View Full Version : Repairing a sight tube holding (essentially) ATF

09-20-2007, 04:46 AM
As stated on another post, the tachometer (manometer) in my new drill press has a broken tube. Right at the bottom, below the sight window, so repair is pretty flexible with respect to appearance. But the oil that goes in it is very expensive if I get the right stuff (about $25 a pint; not break the bank, but give me a break...), or some folks just use Dexron III which is VERY close, but will read a few % lower than the recommended stuff. Plus, if it leaks out, no big deal but the mess...

So, you can no longer order the part (I'm told). <shrug> Not that I would given typical pricing. I can order tube from McMaster, but it has to be tapped with a funky 3/8" fine thread for the feeder. That or make up something else, which is no biggy. Or, maybe I can fix the broken tube sufficiently for my needs? This is what I want to investigate (hopefully with your help).

It has a (roughly) 3/8 ID (measured) and between 7/16 - 1/2 ID (need to measure). It is a straight hard plastic tube, that appears to have been heated and bent almost 90*. It broke in the bend. I'm thinking there are several ways to go.

1) Machine a simple aluminum(?) block to provide a 90* elbow. Easy fit for the tube. Maybe an o-ring groove inside both holes. A countersink on the outer edge. And fill that countersink with common RTV gasket maker. Probably don't even need the o-ring. Tube would just need a little trimming to remove the are distorted by the heat/bend. But the threaded part and all the rest would still be perfectly useful with no additional fuss.

2) Use heavy heat shrink tubing to hold the pieces together and reseal them. Particularly the type with adhesive (or RTV?) inside. But I'm not sure how it will hold up to the fluid. But there is no pressure or stress on this piece, so as long as solvent didn't cause a problem, it should work fine, I think.

Other ideas? If it doesn't work, I'll just go the McMaster route and order some hard tube. I've also got flexible tube I use for fuel cell sight gages, and that could be made work with a supporting back plate. But it's got a yellowish tent, so not totally ideal...

09-20-2007, 08:35 AM
I think I'd be inclined to go the aluminum elbow route. For the seal, if you want to get a bit fancy, taper the hole for the tube and make a screw-in compression ring to compress an O-ring down into the taper and around the tube. Then you shouldn't need any extra sealer at all, and it ought to be solid.

09-20-2007, 02:17 PM
That's a nice idea about the compression ring. In fact, they make fittings like that, and if I could find one the right size, probably cost less than $10...

10-05-2007, 02:05 AM
Just an update.

I looked around for one of those nylon o-ring compression fittings in a 90*. Ace, HD, 2 plumbing stores. They've all got it in 3/8 (OD), found 5/8 OD, found 1/2 straight and "T" (among a few others, but no 90*). Essentially the typical (for me) "they've got everything but what I need". :(

Looked for other viable alternatives, but not wasting too much time or making any specific trips (that's why it took this long). No luck.

Finally got some more time in the shop, so I decided to just make one. Hacked out a piece of 1" x 1" x 3/4" scrap aluminum. Squared it up on the mill, more to clean it up than needing it square. The scrap was from a real scrap yard and looked like it had been through a rock crusher conveyor on the outside. Milled out a 1/4" notch in 1 corner, which by now, left me with 3/4 x 3/4 square cross section 90*. Chucked it in the 4 jaw and turned the stubs to round OD spigots. Just for aesthetics and clearance, the main reason I did it was to drill/bore the holes inside the spigots. Drilled/bored 2 holes to intersection with 0.050 clearance on the 1/2 tube. Then RTV'd in the tubes after cleaning up the broken ends. Should be ready to install tomorrow after curing.

While I was in there, I also made the sight tube externally adjustable and simplified the previously complex removal of the hood/belt guard. In fact, the removal of the belt guard being a pain in the patooey is probably what caused the previous damage. It's still going to be a mess because you can easily spill the trans fluid, but it's going to be MUCH easier than it was.

10-06-2007, 11:03 PM
For you pic junkies, here it is in all it's simple glory. Oh, and that nipple IS crooked, but that's just how the threads were in the sight tube, nothing I could do (well, nothing I felt like doing).

If I had it to do over again, I would have either threaded the aluminum piece and screwed in a brass barb, or started with a brass elbow/90* and bored it for the vertical tube. I'm not real keen on the way it looks, so who knows, maybe I will redo it. The biggest problem is the vertical tube is not only bowed a bit, but has some twist as well. So you can see the edge of the white sprayed back sorta spiraling around into the view port near the top, as it sorta meanders off the left side of the view port. <shrug> I'm probably too lazy to do anything about it as long as it works and doesn't leak (none so far), but who knows.

10-07-2007, 09:03 AM
Real nice machining job, BadDog. So, how does this gage work? Why does ATF cause an error-viscosity or density difference?

Like this great weather?

10-07-2007, 02:05 PM
LOVin' this weather! :D

The reeves drive is adjusted by a crank that also pushes a piston into a cylinder to displace a volume of fluid. The recommended fluid is something like $20+ for red dyed oil that is almost identical to trans fluid, but just slightly off the required specific gravity. One or two leaks, or spilling the tube while servicing the head (it can easily be tipped over when the cowel is off) would get fairly expensive. So yes, there is a slight error. My plan is to measure the rpms, at least at the extremes, and place labels on the opposite side from the original marks. That'll give me what I need to extrapolate easily close enough for my needs. According to discussions on another post, the error will be small and shouldn't be a big issue at all.