D24T.com  

Go Back   D24T.com > Technical Discussion Area > Diesel Engine and Drivetrain

Reply
 
Thread Tools Display Modes
  #1  
Old 08-11-2020, 01:32 PM
ngoma ngoma is offline
Supporting Member
 
Join Date: Nov 2009
Posts: 1,351
Default Injection Pump Pressure Control Valve and Varying Amounts of Electronic Control

Thread Title: Discussion: Injection Pump Pressure Control Valve and Varying Amounts of Electronic Control

This thread is a breakaway discussion of a subject that came up in another topic. Instead of hijacking the other thread (Exhaust leak, loosing boost) we have moved the posts pertaining to this side topic into its own thread.

Here we pick up the beginnings of the discussion:

Quote:
Originally Posted by v8volvo View Post
And one other thing to not forget about is the internal pressure relief valve, screwed into the pump at the drive end from the top... No big deal to fix it, you just carefully rescue all the pieces from the hole and reassemble the valve (being careful to let no dirt inside!) and reinstall. If the basic checks don't give you any progress I would check this next.
Isn't a pressure gauge (and a way to plumb it in before the OUT bolt restrictor orifice) required to reset the internal pressure to spec?

IIRC, it's the IP internal pressure (that this PRV manages) that controls the IP dynamic timing advance.
__________________
1985 744 gle d24t
1985 745 gle d24t
Reply With Quote
  #2  
Old 08-12-2020, 09:11 PM
v8volvo v8volvo is offline
Supporting Member
 
Join Date: Jun 2009
Location: Montana, USA
Vehicle: '86 745, '83 764
Posts: 1,618
Default

Quote:
Originally Posted by ngoma View Post
Isn't a pressure gauge (and a way to plumb it in before the OUT bolt restrictor orifice) required to reset the internal pressure to spec?

IIRC, it's the IP internal pressure (that this PRV manages) that controls the IP dynamic timing advance.
True for sure. Ideally on a calibration bench I suppose.

But if the valve is completely sitting in pieces and doing nothing at all, at least reassembling it is better than nothing I think. Then it can at least have some regulating effect on internal pressure. If the spring and piston are just floating around loose down there in the well then the pressure has to be way off.

In truth I have never known this to happen on any VE pump EXCEPT in TDI engines where for some reason it happens all the time. Maybe the combination of more severe engine vibration and higher pump pressures in those conspires to pull the valves apart more there. But I believe it's theoretically possible in any pump like this. Something basic like a clogged filter still seems far more likely as a cause for driveability symptoms though.

On TDI engines I have always figured you can more or less get away with just slapping that regulator valve back together if it pops apart and not worrying too much about calibration, since they have their feedback timing control capability meaning that the mechanical systems (regulated by internal pressure) only have to get the timing curve reasonably close, then the computer can do the rest. Similar to how they are much more forgiving of sloppily set static injection timing as well. If it gets out of range then a fault code will sound the alarm, too, so there are guardrails to ensure you got it where it needed to be and it generally seems to work out no problem. In short they make it easier to cheat and get away with it.

You're right that conversely with these fully mechanical engines, none of that happens, so like with the pump timing, the pressure curve does need to be dead on and maybe set with a gauge. I hadn't really thought that through until you pointed it out.
__________________
86 745 D24T/ZF 345k lifted 2.5"
83 764 D24T/M46 155k
Reply With Quote
  #3  
Old 08-13-2020, 10:29 AM
ngoma ngoma is offline
Supporting Member
 
Join Date: Nov 2009
Posts: 1,351
Default

Quote:
Originally Posted by v8volvo View Post
In truth I have never known this to happen on any VE pump EXCEPT in TDI engines where for some reason it happens all the time.
Yes agree, not likely unless the pin has been tapped down a bit too far by someone attempting to adjust.

There is experiential evidence that indicates the dynamic timing action weakens over time (internal vane pump wear, general wear, lower modern ULSD viscosity) and therefore, older IPs need Pressure Control Valve recalibration to restore the desired dynamic timing action.

In addition to plumbing in a pressure gauge, one can measure the fuel output quantity from the IP OUT bolt at specific RPMs over time. Careful not to confuse IP RPMs with engine RPMs.

Plenty of interesting reading here:

internal injection pump pressure

Sorry for hijacking this thread. If we continue this topic we should start a new one.
__________________
1985 744 gle d24t
1985 745 gle d24t
Reply With Quote
  #4  
Old 08-14-2020, 07:49 AM
v8volvo v8volvo is offline
Supporting Member
 
Join Date: Jun 2009
Location: Montana, USA
Vehicle: '86 745, '83 764
Posts: 1,618
Default

Quote:
Originally Posted by ngoma View Post
There is experiential evidence that indicates the dynamic timing action weakens over time (internal vane pump wear, general wear, lower modern ULSD viscosity) and therefore, older IPs need Pressure Control Valve recalibration to restore the desired dynamic timing action.
Interesting points and discussion on that other thread. Maybe this accounts for why putting on a freshly rebuilt and calibrated IP from a professional fuel shop on these old mechanical IDI engines often seems to really transform the engine, makes it run cleaner and stronger, etc. While with the electronic feedback control motors like TDIs it makes no noticeable difference I have ever seen in how the engine runs with a new pump vs a quarter million mile pump, unless the old IP was seriously broken in some way.

The slow, gradual internal wear over time makes sense as a reason. Experimenting with a pressure gauge or return volume metering system would be very interesting, especially if also trying some fuels of differing viscosity and lubricity too (like ULSD vs biodiesel).

One thing I have imagined would be fun to attempt as a tuning and calibration technique sometime would be to find a way to install an injector needle lift sensor and crankshaft position sensor from a TDI engine, then get an ECU powered up and running on a bare-bones level and the OBD port wired in. With those pieces you would be able to read actual start-of-injection timing and RPM to see what the real world timing curve is, and adjust accordingly. This same method is often used by the guys doing M-TDI conversions to get their mechanical pumps dialed in to mimic what the electronic pumps achieve with computer control, though easier for them since they're starting with an engine that already has the sensors built in and ready to use. I have to imagine it would be more accurate than any other method, since using actual injector opening moment as the timing reference accounts for wear of all the parts in the system that could affect timing, not just the advance piston and vane pump and PRV, etc, but also the head/rotor, injector nozzle, etc.

Of course even more interesting would be to just put a full electronic control system on, with drive by wire pedal and all, and see what kind of performance gains could be had on an old IDI turbo engine simply by optimizing timing and injection quantity with feedback control, plus response to barometric pressure and intake air/coolant temp, etc, like the newer stuff does. But part of the appeal of these engines is their simplicity and non-reliance on computerized systems, so it's hard to see ever really going there.

You're right that we're hijacking RedArrow's thread here. Maybe we should move these posts to a new thread.
__________________
86 745 D24T/ZF 345k lifted 2.5"
83 764 D24T/M46 155k
Reply With Quote
  #5  
Old 08-14-2020, 11:58 AM
ngoma ngoma is offline
Supporting Member
 
Join Date: Nov 2009
Posts: 1,351
Default

Quote:
Originally Posted by v8volvo View Post
One thing I have imagined would be fun to attempt as a tuning and calibration technique sometime would be to find a way to install an injector needle lift sensor and crankshaft position sensor from a TDI engine, then get an ECU powered up and running on a bare-bones level and the OBD port wired in. With those pieces you would be able to read actual start-of-injection timing and RPM to see what the real world timing curve is, and adjust accordingly. This same method is often used by the guys doing M-TDI conversions to get their mechanical pumps dialed in to mimic what the electronic pumps achieve with computer control, though easier for them since they're starting with an engine that already has the sensors built in and ready to use. I have to imagine it would be more accurate than any other method...
Would the OBD provide a way to connect a VAG-COM to produce a timing curve graph? Displayed on a laptop screen in real time?

Maybe a simpler but sufficient method would be using one of those Snap-on diesel timing kits that use a sonic wave sensor that clamps onto the injector hardline to signal what looks like a common timing light?
__________________
1985 744 gle d24t
1985 745 gle d24t
Reply With Quote
  #6  
Old 08-15-2020, 06:38 AM
v8volvo v8volvo is offline
Supporting Member
 
Join Date: Jun 2009
Location: Montana, USA
Vehicle: '86 745, '83 764
Posts: 1,618
Default

Thanks for moving these, this is an interesting theoretical discussion that deserves its own thread.

Quote:
Originally Posted by ngoma View Post
Would the OBD provide a way to connect a VAG-COM to produce a timing curve graph? Displayed on a laptop screen in real time?
Exactly yes, that's what the mtdi guys are doing. Like what you can do with a full electronic engine, it can do with a mechanical one in terms of data acquisition, even without the computer calling the shots. And you would be able to watch the effect of all the variables also as you played with them -- engine/fuel temperature, playing with smoke screw, playing with IP case pressure, advance piston shims, governor changes, different injector opening pressures, different head and rotor size, etc.

The computer would register all kinds of diagnostic faults also of course, due to the other missing sensors and its inability to control any outputs, but with just the inputs of injector needle lift sensor and crankshaft position sensor it's known (from others' experience doing this process for their M-TDI conversions) that the OBD will still happily read out timing and RPM information.

What those guys then do is take readings of timing curve from their M-TDI engine running on its mechanical pump and compare them to a log file of what an electronic engine would be trying to achieve at the same RPM and load. They then manipulate the static and dynamic settings of their M-TDI configuration to mimic what the computer tries to do, or at least get reasonably close.

For us, the trick would be that we wouldn't have access to a "standard" map to compare to, for these old mechanical engines that never used electronic control. So in theory, in order to establish our own benchmark, I suppose we would have to first get a known perfectly running, freshly rebuilt and calibrated IP installed on a known good condition engine, along with a set of brand new freshly rebuilt injectors calibrated exactly to the factory pop pressure, and then take a full library of timing curve readings from that engine under all the various relevant conditions of temperature, fuel type, etc.

Then once we had that benchmark reference data, we would in principle be able to make running adjustments to our old, worn IP's if we could take timing curve readings from them and play with internal pressure to get them to match up as close as we could to the benchmark. Of course to do so we'd have to get all the timing and communications equipment set up there too.

Sounds like a lot of work and expense. We are probably better off just getting an IP properly rebuilt and professionally calibrated once in a blue moon.

Quote:
Maybe a simpler but sufficient method would be using one of those Snap-on diesel timing kits that use a sonic wave sensor that clamps onto the injector hardline to signal what looks like a common timing light?
Interesting idea, that would certainly be easier to set up, though would lack the ability to give real time communication when driving under load since you'd have to be standing there under the hood holding the light. Maybe could still get close enough though? Or maybe there would be someone clever enough to transform the signal into something that could be used an input for a data log?

I think (?) some of the bench calibration process for setting internal pressures occurs with the pump at a variety of load/injected quantity settings. Maintaining a stable pump RPM to take readings at anything other than idle fuel quantity of course would be pretty tough without being able to drive the car and apply a load to the engine. Maybe having it running on a chassis dynamometer would do the trick though?
__________________
86 745 D24T/ZF 345k lifted 2.5"
83 764 D24T/M46 155k
Reply With Quote
  #7  
Old 08-15-2020, 10:25 AM
ngoma ngoma is offline
Supporting Member
 
Join Date: Nov 2009
Posts: 1,351
Default

Quote:
Originally Posted by v8volvo View Post
For us, the trick would be that we wouldn't have access to a "standard" map to compare to, for these old mechanical engines that never used electronic control.
Might it be possible to extrapolate 4-cyl readings to a 6-cyl theoretical? Maybe not the turbochargers are too different.

Quote:
Originally Posted by v8volvo View Post
Then once we had that benchmark reference data, we would in principle be able to make running adjustments to our old, worn IP's if we could take timing curve readings from them and play with internal pressure to get them to match up as close as we could to the benchmark.
I had an idea for the Pressure Control Valve-- devise a fine thread thumbscrew in the top of it that alters the position of the plunger. Might even be able to run it to the dash panel using a mechanical speedo cable.


__________________
1985 744 gle d24t
1985 745 gle d24t
Reply With Quote
  #8  
Old 08-15-2020, 10:44 AM
v8volvo v8volvo is offline
Supporting Member
 
Join Date: Jun 2009
Location: Montana, USA
Vehicle: '86 745, '83 764
Posts: 1,618
Default

Quote:
Originally Posted by ngoma View Post
Might it be possible to extrapolate 4-cyl readings to a 6-cyl theoretical? Maybe not the turbochargers are too different.
Also I think there may be some meaningful differences in the timing curves required by direct injection vs prechamber injection engines. All the electronic control engines for which existing timing maps can be found are DI since to my knowledge electronic controls were never applied to the old IDI engines (at least for VW -- maybe timing data for an old GM IDI 6.5 TD would be interesting to compare?).

But maybe another option could be to get timing calibration book data from a Bosch service spec for the IDI pumps and build a map from that. I think those give a lot of detail for advance piston displacement at different RPM and load settings and from those data points it might be possible to interpolate and create a decent curve.


Quote:
I had an idea for the Pressure Control Valve-- devise a fine thread thumbscrew in the top of it that alters the position of the plunger. Might even be able to run it to the dash panel using a mechanical speedo cable.


That is a clever idea. It would be something that would basically adjust the preload on the spring inside the valve? Doesn't seem like it would be particularly hard to machine from a spare valve, no? Just drill out the upper center and cut some fine threads into it? Sealing up the thumbscrew might take some creativity I guess.
__________________
86 745 D24T/ZF 345k lifted 2.5"
83 764 D24T/M46 155k
Reply With Quote
  #9  
Old 08-16-2020, 06:41 PM
ngoma ngoma is offline
Supporting Member
 
Join Date: Nov 2009
Posts: 1,351
Default

Yes that requires more creative thought. Also, it would be better to be able to pull the plunger back up (retract) remotely.
__________________
1985 744 gle d24t
1985 745 gle d24t
Reply With Quote
  #10  
Old 09-26-2020, 08:02 AM
v8volvo v8volvo is offline
Supporting Member
 
Join Date: Jun 2009
Location: Montana, USA
Vehicle: '86 745, '83 764
Posts: 1,618
Default

I wanted to share this since it's funny we were just talking about it here recently.

I haven't seen one of these pressure relief valves fall apart in years, but recently I noticed the TDI engine in my Toyota pickup truck was running a little funny, seemed to be smoking more at high RPM and driveability was a little less smooth, just generally seemed off.

After checking other more common causes first (fuel filter, air filter, in-tank fuel lift pump in the particular case of this vehicle), I pulled the pressure relief valve out. The pictures attached show what I found. That's the valve body empty, with the spring, piston, and retainer all having fallen out of it and floating around loose in the well. I had to fish them out with a (clean!!) magnet and put the valve back together. I put on new O-rings too for good luck.

Again for some reason this seems to be a commonly seen problem on TDI engines with the VE pump, yet does not seem to happen to the injection pumps on the older IDI Volvo engines nearly as often. Hard to understand why. Maybe differences in severity or type of vibration, since the vibration is both different in character and also generally far more pronounced on the 4-cylinder direct injection engines than on the comparatively glass-smooth IDI 6-cylinder D24/T. Maybe something about the way the TDI vibrates happens to be a natural frequency for the parts of the valve, and slowly over time the retainer gets pounded out?

Anyway, just a curiosity, and maybe something to check on a D24/T every once in a while if ever chasing a driveability issue where looking at the other usual causes doesn't seem to be giving any progress. I had not checked the valve on this IP yet and had been meaning to anyway, so this was good timing. After putting it back together it runs quite a bit better than it has in a while, so the valve may have been working its way apart for some time.
Attached Images
File Type: jpg IMG_20200923_192324.jpg (302.0 KB, 5 views)
File Type: jpg IMG_20200923_192336.jpg (239.5 KB, 6 views)
File Type: jpg IMG_20200923_193342.jpg (609.2 KB, 3 views)
File Type: jpg IMG_20200923_193349.jpg (274.6 KB, 3 views)
__________________
86 745 D24T/ZF 345k lifted 2.5"
83 764 D24T/M46 155k
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -8. The time now is 01:10 AM.


Powered by vBulletin® Version 3.8.10
Copyright ©2000 - 2024, vBulletin Solutions, Inc.