Jump to content

jericho injector bench build


scooters

Recommended Posts

Hey Guys. I'm so new it's not even funny to old school. I thought id start a thread on a shed project that started out as something simple and has turned into a thing that is now taking over. 

All I wanted to do was test fuel injectors. I'm a mechanic by trade so I have the basics. I get that you can touch a injector on a 12v battery and get it to click. But what if we need to find out things about our injectors?

I could out source them for a quick service and drop $30 per injector for a quick buzz in a ultrasonic bath.... But I think we are only scratching the surface. 

What if we need to find out some data? I think these little clicking givers of fuel are more important than you think. 

 

for a start:

Battery offset 

standard dynamic service of injectors

deviation vs pulse width

injector latency

injector linearity

 

This will get us started. I am going to build a injector test rig that will do far more than a injector test machine. Im at a total loss why a quality test bench like a ASNU costs $12.000 kiwi. It's just out of control and will give no control over injector voltage. 

So with that for a intro i'll just leave a few pics and come back in the weekend.

 

This is the pic that started this all off. It's a big dollar bench that is run on a haltech. It has the grunt to handle injectors of 5000cc it's the balls if you run methanol or if you want to know about injectors. 

What I will build will be first of all the prototype and a proof of concept that I can develop out. 

many more pics to come. 

 

post-21390-0-65276100-1479981069_thumb.jpg

  • Like 7
Link to comment
Share on other sites

  • Replies 65
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

So big tubes eh? 

After much searching I found some tubes. It was nearly the end of the project being that the cost of a glass tube from here in NZ is very expensive....I did not want to use perspex or other plastic due to the fact the fuel will make it yellow off. 

It's about now I need to mention that when you build a test rig of any sort you need to get close to real world conditions. So there is no point shooting water (for example) Specific gravity and viscosity are critical not to skew test results. 

If you are going to do something meaningful might at least try. 

 

After much looking I found some tubes. I will make a bench that has 6 tubes. The tubes are 675mm tall X 75mm wall OD and the wall is 3mm thick. 

 

 

This is awesome! 

 

I built a test bench to measure deadtime and build a compensation table for non linear portion of injector spray.

 

However it was super ghetto compared to this haha.

I can't find the attach button and i'm late for work haha. 

Link to comment
Share on other sites

It's about now I need to mention that when you build a test rig of any sort you need to get close to real world conditions. 

Agreed, I decide to use as much of my car as possible for my test bench.

 

Cars fuel pump, cars fuel lines, cars fuel pressure regulator, cars injector wiring etc.

 

When you look at injector deadtime, people will measure at say 60psi and then drop injector down to 6 volts etc. 

 

However in my car, below 8-9 volts the fuel pump cant keep up so the fuel pressure drops in conjunction with deadtime increasing, so you get even less flow.

 

Although if I was measuring deadtimes through different pressures and had a fuel pressure sensor, then obviously you could work it that way too.

But doing it that way meant I could have a simple volts/deadtime table.

Link to comment
Share on other sites

On the subject of glass I might as well start with that. 

My tubes have a bigger diameter than the tubes in the first photo. Glass tube is a odd beast. Sometimes the smaller the more expensive it is. Also you buy glass tube length x wall thickness. So you can also do yourself in if your not careful. 

It's very important with a rig like this that both ends are parallel ground. If they are not you might as well just throw your glass tube at the floor. Parallel grinding does usually add to the cost. 

 

The reason that I picked tube and not a graduated cylinder is because once you start dealing with 2L of liquid decanting it becomes a issue after each test. It wont be long before my muppet fingers drop it. Also a hollow tube means I have provision for a drain. 

 

As I was saying to will the other night. China seams to value good relationship when dealing with the online market rather than the actual dollar figure. Things like "dear sir" and "please answer me" and "thank you" seam to go a very long way when dealing with them. 

 

So here is a pic of my glass. 

 

Final figures are:

 

675mm Length 

75mm OD 

3mm wall.

 

241116_1077_zpsoz0dyx1k.jpg

 

241116_1076_zpsd86qyajf.jpg

  • Like 2
Link to comment
Share on other sites

Now I have the pictures figured out I might as well go on to my top and bottom plates. 

This was tricky far to tricky for myself and a drill press. It's important for the next part before I go on that I have had outside help. Tony from Entech engine reconditioners gave me a huge hand. When I say huge he did it all for me. So I can't take any credit I just told him what I want and then a few days later I had the alloy back in my hand cut spot on. He is very clever. All cutting I understand was done on a Serdi. A serdi is a machine that is used for cutting valve seats on cylinder heads. 

 

I did buy the o rings first and then gave him a tube for a final test fit. But I think you guys will be able to see how it all ended up. Alloy is a pain it can chatter when being machined. So how it has turned out is nothing short of amazing. 

 

Tony has also built himself a head flow bench. I can't remember what it pulls but it's up there. If you want to talk race heads I can put you in touch with a very good engine builder who has a lot of passion it. 

 

anyway check this out! 

 

241116_1080_zpsopirxdin.jpg

 

241116_1078_zpshokxho9y.jpg

 

241116_1079_zpsexyajocl.jpg

 

Far out what is a better photo share site? 

Photo bucket is going to make me cranky it's so slow. 

  • Like 4
Link to comment
Share on other sites

Agreed, I decide to use as much of my car as possible for my test bench.

 

Cars fuel pump, cars fuel lines, cars fuel pressure regulator, cars injector wiring etc.

 

When you look at injector deadtime, people will measure at say 60psi and then drop injector down to 6 volts etc. 

 

However in my car, below 8-9 volts the fuel pump cant keep up so the fuel pressure drops in conjunction with deadtime increasing, so you get even less flow.

 

Although if I was measuring deadtimes through different pressures and had a fuel pressure sensor, then obviously you could work it that way too.

But doing it that way meant I could have a simple volts/deadtime table.

 

Yea I can agree with most of that. Like you say voltage jumps around all over the place. 8-16 volts would be the min max range. It has a drastic effect on injectors. As does fuel pressure. What I am making gives me total control over fuel pressure and voltage. That way when I put the injector in the test rig I can be sure

that the data I get is correct for characterising the injector. irrespective of the vehicle it's going to sit in. 

We know that there is lag time between a injector turning on and actually giving a linear amount of fuel. I don't like the term "dead time" the injector is not dead in fact the coil may be energised but the pintle may not have moved quite enough. I feel the best way to describe it is Latency. 

 

Did you get your figures by weight? Did you use different fuel pressures? 

Link to comment
Share on other sites

I am in the opposite frame of mind, I dont like the term latency at all because it just implies a delay in spraying, rather than "missing fuel"

This is why I dont like the idea of people trying to work out deadtimes using an oscilliscope or whatever... Who cares what the injector is doing electrically.

 

A little context for my scenario:

I decided to try a staged injection setup for my engine, so that's where it's got primary injectors in normal place down by the head. 

Then a second set of injectors further up the runners, and the ECU can phase between them.

So in other words it crosses over to the outer injectors at high rpm and load.

 

So my problem was that since I did not have accurate enough deadtime info for each set of injectors, phasing between them didnt work particularly well, especially if IAT compensation is applied or whatever. So it was very fiddly to tune.

 

Since this was a single use case for my car only, I tested them at the fuel pressure my car runs at, (Using the exact FPR my car runs, etc) Then used the Injector drivers, power supply, etc that the actual car uses.

 

More information is here, including my dodgy maths excel sheet if you're interested 

https://www.hpacademy.com/forum/general-tuning-discussion/show/calculating-deadtimes-with-a-link-g4-and-injector-test-bench

But the long and short of it is that with a few hours work I found out the deadtimes.
Then established where the non linear portion of injector flow is and built a compensation table for it. Including establishing minimum opening times to spray ANY fuel. (This is different to deadtime due to non linearity at low pw)

My rig was amazingly ghetto, using a plastic graduated cylinder, the fuel rails from my engine, and extension cord for injector signal and fuel supply/return. 

Yet it gave very clear results, since completing this the staged injection runs MUCH better, the car started idling a lot better and the auto tune function in the ECU started getting a tune a lot more accurate a lot quicker. 

It was a very worthwhile exercise.

  • Like 3
Link to comment
Share on other sites

I am in the opposite frame of mind, I dont like the term latency at all because it just implies a delay in spraying, rather than "missing fuel"

This is why I dont like the idea of people trying to work out deadtimes using an oscilliscope or whatever... Who cares what the injector is doing electrically.

 

A little context for my scenario:

I decided to try a staged injection setup for my engine, so that's where it's got primary injectors in normal place down by the head. 

Then a second set of injectors further up the runners, and the ECU can phase between them.

So in other words it crosses over to the outer injectors at high rpm and load.

 

So my problem was that since I did not have accurate enough deadtime info for each set of injectors, phasing between them didnt work particularly well, especially if IAT compensation is applied or whatever. So it was very fiddly to tune.

 

Since this was a single use case for my car only, I tested them at the fuel pressure my car runs at, (Using the exact FPR my car runs, etc) Then used the Injector drivers, power supply, etc that the actual car uses.

 

More information is here, including my dodgy maths excel sheet if you're interested 

https://www.hpacademy.com/forum/general-tuning-discussion/show/calculating-deadtimes-with-a-link-g4-and-injector-test-bench

But the long and short of it is that with a few hours work I found out the deadtimes.

Then established where the non linear portion of injector flow is and built a compensation table for it. Including establishing minimum opening times to spray ANY fuel. (This is different to deadtime due to non linearity at low pw)

My rig was amazingly ghetto, using a plastic graduated cylinder, the fuel rails from my engine, and extension cord for injector signal and fuel supply/return. 

Yet it gave very clear results, since completing this the staged injection runs MUCH better, the car started idling a lot better and the auto tune function in the ECU started getting a tune a lot more accurate a lot quicker. 

It was a very worthwhile exercise.

 

I didn't realise that it was you on that hpacadamy page. I have in fact trawled over this page a few different times. 

I'm glad we both agree that a scope on it's own is a bad way to characterise a injector. Hang time is just as important as the turn on time and needs to be included.

What is also great and I am still very much a noob... is the power of excel. And how once it's set right it's good for all the heavy lifting. 

 

In your case I can understand the frustration of trying to stage injectors when there is no data. They can set up the batt offset table by looking at the air fuel ratios at idle and changing the electrical loading conditions on the vehicle. As i'm sure you understand this gets them in the ball park 

and it's ruff enough to get a semi stable idle and most people are never going to find out how close it is until they get a flat battery.  :-) 

But to stage a second set of injectors in at a nominated boost point (or other) is tuff going as things are moving around to much for air fuels to be useful for finding out about injectors. 

 

It's all good stuff. Finding out how simple the humble injector isn't.

  • Like 2
Link to comment
Share on other sites

EvsyinB.jpg

 

Z2d1woP.jpg

 

Everything about the bench is on the + size. The hoses and fittings are all -8. It's over kill unless someone sends me a set of 2200cc injectors + it's never going to get close to maxing out the lines. 

I have spent a lot of time trying to work out how to go between top feed injectors and side feed. If I am going to go to all this trouble I might as well have it fully functioning. 

 

At the end of this... If it ever runs and checks out and does what I want it to with a fine degree of acuracy I will be putting the call out for you guys to send in your injectors so I can cross check the rig. 

You will end up with nice clean injectors and a detail sheet with all the info you could ever want on your injectors. 

 

so if You read this far keep it in mind. 

  • Like 2
Link to comment
Share on other sites

This looks so good man, I'm excited haha. 

Those are some big heffa tubes, sorry you may have mentioned but how are you going to measure the fuel? By weight? Or will you etch some marks on the tube for how many CCs.
 

In your case I can understand the frustration of trying to stage injectors when there is no data. They can set up the batt offset table by looking at the air fuel ratios at idle and changing the electrical loading conditions on the vehicle. As i'm sure you understand this gets them in the ball park 

and it's ruff enough to get a semi stable idle and most people are never going to find out how close it is until they get a flat battery.   :-)

But to stage a second set of injectors in at a nominated boost point (or other) is tuff going as things are moving around to much for air fuels to be useful for finding out about injectors. 

It still doesnt perfectly blend, like if I set a 30/70 injector ratio and get the AFR stable. Then change it to say 70/30. It will not be correct. 
I think part of the reason for this is the reason why staged injection works at all - The amount of fuel that gets atomized and properly combusted changes depending on injector placement which alters the AFR at the exhaust even if the total volume of fuel delivered was identical. 

Even just adjusting the injector timing changes the AFR, when you're only spraying for a very small duration with high CC injectors. (440 and 1000cc in my case)

  • Like 1
Link to comment
Share on other sites

This looks so good man, I'm excited haha. 

Those are some big heffa tubes, sorry you may have mentioned but how are you going to measure the fuel? By weight? Or will you etch some marks on the tube for how many CCs.

 

I'm real glad you enjoy the project. Ha yea no diddling around with 160cc tubes when you are running a 2200cc injector. My idea behind the tubes is the more times the injector can shoot the greater the accuracy when I crunch the numbers. I have no issues letting it bench for a few minutes

Secondly it becomes very clear to the eye when there is a fueling error with a injector. commencing demonstration.. Also I get to see the spray in a nice large area not a tiny piss pot cylinder. 

RsZJrOo.jpg

To work it all out if fairly straightforward. Firstly you find the radius Inside the tube & do a pi calculation and work out how many cc per mm the fluid climbs the tube. From there you stick a ruler on the base and measure how far up the fluid goes. 

Secondly add the cc amount caught in the bottom above the tap valve. The trigger unit for the injectors has a counter but i'll get to that later. 

I was wanting to use a flow meter. But I have since found them to have + or - x% error. They are also very expensive. So for now its a steel rule. 

I will also cross check my results by weight. When you are shooting for good data and at such small pulse widths 2ms and under by weight is really the only way to cross check results. 

 

 

 

 

 

It still doesnt perfectly blend, like if I set a 30/70 injector ratio and get the AFR stable. Then change it to say 70/30. It will not be correct. 

I think part of the reason for this is the reason why staged injection works at all - The amount of fuel that gets atomized and properly combusted changes depending on injector placement which alters the AFR at the exhaust even if the total volume of fuel delivered was identical. 

Even just adjusting the injector timing changes the AFR, when you're only spraying for a very small duration with high CC injectors. (440 and 1000cc in my case)

hmmmm.. I never thought of that... the secondary injectors sit in a different part of the manifold. 

I suspect spray pattern and atomisation will also be vastly different between injector?  

I really have not had much to do with staging injectors.. other than staging fuel pumps.. that is for fuel temp control more than any other purpose. 

Why run 3 pumps and boil your fuel with returning heat when 1 is plenty. 

Link to comment
Share on other sites

 They can set up the batt offset table by looking at the air fuel ratios at idle and changing the electrical loading conditions on the vehicle.

The reason I dont like this idea is that if you are spraying a lot of fuel through the injector (say 6ms) then the deadtime is only a small portion so you wont get it very accurate deadtime.

 

Then if you spray at a very low amount (say 1ms) then although your deadtime might now be say 50% or more of total spray time you are quite likely in the non linear region of injector flow. So the number you get wont be representative of the actual deadtime at high fuel flow.

 

Like looking at this injector if I exclude the deadtime it's not really linear until 2.2ms which is way above idle.

This is where taking the time with a test rig to run all of the different tests really earns its keep.

 

djek3iyq.qtj.jpg

Link to comment
Share on other sites

The reason I dont like this idea is that if you are spraying a lot of fuel through the injector (say 6ms) then the deadtime is only a small portion so you wont get it very accurate deadtime.

 

Then if you spray at a very low amount (say 1ms) then although your deadtime might now be say 50% or more of total spray time you are quite likely in the non linear region of injector flow. So the number you get wont be representative of the actual deadtime at high fuel flow.

 

Like looking at this injector if I exclude the deadtime it's not really linear until 2.2ms which is way above idle.

This is where taking the time with a test rig to run all of the different tests really earns its keep.

 

 

Clearly you understand all about it. it's kind of the whole point of the test rig. I will eventually get my hands on some injectors that have data. (like injector dynamics) then I will run them up and see if I can come up with the same figures as injector dynamics. 

That to me will be proof of concept and the rig then becomes useful to a handful of people that want to do things right. 

It's my gut feeling that everything in the red is about were most people that tune get to when dealing with injectors. It's very loose and there about way to go.

I really just want to give the most spot on data I can and supply the most info people could ever want  about their injectors. 

 

 

jOQHkO9.jpg

 

 

But at the moment it's the build phase and we are on the 

journey If you have any injectors floating around you want looked at for any reason id be only to happy to help anyone on old school for free. 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...