Yes, I've done everything in thread, twice at least, for three years now. Delivery valves did have some mild corrosion, I replaced with new from the dealer along with delivery valve springs & crush washers. I even sent my injection pump for a rebuild to a place Bosch North America recommended to me, D&W Diesel which was a nightmare. They were horrible to deal with and I don't believe they did a rebuild. Still had the original delivery valves... (DON'T USE D&W!)
Compression is very good 360 to 400+ Chasing this I even got a rebuilt head from Metric Motors, after pump rebuild and before new clear hard lines that showed the bubbles.

I don't have an air leak I can find, went over every inch. replaced all lines except undercar steel. I have exactly what your video shows, massive fine bubbles from pump when revved and an intermittent in intensity rough hot idle. No air seen before or entering lift pump

Guys, could it be the MW injection pump itself? Some kind of cavitation?

Exactly what I’ve been wondering.

__________________
1981 240D 4sp manual. Ivory White.

There seem to be five (?) topics on this forum addressing the same issue. So far, no one has found a fix to the exact problem; the banjo bolt fuel return is leaking air and there is no sign of air entering the injector pump.

I think we can solve this!

1. Bosch injector pump service manuals have a section on plugging the fuel bypass and all ports. To check for air bubbles, it must be pressurized, then placed in an oil bath. This should show you if there is an opening outside. It is possible, however, that an o-ring could be faulty somewhere on the delivery valve body holder.

2. When I prime the banjo bolt, I can see air coming out. Air comes out, but not as much as when revving it. Despite my best efforts, I do not get the banjo bolt "squeal." I think I should get it. Over time it does encounter some resistance, not all of the air is expelled, but some of it. I can keep that going with air coming out the bypass. Mind you, no air is coming from the supply.

4. There is a possibility that the air is coming from negative pressure. What processes would cause that, and where are the possible points of air entry?

5. Cavitation? It might be possible. Is cavitation a dynamic phenomenon? It is my understanding that bubbles grow and then collapse. I see air bubbles without flow.

__________________
Theseus

Instragram: doss_project
1984 300SD - 930 Blue, OM 617
Murfreesboro, TN

Theseus owned a ship in Greek mythology, and when a part of the ship needed replacing, it was replaced. Eventually, everything was replaced. Is it the same ship? (I own the car of Theseus. )

Just thinking out loud. Since the fuel supply to the injection pump runs ~12 psig, there shouldn't be any place in the fuel passages within the I.P. that is below atmospheric pressure to suck in air. Indeed, if there were bad seals, you should see fuel dribble out, though it could leak into the IP crankcase (flows to engine oil pan). But, if there was a blockage plus a bad seal, it might suck air on say a piston refill cycle. But the bad seal would also have to act like a check valve, allowing air to get sucked in but not fuel to dribble out. I would think that if the IP was susceptible to letting air into the system, this would be a known problem after 6 decades of Bosch IP use.

People here have long fretted over air in the injection system, but I haven't heard of anyone proving that causes a rough idle. If too much air, the pressure in the injection line will not reach 1950 psig to pop open the injector. Try loosening the nut at the top of each injector, one at a time, and verify a little fuel dribbles out with no puff of gas, and each one should slow and roughen the idle by about the same amount. If that proves OK, I would move on from worrying about the gas in the injection return hose and try other things like verifying the injectors with a pop tester, especially for good spray pattern, and verify injection timing, as both of those can greatly affect idle. If you ever drive out to CA, try the "renewable diesel" you can buy at Propel and Union76 stations. My engines idle much smoother on them, due to the high cetane rating (~75), so don't neglect your fuel quality.

__________________
1984 & 1985 CA 300D's
1964 & 65 Mopar's - Valiant, Dart, Newport
1996 & 2002 Chrysler minivans

Speaking as a mechanical engineer, cavitation is greatly misunderstood. It is liquid vaporizing temporarily due to a low pressure region, such as behind a spinning pump impellor. It quickly condenses back to liquid as the vapor bubbles reach a higher-pressure region (releasing energy in a small point to make noise and even erode metal). Cavitation does not form air bubbles.

There is a related issue where dissolved air can be drawn out of solution by applying a vacuum (or lower pressure region) to the liquid. That can draw air bubbles out of solution. But, there would be no excess dissolved air in the fuel drawn from the fuel tank. If the fuel tank was pressurized to say 1000 psig, it could have appreciable dissolved air, but unless testing liquid rocket engines or working in a chemical plant (me), you wouldn't run into that.

__________________
1984 & 1985 CA 300D's
1964 & 65 Mopar's - Valiant, Dart, Newport
1996 & 2002 Chrysler minivans