milehighassassin
Top Post Dawg
Excess fuel is pumped to help with cooling for the high pressures.weedeater said:
A4 PD BEW fuel cooling
- Thread starter DanG144
- Start date
joshgb
Veteran Member
But only at the injectors themselves. Fuel running past them is not recirculated into the system.weedeater said:
Pressures of what? The "excess" fuel does not flow past the high pressure unit injectors.milehighassassin said:
DanG144
Top Post Dawg
Josh, my understanding for the PD is different than yours. The entire fuel rail system in the head does recirc. It recircs 1.2 lpm at 903 rpm. There is a back-pressure regulator that holds about 15 psig on the rail.
And since the tandem pump is a positive displacement pump, I was expecting the recirc flow rate to linearly increase with rpm until the in tank lift pump could not keep up with the tandem pump. Then flow would pretty much level off.
I think the fuel not used in the injectors is only a very small component of the recirc flow.
Dan
And since the tandem pump is a positive displacement pump, I was expecting the recirc flow rate to linearly increase with rpm until the in tank lift pump could not keep up with the tandem pump. Then flow would pretty much level off.
I think the fuel not used in the injectors is only a very small component of the recirc flow.
Dan
whitedog
Veteran Member
Dan, do you know where this back pressure regulator is locatated?
DanG144
Top Post Dawg
It is built into the return rail exit point, which is built into the fuel pump portion of the tandem pump.
It is a simple ball and spring back pressure regulator, I believe. I will look for pics or diagrams.
Dan
It is a simple ball and spring back pressure regulator, I believe. I will look for pics or diagrams.
Dan
DanG144
Top Post Dawg
http://pics.tdiclub.com/data/516/tdi-technik-eng.pdf look on page 11.
I have not had my tandem pump off, but this looks identical.
Dan
I have not had my tandem pump off, but this looks identical.
Dan
whitedog
Veteran Member
OOOps wrong book Dan. 
I have the right book here somewhere...
I have the right book here somewhere...
whitedog
Veteran Member
AHA! That's the outlet from the tandem pump. I see it now.
DanG144
Top Post Dawg
Correct, all you see is the nipple from the outside. You can also see when looking at the pump externals, that this section of the tandem pump is really just an outlet port that bolts up to the head return fuel rail.
Dan
Dan
joshgb
Veteran Member
Your explanation makes more sense than my garbled supposition. I have had the tandem apart and indeed there is a ball and spring regulator; my brain must not have been working at the time.
Still, something does change in the pressure reading as rpms stay constant and road conditions change. Miniscule though it is in comparison to the fuel flowing in the recirc loop, it pretty much has to be injector demand, no?
Still, something does change in the pressure reading as rpms stay constant and road conditions change. Miniscule though it is in comparison to the fuel flowing in the recirc loop, it pretty much has to be injector demand, no?
joshgb
Veteran Member
Though this picture does not show the regulator, it does show the outlet and return ports for the tandem pump to cylinder head connection. My stupid; I owe everyone some vacuum readings.
whitedog
Veteran Member
Dan, do you have a picture handy of the distribution pipe? If not, I can scan one.
milehighassassin
Top Post Dawg
Injectors are regulated by the camshaft. Excess fuel does cool the injectors for the pump is always on.joshgb said:
DanG144
Top Post Dawg
Nope, I do not have one handy.whitedog said:
Dan
DanG144
Top Post Dawg
This is a post where I was asking for data. It is long, so read on down to the list of requested data.
What I would like to have is flow rate down the return line at 4500 rpm, as well as pressure at the outlet of the fuel filter, on the supply to the engine. I plan to direct my return flow to a calibrated fuel container, and I plan to do this test with a tank at least 3/4 full, to make sure my electric pump does not starve for fuel.
http://forums.tdiclub.com/showthread.php?p=2333343#post2333343
Josh, don't sweat it, you will undoubtedly update me on something later. I still have a lot to learn. I would not have been too surprised to find out you were right and I was wrong, in this case.
Dan
What I would like to have is flow rate down the return line at 4500 rpm, as well as pressure at the outlet of the fuel filter, on the supply to the engine. I plan to direct my return flow to a calibrated fuel container, and I plan to do this test with a tank at least 3/4 full, to make sure my electric pump does not starve for fuel.
http://forums.tdiclub.com/showthread.php?p=2333343#post2333343
Josh, don't sweat it, you will undoubtedly update me on something later. I still have a lot to learn. I would not have been too surprised to find out you were right and I was wrong, in this case.
Dan
whitedog
Veteran Member
And here is the tandem pump diagram.
Last edited by a moderator:
joshgb
Veteran Member
With a nearly new filter, and a fuel temperature of 70C, I get the following pressure readings in neutral, not moving:
6psi at idle
5psi at 3000 rpms
4psi at 4500 rpms
Why is the pressure still positive? For starters the cam is at 1/2 the engine speed, no? So the tandem pump is spinning at 2250 when the reading is 4psi. That doesn't work out with your numbers as far as flow of the tandem pump overcoming the lift pump. All I can say though is that, empirically, the pressure drops with engine load at the same rpm. In other words, if I were going 4500 rpms down the highway I know for sure that my gauge would be closer to 0.
And looking at Whitedog's pump diagram, does it not make sense that back pressure in the injector supply and return line would prevent the tandem pump from overcoming the lift pump (assuming no restriction or other vacuum creating conditions in the fuel filter or supply line) and creating a vacuum condition?
6psi at idle
5psi at 3000 rpms
4psi at 4500 rpms
Why is the pressure still positive? For starters the cam is at 1/2 the engine speed, no? So the tandem pump is spinning at 2250 when the reading is 4psi. That doesn't work out with your numbers as far as flow of the tandem pump overcoming the lift pump. All I can say though is that, empirically, the pressure drops with engine load at the same rpm. In other words, if I were going 4500 rpms down the highway I know for sure that my gauge would be closer to 0.
And looking at Whitedog's pump diagram, does it not make sense that back pressure in the injector supply and return line would prevent the tandem pump from overcoming the lift pump (assuming no restriction or other vacuum creating conditions in the fuel filter or supply line) and creating a vacuum condition?
Last edited:
DanG144
Top Post Dawg
Josh,
Excellent points and questions.
I don't know. It may turn out that you are correct. This is just the type of information we need to figure this out.
Dan
Excellent points and questions.
I don't know. It may turn out that you are correct. This is just the type of information we need to figure this out.
Dan
joshgb
Veteran Member
I have been periodically baffled about PD fuel system theory since the first time I delved into it and was told to expect mostly vacuum pressures because the lift pump was really only a priming pump. I know my system just does not work that way. In fact, if I pull the juice on the lift pump at idle, the engine dies.
I guess my only practical suggestion, as it applies to your situation, is that if I read the diagram correctly and the entire output of the tandem pump is fed into a variable pressure (14.5-109 psi) circuit, it might be tough sledding to figure out the theoretical return flow of the pump. Even if the injector circuit is dedicated and a much bigger stream just runs through the pump and back to the tank, it would still do some non-linear things to the overall flow.
Perhaps a more practical solution is to compare the line pressures of the "overheating" PD system to one of "normal" pressure at the same temperatures and same place in the line?
That is, more or less, why I installed a dedicated pressure gauge in my fuel line—it enables me to see poor filter performance before any real world problems exist. One thing I have learned, by the way, is that given how the fuel heats up and changes density/pressure it is much more practical to consider rapid pressure drop under engine acceleration as the sign of a restricted system than to go by absolute pressure benchmarks.
I guess my only practical suggestion, as it applies to your situation, is that if I read the diagram correctly and the entire output of the tandem pump is fed into a variable pressure (14.5-109 psi) circuit, it might be tough sledding to figure out the theoretical return flow of the pump. Even if the injector circuit is dedicated and a much bigger stream just runs through the pump and back to the tank, it would still do some non-linear things to the overall flow.
Perhaps a more practical solution is to compare the line pressures of the "overheating" PD system to one of "normal" pressure at the same temperatures and same place in the line?
That is, more or less, why I installed a dedicated pressure gauge in my fuel line—it enables me to see poor filter performance before any real world problems exist. One thing I have learned, by the way, is that given how the fuel heats up and changes density/pressure it is much more practical to consider rapid pressure drop under engine acceleration as the sign of a restricted system than to go by absolute pressure benchmarks.
DanG144
Top Post Dawg
Josh,
Can you tell me the lift pump part number for your car? If it is the B revision pump, then you might have an air leak on your fuel suction or return path. Otherwise I don't think it will immediately die.
Most of the time the 109 psig tandem pump discharge pressure regulator will not be opening. It will be needed if the screen plugs up or the back pressure regulator fails on the return line, or perhaps at high RPM.
Does anyone have the banjo bolt and seals part numbers needed to check out the tandem pump's operation?
Figuring the actual flows and pressures is problematic.
The positive displacement pump styles used in the lift and tandem pumps will put out a constant volume of flow for each revolution of the pump, if their suction stays flooded. The volume of flow from these pumps is not dependent upon discharge pressure (If the speed stays constant), until you reach extremely high pressures, where the leakage through the seals becomes the overriding factor.
But then the pressure regulators needed for these pumps throw in complications.
The speed of the electric motor driving the lift pump can be considered constant (for our purposes here), but the speed of the cam driven pump varies with engine speed. It's flow potential varies directly with engine speed. Double the speed, double the flow.
The backpressure that the tandem pump sees is a function of the restrictions down the system - the orifices and backpressure regulators, head losses due to friction, and so forth.
The pressure check for the tandem pump in the Bentley's manual is done at 1500 rpm and they expect 51 psi. The reason this is a valid test is that MOST of the components that vary the back pressure are located in the tandem pump assembly (two pressure regulators, orifices, screen, pump itself.) The remaining components are supposedly fixed orifices in and around the injectors, except for the amount of flow going through the injectors into the cylinders. And the amount of flow going through the injectors to maintain 1500 rpm will be about the same for most engines, when unloaded.
I would love to know how many tandem pumps were replaced because the lift pump had failed, or the fuel filter was plugged, though.
The flow through the injectors into the cylinders must be a larger component of the total than I had first figured, or at least more critical. My PD when stock would only pull about 35 liters per hour at max throttle. Call it 0.6 liters per minute. My lift pump produces about 3.2 liters per minute, leaving up to 2.6 liters per minute recirc flow available before the tandem pump suction starts going into a vacuum, even at full throttle.
Dan
Can you tell me the lift pump part number for your car? If it is the B revision pump, then you might have an air leak on your fuel suction or return path. Otherwise I don't think it will immediately die.
Most of the time the 109 psig tandem pump discharge pressure regulator will not be opening. It will be needed if the screen plugs up or the back pressure regulator fails on the return line, or perhaps at high RPM.
Does anyone have the banjo bolt and seals part numbers needed to check out the tandem pump's operation?
Figuring the actual flows and pressures is problematic.
The positive displacement pump styles used in the lift and tandem pumps will put out a constant volume of flow for each revolution of the pump, if their suction stays flooded. The volume of flow from these pumps is not dependent upon discharge pressure (If the speed stays constant), until you reach extremely high pressures, where the leakage through the seals becomes the overriding factor.
But then the pressure regulators needed for these pumps throw in complications.
The speed of the electric motor driving the lift pump can be considered constant (for our purposes here), but the speed of the cam driven pump varies with engine speed. It's flow potential varies directly with engine speed. Double the speed, double the flow.
The backpressure that the tandem pump sees is a function of the restrictions down the system - the orifices and backpressure regulators, head losses due to friction, and so forth.
The pressure check for the tandem pump in the Bentley's manual is done at 1500 rpm and they expect 51 psi. The reason this is a valid test is that MOST of the components that vary the back pressure are located in the tandem pump assembly (two pressure regulators, orifices, screen, pump itself.) The remaining components are supposedly fixed orifices in and around the injectors, except for the amount of flow going through the injectors into the cylinders. And the amount of flow going through the injectors to maintain 1500 rpm will be about the same for most engines, when unloaded.
I would love to know how many tandem pumps were replaced because the lift pump had failed, or the fuel filter was plugged, though.
The flow through the injectors into the cylinders must be a larger component of the total than I had first figured, or at least more critical. My PD when stock would only pull about 35 liters per hour at max throttle. Call it 0.6 liters per minute. My lift pump produces about 3.2 liters per minute, leaving up to 2.6 liters per minute recirc flow available before the tandem pump suction starts going into a vacuum, even at full throttle.
Dan
Sooch
Veteran Member
Dan,
Just wanted to give you an update that I promised a few weeks ago. I had mentioned that I would take a look at my lift pump at the ON GTG, which I did. The return line was clear since I could easily blow through it. I actually checked to see if I could easily blow through the return line from the filter back, once I reinstalled the lift pump, and I could. So there isn't any significant restriction there on my car. We did notice that the thermo-tee on the filter was always open, but that "should" theoretically keep fuel temps lower since fuel is always going back to the tank instead of recirc. I will be replacing it soon.
We also found that my fuel temp sensor is suspect. It reads low resistance even when the fuel is relatively cool, which the ECU interprets as a high temp. So it may be indicating a false positive. I will be replacing that soon also. Has anyone measured actual fuel temps with an accurate external sensor, or are we trusting the factory sensor for our readings?
I also found that there is no way to mount a pump alongside the lift pump canister and get it back into the tank, for supplying cool tank fuel directly into the canister. No room to do it. However, it seems that the actual pump inside the lift pump canister is nothing more than a lower flow, low pressure Walbro style pump. Do you think if that pump were replaced with something that could flow more, it would help with high temps, or would weird things begin to happen at the tandem pump if it received say 50psi from the lift pump?
Just wanted to give you an update that I promised a few weeks ago. I had mentioned that I would take a look at my lift pump at the ON GTG, which I did. The return line was clear since I could easily blow through it. I actually checked to see if I could easily blow through the return line from the filter back, once I reinstalled the lift pump, and I could. So there isn't any significant restriction there on my car. We did notice that the thermo-tee on the filter was always open, but that "should" theoretically keep fuel temps lower since fuel is always going back to the tank instead of recirc. I will be replacing it soon.
We also found that my fuel temp sensor is suspect. It reads low resistance even when the fuel is relatively cool, which the ECU interprets as a high temp. So it may be indicating a false positive. I will be replacing that soon also. Has anyone measured actual fuel temps with an accurate external sensor, or are we trusting the factory sensor for our readings?
I also found that there is no way to mount a pump alongside the lift pump canister and get it back into the tank, for supplying cool tank fuel directly into the canister. No room to do it. However, it seems that the actual pump inside the lift pump canister is nothing more than a lower flow, low pressure Walbro style pump. Do you think if that pump were replaced with something that could flow more, it would help with high temps, or would weird things begin to happen at the tandem pump if it received say 50psi from the lift pump?
joshgb
Veteran Member
You can fit a pressure tester on the tandem pump without a banjo set up but I don't remember the size.
Last edited:
whitedog
Veteran Member
Dan, I have an M12X1.5MM to -4 SAE 37* fitting adapter I bought locally. It has an Oring that is surrounded by a steel washer. This gives a good, repeatable seal. It is necessary to install an elbow on that so my hose will fit. I'm assuming tha banjo fitting is VAS 5187.
DanG144
Top Post Dawg
Sooch,
Thanks for the feedback.
A higher flow lift pump, would not hurt anything, and would keep the tandem pump fully supplied with fuel, and this MAY (probably would?) increase recirc flow rates and keep the fuel cooler at higher RPM. The Pierburg pump that is installed seems to be a rotary positive displacement pump, not a solenoid style pump, such as the Walbro's appear to be (going by other's descriptions - I have never seen one.)
It would be important to keep all of the design features of the present fuel canister assembly, or accept the loss of some capabilities.
I would keep the pressure regulator that would hold the maximum pressure at about 15 psig to the suction of the tandem pump.
I would keep the eductor, and inlet check valve if at all possible. This actually does two functions; it keeps the canister full of fuel to help handle angles at very low fuel levels, and it pulls in excess cool fuel (or at least fuel at tank temperatures, rather than recirc line temperatures), allowing the warm return fuel to spill out the top of the canister, rather than going down to the pump suction.
Another possibility would be to install a bypass suction line, so that at really high demand, the tandem pump could pull fuel from the tank itself. I think this line would entail a third fuel line connection on top of the pump assembly, and a line going through the canister to pull from the tank at the bottom of the canister. This line would bypass the spring loaded check valve built into the top of the assembly, and would have to have a simple diaphragm check valve in it. This approach would mean that the PD owner would have to be as cautious of air inleakage as the ALH owner, or he could entrain air at high engine speeds and loads. This is the approach I was thinking of for myself. The most problematic area of this is liable to be the bulkhead penetration and sealing for the third line on top of the fuel pump plate.
A variation of the bypass line, would be to put an external (maybe Walbro or Carter solenoid style pump) that would, because of its design, only draw power when pressure dropped off in the filter supply line - normally at high RPM and high flow. This puts the two pumps in parallel, without having to put in another in-tank pump.
If I was racing, or really interested in high performance I would go with a system that maintained a positive pressure into the tandem pump, at all engine speeds at full fuel demand. This would mean higher electrical power consumption, and some loss in mileage - probably very minor. Heavier wiring may be required to the fuel pump. I do not know if the fuel pump relay can handle the increased current flow.
Remember that the flow through the orifices will vary with the square of the differential pressure across them. So for instance the minimum acceptable 51 psi tandem pump discharge pressure at 1500 rpm, and the 14.5 psi backpressure on the fuel rail give you a 36.5 psi difference across the orifices. The maximum pressure out of the tandem pump is 109 psi, giving you a maximum differential pressure of 94.5 psi. This works out to a maximum flow that is only 1.6 times the available flow at 1500 rpm. So there are limits to the benefits of maximizing suction pressure to the tandem pump.
Heat load from increased combustion may simply increase faster than we can increase fuel flow for cooling.
I usually cruise at speeds (73mph indicated, 70 actual) and power levels that do not cause any elevated fuel temperatures, so this is all a labor of curiosity for me - I have no performance issue that needs correction. This means I am not pursuing this with my usual goal oriented single mindedness (more descriptive and less kind words are often used by observers), or even average human vigor.
Dan
Thanks for the feedback.
A higher flow lift pump, would not hurt anything, and would keep the tandem pump fully supplied with fuel, and this MAY (probably would?) increase recirc flow rates and keep the fuel cooler at higher RPM. The Pierburg pump that is installed seems to be a rotary positive displacement pump, not a solenoid style pump, such as the Walbro's appear to be (going by other's descriptions - I have never seen one.)
It would be important to keep all of the design features of the present fuel canister assembly, or accept the loss of some capabilities.
I would keep the pressure regulator that would hold the maximum pressure at about 15 psig to the suction of the tandem pump.
I would keep the eductor, and inlet check valve if at all possible. This actually does two functions; it keeps the canister full of fuel to help handle angles at very low fuel levels, and it pulls in excess cool fuel (or at least fuel at tank temperatures, rather than recirc line temperatures), allowing the warm return fuel to spill out the top of the canister, rather than going down to the pump suction.
Another possibility would be to install a bypass suction line, so that at really high demand, the tandem pump could pull fuel from the tank itself. I think this line would entail a third fuel line connection on top of the pump assembly, and a line going through the canister to pull from the tank at the bottom of the canister. This line would bypass the spring loaded check valve built into the top of the assembly, and would have to have a simple diaphragm check valve in it. This approach would mean that the PD owner would have to be as cautious of air inleakage as the ALH owner, or he could entrain air at high engine speeds and loads. This is the approach I was thinking of for myself. The most problematic area of this is liable to be the bulkhead penetration and sealing for the third line on top of the fuel pump plate.
A variation of the bypass line, would be to put an external (maybe Walbro or Carter solenoid style pump) that would, because of its design, only draw power when pressure dropped off in the filter supply line - normally at high RPM and high flow. This puts the two pumps in parallel, without having to put in another in-tank pump.
If I was racing, or really interested in high performance I would go with a system that maintained a positive pressure into the tandem pump, at all engine speeds at full fuel demand. This would mean higher electrical power consumption, and some loss in mileage - probably very minor. Heavier wiring may be required to the fuel pump. I do not know if the fuel pump relay can handle the increased current flow.
Remember that the flow through the orifices will vary with the square of the differential pressure across them. So for instance the minimum acceptable 51 psi tandem pump discharge pressure at 1500 rpm, and the 14.5 psi backpressure on the fuel rail give you a 36.5 psi difference across the orifices. The maximum pressure out of the tandem pump is 109 psi, giving you a maximum differential pressure of 94.5 psi. This works out to a maximum flow that is only 1.6 times the available flow at 1500 rpm. So there are limits to the benefits of maximizing suction pressure to the tandem pump.
Heat load from increased combustion may simply increase faster than we can increase fuel flow for cooling.
I usually cruise at speeds (73mph indicated, 70 actual) and power levels that do not cause any elevated fuel temperatures, so this is all a labor of curiosity for me - I have no performance issue that needs correction. This means I am not pursuing this with my usual goal oriented single mindedness (more descriptive and less kind words are often used by observers), or even average human vigor.
Dan
whitedog
Veteran Member
Since this same system runs higher HP in Europe and higher HP equates to more fuel consuption, it would seem that even though the fuel pressure drops into the vacuum range on these engines, couldn't we assume that the tandem pump will pull sufficient fuel at higher RPMs with North America tuning? I'm just wondering if this reengineering for a stock car is actually needed.
DanG144
Top Post Dawg
No telling yet; that is what we want to find out.
The musings in my previous post were early thoughts.
A driver of a stock car here started complaining of loss of power, it turns out that it was due to high fuel temperatures causing the ECM to retard timing. He notices it in the summer.
I never noticed anything on my RC 1+ tuned car.
I do not drive as hard as he does, except once a week.
That is what started this whole rigamarole.
Several other people have posted of similar instances.
We do not know yet if it plugged orifices, screens, worn tandem pumps, worn lift pumps, bad sensors have been ruled out by replacements.
So, for me, this is a back-burner medium interest issue. It seems a few others have more urgency, so I thought to share some brain storming.
That's about it.
Dan
The musings in my previous post were early thoughts.
A driver of a stock car here started complaining of loss of power, it turns out that it was due to high fuel temperatures causing the ECM to retard timing. He notices it in the summer.
I never noticed anything on my RC 1+ tuned car.
I do not drive as hard as he does, except once a week.
That is what started this whole rigamarole.
Several other people have posted of similar instances.
We do not know yet if it plugged orifices, screens, worn tandem pumps, worn lift pumps, bad sensors have been ruled out by replacements.
So, for me, this is a back-burner medium interest issue. It seems a few others have more urgency, so I thought to share some brain storming.
That's about it.
Dan
whitedog
Veteran Member
Although barn storming is more fun, brain storming is more useful. (and lets not even get in to Brian storming)
Really, that's where I am too; just throwing out ideas and trying to get some useful information in and out.
Really, that's where I am too; just throwing out ideas and trying to get some useful information in and out.
Regarding the existence of J445 and V166:
These were installed in Passat PD engines manufactured prior to 1/2002. Later versions have the cooler under the car chassis. There are as many as 6 hoses connecting to the top flange of the fuel filter (diesel from/back to tank, diesel to/from the engine and coolant to/from auxiliary radiator).
The J445 is located inside the ECU's box. The V166 is located at the auxiliary radiator in front of the passenger side front wheel (approx. same location as intermediate cooler on the driver side). Using the Output command of a tester, the J445 clicks and the V166 is audibly heard to turn on and off.
These were installed in Passat PD engines manufactured prior to 1/2002. Later versions have the cooler under the car chassis. There are as many as 6 hoses connecting to the top flange of the fuel filter (diesel from/back to tank, diesel to/from the engine and coolant to/from auxiliary radiator).
The J445 is located inside the ECU's box. The V166 is located at the auxiliary radiator in front of the passenger side front wheel (approx. same location as intermediate cooler on the driver side). Using the Output command of a tester, the J445 clicks and the V166 is audibly heard to turn on and off.
DanG144
Top Post Dawg
Excellent!
Thanks for the education. I thought maybe they had only been used on industrial engines.
Thanks for the education. I thought maybe they had only been used on industrial engines.
Redneck Truck
Gone, but not forgotten
- Joined
- May 3, 2009
- Location
- Plano, TX
- TDI
- 2008 Touareg V10, 2001 Audi TTQ Roadster TDI 6-speed, 2005 Jetta Wagon TDI 6-speed
Hate to bring up old information, but I'm chasing my tail with my fuel temps over 90C and am curious if you ever resolved the problem.