CP3 Pump Install Experiments.**Big Picts**

[2micron]

Rail pressures

Sure thing Veedub, Give me a couple of days. I have data from "free revving" the engine at various RPM's, read directly from the rail pressure sensor, via voltmeter. I am planning a run in the next couple of days with the new pump and will post comparisons from each pump, via VCDS.
.
I don't have a dyno, so it is hard to "load" the engine, while stationary. Any suggestions how to load the engine, with the lawnmower attached to the side? Thanks,
All the best,

 

[IFRCFI]

Stupid question. If a CP3 pump runs these engines why did they develop/use a new pump?

Marketing + Accounting - Engineering = Car Company.

In the end, it's all about cost.......

 

[TDIMeister]

Very nice work! Maybe I missed it somewhere above, but what drive ratio are you going to end up using in the engine?

 

[GoFaster]

I think the ratio between lawnmower engine and Jetta engine would count as "variable"

 

[Blackheart]

Watching in fascination, laced with hope!

 

[TDIMeister]

Oh no, the tragedy! Drivbiwire said a TDI would never work properly unless the HPFP pump pulses were syncronized with injection events.

 

[CNDTDI]

good for you 2micron, if you need some machining done let me know i could do some oddball things for you.

by the way love your snowblower

 

[2micron]

Common Rail Mythbusters

Very nice work! Maybe I missed it somewhere above, but what drive ratio are you going to end up using in the engine?

Thanks TDImeister, that means a lot to me coming from you! Pump ratio will be slightly less than 50% of crank shaft speed for stock engine flow rates.

I think the ratio between lawnmower engine and Jetta engine would count as "variable"

The 1950's Gravely Lawn Tractor simply can not be timed or set up to develop a set "Ratio" to the VW engine. It will happily run at any speed, feeding the VW, like a patient in a hospital bed with an IV.

Watching in fascination, laced with hope!

Thanks Blackheart!!

Oh no, the tragedy! Drivbiwire said a TDI would never work properly unless the HPFP pump pulses were syncronized with injection events.

I respect Drivbiwire's knowledge and comments. He has probably been playing with diesels before I was born. However, My experiments and research has assured me the Pump may not have to be timed to the Injection Events.

Here's where we can all play Mythbusters:
Myth #1.) The Pump "pulses" have to be timed to the injection event?
.
FACTS: - The CP3 "Rear Gear" pump simply has no key, marks, pin, or anything to index a pulley or gear to the crankshaft, when it is installed on the Cummins 6.7 Engine.
- A Cummins "Rear Gear" 4.5 Common Rail 4 cylinder runs this same Pump, driven at 1:1 ratio with the crank. There is no way it can be Timed to the injection events.
- During this Experiment, countless hours were logged, performing precise Vibration Analysis on various parts of the Rail, both on the Test Bench and the car. The results were repeatable and the same across the board:

Vibration Monitoring on the Rail

-The results? Rail Vibrations are on average, 1.5mm/s (mm per second) less with the 3 Piston CP3 Pump at all RPM's, compared to the CP4.1 single Piston Pump.
.
Myth #2.) - CP3 Pump maximum pressures are different than the CP4.1:
FACTS: - both the CP4.1 and the CP3 pumps have no internal High Pressure Relief Valve. If you were to "Dead Head" the pumps, they would build pressure until the Driven motor stops, or something breaks. Think of your grease gun developing 10,000 psi, at no stroke, with you pushing as hard as you can until the handle bends, or the Zirk fitting connection leaks.
- On our cars, the Fuel Rail Control Solenoid buzzes at 1000Hz, while the ECM varies the duty cycle to control Pressure.
.
Lets all play.
With the Facts given, who wants to help Bust Myths 1 and 2??
.

good for you 2micron, if you need some machining done let me know i could do some oddball things for you.
by the way love your snowblower

Thanks CNDTDI, A fellow Machinist a couple hours away is great to know! Certainly will one day require some help or suggestions.
Update from the Weekend:
Had great fun running the car and Lawnmower again. Did more testing and Experimenting Via VCDS logging suggestion from:

Proof of concept is great. Can we see some VCDS graphs of rail pressure?

Thanks to Karter59 and Tyrel for the VCDS and hardware!!
I will post a couple of Graphs, showing the comparison between Actual Rail Pressure and Specified Rail Pressure.
All the best,

 

[kjclow]

I guess all you have to do is mount the lawn mower in the back of the JSW and your set! Of course, you may want to pop out the rear hatch window so you don't gas yourself.

Seriously, keep up the good work. We can all benefit from it, even VW.

 

[storx]

The cp3 also comes on 8.3 cummins in semi trucks...

Good job so far on the work... If I was closer I would lend a hand. I can't wait tell you get this all figured out cause there are so many company's out there that sell upgraded cp3 pumps that double its flow capability and such... Friend has a dragon cp3 pump on his cummins an it can feed his compound turbo setup without issues and he is putting down around 1200 torque daily

 

[TDIMeister]

Do you have any guesses based on the lawnmower driving the pump what the power draw was?

Perhaps the choice was probably made from a desire for redundancy in case of a belt break, but I would seem that it would be much simpler to mount the CP3 pump essentially where the original pump was (moved only enough so that the same original belt length could be used); use a 1:2 VE injection pump toothed sprocket and drive the CP3 pump with the original timing belt without having to run a separate serpentine belt and additional fabricated parts. Unless I'm missing something from packaging constraints (limitation of not being there to see the setup in person), this approach would seem to be easier to realise, especially if others might want to replicate the CP3 retrofit.

Not knocking your idea - I love the overall concept! Just throwing questions to make a better end result!

P.S.: For the sake of historical accuracy -- and Louis is free to dispute this -- but the idea to use a CP3 pump in a TDI has its roots even before it was conceived as a replacement for the 2.0 16V common rail engine. Back in 2009, I had done engine simulations and proposed a 500 HP build based on a 5-cylinder VE TDI. This was scaled and morphed to a 400- and then 450 HP 4-cylinder TDI concept when Louis approached me to realise this in his project build. Up to this point, in early 2010, my simulation models still assumed an 8V TDI head and I knew that the biggest challenge to achieving the HP goal was fuel atomization quality and smoke-limited injection quantity, so Louis and I looked at every way possible to outfit the 8V TDI head with common rail based on a Frankenstein of parts from the VAG parts bin and other different OEMs (VW Crafter injectors, Bosch EDC15/16C from some other application and CP3 pump from plentiful Duramaxes/Cummins). The decision was made to abandon the 8V CR TDI idea for Louis' project when we failed to come up with a proven, working solution and it was much easier (and ultimately cheaper, when Louis found and bought a burned CBEA engine) to go with an OE 16V CR TDI base for his project. That's where histories converge with the idea to use the CP3 pump to have sufficient fueling for a monster TDI build. I have since revived a project to retrofit a CR injection system into the 8V VE TDI and it will also use a CP3 pump. Just to clear up the air...

 

[greengeeker]

Perhaps the choice was probably made from a desire for redundancy in case of a belt break, but I would seem that it would be much simpler to mount the CP3 pump essentially where the original pump was (moved only enough so that the same original belt length could be used); use a 1:2 VE injection pump toothed sprocket and drive the CP3 pump with the original timing belt without having to run a separate serpentine belt and additional fabricated parts. Unless I'm missing something from packaging constraints (limitation of not being there to see the setup in person), this approach would seem to be easier to realise, especially if others might want to replicate the CP3 retrofit.

1) The overall diameter of the CP3 pump is larger and will not allow a person to tuck it as close to the block so the shaft centerline is going to need to move away from the block.
2) Of course the 1:2 sprocket is much larger than the 1:1 used with the CP4.2 in the tdi.

#1 + #2 => much longer timing belt required. I think we would almost have to dive into the v6 tdi parts bin to get a longer belt as the CBEA/CJAA belt is friggin long as it is.

I agree that for simplicity's sake the HPFP should still be driven by the timing belt instead of an additional accessory belt.

 

[2micron]

Horsepower to drive the pump.

Do you have any guesses based on the lawnmower driving the pump what the power draw was?

Perhaps the choice was probably made from a desire for redundancy in case of a belt break, but I would seem that it would be much simpler to mount the CP3 pump essentially where the original pump was (moved only enough so that the same original belt length could be used); use a 1:2 VE injection pump toothed sprocket and drive the CP3 pump with the original timing belt without having to run a separate serpentine belt and additional fabricated parts. Unless I'm missing something from packaging constraints (limitation of not being there to see the setup in person), this approach would seem to be easier to realise, especially if others might want to replicate the CP3 retrofit.

Not knocking your idea - I love the overall concept! Just throwing questions to make a better end result!..

Sure thing Meister!
Horsepower to drive the pumps can easily be figured out, based on bore, stroke, rpm, pressure and flow. This "theoretical" calculation does not account for losses in pump efficiency or drive losses. (adds about 15%)
The numbers below are based on 26,000 psi at full pump flow.
CP4.1@ 4000rpm consumes 4.49Hp
CP3 @ 2000 rpm consumes 7.56 Hp.
In real life, the numbers are slightly lower, based on the lawnmower test runs.
In 1952, the Gravely tractor was rated at 6.6Hp. During these tests, the lawnmower could power each pump to full pressure, only working slightly more with the CP3 pump. Estimate approx 1.5 Hp more for the new pump at 1/2 crank speed.
.
As for the mounting and the drive system, the pump simply will not fit between the timing belt sprockets and the EGR intake port. The pump is mounted in the same location as the original, just more in alignment with the crank pulley.
The other main reason, as you note is the extra load is not favorable on the timing belt.
I am excited to see your project and progression and hope this information and testing helps!
Between you and TDISyncro, credit must be given for the whole CP3 pump idea, I didn't realize you were involved years ago!!
TDISyncro has also been instrumental with this project, helping me with critical parts and measurements. When he can, he will chime in and add his valuable contributions.
All the best!!

 

[2micron]

1)

I agree that for simplicity's sake the HPFP should still be driven by the timing belt instead of an additional accessory belt.

Good thought greengeeker, it is not really simpler though. I dropped the idea months ago for these reasons:
- EGR / intake port would have to be modified to allow the pump to line up with the timing belt sprockets.
- the slight increase in HP required to drive the CP3 would not be favorable on the timing belt.
- if the pump were to lock up for whatever reason, this project would be a total loss
.
I'm not saying it is not possibe to drive the pump with the timing belt, but for the experiment and prototype, an additional belt is easier and much safer.
The benefits of the additional drive belt set up also allow easy modifications to pump speed ratios, in order to develop an ideal relationship between the power required to turn the pump and the actual "required" flow rate. The speed of the pump for the experiment will be less than 1/2 crank speed.
storx, thanks for the comments and posting the crazy information on modded pumps in the big truck world!
All the best,

 

[greengeeker]

I'm not saying it is not possibe to drive the pump with the timing belt, but for the experiment and prototype, an additional belt is easier and much safer.

I agree, for an experimental prototype the additional belt is the way to go. I'm thinking ahead to when this makes it to kit form

 

[TDIMeister]

Sure thing Meister!
Horsepower to drive the pumps can easily be figured out, based on bore, stroke, rpm, pressure and flow. This "theoretical" calculation does not account for losses in pump efficiency or drive losses. (adds about 15%)
The numbers below are based on 26,000 psi at full pump flow.
CP4.1@ 4000rpm consumes 4.49Hp
CP3 @ 2000 rpm consumes 7.56 Hp.
In real life, the numbers are slightly lower, based on the lawnmower test runs.

In 1952, the Gravely tractor was rated at 6.6Hp. During these tests, the lawnmower could power each pump to full pressure, only working slightly more with the CP3 pump. Estimate approx 1.5 Hp more for the new pump at 1/2 crank speed.

I am quite certain that each successive new generation of pump takes less power for a given flow and pressure, and with fewer pistons also be cheaper to manufacture. It would be useful to consider calculating specific pump power by normalising it against flow rate and pressure, e.g. kW/(100 bar*g/s) or HP/(1000 PSI*cc/min). I guess both rail pressure and fuel flow rate can be interrogated in VCDS. I prefer working exclusively in metric so I'd choose something like the former form. I would be curious then to know what the real difference in drive power draw is for a given fuel flow and what impacts it would have on timing belt life. Also, as I mentioned in this post, drive torque fluctuations can be reduced by using a sprocket with sufficient inertia so that it acts like a flywheel.

As for the mounting and the drive system, the pump simply will not fit between the timing belt sprockets and the EGR intake port. The pump is mounted in the same location as the original, just more in alignment with the crank pulley.

I understand what you're saying, and it took an old picture from Louis to drive the point home. I will now humbly scurry back to my armchair

My line is always open to you if you wish to contact me.

 

[GoFaster]

The CP3 pump with 3 pistons in it should have much more steady output flow than the single-piston CP4.1 pump. It doesn't surprise me at all that the synchronization to the cam/crank doesn't matter. This means you can drive it with a serpentine or V-belt independently of the timing belt at whatever ratio you choose, and it frees up the location of the pump.

 

[birkie]

Very impressive work! I'm especially glad that you confirmed that "timing" of the pump is mostly irrelevant as long as it has enough output for the conditions (at least for the 3-piston CP3. Have you driven it externally with a CP4.1?)

The interesting part was that the new pump takes the same horsepower to develop the same flow rates and pressures. It is much more efficient and runs considerably cooler.

Ah, I think I read this the wrong way at first (i.e with the CP3 as the "new" pump to this engine). Are you saying is that the CP4.1 is more efficient, and runs cooler than the CP3? This would agree with your observations that the CP3 requires more horsepower to reach a given pressure and flow rate.

 

[TDIMeister]

The above statement is contradictory. If it takes the same power to develop the same flow rate and pressure, by definition the efficiency is the same and so should operating temperature. Temperature will also be affected if there is a return fuel flow to carry heat away. The latest injection systems are return-less.

Eager to hear a clarification.

 

[2micron]

Clarifications:

Sorry for the confusion, I hope this will clarify the concerns:

The interesting part was that the new pump takes the same horsepower to develop the same flow rates and pressures. It is much more efficient and runs considerably cooler.

To achieve the same flow rates, the two pumps are run at different rpm’s because they have different displacements. The CP3 runs at approx 30% less rpm to develop the same flow rate as the CP4.1. During these runs, the horse power required to turn the pumps is the “Same”. Bold, Exaggerated Statement? Yes, perhaps, but here is my reasoning:
Let’s face it; I can’t measure the Lawnmower horsepower accurately. The 6.6 HP Gravely seems to run most efficiently at 2850 rpm’s. The 2 pumps are set up to run at the 30% difference in rpm’s as noted, while maintaining 2850 RPM at the Lawnmower, via different pulley sizes. (See below pictures)

Stock CP4.1 Pump


.
New CP3 Pump with 30% less RPM.


The Machine is quite sensitive to a load with its low Horsepower, so it is very easy to notice a drop in RPM. My entire “Same Horsepower” statement is based on a “non scientific” feel for throttle position, sound and actual engine RPM. Apologies if I exaggerated. The more accurate statement would have been “The interesting part was that the new pump takes ABOUT THE same horsepower to develop the same flow rates and pressures.” I was so excited the new Pump would even react similar to the original Pump; I never or have yet to look for a more accurate way to measure driven horsepower.
The other statement “During these tests, the lawnmower could power each pump to full pressure, only working slightly more with the CP3 pump. Estimate approx 1.5 Hp more for the new pump at 1/2 crank speed.” **This is based on ½ crank speeds, as a suggestion for TDIMeisters VE Sprocket driven Pump Project. *** And again, is based on the same principles as above, for the new CP3 Pump.

Ah, I think I read this the wrong way at first (i.e with the CP3 as the "new" pump to this engine). Are you saying is that the CP4.1 is more efficient, and runs cooler than the CP3? This would agree with your observations that the CP3 requires more horsepower to reach a given pressure and flow rate.

I must have really confused you here!!!
The CP3 (new) pump is more efficient and runs cooler than the (old) CP4.1 Pump.

The above statement is contradictory. If it takes the same power to develop the same flow rate and pressure, by definition the efficiency is the same and so should operating temperature.

Eager to hear a clarification.

Yes, I agree. The new pump achieves the same flow rates and pressures, but this at 30% less RPM’s and the Revised Statement “takes ABOUT THE same horsepower “ Apologies again for the exaggerated statement. The new CP3 Pump has a much greater surface area and 3 separate heads and check valves, to help individually dissipate heat, compared to the CP4.1.

Temperature will also be affected if there is a return fuel flow to carry heat away. The latest injection systems are return-less.

Eager to hear a clarification.

Not sure about the latest injection system, but these two Pumps have a very healthy Return system. The small VW Bus fuel tank would get significantly warmer with the return flow from the CP4.1 compared to the CP3. I did not realize the Latest systems were “Returnless”
VW BUS Fuel Tank:

Thanks for the healthy scrutiny, hope I clarified and apologize for any confusion!!! I will revise the line from the original post:"The interesting part was that the new pump takes the same horsepower to develop the same flow rates and pressures." to “takes ABOUT THE same horsepower “
All the best,

 

[TDIMeister]

Thanks for the clarifications. I'm all the more encouraged that your project and adapting the CP3 pump into the TDI in general will be successful.

 

[red64chevelle]

I can't wait for you to be driving around with a CP3 pump! I can only hope that you will have it all figured out by the time my Golf goes out of warrantee. I would definitely look at installing one of these setups before my CP4.1 eats it-self.

 

[BluGraphTDI]

I guess all you have to do is mount the lawn mower in the back of the JSW and your set! Of course, you may want to pop out the rear hatch window so you don't gas yourself.

Would we consider this a gas/diesel hybrid?

This project is very cool, hope this moves along quickly or VW takes notice..

 

[storx]

So who is testing first prototype??

 

[storx]

i was just reading through this again and i relized an post i made earlier never got posted. We have an tactical peace of equipment at work that utilizes an small diesel engine for cooling of the equipment portable and i noticed that the injection pump on it was not mounted on the engine but mounted on its own area belted up with the mechanical fan blade assembly being turned by an small electric motor. I am not sure how practical this is for us but its an option. The pump was probably moved because it was an off the shelf motor that didnt fit in the box so they moved stuff to make it fit in the spot it needed to.

 

[2micron]

VCDS Rail Pressure Graph

Proof of concept is great. Can we see some VCDS graphs of rail pressure?

Sure thing VeeDubTDI:
.

This graph is developed from VCDS data. The pump was run at a constant 1250 rpm, while the VW engine "Free Revs" from idle to near red line.
It is amazing to see how quickly the "actual" pressure follows the "specified" pressure. (Time stamp is in seconds)
.
Storx- if possible, please post a picture of the "remote pump" on your tactical piece of equipment. This is an interesting idea, with an electric motor driven pump!
Good progress again this week, still working out the drive belt and tensioners.
Thanks again for all the comments and thoughts!! Please keep them coming and feel free to post!!
All the best,

 

[VeeDubTDI]

That graph is outstanding. Very impressive work!

 

[TDIMeister]

Nice work, but to be clear, VCDS would not have the sampling frequency to capture pressure wave fluctuations within individual injection events in the rail.

 

[storx]

Wow nice graph... Knew it work... All it does is supply pressure nothing to complicated about it

 

[N2UADTDI]

2Micron - does the fuel metering valve work the same way on both pumps?