the collar, set it, take thirty cuts into scrap, and re-measure. A collar that creeps will drift through a production run.
- Check presentation as your customers will receive it: pieces retained, no contact between cones during transit, rust protection present on uncoated steel for sea freight.
16. Troubleshooting
| If this is happening | Check first | Then | If still unresolved |
|---|---|---|---|
| Head will not sit flush however deep you go | Whether the bit angle matches the screw standard | Split a test cone and lay the screw in it | You have the wrong angle. Replace the tool; no technique compensates |
| Cone comes out with visible facets | Pilot, feed rate, runout — in that order | Irregular-spacing geometry | Move to a cross-hole form and finish the last pass slowly by hand |
| Cone is consistently too large at the mouth | Depth setting against the table in section 4 | Whether the collar is drifting | Change to a cage stop and re-set against scrap every set-up |
| Burning at the mouth | Rim sharpness | Surface speed relative to the widest point of the cone | Replace the cone; check that the pilot is pre-drilled so the apex is not rubbing |
| Screws loosen over time | Whether the head was crushed in rather than seated | Pilot diameter against the species | Correct angle, then re-set the stop for a true flush seat |
| Pilot and cone not concentric | Chuck condition and any extension in the line | Whether the pilot is being drilled square | Use a guide or a drill press; retire worn extensions |
17. Applications
- Cabinet and carcass assembly — the highest-volume use, and the one where a wrong angle gets discovered after the finish is on.
- Joinery and trim — architrave, skirting and window boards, where a proud head prevents a joint closing.
- Decking and exterior carpentry — where flush seating is a snag hazard as much as an appearance issue, and where combination units pay for themselves fastest.
- Flat-pack and metric hardware installation — the single biggest source of 90° requirements in traditionally inch markets.
- Furniture with hidden fixings — where deliberate controlled over-sink plus a plug is the specified finish, which makes accurate depth control a requirement rather than a nicety.
- Veneered and laminated panel work — where the whole tolerance budget is the thickness of the facing veneer.
- Shopfitting and exhibition build — rapid assembly in mixed board materials, usually one gauge repeated hundreds of times.
- Door and hardware hanging — hinge and plate screws that must sit flush against hardware with no gap.
18. Frequently asked questions
Should I buy an 82° or a 90° countersink for wood screws?
It depends entirely on the screws, not on where you live. Inch-series flat head wood screws made to ASME B18.6.1 are 82°, and practically every flat head wood screw sold in North America follows it. ISO metric countersunk heads covered by ISO 2009, ISO 7046 and ISO 10642 are 90°, and that includes most flat-pack hardware, European cabinet fittings and machine fasteners. If your fasteners are mixed you need both angles, and the honest answer for most export programmes is a set containing both rather than a bigger set of one.
What actually happens if I use a 90° countersink with an 82° screw?
The head stands proud and the load path collapses onto a ring. An 82° cone is steeper and therefore deeper than a 90° cone of the same mouth diameter, so when you cut the shallower cone the screw bottoms out against the small end before its rim reaches the surface. On a #8 flat head the difference is roughly 0.6 mm, enough to catch a cloth, stop a drawer closing, or hold a mating part off its seat, and all the clamping force travels through a narrow ring instead of the whole conical surface. No depth adjustment fixes this: sinking deeper moves the contact line, it does not create cone contact.
How deep should a countersink be?
Depth is not a preference, it is fixed by geometry. For a flush seat, depth equals the head radius divided by the tangent of half the cone angle, which is the same as head diameter multiplied by a constant for that angle: 0.575 at 82°, 0.500 at 90°, 0.420 at 100° and 0.866 at 60°. An 8.4 mm head seated at 82° must be 4.8 mm deep; the same head at 90° only 4.2 mm. Anything past that figure over-sinks the cone.
How much damage does over-sinking do?
Far more than the extra depth suggests, because deepening the cone widens its mouth at the same time. Every millimetre you sink past flush widens the visible mouth by 1.74 mm at 82°, 2.00 mm at 90° and 2.38 mm at 100°. A #8 head is 8.4 mm across, so one millimetre too deep turns an 8.4 mm seat into a 10.2 mm crater, and two millimetres too deep into nearly 12 mm. On solid timber you simply lose holding material. On veneered board you cut through a face veneer that may be 0.6 mm thick and expose a core of another colour, which no amount of sanding removes.
Why does my countersink leave a rough, faceted cone?
That is chatter, and in wood it is usually not primarily a flute-count problem. Three causes account for most of it: there is no pilot hole guiding the cone, the feed is too light so the edges rub rather than shear, and there is runout in the drive line from a worn chuck, a bent shank or a quick-change extension. Fix those first. If trouble persists, choose a cutter with uneven flute spacing or a single cutting edge — irregular spacing stops each edge re-entering the cut at the same phase, which is what lets the vibration build into a polygon.
Is a pilot and countersink combination bit better than two separate tools?
For repetitive driving of one screw gauge into one material, yes, and it is not close: one plunge, perfect concentricity between hole and cone, and a depth collar that makes every seat identical. But the combination locks the pilot diameter to the unit, and the correct pilot diameter changes with wood density, being larger in hardwood than in softwood. So a combination set up for softwood is wrong the day it meets oak, and correcting it means replacing the unit rather than swapping a twist drill.
How can I check the angle of a countersink that has no marking?
Cut it and measure it — you only need two numbers. Bore a pilot so the cone runs true, cut into scrap, then measure the mouth diameter and the depth. Divide half the mouth by the depth to get the tangent of half the angle, then double the arctangent. A cone 10.0 mm across and 5.0 mm deep is exactly 90°; one 9.0 mm across and 5.2 mm deep works out at about 82°. Alternatively, split the scrap lengthwise through the cone and lay a known-angle screw into the half-cone, looking for a light gap.
Can I use a wood countersink bit on metal?
Occasionally on thick mild steel at low speed, but treat that as an exception rather than a capability, and never on thin sheet. Wood cones are ground with generous rake and clearance so they slice fibrous material freely; that same geometry digs in on hard, thin material, and a grab at the mouth of a thin sheet will wrench the drill and tear the workpiece. Use a countersink ground for metal, clamp the work, and if you do both routinely keep separate clearly identified tools.
What exactly is in a six-piece countersink set?
Whatever the buyer specified, which is the point. A countersink set is a matrix of angle against body diameter, not a ladder of diameters, so six pieces can mean six diameters of one angle or three diameters of two. For an inch market the useful run is 82° cones sized to cover #4 through #14 flat heads, roughly 6, 8, 9.5, 12 and 14 mm body diameter. For a metric market the natural run is the DIN 335 form C ladder of 6.3, 8.3, 10.4, 12.4, 16.5 and 20.5 mm at 90°. If your customers see both standards, three diameters in both angles beats six diameters in one.
Why does my countersink burn a dark ring at the mouth of the cone?
Because cutting speed along a cone is not constant: it is zero at the apex and maximum at the widest point, so the rim does nearly all the work and generates nearly all the heat. Burning at the mouth means the rim has lost its edge or the surface speed there is too high for the material, and over-sinking makes it worse by engaging the fastest part of the cone deeper. Sharpen or replace before increasing effort, and keep the pilot pre-drilled so a barely cutting apex is not also rubbing.
Will using my drill’s fastest speed make the cone smoother?
No, and it does the opposite at the rim. Speed has to fall as diameter grows, because the rim is the fastest-moving part of the tool and it is the part that burns and dulls first. We deliberately do not publish a speed table here: the general method for large wood cutting diameters is set out on our Forstner drill bit set page and the metric speed formula on our HSS twist drill set page. Both apply here, both need the rim diameter rather than the nominal size as the input, and both will tell you that slower than instinct is usually right for a cone.
MOQ, lead time, OEM and samples?
Countersink sets are configured rather than picked from a fixed catalogue, so commercial terms follow the configuration. Indicative minimum order quantity starts from 300 sets for a stock angle and diameter combination, and from 1,000 pieces for an angle outside our standard range, a private tool profile or private label packaging. Indicative lead time is 15 to 25 days for stock configurations and 30 to 45 days for OEM production. Samples are available for testing after terms are confirmed, with the sample cost credited against a subsequent order. Please confirm current figures with us before publishing them.
OEM & Wholesale
This is the most configurable product in the wood drilling range, which is exactly why it generates complaints when it is sold as a commodity. We support:
- Angle specified per line, per market — 82°, 90°, or both in one order, marked on every piece.
- Body diameter lists built around the fasteners your customers buy, rather than around a piece count.
- Cutter form selection by application — irregular-spacing multi-flute for general joinery, cross-hole for visible hardwood and veneer, combination units for production volumes.
- Depth control supplied with the tool, which converts a cone into a repeatable system and materially reduces returns.
- Shank form matched to your chucks and machines — round, three-flat, or 1/4″ quick-change hex.
- Angle verification documentation on request, using the two-measurement method described on this page.
- Private label and private mould packaging, from hanging cards to blow-moulded cases.
Our factory is ISO 9001:2015 certified as a quality management system. Please ask us for the certificate scope and number before reproducing that claim as a product certification.





