Diff Welding (Tig)

[tptrsn]

Just wondering if anyone might have any input on this, I have been able to get some pretty nice, beefy tig welds onto the junction of the pinion shaft and spider gears. This should lock up the diff unless the pinion shaft twists in two. Haha... That said, the pinion shaft looks pretty beefy (E36 328i Auto, 2.93 gears diff), so I'm thinking I am probably good to go.

Anyone have any thoughts/experiences on point to share?

- - - Updated - - -

Here's what I have so far on both sides of the pin (and obviously both ends). It really wasn't easy to get in there for tigging, and it REALLY wasn't easy to get everything cleaned up:

_1200930-2.jpg

Unless someone has a solid reason why this is likely to fail without much abuse, I think I'm going to give it a shot and see what it will take. It will be fun to find out.

- - - Updated - - -

I said F-it and just plated in both sides. It's not pretty, and the camera couldn't get a perfect angle on it with good lighting to see what is really going on with the welds, but looking at it by eye with my cheaters on, I am 100% confident that the welds are totally solid, and now I know that my fabbing won't break. The 220k miles old 328i auto axles might, but c'est la vie. Lol

_1200931.jpg

[jalopi]

it probably won't break, but i've only ever seen diffs break when the job didn't include putting bolts or metal plate in the 8 corners where the gears meet

[tptrsn]

Stick bolts in the corners huh... GENIUS! Too late now for this one though I guess. But I can't see how anything is going to possibly move now that I put the plates in each side. One good thing about using a tig for this is that you really know how much penetration you're getting. As opposed to one guy one LS1tech who's migged diff went from welded back to open! Haha... I had no idea that could really happen without breaking them.

[jalopi]

the problem is less so mig vs tig, it's people trying to weld gears together with low amp 120V machines. unless you preheat the gears with a torch you're taking a gamble on good penetration

though after finally learning how to tig recently, i know where you're coming from. it's nice to be able to apply as much heat as you want without adding in unnecessary filler

[rpm620]

They will break eventually without plate.. ive never heard of the bolt thing before but im sure it works. I used a Miller Autoset 140 (120v) to weld my 3.91 Diff and it got great penetration as well adding some 3/16" plate on both sides. No issues.

[Novablue454]

This is how I did mine, 1/4" steel plate cut to fit inside the gears


Then I used a 220v Mig with .030 set on the max heat setting (IIRC) and just melted everything to the plate. Been holding up for about a season now to LS power.

[TheJuggernaut]

This may be of instructional value

https://www.facebook.com/mike.navarr...62126723816281

[protomor]

I'd love to do a clutch type on my E46. But something tells me I'm gonna have to weld it

[bennyfizzle]

too much thinking. stick weld the piss out of it and make sure it still spins. diffs are cheap.

[Piner]

What amperage is your Tig? I've never used mine to weld a diff I have always just stick welded them with nickle rod.

[tptrsn]

That instructional video was GREAT!! Lol

My little tig welder is a 2016 AHP AlphaTig 200X. It's phenomenal, and so far I've been able to get by on 110V for everything. This welder will do a little over 150 amps on 110V, which was plenty of power, but I did pop the breaker to my garage a few times while welding this diff. I guess I'd have to bite the bullet and get some 220V out there if I was going to do much heavy stuff.

- - - Updated - - -

By the way, I used 3/32" ER70S-2 filler and I probably would have been a little better off with 1/8" filler. Also, I did see a place or two where there were some tiny cracks in a couple of the first pass welds. I'm not sure if that is because of different metals between the gears, or if it's just because gear oil was getting pulled up into the puddle. A guy on yellowbullet said 309 filler is the way to go for a diff, maybe he's right.

[Piner]

I have a Powertig 185 DV, Its hooked up to 220v atm. Thats pretty wild I would have assumed 200amps would have been needed to get that sort of heat into it. That's damn good for 150 amps. Did you do any preheating or just went straight at it?

I have heard mixed reviews about 309 when it comes to structural stuff. It makes beautiful looking welds due to it being stainless but since its a dissimilar metal it has a tendency to crack in a high stress environment.

[tptrsn]

No preheating, I just went at it and it welded fine. On one hand I would say I had plenty of power with ~150 amps, but on the other hand, I did run flat out for several seconds at a time quite a few times. I guess maybe that indicates I didn't really have "plenty" of power, more like "just enough" power maybe. Lol

Thinking about what the guy on YB said, I checked all of the gears and the pinion shaft with a magnet before I started welding, and they were all very magnetic. I have had some stuff crack on me with 309 when I was using it to weld mild steel to mild steel just to see how well it would work. It seems like the 309 puddle cracks a lot as it cools unless one of the metals being welded is some grade of stainless.

Because of that, I decided to use ER70, and the largest ER70S-6 I had on hand was 1/16" so I just used ER70S-2 I had that was 3/32". From what I've seen the only significant advantage of ER70S-6 over ER70S-2 is a better back side of the weld (un-backpurged). and there really wasn't much backside to any of these welds anyway.

[Piner]

Well I'll have to give it try the next one I weld. I'll preheat it a bit with my MAP torch and throw it on 185 amps and should be alright.

[tptrsn]

With 185 amps available, you'll definitely be able to just hit it without any pre-heat. I think the pre-heat is really only something that might be necessary if you're welding the steel gears to the cast iron carrier. Well, I think it's cast iron...?

There is one other little issue though: While welding the first side of the gears, the heat of the welding caused some of the gear lube to seep out of the back/bottom side of the gears, which made them messy for welding the second side. Maybe it's because I had loaded up the case with argon by welding the first side, but when I tried to use a propane torch to burn away that gear lube mess for cleaner welding, I couldn't keep the torch lit... I don't know if it's because the argon was sitting down in there keeping out the O2, or if the flame was just burning up all of the available O2 in the case, or what.

I didn't think about it much at the time though, I just said F-it and started welding again. With enough filler added everything smoothed right out fast.

[Piner]

Yeah I actually forgot that I have some 1/8 312-16 stick rods laying around. I may just bang the flux off of them and use those as filler. 312 is supposedly really awesome on hardened steel and is very resistant to cracking.

[JesterMX6]

idk about all that. this is how mine is welded and it's held up to 2 drifting seasons and a year of daily driving with lots of tight radius parking


2015.07.27 welded 3.91 diff by George, on Flickr

[bimmerlust]

idk about all that. this is how mine is welded and it's held up to 2 drifting seasons and a year of daily driving with lots of tight radius parking

Fair enough but most people with a welder have some plate around and it's cheap insurance. That said I agree your method is fine if the welds are properly made.

[tptrsn]

Well, that sure doesn't look like it would let anything move! Was that stick or mig?

[Piner]

I would Guess stick by the looks of it. Nicely done if it is.

[JesterMX6]

Mig

[Piner]

So I ended up with a spare scrap diff that didnt have to go back into anything and figured I'd try TIGing it for practice. Didn't want to use up any of my steel to plate it since its not going to be used for anything but practice.

I used 1/8" 312 stainless rod at about 170amps. Not perfect but just sorta seeing what worked and what didn't. I probably should have switched to a smaller cup to get in the tight spots.

[tptrsn]

Looks good, like it would work!!

[jalopi]

that aint coming apart any time soon

[tptrsn]

it probably won't break, but i've only ever seen diffs break when the job didn't include putting bolts or metal plate in the 8 corners where the gears meet

Add me to the list of the broken!! My 1/8" plate was not substantial enough, and eventually fatigued and broke away on both sides of the diff. It left enough material remaining that the diff action only allowed about 1/5 rotation on one tire with the other tire held in place, but it was broken. Photos forthcoming.

This time I'm going to weld the 8 corners where the gears meet, and probably work on welding side gears to the case. I love this little sh*tbox, it puts up with so much nonsense from me. Lol