OM606: Fixing the overflow return valve design
OM606: Fixing the overflow return valve design
The stock OM606 fuel system has bubbles flowing through it whenever the RPM's are elevated. Go ahead and check: shine a flashlight through the pipe going into the fine fuel filter and rev the engine to 2500RPM.
The problem is that MB miss-designed the overflow relief valve that returns fuel to the tank. This valve is supposed to do two things: 1) prevent the reverse of fuel; and 2) regulate the pressure in the IP.
There are several possible ways of designing this valve. The most common designs use a flow restriction or a calibrated opening pressure. Unfortunately what got installed in the OM606 (I have two of them) was a simple one-way valve that opens at a very low pressure (you can blow through it). Absent a stiff spring or narrow orifice, the lift pump has no problem moving a lot of fuel through the pump. Unfortunately this doesn't equate to a properly pressurized pump. Instead the lift pump moves so much fuel that it pulls a near vacuum on the intake side. This in-turn pulls air into the lines. That air gets turned into micron sized bubbles by the fine fuel filter and then gets fed to the injection pump. The end result is that the air is pulled into the plungers where it reduces injection quantity and effectively delays start of injection (the air has to be compressed first before the pressure rises to the point that the injector pops).
The original overflow valve has a 2.5mm entry opening with a shallow 1.5mm exit hole into the banjo fitting. Hardly any restriction at all. I opted to drill and tap the entry for M5x0.8 and then to make a 0.5mm orifice that threads into that.
The first picture below shows the original diss-assembled part. I wanted to understand the original function of the valve so I used a punch to push out the BB seal. It turns out that the internal valve is a piece of plastic pushed up against the backside of the entry hole with a very fine spring. The plastic plunger was badly warped by diss-assembly.
The second picture show the parts I put back in. The smaller BB and aluminum shaft are crude replacement parts I hacked together, the spring is original. The only real improvement to the valve is the brass 0.5mm orifice.
The last picture shows a properly made part (stock valve, drilled & tapped, SS bolt, drilled for orifice and parted on a lathe) that has a much longer 0.7mm orifice that is properly centered. The larger orifice is less likely to be clogged while the longer path should provide roughly the same restriction as the smaller-but-shorter prototype.
The end result is no bubbles in the lines at any RPM and better top end power. Power increases are easily measured by the up hill speed of my ~6000lb G-wagen on well known hills.
This how-to might help those of you working with the stock OM606 electronically controlled pump. It may apply to the NA system and the OM603, but I don't have one to check out.
The stock OM606 fuel system has bubbles flowing through it whenever the RPM's are elevated. Go ahead and check: shine a flashlight through the pipe going into the fine fuel filter and rev the engine to 2500RPM.
The problem is that MB miss-designed the overflow relief valve that returns fuel to the tank. This valve is supposed to do two things: 1) prevent the reverse of fuel; and 2) regulate the pressure in the IP.
There are several possible ways of designing this valve. The most common designs use a flow restriction or a calibrated opening pressure. Unfortunately what got installed in the OM606 (I have two of them) was a simple one-way valve that opens at a very low pressure (you can blow through it). Absent a stiff spring or narrow orifice, the lift pump has no problem moving a lot of fuel through the pump. Unfortunately this doesn't equate to a properly pressurized pump. Instead the lift pump moves so much fuel that it pulls a near vacuum on the intake side. This in-turn pulls air into the lines. That air gets turned into micron sized bubbles by the fine fuel filter and then gets fed to the injection pump. The end result is that the air is pulled into the plungers where it reduces injection quantity and effectively delays start of injection (the air has to be compressed first before the pressure rises to the point that the injector pops).
The original overflow valve has a 2.5mm entry opening with a shallow 1.5mm exit hole into the banjo fitting. Hardly any restriction at all. I opted to drill and tap the entry for M5x0.8 and then to make a 0.5mm orifice that threads into that.
The first picture below shows the original diss-assembled part. I wanted to understand the original function of the valve so I used a punch to push out the BB seal. It turns out that the internal valve is a piece of plastic pushed up against the backside of the entry hole with a very fine spring. The plastic plunger was badly warped by diss-assembly.
The second picture show the parts I put back in. The smaller BB and aluminum shaft are crude replacement parts I hacked together, the spring is original. The only real improvement to the valve is the brass 0.5mm orifice.
The last picture shows a properly made part (stock valve, drilled & tapped, SS bolt, drilled for orifice and parted on a lathe) that has a much longer 0.7mm orifice that is properly centered. The larger orifice is less likely to be clogged while the longer path should provide roughly the same restriction as the smaller-but-shorter prototype.
The end result is no bubbles in the lines at any RPM and better top end power. Power increases are easily measured by the up hill speed of my ~6000lb G-wagen on well known hills.
cars? what cars? WHERE???
randomdude
07-06-2011, 09:21 PM #2nice fix! too bad i cant take advantage of it lol.
cars? what cars? WHERE???
1982 300D Turbo ... 3,6xxlbs, No fan, No AC, Hood Stack, No ALDA, No rear bumper and stuffed front, A/W Intercooled, Injectors by Greezer and HUGE Pre-Chambers with help from OM616 & Simpler=Better, Fuel Cranked up, 60 Trim Compressor wheel, EGT, EMP, Boost 50" Rigid Radius bar on roof Aux tank for a total of 48 Gal Of Diesel! Odyssey PC-1750 Battery in trunk, 27"x8.5"/R14 Maxxis BigHorn Mud Terrains, In June '14 issue of Off Road Mag
AX15 Jeep Trans swap in progress....
Captain America
07-06-2011, 11:24 PM #3That same thing might happen on the 617 engines as well
1982 300D Turbo ... 3,6xxlbs, No fan, No AC, Hood Stack, No ALDA, No rear bumper and stuffed front, A/W Intercooled, Injectors by Greezer and HUGE Pre-Chambers with help from OM616 & Simpler=Better, Fuel Cranked up, 60 Trim Compressor wheel, EGT, EMP, Boost 50" Rigid Radius bar on roof Aux tank for a total of 48 Gal Of Diesel! Odyssey PC-1750 Battery in trunk, 27"x8.5"/R14 Maxxis BigHorn Mud Terrains, In June '14 issue of Off Road Mag
AX15 Jeep Trans swap in progress....
I think this is not a good idea:
you want a certain flow through the IP for cooling and getting out air.
The valve should open at about 0.5-0.8 bar and should be able to flow the quantity the lift pump delivers.
In case the lift pump can not complete its stroke for pushing fuel (this delivery stroke is done by a spring, the suction stroke is done by force of the cam) you risk the lift pump plunger shaft floating on the cam.
Tom
Hi,
I think this is not a good idea:
you want a certain flow through the IP for cooling and getting out air.
The valve should open at about 0.5-0.8 bar and should be able to flow the quantity the lift pump delivers.
In case the lift pump can not complete its stroke for pushing fuel (this delivery stroke is done by a spring, the suction stroke is done by force of the cam) you risk the lift pump plunger shaft floating on the cam.
Tom
(07-06-2011, 11:29 PM)tomnik Hi,
I think this is not a good idea:
you want a certain flow through the IP for cooling and getting out air.
The valve should open at about 0.5-0.8 bar and should be able to flow the quantity the lift pump delivers.
In case the lift pump can not complete its stroke for pushing fuel (this delivery stroke is done by a spring, the suction stroke is done by force of the cam) you risk the lift pump plunger shaft floating on the cam.
Tom
Tom,
I understand the concern when changing MB's handiwork. However I think this is an OK change.
There is still plenty of flow through the system. I'll have to measure it directly, but an educated guess based on pressure, viscosity and orifice size would be around 1L/M. Easily enough to carry away heat. One good thing is that it the cooling is still strong at low RPM where air flow through the engine compartment may not be very high.
The speed of air purging should be unchanged since air has very low viscosity and will move through the orifice easily. In any case the system seemed to self-purge at the usual pace. Putting in a true pressure valve would definitely make purging air difficult. I spent quite a bit of time trying to find a pressure valve that fit this location. All the valves that truly regulated pressure had a 14mm thread rather than the 12mm thread that the OM606 pump uses.
It is possible that the lift pump will float, but that is probably happening even with the standard setup. Based on it's diameter and stroke, the lift pump would have to move >4L/M of fuel at (5000RPM) to prevent float. In any case, the lift pump is designed to float. It is designed just like fuel pumps from the old carburetor days. Once these pumps raised the fuel pressure to the level of the spring tension then they would float until the pressure dropped again.
(07-06-2011, 11:29 PM)tomnik Hi,
I think this is not a good idea:
you want a certain flow through the IP for cooling and getting out air.
The valve should open at about 0.5-0.8 bar and should be able to flow the quantity the lift pump delivers.
In case the lift pump can not complete its stroke for pushing fuel (this delivery stroke is done by a spring, the suction stroke is done by force of the cam) you risk the lift pump plunger shaft floating on the cam.
Tom
Tom,
I understand the concern when changing MB's handiwork. However I think this is an OK change.
There is still plenty of flow through the system. I'll have to measure it directly, but an educated guess based on pressure, viscosity and orifice size would be around 1L/M. Easily enough to carry away heat. One good thing is that it the cooling is still strong at low RPM where air flow through the engine compartment may not be very high.
The speed of air purging should be unchanged since air has very low viscosity and will move through the orifice easily. In any case the system seemed to self-purge at the usual pace. Putting in a true pressure valve would definitely make purging air difficult. I spent quite a bit of time trying to find a pressure valve that fit this location. All the valves that truly regulated pressure had a 14mm thread rather than the 12mm thread that the OM606 pump uses.
It is possible that the lift pump will float, but that is probably happening even with the standard setup. Based on it's diameter and stroke, the lift pump would have to move >4L/M of fuel at (5000RPM) to prevent float. In any case, the lift pump is designed to float. It is designed just like fuel pumps from the old carburetor days. Once these pumps raised the fuel pressure to the level of the spring tension then they would float until the pressure dropped again.
1 Guest(s)
1 Guest(s)