Ultima 90 Question

[Snakexor]

I would rather not lap a $50 hsf, so the question is: Is the Ultima 90's base similar to the warpedness of the TRUE120?

 

[BonzaiDuck]

You're lucky for crossing paths with me.

I. MY THERMALRIGHT EXPERIENCES AND THE NEED FOR A LAPPING JIG OF SOME SORT
II. THE MYTHICAL RISKS, AND THE REAL REASON TR VOIDS WARRANTY FOR LAPPING

I. THE LAPPING JIG
Starting last May, a small group of us here had heated discussions on the ThermalRight coolers, and I published a lot of material here with bar-graphs sampled from CoreTemp logs on the Ultra-120-Ex and the use of Diamond thermal paste.

Having a spare motherboard -- same model as the first -- I've got it destined for a smaller case, and needed a smaller cooler than the Ultra-120. So I bought an Ultima 90, lapped it, and installed it in my "test case" prior to the final build of the system.

The convex ridge or bowed curvature in the U-120-Ex's base is, indeed, built in to the Ultima 90's design. However, it is a shallower convexity. It's peak along the ridge is less than a millimeter -- I think it was close to 0.5 mm -- above the low edges of the heatsink base.

For lapping the Ultima 90, I had some worries and misgivings. I discussed this issue with the tech-support at SVC -- Silicon Valley Compu-cycle -- who sell "custom-lapped" U-120-Ex's. I speculated that they built a jig to do it, and that was confirmed.

The reason you'd want a jig of some sort is that you want to grind down the ridge evenly so that the resulting surface remains perfectly perpendicular to the vertical of the heatpipe-cooler's design. since there is a ridge, it would not be hard to imagine that lapping the base might result in a not-so-perpendicular surface, and you'd be struggling to keep the cooler perfectly upright until a flat surface began to widen on the cooler's base.

But it's very easy. At auto-parts stores, you can buy sets of clamps for $10 that include small clamps built of composite and light metal, with plastic pads at the clamping points.

I cut two pieces of white-pine-wood -- 6" x 1/2" x 1/2" -- ran the clamp through the cooler's heatpipes, and clamped the wood pieces on each side of the cooler's base. Frankly, a hardwood like oak or maple would be even better, but the pine -- to make a rhyme -- worked fine -- and lapped evenly along the ridge-line.

Once I could see the flat surface widening evenly along the line following the convex ridge, I removed the clamp and wood blocks, and proceeded to finish the job so as to grind off the nickel plating on the entire surface of the heatsink base surface. And frankly, I think it took less than a half-hour's work.

II. THERMALRIGHT'S JUSTIFICATION FOR VOIDING THE WARRANTY, AND THE REAL REASON BEHIND THEIR POLICY

Discussion with SVC and ThermalRight made us tentatively conclude that TR may have worried about the nickel-plating, corrosion after compromising the nickel-plate, and the possible worries about the integrity of heatpipe solder-points on the heatsink base.

ThermalRight only offered to say that "lapping the base might compromise or damage these solder joints." Without further information, the rest of us speculated about the Galvanic Table of metals, and the possibility of corrosion.

I have since revised my views on this.

The reason ThermalRight voids the warranty for lapping the heatsink base is this: They make a reasonable assumption that some hoople-heads and noobies will grab the heatsink by the pipe-and-fin assembly as they move it back and forth across the sandpaper. Obviously, this would put stress on the heatpipe solder joints -- and it is the most obvious risk for which ThermalRight has determined to void the warranty.

But I must emphasize here -- there is no need to worry about a voided heatpipe-cooler warranty, unless ThermalRight had created the defect in the assembly at time of manufacture, or the unit was damaged in shipping to the buyer. Obviously, if the damage were caused by ThermalRight, the company would not be able to guarantee or prove this to themselves if the unit had been lapped, and they want to minimize RMA returns to damage that is solely their responsibility.

FIRST, the expected MTBF or life-span of a heatpipe by itself, according to a white-paper published by the first manufacturer of heatpipes under NASA contract -- is 1 Million Years -- barring it's being run over by a car, hit by a meteorite, or used as a murder weapon.

SECOND, if you only grasp the cooler by its base-plates while you lap it, there is no risk of damaging it.

 

[DerwenArtos12]

Originally posted by: Snakexor
I would rather not lap a $50 hsf, so the question is: Is the Ultima 90's base similar to the warpedness of the TRUE120?

There have been cases i've seen of the ulitima 90 having the same issues.