Sum Total of All Knowledge

Tackling the Vacuum system on a 300D

I think I have enough material to put together a video on this, but I think it's worth logging the work so far on the Mercedes 300D vacuum system.

The first sign that things weren't so great with the entire system showed up immediately: the door plungers weren't going up and down. So I knew there was something not right. I just didn't know how deep the problem was. The truth was that the pump was not performing anywhere close to normal, so all the downstream components were either substandard or disconnected as a symptom of that.

I didn't' really have any testing equipment so I went out and bought a pressure/vacuum gauge from Harbor Freight (this is a nice big gauge and it's worked well so far. I've measured vacuum and boost pressure and I feel it's a good quality tool) and a vacuum hand pump with a gauge, also from Harbor Freight (I've been less happy with this- it simply doesn't hold vacuum very well after about a year, and it has a sliding ring on the front that switches it from pressure to vacuum. It's easy to knock or shove this in the wrong position, it doesn't stay in either position very firmly, and I suspect all the leaks are coming from that part of the pump. I'm looking for a better pump that's less than $100. McMaster sells one, I am tempted...)

The first thing I tried was measure vacuum from the main line connecting the pump to the brake booster. 1984 has two ports coming from the pump and both of them looked suspect- swaddled in black cloth tape and sticky epoxy. On startup, the gauge swung wildly from plus 10 to minus 15, eventually 'settling down' to vaguely negative territory as the check valves downstream built up a little resistance. But it still flipped on engine engine rotation from -5 to -15. That was an eye opener. I was kind of amazing that anything was working- especially the brakes.

I researched the issue enough to figure out that probably both the pump and the main check valve directly after the pump were shot. I ordered a good used check valve from Kent Bergsma, and removed the pump. Part of the challenge is removing it without stripping the socket head screws holding the pump to the engine. I cleaned them thoroughly beforehand with solvents and q-tips to prevent this problem.

I also bought another pump on Ebay and went through it. That was 150 bucks. The check valves and the roller bearings are what fail on these. The old pumps have two check valves, and for my car, the newer pump has three check valves. Rebuild kits are no longer made for them. But the secret trick is that if you can save the delicate o-ring and re-use it, the check valves for the new and old pump are identical, it just uses two valves instead of three. So I ordered two of the old kits to get 4 valves. The piston and sleeve on the Ebay pump weren't great. And the roller cam and bearing were sloppy.

When I finally got the nerve to dig in and remove the pump, I discovered the pump on the car was actually in much better condition. The piston seal was tight. The roller bearing was very good, and the o-ring was good too. So I decided to look at the eBay pump as a learning experience and set about rebuilding the original pump from the car. I think one of the check valves might have been installed backwards. It took a bit of clear thinking to 'see' the airflow through the entire pump to make sure the new ones were installed facing the right way. There is some finicky pressing you have to do with the piston spring. Kent Bergsma makes what appears to be a pretty solid rebuild kit for these pumps. which is why Kent's kit comes with a little wooden panel and some hardware. After doing the job on my own, in retrospect I think for about 60 bucks it's well worth the time he put into finding all the little details and tools that make the job actually work smoothly. I reassembled the pump.

I installed the used big check valve and put nev-r-seize on the the threads. Not sure if this was a good idea or not but whatever.

I also ordered a new main line from the pump to the brake booster. I ordered this from carsoemparts in Lithuania. I thought it was worth a shot. It's been nearly a year and it never arrived. I'm just too lazy to follow up on it and there's something fishy about how the item number is no longer valid. This vacuum line is often listed as available in internet auto parts suppliers sites, but when you try to check it out and ship it, it seems to vanish or it's not in stock. I think that this line could easily be custom fabricated. These lines are (or were) apparently made by a company called Festo. I haven't reserached this yet but it could be a solution to another part that seems to be gone forever.

I carefully cleaned the mating surfaces of the pump and the engine. For the engine, it's extremely tight quarters down there. To prevent contamination, I put a rag around the internals and zip-tied around the perimeter to keep it in place. I put a single edge razor blade on a pair of vise-grips and scraped it clean. That, combined with some work with a dremel and scotch-brite on a die grinder got the surface to a properly clean state. Very frustrating work.

With the pump reinstalled, I tested again and the vacuum reading was greatly improved.

I moved on to the next part of the system, downstream from the pump. With the main hose being not so hot, I have since improved it by reaming out the ports, then epoxying a new tap onto it. This lasts a while, then pulls out again. The epoxy just doesn't hold onto that plastic. Lately I tried a different tactic- I tapped the hole and made a new, threaded connector and screwed it into the hole. There isn't much left to work with, but having a more mechanical grip on the parts seems to be working, and the vacuum is better than ever.

I replaced a check valve by the firewall, and started going through the connections. Lo and behold there was a hidden surprise- someone plugged the entire system leading into the engine compartment with a screw. With such poor vacuum coming from the pump, I guess I don't blame them for still wanting brakes instead of fancy door locks.

Now that I had stronger vacuum, I removed the plug and the locks were working again. Very satisfying. Though they'd only hold vacuum for about 5-10 minutes after I turned off the car and locked it. Still working on that one...

The current state of things is that I've been focusing on the transmission shifting, which I'l go into more detail with in another post. Also the 3/2 valve is completely missing from my car- something I didn't really put together until recently. I just thought maybe this year didn't have that because there were so many variations. But the entire system that controls the EGR is complex and most of it missing. Over the winter I personally removed the EGR valve and blocked the ports with plates I made on the milling machine. I also plugged the lines to the switches. There may be a point where I put the entire thing back together as designed. One of the things that Kent Bergsma discusses in his Transmission Tuning manual is that while many folks think you should just tear out the 3/2 valves and all the EGR stuff, the other side of the argument is that there are subtleties of how vacuum signals are sent to the injection pump that are part of the EGR system, and that should probably stay in place if you want to properly tune up the car: it all affects the shift points as designed when it's all running as it left the factory

While that might be true, I do wonder how a Euro-spec 300D looks under the hood and how much simpler and perhaps more efficient and powerful it is compared to the California and Federally mandated rules that they had to adhere to.

For the gassers in particular, the V8's- the hoops they had to jump through with the detuning of the 560SEL for instance, it seems ridiculous.

As to the adventures with adjusting the transmission shifting, that is a rabbit hole that I have fallen completely down into, and I'll discuss more in the upcoming posts.

Mercedes W123 300D air filter proper part numbers

Correct air filter part number for a 1984 Mercedes is MANN C 31 190. This fits the range for years from 1977 to 1983.

There is a smaller version of the assembly which partially overlaps year range (1978-1985)- this is confusing. The part number is MANN C 30 122.

Mercedes Hood hinges

As wonderful as the W123 chassis is supposed to be, it has its flaws. Maybe I've lived a sheltered life but I've never heard of a car's hood hinges disintegrating before. To be precise, it's not the hinges but the attach points in the wells that carry the hinges and springs. These are well-known trouble spots that collect water and leaves and other junk that falls into them, blocking the drains and rusting the holes that the hinge pins fit into. One of the things that IS truly great about the W123 is how the hood can lift up like a normal hood, then you flip these little levers on the hinge corners and the hood can continue lifting all the way to a vertical position, making access to the engine much easier. In my case, The hinges themselves are distorted, and the levers the release the hood to the vertical position are twisted and broken off. Also, my hinge attach points are suffering. Someone has been in there already, stick welding some metal to reinvigorate the receiver holes. The hinges are so bend that they keep popping out of the holes and the springs yank them back into the wells.

I found a much better set at the LKQ Pick & Pull near here in Wilmington, CA. today I picked up some custom mixed touch-up paint today at Stevenson's in Carson. Tomorrow I hope to remove the hood and install the better versions. The paint code for the car is 623, or Hellelfenbein (light Ivory).

A setback, or two...

My case is cracked.

I took the case halves, the crank and the cam down to Old Speed in Paramount here in LA. Russ was very up front- you might even say blunt- that the case was no good. He passed a torch over the area where they usually crack and sure enough, it was cracked. When you put a flame to it, if there's any oil or moisture in there, it will bubble out or at least seep from the crack visibly.

It's only very faintly visible in this still frame- and honestly it was difficult to see in person.

An option is to just use the case anyway. But it had already been line bored several times- to the edge of what is allowable. I left his shop feeling pretty down. He suggested that we use a good case that he has in his stock of used engines. That's a possibility, but he wanted to build the engine all the way up, bench test it and then install it. I know that's a really good option but it comes at a pretty high price. I believe over 5,000 or more.

At any rate it clipped my wings as far as momentum on this project.

Something else that revealed itself was just how chewed up the front quarter panel is on the body. I was happy to finally get it up on the rolling stand. I had grand plans to make a rotisserie but I didn't like the idea of my self-engineered contraption and substandard welding failing and crushing some poor guy in a body shop down in Wilmington. I just removed the complexity and welded the supports to the verticals and it seems overbuilt enough, and high enough that I can get all around it an under it.

But the quarter panel... I've seen some really beautiful threads on TheSamba with folks carefully cutting out the old panel, grinding all the spot welds clean and installing another. I think in a different world I could spend 8 days grinding and prepping and fitting, but even then, I don't have a TIG welder.

I've made some attempts to cut apart and re-weld pieces with the MIG but even with the .023 wire, it's not pretty. It dumps so much heat into the parts, and it blows right through the thin sheet metal. The trick is to just pulse it together with tacks, then add more in between, then more until the seam is closed. But it's so clumsy and ugly compared to what a TIG welder can do.

So the pieces sort of knock around the garage, and the engine parts knock around the garage becuase I can't bear to just throw them away.

But the time will come... I'll get a TIG, and I'll buy a new engine kit, and the bug will continue....