Type 4 oil pump rebuild, Comparison of OEM vs. Samba CNC pump Options

[Superhawk996]

Playing with the idea of what it will take to rebuild a type 4 pump.

The pump that came out of my 2.0L engine is trashed from pumping rust. Both the gears and the housings are seriously scored.



I figured I'd roll the dice and see what might be on Ebay. Took a gamble and bought a used pump hoping I might get some core parts that are useful.



I did end up with a good housing outer. However, the gears are pitted and it the inner housing plate either has a serious casting flaw, or some sort of galvanic corrosion that pitted it. I always struggle with this kind of thing. It really would have minimal effect on the pump and it probably worked OK for a lot of years. However, I'd just like to know that it is perfect going before going into a engine.





It looks like new inner plates are available but at a cost of $150+.

I decided that since I have a donor "scrap" pump, I have nothing to lose by experimenting with machining it.

[Superhawk996]

(IMG:style_emoticons/default/agree.gif)

Been there before.

https://www.thesamba.com/vw/forum/viewtopic.php?t=699156

I will definitely use the site as a reference. It has great information and some useful techniques and I really like his gear locking fixture.

"Blueprinting" is overkill in the sense that I'm not looking for perfectly lapped surfaces.

Don't get me wrong, blueprinting has it's place on race engines and other really high dollar machinery that runs at high RPM.

Even when these Type 4 pumps were new, there are machine tool marks evident in both housings as well as on the gear ends. There is sufficient overcapacity built into all mass produced oil pumps to account for these manufacturing imperfections.

Likewise all manufactured components have a tolerance from nominal. Without the actual part blueprint (or CAD in modern manufacture) there is no blueprinting going on in the sense that the part is being brought to nominal dimension, balance, surface finish, etc.

So for my purposes, I'm mainly interested in obtaining fully functional, durable, and usable parts without being overly concerned about "blueprinting" and having highly lapped surfaces.

Again, to be clear - I'm not talking down that Samba post, and, thank your for originally posting it. It is a great resource.

What I am very interested in is building up a means to test and compare pumps though to ensure that on any pump that gets mix & matched parts ends up fully functional and that functionality can be proven.

I have some ideas on this test fixture -- I'll post as it evolves.