The process is a straight forward one:

1. check your timing of the valve (valve and installation data necessary)
2. decide a rational EOI. A very conservative EOI would be scavenging TDC, a more progressive one 90° after scavenging TDC
3. decide EOI control in the SW of your ECU
4. a hint for the load and engine speed relation: the lower, the more advanced for both
5. run a WOT 1500-redline and an mix of dynamic and stationary (10 s) drive around and log it
6. check your log data to find out if lambda did change somewhere,
   • if, find out why
   • if not, retard EOI
7. compare the WOT curves and fuel consumption on your variations and find the best set

That's it.

 

That's what I thought but I believe I saw something from lotuselise saying there not good. So wanted to confirm.

Exactly what @Lotus wrote...

...they are good injectors, but as pintle “single hole” injectors provide a poor spray pattern and inefficient fuel air mix creation. Combustion efficiency is not as high as with the finer spray multi hole injectors provide.

Warning! Following review might trigger your feelings ...that's why I call they the worst injector mis-modification. We didn't went through centuries developing fine droplet exhaling carburators with high homogenization rate to introduce fine and free dose-able EFI systems to go back to pre-1920'ies highly stratified fuel-air-mixtures. Beside that the O-rings, they come with, didn't fit well to seal. For me these are a waste of money, these guys get a D from me for modificating the superior BOSCH to functional crap. The cost me 7 hp at peak on my DAMPFHAMMER (attention, not comparable to OE injector location) compared to the RDX or EV14 injectors.

The RDX, the Magneti Marelli Piko, the BOSCH EV14 and other multi-hole injectors are worth to check in your basket.

 

Exactly what @Lotus wrote...



Warning! Following review might trigger your feelings ...that's why I call they the worst injector mis-modification. We didn't went through centuries developing fine droplet exhaling carburators with high homogenization rate to introduce fine and free dose-able EFI systems to go back to pre-1920'ies highly stratified fuel-air-mixtures. Beside that the O-rings, they come with, didn't fit well to seal. For me these are a waste of money, these guys get a D from me for modificating the superior BOSCH to functional crap. The cost me 7 hp at peak on my DAMPFHAMMER (attention, not comparable to OE injector location) compared to the RDX or EV14 injectors.

The RDX, the Magneti Marelli Piko, the BOSCH EV14 and other multi-hole injectors are worth to check in your basket.

Hey lotus, sorry to bother you. I've been reading the whole form of trying to find the right injector and the right Delta gamma angles that you guys are talking about. I'm not educated enough to really understand it. I was wondering if you can point me in the right direction for a good set of multi-hole injectors like the boss ev14s you were talking about for a semi-built setup. Or if you had a possible part number or link you could send me that was a stock RSX plug style injectors. Thanks in advance

 

they are good injectors, but as pintle “single hole” injectors provide a poor spray pattern and inefficient fuel air mix creation. Combustion efficiency is not as high as with the finer spray multi hole injectors provide.
These spray with multiple cones into the bifurcated intake port vs a single cone onto the divider.
Lower mpg and less power is the result.
This is the price you pay for the higher flow rate.
I run a set of ID1050x myself due to SC and e85.

If you don’t need the flow rate these injectors provide, you‘re better off with multihole injectors.
The 2003-2009 Prius, my daily, uses Denso 12 hole injectors to maximize efficiency and lowest emissions as homogenous mix offers best performance on both fronts.

 

I found Bosch ev14 600 cc 6 hole nozzle injectors for $53 a piece, way cheaper then ID and better spray pattern I assume.

 

Direct bolt up meaning no spacers required and they come with adapter harnesses for them no extra cost

 

Those are good and “drop in” however if you compare the spray pattern of that injector as a 6 hole injector, to a 12 hole one such as a 123 128 or 270 for example, you will see how the fuel spray pattern and atomization is not optimal for a head such as a K series one, in the OEM injector location. Not all of the Bosch EV14 injectors have the same spray pattern or cone formation at the same distance as each other. Bosch does not share this information, sometimes you can get it from motorsports if you are nice to them. If you put several of the Bosch EV14 injectors on a injector dyno rail and test them, and measure the spray pattern, width and length, you can find the ones that are optimal for the 4 valve head type, and or work it backwards and find the correct injector location relative to the valve based on spray pattern.

It’s another reason why single cone injectors on a 4 valve head with close proximity to the valves are never a good match. You force fuel onto the walls vs on the back of the valve Head. If you compare an ID650 to a 0258158123, the cone vs direct dual cone fine spray is vastly different and the dyno results are real, as well as the QTY of fuel used.

 

Those are good and “drop in” however if you compare the spray pattern of that injector as a 6 hole injector, to a 12 hole one such as a 123 128 or 270 for example, you will see how the fuel spray pattern and atomization is not optimal for a head such as a K series one, in the OEM injector location. Not all of the Bosch EV14 injectors have the same spray pattern or cone formation at the same distance as each other. Bosch does not share this information, sometimes you can get it from motorsports if you are nice to them. If you put several of the Bosch EV14 injectors on a injector dyno rail and test them, and measure the spray pattern, width and length, you can find the ones that are optimal for the 4 valve head type, and or work it backwards and find the correct injector location relative to the valve based on spray pattern.

It’s another reason why single cone injectors on a 4 valve head with close proximity to the valves are never a good match. You force fuel onto the walls vs on the back of the valve Head. If you compare an ID650 to a 0258158123, the cone vs direct dual cone fine spray is vastly different and the dyno results are real, as well as the QTY of fuel used.

Got a link to where you'd buy them from?

Be about perfect for when I'm around 280'ish whp.

 

Well, I decided to go with a set of the 0280158235 injectors. The retailer linked there offers them in +/-1% flow balanced form for a reasonable price, and the whole set is about $230 to my doorstep with that service included after using a coupon code that can be found on google.

A couple of independent places have published dead time data for these, and they're both way way closer than anything I've managed to find for my current RDX injectors so hopefully it's good enough. I attached the more comprehensive of the two for posterity, and the other is here.

-Matt

 

Do you need the dead times pressure referenced in the ecu if you're running fuel pressure regulator referenced to manifold pressure? You'll be operating the injectors at a fixed pressure drop through the entire operating range.

If fuel pressure is referenced to MAP, the diffferencial pressure between fuel rail and inlet is constant. This leads to the dead times being constant as well. Outside pressure is irrelevant for the injectors. What influences the dynamics is the force keeping them closed. And this directly proportional to the pressure difference.

So it should not require moire than a 1D table with dead times over voltage populated with the corresponding data from the ID website for that particular injector

I finally got around to hooking up the vacuum reference, and I'm kicking myself for not doing it earlier. All I did was hook up the hose and flip NSP from fixed pressure to manifold reference and drivability instantly improved and fueling accuracy is no worse than it was. If anything, during throttle tip in and low load it's better!

No idea when the injectors might arrive, but I'm looking forward to that even more now 🤣

--Matt

 

I finally got around to hooking up the vacuum reference, and I'm kicking myself for not doing it earlier.

The measurement chain and the injector duration spectra are proviting from the constant differential pressure over injector, referenced to MAP- and fuel distribution-side. I am looking forward to see your new injectors running. Do you a 1500-redline WOT to compare these two sets of injectors?

 

Great thing about this forum is if someone doesn't know they don't speculate. They let lotus or lotus Elise, answer. That's why there's great information here and not a lot of false information.

 

I got my car up and running on the 0280158235 injectors. It's like 30 degrees and snowing, so I didn't get further than a couple laps around the block roughing things in. The car starts/runs idles really well.

Tying back to an earlier discussion in this thread, it's cool to be able to see more specifically what's going on with this stuff in greater detail. I definitely have to correct the fuel mapping, but it's comforting to know that I can see a bunch more data with extra sensors on. It was just guessing before, but now it can know.

--Matt

 

Bosch 0 280 158 298 from Ballenger were direct bolt in. Also even though the website said they were 6 hole they were 16 hole version.

 

I've invested about an hour, but I couldn't find where your engine made the power between VTEC (5100 rpm) and 7000 rpm, I miss up to 100 Nm at 5100 rpm and still 8 Nm at 7000 rpm, beyond that its within 4 Nm. Following graph shows the pressure tracers at different locations at 8500 rpm:

Blue - exhaust at valve seat
Red - intake at valve seat
Green - cylinder pressure
Pink - collector exhaust
Black - intake at flange level
EVO = exhaust valve opening
EVC = exhaust valve closing
IVO, IVC accordingly for intake

The cause of your pressure oscillations is likely exhaust driven and communicates through scavenging phase into the intake phase and causes there a communicating chamber in the plenum.

Black - intake valve seat for reference
Red, Blue, Dark Green and Pink are plenum positions from left to right plus, pink is nearest to MAP sensor position
Orange, Light Green - intake path

Pink shows from, only there, 270° to 630 ° every 72° a little peak of around 20 to 80 mbar, which can be measured by the sensor as it is 2 to 8 kPa. Maybe not everyone of these, but the higher ones at least 3 events with more than 4 kPa, which is about 4 % of the MAP value (= around 100 kPa). Pink is a superpositioned wave like all others, but Pink is the one nearest to the MAP sensor location.

Now to my evaluation of that. The basis of the intake plenum pressure oscillation is induced by the resonance of the exhaust system, which communicates through the scavenging phase. The header is quite efficient, the scavenging phase is accordingly efficient too. The found superpositioned pressure wave at MAP sensor location shows an unregular frequency of at least 3 measureable peaks at 8500 rpm, which likely are shown only as one at 8500 rpm, as the pressure sensor isn't that fast. That would mean we should see an +- 6 kPa oscillation (peak to peak 12 kPa@8500 rpm). A measurement at 8500 rpm and 1000 Hz would validate it.

I will not simulate idle, as this is almost impossible without information about the combustion. I don't see actually the torque curve the dyno have shown, I also can't improve combustion that far to get in the near of this. There is something in the exhaust system, which I still have not transferred into my simulation, which contributes essential on the left corner of high speed cam (5100-6500 rpm). I can try to phase the exhaust more in, as actually exhaust cam center is set to an resulting 4.1 mm lift@TDC. If I retard it, I would improve the low end torque.

 

I've invested about an hour, but I couldn't find where your engine made the power between VTEC (5100 rpm) and 7000 rpm, I miss up to 100 Nm at 5100 rpm and still 8 Nm at 7000 rpm, beyond that its within 4 Nm. Following graph shows the pressure tracers at different locations at 8500 rpm:

View attachment 120675

Blue - exhaust at valve seat
Red - intake at valve seat
Green - cylinder pressure
Pink - collector exhaust
Black - intake at flange level
EVO = exhaust valve opening
EVC = exhaust valve closing
IVO, IVC accordingly for intake

The cause of your pressure oscillations is likely exhaust driven and communicates through scavenging phase into the intake phase and causes there a communicating chamber in the plenum.

View attachment 120676

Black - intake valve seat for reference
Red, Blue, Dark Green and Pink are plenum positions from left to right plus, pink is nearest to MAP sensor position
Orange, Light Green - intake path

Pink shows from, only there, 270° to 630 ° every 72° a little peak of around 20 to 80 mbar, which can be measured by the sensor as it is 2 to 8 kPa. Maybe not everyone of these, but the higher ones at least 3 events with more than 4 kPa, which is about 4 % of the MAP value (= around 100 kPa). Pink is a superpositioned wave like all others, but Pink is the one nearest to the MAP sensor location.

Now to my evaluation of that. The basis of the intake plenum pressure oscillation is induced by the resonance of the exhaust system, which communicates through the scavenging phase. The header is quite efficient, the scavenging phase is accordingly efficient too. The found superpositioned pressure wave at MAP sensor location shows an unregular frequency of at least 3 measureable peaks at 8500 rpm, which likely are shown only as one at 8500 rpm, as the pressure sensor isn't that fast. That would mean we should see an +- 6 kPa oscillation (peak to peak 12 kPa@8500 rpm). A measurement at 8500 rpm and 1000 Hz would validate it.

I will not simulate idle, as this is almost impossible without information about the combustion. I don't see actually the torque curve the dyno have shown, I also can't improve combustion that far to get in the near of this. There is something in the exhaust system, which I still have not transferred into my simulation, which contributes essential on the left corner of high speed cam (5100-6500 rpm). I can try to phase the exhaust more in, as actually exhaust cam center is set to an resulting 4.1 mm lift@TDC. If I retard it, I would improve the low end torque.

🤯 A lot of this went over my head, but wow!

The exhaust has 3 resonators and one Helmholtz resonator near the end. The first resonator is an
QTP AR3 muffler, next a Thrush 2421 glasspack, next the Helmholtz branches of at 90° right before the Flowmonster 12867-FM at the end.

I have a pull (a few, probably) to 9k with 1K sampling if you want to see it. I know you burned a lot of time, and it's interesting to see data I don't understand, but if there's a final "what on earth" answer hiding in there I'm happy to share. 😅

Here's a second data point on power output, from 2 3rd gear pulls passed into Virtual Dyno with an accurate vehicle weight on flat ground:

--Matt

 

Anything from injector dynamics. They're matched sets with dead times measured for each set.

 

I have no opinions since my experience is limited. My matched set of 1000cc injectors from DW seems to work well. Also EV14 injectors, I have just under 80k miles on them.

 

I am running 60lb Bosch 630cc

 

I like BOSCH as a brand. I recently purchased some of there windshield wipers!

 

Them 2 icon wiper blades cost as much as 2 injectors lol

 

I haven't been fortunate enough to work on a PFI only engine, only PFI+DI, and they only use PFI for 10-20% of the fuel at WOT. That said, except for a few very specific engines and circumstances, the port injectors are spraying ONLY while the intake valve is closed, they're timed to complete injecting just before the intake valve opens. The dual cone pattern is critical to distribute fuel onto the back of both valves equally and keep it off the port walls as much as possible. Fuel on the port wall causes all sorts of tranisent/event-to-event fueling problems. It was counter intuitive to me at first, but two main things are happening:

First, heat from combustion is going into the intake valves which have poor cooling from the small contact patch at the seat of materials that are durable, but that don't conduct heat very well. That heat goes into the fuel, helping vaporize it and simultaneously cooling the valve.

Second, the pressure in the combustion chamber at the end of the exhaust stroke is quite high (relative to the pressure at the intake port), and at/near TDC there is no piston movement to draw in fresh intake air. When the intake valve cracks open, that pressure blows the fuel off the back of the valve and into the intake port, helping to atomize it.

I've got some PFI videos looking at the back of the intake valves somewhere, I'll see if I can't get them uploaded later.

 

...except for a few very specific engines and circumstances, the port injectors are spraying ONLY while the intake valve is closed...

That's the task of the tuner to optimize this. Injector size, fuel pressure and EOI are the parameters to be optimized for an optimal mixture, at best injection location too. There is no cause to fill up the port with small injectors and retarded EOI's.

 

Full load, injection starts on closed valve and continues during intake stroke

Full injection on closed valve, backflow when it opens

 

Thanks for the link, @Andy Midas! It raises a concern to me, though, which is that I really want to have accurate as possible flow rate and dead time values for a range of fuel pressures. I'm intending to vacuum reference my fuel pressure, as well as to use the fuel pressure as an axis for flow rate and dead times, so having all of this data is critical.

I see that some injector manufacturers, like Injector Dynamics, supply robust documentation that answers all of my questions here. I don't see the same for the Bosch Motorsport injectors, though, which seems really weird for Bosch. Are they leaving the work of doing this measurement to third-parties to measure and supply? Are there other third-party suppliers of these injectors that provide that level of detail, but sell less huge injectors than ID1050x injectors (hopefully also at a lower price point)?

--Matt

 

Do you need the dead times pressure referenced in the ecu if you're running fuel pressure regulator referenced to manifold pressure? You'll be operating the injectors at a fixed pressure drop through the entire operating range.

 

Mmmm

 

No data that's why I said believe.

 

The BOSCH 0280158235 injector...

...no bad 100 % duty cycle spray pattern, if it is like this one.

 

Hopefully you mean not bad

 

Oh, thanks for the hint. Yes, I meant NOT bad in the meaning of looking interesting and promising. Depending on the engine size, fuel kind type and power level it opens up, due to the 780 ccm, the option to play with it as a sort of DI and PFI injector by variing the EOI timing and fuel pressure.

One note here, depending on the application, K-series are either more high engine speed related, but quite often just driven in a DD application, the effect of spray pattern, its timing and its duration the size and fuel pressure drop over the injector can make a difference in driving quality and fuel efficiency. What may help to improve power output up top may not help down low and especially in part load. We came out of an age where carburators ruled the engine world, which was most of the time a better fuel-air-mixture environment. With todays EFI systems, it is very easy to fall back into stone age of fueling (just remember single beam sprayers from ID and such kind of), implementing them into stock injector positions. Any change of that position can change the fuel-mixture-efficiency again, which is valid for a wide range of engine speeds and loads. That's why, anytime you read about the best injector, it may be worth to read also in which application, relative position to the air flow and distance to the intake valve, fuel type, fuel pressure, EOI, ..., as well as temperature environment it was measured.