Chasing a 0.0007 in wobble on a 2008 Haas VF-2 spindle and I’m getting conflicting readings between a Blake coax and a Mitutoyo 0.0005 in test indicator (about 0.0003 in apart). Which tool do you trust for final tram, and any trick to keep the mag base from flexing on the table while nudging the head? Trying to decide if I’m looking at face runout or a slightly bent drawbar before first shift.
, I’d trust the Mitutoyo for final — the Blake can lag/stick and 0.0003" spread isn’t shocking if the lever angle or rpm is low. Kill mag-base flex by parking the Noga on a 1–2-3 block and toe-clamping the block to the VF-2 table, or put the TI on a spindle bar and sweep a ground ring to separate tram from face runout. Are you seeing that “0.0007” at about 5" radius with a known-good holder/pull stud?
On my VF‑2 I quit chasing ‘0.0007’ once I bolted the indicator to a T‑slot stud and swept the table with the Mitutoyo in a rigid tram bar — ditching the mag base made the last 0.0002 go away. If you’re worried about face runout vs a bent drawbar, pop in a CAT40 test bar and compare TIR at 1" and 6" radius; if it grows with radius it’s tram, if it stays but rotates it’s the nose. Want the stud/arm dims I used so it doesn’t wiggle when you bump the head?
I finish with the Mitutoyo on a short, rigid tram arm and sweep a hardened ring on parallels, keeping the lever about 12–15° to kill cosine error — that’s where the Blake can look off by about 0.0003 in. @rsmith84’s stud-mount idea plus a light preload from a screw jack under the arm keeps the setup from creeping; what’s your drawbar pull force reading?
Quick fix on your 2008 VF-2: stick the test indicator in an Indicol on the spindle and sweep a hardened ring 6–8 in out after a 10‑min warmup — it repeats better than the Blake for me. To keep the stand from squirming while you bump the head, bridge a small plate across two T-slots and clamp the arm to that. If the ring sweep and a light kiss on the spindle face don’t agree, you’re probably seeing “face runout”; cycle the clamp a few times and see if the pattern shifts — @michael_lew32 is right about coax lag.