Common Rail injectors

TDIA3

Well-known member
Joined
Sep 17, 2015
Location
VA
TDI
2011 A3 TDI 2260vk
This is a long shot but i want to get some larger injectors have my extra set extruded>
Are there any leads to procuring some?
thanks
 

turbobrick240

Top Post Dawg
Joined
Nov 18, 2014
Location
maine
TDI
2011 vw golf tdi(gone to greener pastures), 2001 ford f250 powerstroke
I would contact Les at GDM. I believe he has had some piezo injectors extruded.
 

gloaming

Veteran Member
Joined
Nov 3, 2012
Location
Port Fishington, Philadelphia
TDI
2010 Golf 6MT CR170 (Sold); 2004 R32 (not a TDI)
Have you run out of fuel on stock injectors? Or do you just want to go bigger just because? Just curious, because I have yet to hear of anyone running out of fuel and needing bigger injectors on a commonrail. FYI, upgraded fuel pump should be a consideration if you don't already have one.
 

shortysclimbin

Veteran Member
Joined
Oct 15, 2004
Location
Virginia currently
TDI
Kubvan, mk2 golf, mk6 golf
I have about 5 sets on my shelf from cars I have parted out.. I haven't gotten around to listing them yet. I'll send you a pm. I also have a place that can extrude them too.
 

1972SS454

Veteran Member
Joined
Aug 27, 2010
Location
Fernandina Beach, FL
TDI
2009 Jetta TDI, 99.5 Jetta TDI
Have you run out of fuel on stock injectors? Or do you just want to go bigger just because? Just curious, because I have yet to hear of anyone running out of fuel and needing bigger injectors on a commonrail. FYI, upgraded fuel pump should be a consideration if you don't already have one.
He's running a 2260vk with CP3
 

Drivbiwire

Zehntes Jahr der Veteran
Joined
Oct 13, 1998
Location
Boise, Idaho
TDI
2013 Passat TDI, Newmar Ventana 8.3L ISC 3945, 2016 E250 BT, 2000 Jetta TDI
and ruining the injectors... Extrude Honing a TDI nozzle is a TOTAL fail.

Good luck with that meltdown (If it even runs afterward... and it won't).
 

KERMA

Vendor , w/Business number
Joined
Sep 23, 2001
Location
here
TDI
99 beetle and 04 jetta
IMA/IVA is pretty important on the common rail cars.

You can definitely tell the difference between a loaner ecu and the one that came with the car- even with the same tune (or stock for that matter)
 

turbobrick240

Top Post Dawg
Joined
Nov 18, 2014
Location
maine
TDI
2011 vw golf tdi(gone to greener pastures), 2001 ford f250 powerstroke
I know Les had some injectors extruded. I'm not sure how it worked out for him though. I think he may have had issues getting them dialed in, and was making much more smoke than he would like.
 

v8 coupe

Veteran Member
Joined
Sep 3, 2008
Location
bloomington, mn
TDI
09 rabbit 2.5L Gas
I'm sorry, but exactly why is it not possible to over size these injectors? Duramax guys do, Cummings guys do, hell even Power stroke guys do so what really separates our injectors from theirs? Duramax and Power stroke both run CP4.2 pumps at higher rail pressures then we do.
 

KERMA

Vendor , w/Business number
Joined
Sep 23, 2001
Location
here
TDI
99 beetle and 04 jetta
Why don't 2004 dodge and 2003 duramax require this? EDC16, solenoid injectors, different certification requirements, etc etc so IOW they are totally different animals. This is not your daddy's common rail.

Don't believe me? Ok, try this yourself, and tell me what you get:

VCDS adaptation channels 71,72,73,74. Write down the IMA/IVA value for each cylinder. Then swap around the values between cylinders or even make them all the same. Or even grab a value from a different car to *really* prove that "it simply doesn't matter."

tell us if it affects how the car runs.

If you cant tell any difference, then have a ball. ;)

Those adaptation values can trim the fuel calibration by a number of mg/r in each cylinder individually. If you think that doesn't matter then... ok I guess.

But beware: once you change the values the car will never run quite the same as it did. The more miles on the injectors, the more pronounced this effect will be. That's because the values are only valid for new injectors. The ecu adapts and adjusts the calibration over time in the eeprom based on a number of factors to account for wear and such. When you enter a value, this history gets reset, as if the injector is brand new again. Then the only way to get it running like before, is to tell the ECU the correct IMA/IVA values- and on used injectors this must be measured and generated using a very specific bosch test routine. (The other way to get it running like before, is to do a 100% back up of the ecu so it can be restored to the same state as it was before the values were monkeyed with... and which method do you think I used when I did this myself :))

Go ahead: do it and report back. ;) You may get lucky and it will be somewhat like a gasser running too lean or something subtle like that. Or maybe that sissy stuff doesn't matter as long as there's more power.
 

Drivbiwire

Zehntes Jahr der Veteran
Joined
Oct 13, 1998
Location
Boise, Idaho
TDI
2013 Passat TDI, Newmar Ventana 8.3L ISC 3945, 2016 E250 BT, 2000 Jetta TDI
I would love to know how calibration codes are being generated for these.
We can generate the codes, that's easy.

The hard part is extrude honing ruins the nozzle resulting in such gross deviations in raw flow for each nozzle, new codes are unable to span the deviation spectrum for proper control.

Extrude honing will cause a net error nozzle to nozzle of around 10-20%. Coding can only compensate for 2% at the extreme.

Ideally, less than 1% is needed to provide a service life of 100-150K.

All the new nozzles we are working with or on are spec'd to .25% not only making nozzle correction values valid longer, but also make initial setup and calibration more straight forward.

I still laugh when anyone tries to apply truck logic to an automotive application... You can't.
 

Drivbiwire

Zehntes Jahr der Veteran
Joined
Oct 13, 1998
Location
Boise, Idaho
TDI
2013 Passat TDI, Newmar Ventana 8.3L ISC 3945, 2016 E250 BT, 2000 Jetta TDI
Those adaptation values can trim the fuel calibration by a number of mg/r in each cylinder individually. If you think that doesn't matter then... ok I guess.
But beware: once you change the values the car will never run quite the same as it did. The more miles on the injectors, the more pronounced this effect will be. That's because the values are only valid for new injectors. The ecu adapts and adjusts the calibration over time in the eeprom based on a number of factors to account for wear and such. When you enter a value, this history gets reset, as if the injector is brand new again. Then the only way to get it running like before, is to tell the ECU the correct IMA/IVA values- and on used injectors this must be measured and generated using a very specific bosch test routine.
To add to that,

The ECU is hard coded with that engines injector requirement these are fairly simple values. The complexity is how the engine analyzes injector performance and determines injector performance drift, in real time...

So even if you manage to change that hard coded data, you are back to the strict production tolerance required to control the injection system.

If at anytime the ECU detects an injector out of tolerance, the ECU will either detune the motor and throw multiple control codes, or it will be unable to control it all together resulting in the engine not running.

Tnese engines are far beyond the point where you can sneak in a hardware change without the ECU detecting it, borderline impossible without extensive software and perhaps even other significant hardware changes.
 
Last edited:

DPM

Top Post Dawg
Joined
Mar 16, 2001
Location
Newtownards, N. Ireland
TDI
2019 Rav4 AWD Hybrid, Citroen C4 BlueHDI
do Bosch not carry out ongoing injector calibration? Denso in the Subaru carries out cylinder balance/ injector response checks at idle every now and then, it's noticeable as the normally glass-smooth idle is replaced by a misfire-like hunting occasionally.
 

turbobrick240

Top Post Dawg
Joined
Nov 18, 2014
Location
maine
TDI
2011 vw golf tdi(gone to greener pastures), 2001 ford f250 powerstroke
So are there options for higher flowing piezo injectors? Or would solenoid injectors be better? I know Turbojohan did a build up w solenoid injectors. I think he used a head and ecu that was set up for the solenoid injectors though.
 

v8 coupe

Veteran Member
Joined
Sep 3, 2008
Location
bloomington, mn
TDI
09 rabbit 2.5L Gas
Why don't 2004 dodge and 2003 duramax require this? EDC16, solenoid injectors, different certification requirements, etc etc so IOW they are totally different animals. This is not your daddy's common rail.
Don't believe me? Ok, try this yourself, and tell me what you get:
VCDS adaptation channels 71,72,73,74. Write down the IMA/IVA value for each cylinder. Then swap around the values between cylinders or even make them all the same. Or even grab a value from a different car to *really* prove that "it simply doesn't matter."
tell us if it affects how the car runs.
If you cant tell any difference, then have a ball. ;)
Those adaptation values can trim the fuel calibration by a number of mg/r in each cylinder individually. If you think that doesn't matter then... ok I guess.
But beware: once you change the values the car will never run quite the same as it did. The more miles on the injectors, the more pronounced this effect will be. That's because the values are only valid for new injectors. The ecu adapts and adjusts the calibration over time in the eeprom based on a number of factors to account for wear and such. When you enter a value, this history gets reset, as if the injector is brand new again. Then the only way to get it running like before, is to tell the ECU the correct IMA/IVA values- and on used injectors this must be measured and generated using a very specific bosch test routine. (The other way to get it running like before, is to do a 100% back up of the ecu so it can be restored to the same state as it was before the values were monkeyed with... and which method do you think I used when I did this myself :))
Go ahead: do it and report back. ;) You may get lucky and it will be somewhat like a gasser running too lean or something subtle like that. Or maybe that sissy stuff doesn't matter as long as there's more power.
Did I say they don't?
I was mainly referring to the LML duramax which is running a CP4.2 pump @2000bar rail pressure or ~29,000psi. A 200bar or ~3,000psi more then we are.
I do understand the complexity of these engines but what is the down side to taking a set of injectors and having them upgraded? Yes so far the hpfp is the limiting factor for power, but if that is overcome what is next? I understand all the work required in designing, testing, tuning, and marketing something like this. I'm not saying the majority or even more than a few are really going far enough to really justify stuff like this. From what I have been reading it's even a debated topic on the LML trucks which do run an EDC17 ecm/ecu like ours IIRC.

what is the limitations of our stock injectors? 300hp, 350hp, 400hp? If you are willing to live with the lag of a big vnt turbo, say gtb30** size do we have enough fuel for this? I'm sure at some point these engines being as plentiful as they are, they will get to the point somebody will do something crazy like this. I also have gathered that there are not many places that can flow test these piezo injectors like Bosch does. Stock the Trucks have adequate fueling for more power, but at some point even their injectors reach a point where injection windows start getting bigger than ideal. Bigger injector nozzles are available from a few sources setup and calibrate to be within 1% IIRC.

So what I am getting at besides the complexity of tuning these edc17 ECUs what is keeping someone like myself from having a set of these injectors edm'd or extrude honed?
 

hatemi

Veteran Member
Joined
Aug 25, 2005
Location
Finland
TDI
Audi A6 4F 3.0TDI
As usual DBW will not be a happy camper untill Bosio makes the upgraded nozzles...

You can get some minor power gains with stock nozzles but lets face it. They are just small. On every stock car. Or truck. LML and new Powerstroke are within few percent of what Audi 3.0TDI injectors flow. I'm not sure on if US spec 2.0TDI have smaller injectors or not, but I have a guess they are smaller. And even if they were same size as V6 TDI they have limits.

Lets assume that the injectors flow around the same ammount as V6, then we can do rough calculation that they are able to make 65-70hp/ cylinder at 4000rpm with high EGT and long duration. With 1800bar pressure. With 2000bar you can get few percent more. So we are at roughly 75hp/injector thus 300hp. If you push them more, EGTs will rise and top power will drop to lower RPM. Thats just physics.

Se what are we left with? 300hp four cylinder engine. Well that has been seen before. Few times actually.

And after 300hp you are pretty much forced to find more fuel one way or another. And maybe sacrafice some smooth idle and "clean" diesel eco friendliness in the process. And eventually you might even blow up and engine too. There seems to be upgrades available for LML and PS. Injectors are basically same. Might even bolt on. Technology is the same. Ecus are all EDC17.
 

v8 coupe

Veteran Member
Joined
Sep 3, 2008
Location
bloomington, mn
TDI
09 rabbit 2.5L Gas
That is my exact point that if the hpfp is addressed and can provide sufficient flow at 1800bar or even better at 2000bar what is stopping us? I understand bigger injectors take modifying the entire fueling algorithm to compensate for the increase, but can that be done? From my understanding our EDC17 ecus are basically the same hardware as Bosch Motorsports Common rail standalone ECU. So I am not convinced there isn't a way to properly modify the ECU to deal with 33%, 44%, or even 120% oversized injectors.

I'm also not sold on the extrude honing ruins the nozzles, because if done properly it won't, but edming is still better. So does any one know the stock size and number of holes in our stock nozzles?
 

turbobrick240

Top Post Dawg
Joined
Nov 18, 2014
Location
maine
TDI
2011 vw golf tdi(gone to greener pastures), 2001 ford f250 powerstroke
Some fuel injection specialists also use edm and extrude honing in conjunction. I also think there are new extrude honing technologies that offer much better precision than older tech.
 

v8 coupe

Veteran Member
Joined
Sep 3, 2008
Location
bloomington, mn
TDI
09 rabbit 2.5L Gas
So let's do a hypothetical build here:

So let's take a mk6 cr140 engine throw a set of Rosten rods in it, get a set of ARP head studs, and switch the HPFP to a CP3. Now that the basics are addressed let's moved to the next ingredient in the puzzle the Turbocharger, sure a GTB2260vk has been done bigger hybrids are available, but let's take it up to a huge for a 2.0L turbo a S300v. Now that might seem like a huge step, but it'll happen eventually, would we go to 30%, 45%, 60%, or 120% oversized injectors?

At what point do we need to stop bandaging the ECUs and actually completely unlock them to actually modify the full fuelling maps?
 

hatemi

Veteran Member
Joined
Aug 25, 2005
Location
Finland
TDI
Audi A6 4F 3.0TDI
When I'm upgrading my injectors (BMW) I will aim for +100% flow. Enough fuel with under 1ms injection duration.
 

Rub87

Top Post Dawg
Joined
Dec 10, 2006
Location
Belgium
TDI
Ibiza '99 90HP
acutally if you EDM and nozzle to nozzle deviation is below +-2% and shot to shot for small (pilot) quantities are below +-0.5 mg/stroke and one takes the time and effort to the make a corresponding duration map the deal is sealed. even without IMA codes after some time the ecu will learn the small quantities away. there might be more combustion noise from less accurate pilot quantities but as Hannu said 300 hp 4 cyl and 450hp v6 give or take is about the max one can run with 2000 bar. after that the egt/lambda tradeoff is quite bad
 

Drivbiwire

Zehntes Jahr der Veteran
Joined
Oct 13, 1998
Location
Boise, Idaho
TDI
2013 Passat TDI, Newmar Ventana 8.3L ISC 3945, 2016 E250 BT, 2000 Jetta TDI
acutally if you EDM and nozzle to nozzle deviation is below +-2% and shot to shot for small (pilot) quantities are below +-0.5 mg/stroke and one takes the time and effort to the make a corresponding duration map the deal is sealed. even without IMA codes after some time the ecu will learn the small quantities away. there might be more combustion noise from less accurate pilot quantities but as Hannu said 300 hp 4 cyl and 450hp v6 give or take is about the max one can run with 2000 bar. after that the egt/lambda tradeoff is quite bad
Try 10%....

Raw (Advanced) EDM is never used on quality nozzles, flow setting is required to achieve less than 2% flow tolerance.

Hole size alone never determines flow volume...
 
Last edited:

Drivbiwire

Zehntes Jahr der Veteran
Joined
Oct 13, 1998
Location
Boise, Idaho
TDI
2013 Passat TDI, Newmar Ventana 8.3L ISC 3945, 2016 E250 BT, 2000 Jetta TDI
I think we just need to know how the calibration codes are developed and the rest will become academic. Knowing the meaning of the characters in the codes would be the first step. There has to be a value for output vs voltage/current and a value for response time. If you have a reasonably well balanced set of nozzles it should be no problem to check cylinder contribution values with VCDS and adjust the IMA/IVA values to compensate for the deviation (within reason).

I doubt that the bosch dealers even know how the IMA/IVA values are derived. I imagine the test rig just spits them out and they tag the injectors. Lets face it, those guys are not all scientists. You're probably not going to get the info from them in any case, so you best start figuring it out yourself...

Given that the CR tdi uses cylinder pressure sensors and has a crank position sensor with sufficient resolution to determine individual cylinder firing acceleration, I doubt that there would be any need for the ECM to "age" the injector response and output values. It should be possible to derive trims for those values based on sensor inputs. I would expect the "learned" values lost by re-adapting the values would be "re-learned" within a short time. Open loop time/cycle count adaptation of that nature is a terrible "loose end" in any control scheme. Unless those piezo elements are predictable like nothing else on earth, I cannot imagine bosch making an error in judgement of that magnitude. Anything is possible though...

In any case, I have swapped used CR injectors and entered the codes with no ill effects. I imagine that the aging process is consistent anyway, so swapping full sets of 100k nozzles for sets of 50k ones would just result in a slightly lower or higher output across the board and would not likely result in any significant running problems.

I say go ahead and stick on some extrude honed or EDM enlarged nozzles and see what happens. Worst case, you lose an engine. If you are topping 300hp, that's pretty much expected. Nobody is going to die. Best case you have some new info to share with us...:D
Re-read my post, IMA coding is a trim value, it does not determine min/max injector control ranges.

And yes, for those of us working on CR systems we do know exactly how they are generated and determined.

My shop has a dedicated station for generating the coding for Bosch, Siemens, Delphi, CAT, and Denso systems.
 
Last edited:

Rub87

Top Post Dawg
Joined
Dec 10, 2006
Location
Belgium
TDI
Ibiza '99 90HP
Does your shop also have the ability to generate a duration map? Normally Bosch supplies it and then it gets fine tuned on the engine as well as the pressure wave correction.

It doesn't need to be in the accuracy magnitude of stock, but lets say if the pilot and low load area below 5mg is covered with good accuracy and some well thought points in the useable map range it would be good.
 

Drivbiwire

Zehntes Jahr der Veteran
Joined
Oct 13, 1998
Location
Boise, Idaho
TDI
2013 Passat TDI, Newmar Ventana 8.3L ISC 3945, 2016 E250 BT, 2000 Jetta TDI
We get updates for every CR/CRI injector made, the data for each of those part numbers lists the control ranges and tolerances.

If an injector falls outside that range, we can make adjustments to get it back into that required range and then re-assign a new IMA code for the new trim characteristics.

If you are making a performance CR injector, pick the nozzle, we would plot the min max ranges for the multiple test steps and then in conjunction with the tuner generate the ECU coding. Then with each subsequent injector assign IMA coding so that all future injectors function identically.

We can plot full graphs automatically, plug in the injector, grab a coffee and there you have it.

Taking it a step further we can also analyze raw nozzle flow as produced, so we can check actual production tolerances without variables associated with the nozzle holder injecting.


 
Last edited:

CNGVW

Vendor , w/Business number
Joined
Sep 15, 2009
Location
Bob Mann Auto, 111 High St, Pembroke MA 02359
TDI
Many TDI Jettas and a Beetle Race car run 2010 jetta tdi cup car build roadrace
If you need to push it on fuel added you should take a look at what I do with CNG.
the red is with just 20% added on the ALH

here is what I run as a system

I have track tested and Dyno all last year with it and it is very controllable.
So till the injectors are up to the grade you wish for this will get you there.
 

Rub87

Top Post Dawg
Joined
Dec 10, 2006
Location
Belgium
TDI
Ibiza '99 90HP
cng can knock if you inject too much, it doesn't burn rich but does burn quite easily lean, so lets say if you come below lambda ~2 it might ignite somewhere far before TDC and create unseen level of stress. for moderate gains although as you pointed out its a good fuel although it might be hard to control the rate of energy release.

now gets back to commonrail nozzles. DBW, could you maybe find in the Bosch catalog a nozzle with similar length (few thenths difference could be compensated with washer thickness) and spray angle that flows a considerable amount more? this would do away all the concern caused by nozzle to nozzle deviation when modifying a existing nozzle.
 
Top