For those who use the E8 as a methanol/H20 injection nozzle driver.
03-17-08, 01:16 PM
#1
For those who use the E8 as a methanol/H20 injection nozzle driver.
I recently purchased a FJO AI kit and have it installed in the car. What makes it different then most AI kits is that I am using the Haltech E8 to deliver the injection. This is accomplished by a supplied FJO injector driver connected to the nozzle and associated pumps and lines. This was a simple hook up. Basically one PWM and your normal power and ground wires.
Connecting it this way I am using the haltech generic pulse dc output. So I have a 32x32 map of injection.
What I noticed is that I cannot get my FJO nozzle to reliable inject below 10%. The only way to accomplish this is to turn the duty cycle down from the recommended 30 to 50hz to about 10 to 15 hz. (Below 10% duty it does not even operate/inject anything)
Apparently the nozzle is not that linear below 10% but whats the difference if I cannot even get the nozzle to fire below this?
My questions is to anybody out there with experience (That is if anyone has tried it this way before) is what hz for the solenoid I should use.
Another question is it normal that at 40% duty output to have so much water come out or is it because I have the frequency set so low.
What I can tell you the noise coming out of the solenoids sounds like they are going to blow! They are loud.
Also firing the injectors without air flow through the piping makes it look like instead of misting its spitting. Must be all of the water collecting on the walls and then getting blown out?
Ian
I will be getting the car tuned in april. So I will have a comparision to my 408rwhp none AI.
Connecting it this way I am using the haltech generic pulse dc output. So I have a 32x32 map of injection.
What I noticed is that I cannot get my FJO nozzle to reliable inject below 10%. The only way to accomplish this is to turn the duty cycle down from the recommended 30 to 50hz to about 10 to 15 hz. (Below 10% duty it does not even operate/inject anything)
Apparently the nozzle is not that linear below 10% but whats the difference if I cannot even get the nozzle to fire below this?
My questions is to anybody out there with experience (That is if anyone has tried it this way before) is what hz for the solenoid I should use.
Another question is it normal that at 40% duty output to have so much water come out or is it because I have the frequency set so low.
What I can tell you the noise coming out of the solenoids sounds like they are going to blow! They are loud.
Also firing the injectors without air flow through the piping makes it look like instead of misting its spitting. Must be all of the water collecting on the walls and then getting blown out?
Ian
I will be getting the car tuned in april. So I will have a comparision to my 408rwhp none AI.
03-17-08, 01:50 PM
#2
registered user
iTrader: (1)
Join Date: Jan 2006
Location: San Diego
Posts: 2,469
Likes: 0
Received 0 Likes
on
0 Posts
I might well be the only (other) person doing this.
My recomendation is for you to set the freq at what they recomend, or between 30 and 50. Don't go to high, at an extremley high freq, a PWM will operate like a proprtional valve, and you don't really want that.
It will also tend to overheat.
Bassicly you want it in the range of a TB fuel injector freq.
Next, why does it matter if there is no fuel being injected below 10, or 15 or 20 or 30 etc % DC.
Even if your start duty cycle is 50%, you will can still have a perfectly steady increase in injected fuel between 50 and maybe 85%. With anything over 80 likley being allmost the same flow a 100% and anything under 50 kind of dribbling or not flowing at all.
Start the map slight above the point where fuel starts comming out steadily.
The FJO controller will likley be doing the same thing, its just not showing you that in its menu, basicly (eg.. when its showing you 1% is really at 51%, does this make sence?)
I would recomend that you get a large!! graduated cyl and measure the output of the setup at everything from the minimum usable, to 100% in 5% increments, and then plot it on a graph (excel is good for this). Use the haltech data logging to log the actual time that water was injected to make the CC calculation.
This will give you a basis for setting the system up.
My recomendation is for you to set the freq at what they recomend, or between 30 and 50. Don't go to high, at an extremley high freq, a PWM will operate like a proprtional valve, and you don't really want that.
It will also tend to overheat.
Bassicly you want it in the range of a TB fuel injector freq.
Next, why does it matter if there is no fuel being injected below 10, or 15 or 20 or 30 etc % DC.
Even if your start duty cycle is 50%, you will can still have a perfectly steady increase in injected fuel between 50 and maybe 85%. With anything over 80 likley being allmost the same flow a 100% and anything under 50 kind of dribbling or not flowing at all.
Start the map slight above the point where fuel starts comming out steadily.
The FJO controller will likley be doing the same thing, its just not showing you that in its menu, basicly (eg.. when its showing you 1% is really at 51%, does this make sence?)
I would recomend that you get a large!! graduated cyl and measure the output of the setup at everything from the minimum usable, to 100% in 5% increments, and then plot it on a graph (excel is good for this). Use the haltech data logging to log the actual time that water was injected to make the CC calculation.
This will give you a basis for setting the system up.
I recently purchased a FJO AI kit and have it installed in the car. What makes it different then most AI kits is that I am using the Haltech E8 to deliver the injection. This is accomplished by a supplied FJO injector driver connected to the nozzle and associated pumps and lines. This was a simple hook up. Basically one PWM and your normal power and ground wires.
Connecting it this way I am using the haltech generic pulse dc output. So I have a 32x32 map of injection.
What I noticed is that I cannot get my FJO nozzle to reliable inject below 10%. The only way to accomplish this is to turn the duty cycle down from the recommended 30 to 50hz to about 10 to 15 hz. (Below 10% duty it does not even operate/inject anything)
Apparently the nozzle is not that linear below 10% but whats the difference if I cannot even get the nozzle to fire below this?
My questions is to anybody out there with experience (That is if anyone has tried it this way before) is what hz for the solenoid I should use.
Another question is it normal that at 40% duty output to have so much water come out or is it because I have the frequency set so low.
What I can tell you the noise coming out of the solenoids sounds like they are going to blow! They are loud.
Also firing the injectors without air flow through the piping makes it look like instead of misting its spitting. Must be all of the water collecting on the walls and then getting blown out?
Ian
I will be getting the car tuned in april. So I will have a comparision to my 408rwhp none AI.
Connecting it this way I am using the haltech generic pulse dc output. So I have a 32x32 map of injection.
What I noticed is that I cannot get my FJO nozzle to reliable inject below 10%. The only way to accomplish this is to turn the duty cycle down from the recommended 30 to 50hz to about 10 to 15 hz. (Below 10% duty it does not even operate/inject anything)
Apparently the nozzle is not that linear below 10% but whats the difference if I cannot even get the nozzle to fire below this?
My questions is to anybody out there with experience (That is if anyone has tried it this way before) is what hz for the solenoid I should use.
Another question is it normal that at 40% duty output to have so much water come out or is it because I have the frequency set so low.
What I can tell you the noise coming out of the solenoids sounds like they are going to blow! They are loud.
Also firing the injectors without air flow through the piping makes it look like instead of misting its spitting. Must be all of the water collecting on the walls and then getting blown out?
Ian
I will be getting the car tuned in april. So I will have a comparision to my 408rwhp none AI.
03-17-08, 02:45 PM
#3
I might well be the only (other) person doing this.
My recomendation is for you to set the freq at what they recomend, or between 30 and 50. Don't go to high, at an extremley high freq, a PWM will operate like a proprtional valve, and you don't really want that.
It will also tend to overheat.
Bassicly you want it in the range of a TB fuel injector freq.
Next, why does it matter if there is no fuel being injected below 10, or 15 or 20 or 30 etc % DC.
Even if your start duty cycle is 50%, you will can still have a perfectly steady increase in injected fuel between 50 and maybe 85%. With anything over 80 likley being allmost the same flow a 100% and anything under 50 kind of dribbling or not flowing at all.
Start the map slight above the point where fuel starts comming out steadily.
The FJO controller will likley be doing the same thing, its just not showing you that in its menu, basicly (eg.. when its showing you 1% is really at 51%, does this make sence?)
I would recomend that you get a large!! graduated cyl and measure the output of the setup at everything from the minimum usable, to 100% in 5% increments, and then plot it on a graph (excel is good for this). Use the haltech data logging to log the actual time that water was injected to make the CC calculation.
This will give you a basis for setting the system up.
My recomendation is for you to set the freq at what they recomend, or between 30 and 50. Don't go to high, at an extremley high freq, a PWM will operate like a proprtional valve, and you don't really want that.
It will also tend to overheat.
Bassicly you want it in the range of a TB fuel injector freq.
Next, why does it matter if there is no fuel being injected below 10, or 15 or 20 or 30 etc % DC.
Even if your start duty cycle is 50%, you will can still have a perfectly steady increase in injected fuel between 50 and maybe 85%. With anything over 80 likley being allmost the same flow a 100% and anything under 50 kind of dribbling or not flowing at all.
Start the map slight above the point where fuel starts comming out steadily.
The FJO controller will likley be doing the same thing, its just not showing you that in its menu, basicly (eg.. when its showing you 1% is really at 51%, does this make sence?)
I would recomend that you get a large!! graduated cyl and measure the output of the setup at everything from the minimum usable, to 100% in 5% increments, and then plot it on a graph (excel is good for this). Use the haltech data logging to log the actual time that water was injected to make the CC calculation.
This will give you a basis for setting the system up.
Thanks for the reply slo.
I was just concerned that there was no flow below 10% duty cycle and maybe the tuner would think that he was injecting something and in fact was not. In fact I was hoping that there would be a fine mist at below 10% duty cycle. I have not seen any other AI kit so I am not sure what the spray patterns of other kits look like.
Can a tuner be that good that he can tell if a little bit of water is being injected?
I may stay around 30hz and see what that looks like. It doesn't take long for everything to get soaked under the engine bay at 35% duty cycle
I have a bucket under the FMIC tubing as its spraying but it still sprays around like crazy.I admire the fact that people are using AI. Its just another "thing" to be worried about on the car. LOL
03-17-08, 03:20 PM
#4
registered user
iTrader: (1)
Join Date: Jan 2006
Location: San Diego
Posts: 2,469
Likes: 0
Received 0 Likes
on
0 Posts
Just tell your tuner that the water or mix starts injecting at 10 or 15 or whatever percentage you begin getting a usable flow.
Mine doesn't start injecting usable flow untill 30% but is very progressive from 30% on up, so I start injecting around 30%.
Problem solved.
Mine doesn't start injecting usable flow untill 30% but is very progressive from 30% on up, so I start injecting around 30%.
Problem solved.
Thanks for the reply slo.
I was just concerned that there was no flow below 10% duty cycle and maybe the tuner would think that he was injecting something and in fact was not. In fact I was hoping that there would be a fine mist at below 10% duty cycle. I have not seen any other AI kit so I am not sure what the spray patterns of other kits look like.
Can a tuner be that good that he can tell if a little bit of water is being injected?
I may stay around 30hz and see what that looks like. It doesn't take long for everything to get soaked under the engine bay at 35% duty cycle I have a bucket under the FMIC tubing as its spraying but it still sprays around like crazy.
I admire the fact that people are using AI. Its just another "thing" to be worried about on the car. LOL
I was just concerned that there was no flow below 10% duty cycle and maybe the tuner would think that he was injecting something and in fact was not. In fact I was hoping that there would be a fine mist at below 10% duty cycle. I have not seen any other AI kit so I am not sure what the spray patterns of other kits look like.
Can a tuner be that good that he can tell if a little bit of water is being injected?
I may stay around 30hz and see what that looks like. It doesn't take long for everything to get soaked under the engine bay at 35% duty cycle
I have a bucket under the FMIC tubing as its spraying but it still sprays around like crazy.I admire the fact that people are using AI. Its just another "thing" to be worried about on the car. LOL
03-17-08, 07:04 PM
#5
FD pro licensed driver
iTrader: (3)
Join Date: Feb 2006
Location: Renton/Bellevue/Seattle WA
Posts: 2,897
Likes: 0
Received 2 Likes
on
2 Posts
Awesome guys. I can't wait to get my e11v2 to control my meth/water injection! I'm glad it is as easy as it should be! Is there any other tricks/tips you guys can think of?
Slo, was this the writeup on how to use a haltech to control meth injection that you were speaking of?
Thanks
~Tweak
Slo, was this the writeup on how to use a haltech to control meth injection that you were speaking of?
Thanks
~Tweak
03-17-08, 07:18 PM
#7
Awesome guys. I can't wait to get my e11v2 to control my meth/water injection! I'm glad it is as easy as it should be! Is there any other tricks/tips you guys can think of?
Slo, was this the writeup on how to use a haltech to control meth injection that you were speaking of?
Thanks
~Tweak
Slo, was this the writeup on how to use a haltech to control meth injection that you were speaking of?
Thanks
~Tweak
It was such an easy hook up. The hardest thing was to install the pump.
What you may have to do is get creative on some sort of safe function. IE switch maps when fluid is low.
Ian
Trending Topics
03-17-08, 08:18 PM
#8
registered user
iTrader: (1)
Join Date: Jan 2006
Location: San Diego
Posts: 2,469
Likes: 0
Received 0 Likes
on
0 Posts
Not exactly a write up but lots of info here:
https://www.rx7club.com/auxiliary-injection-173/cooling-mist-s-hsv-installed-working-off-haltech-pwm-716541/
It really is very easy... The bottom line is that you simply can't exceed around 1 amp from the haltech's PWM.
If you have an extra E11 injector output then the limit might be higher (call haltech and ask).
So you have 2 approaches, PWM the pump through a high speed relay, or PWM an HSV, either through a high speed relay/ injector driver or directly if its draw is low enough.
The HSV approach is the best way to go.
Its also possible to do variable water/alcohol injection without any electronic controls using a fuel system with a return and a variable rate fpr.
https://www.rx7club.com/auxiliary-injection-173/cooling-mist-s-hsv-installed-working-off-haltech-pwm-716541/
It really is very easy... The bottom line is that you simply can't exceed around 1 amp from the haltech's PWM.
If you have an extra E11 injector output then the limit might be higher (call haltech and ask).
So you have 2 approaches, PWM the pump through a high speed relay, or PWM an HSV, either through a high speed relay/ injector driver or directly if its draw is low enough.
The HSV approach is the best way to go.
Its also possible to do variable water/alcohol injection without any electronic controls using a fuel system with a return and a variable rate fpr.
Awesome guys. I can't wait to get my e11v2 to control my meth/water injection! I'm glad it is as easy as it should be! Is there any other tricks/tips you guys can think of?
Slo, was this the writeup on how to use a haltech to control meth injection that you were speaking of?
Thanks
~Tweak
Slo, was this the writeup on how to use a haltech to control meth injection that you were speaking of?
Thanks
~Tweak
Last edited by slo; 03-17-08 at 08:27 PM.
03-19-08, 07:28 AM
#10
Senior Member
Join Date: Sep 2006
Location: uk
Posts: 402
Likes: 0
Received 0 Likes
on
0 Posts
I recently purchased a FJO AI kit and have it installed in the car. What makes it different then most AI kits is that I am using the Haltech E8 to deliver the injection. This is accomplished by a supplied FJO injector driver connected to the nozzle and associated pumps and lines. This was a simple hook up. Basically one PWM and your normal power and ground wires.
Connecting it this way I am using the haltech generic pulse dc output. So I have a 32x32 map of injection.
What I noticed is that I cannot get my FJO nozzle to reliable inject below 10%. The only way to accomplish this is to turn the duty cycle down from the recommended 30 to 50hz to about 10 to 15 hz. (Below 10% duty it does not even operate/inject anything)
Apparently the nozzle is not that linear below 10% but whats the difference if I cannot even get the nozzle to fire below this?
My questions is to anybody out there with experience (That is if anyone has tried it this way before) is what hz for the solenoid I should use.
Another question is it normal that at 40% duty output to have so much water come out or is it because I have the frequency set so low.
What I can tell you the noise coming out of the solenoids sounds like they are going to blow! They are loud.
Also firing the injectors without air flow through the piping makes it look like instead of misting its spitting. Must be all of the water collecting on the walls and then getting blown out?
Ian
I will be getting the car tuned in april. So I will have a comparision to my 408rwhp none AI.
Connecting it this way I am using the haltech generic pulse dc output. So I have a 32x32 map of injection.
What I noticed is that I cannot get my FJO nozzle to reliable inject below 10%. The only way to accomplish this is to turn the duty cycle down from the recommended 30 to 50hz to about 10 to 15 hz. (Below 10% duty it does not even operate/inject anything)
Apparently the nozzle is not that linear below 10% but whats the difference if I cannot even get the nozzle to fire below this?
My questions is to anybody out there with experience (That is if anyone has tried it this way before) is what hz for the solenoid I should use.
Another question is it normal that at 40% duty output to have so much water come out or is it because I have the frequency set so low.
What I can tell you the noise coming out of the solenoids sounds like they are going to blow! They are loud.
Also firing the injectors without air flow through the piping makes it look like instead of misting its spitting. Must be all of the water collecting on the walls and then getting blown out?
Ian
I will be getting the car tuned in april. So I will have a comparision to my 408rwhp none AI.
I can help you a little.
Without knowing the actual opening time and shutting time of the FJO valve, it is not easy to translate the flow from the duty cycle % against a 32x32 map grid.
Based on the valve is opening reliably at 10% DC at 10Hz or 100mS cycling time. I can assume the valve needs a 10mS pulse to open it fully. Shutting time will be about the same normally.
Your real control mapping range at:
- At 10Hz: it will be 10% to 90%
- At 15Hz: it wil be 15% to 85%
- At 30Hz: it will be 30% to 70%
- at 40Hz: it will be 40% to 60%
- at 50Hz: it will be no control at all.
Overflow at 40%+ DC:
This can be a CV (flow factor) mis-match between the valve and the nozzle. If the maximum flow of the valve is much greater than the nozzle flow, you will "DC% vs actual flow" linearity
Loss of flow continuity (smoothness) due to slow reacting valve:
Stretching the reasoning further, let say you are pulsing at a much lower frequency of 0.5 Hz. Duty cycle is set to 50%.
The nozzle to give full flow for 1s and nothing for the next second. Although over time (minutes), the system will give you an accurate re-production of %DC. But meanwhile you engine will hate it. The slower the valve, the more abruptness to the flow.
What to do:
Your best bet is doing what slo mentioned, offset the DC% mapping point to suit or get a faster valve. Other minor problem such as the hystersis of the magnetic circuit affecting rise and decay of magnetic flux stored in the valve core. This effect is commonly known as B-H curve. If you are operating in a small section of the curve, as suggested above somewhere between 40%-60% DC (cycle time=40Hz), you can not going to get any accuracy at all or even seeing any form of control at all - ending up to be a on/off valve.
Last edited by Richard L; 03-19-08 at 07:44 AM.
03-19-08, 01:07 PM
#11
registered user
iTrader: (1)
Join Date: Jan 2006
Location: San Diego
Posts: 2,469
Likes: 0
Received 0 Likes
on
0 Posts
Very interesting.
I still think testing it with a graduated cylinder is the way to go.
I have noticed that PWM valves seem to need shorter pulses with increasing freq as they head tword being virtual proportional valves.
That might be due to the magnetic effect you describe below.
Why is it that the valve stops being controllable (per your chart below) when you have exactly the opening pulse (or closing pulse) time left.
I guess what I am saying is your not at 100% if the valve is opening an closing. You can still get some proportional control out of the valve when its not fully closing
Right?
I still think testing it with a graduated cylinder is the way to go.
I have noticed that PWM valves seem to need shorter pulses with increasing freq as they head tword being virtual proportional valves.
That might be due to the magnetic effect you describe below.
Why is it that the valve stops being controllable (per your chart below) when you have exactly the opening pulse (or closing pulse) time left.
I guess what I am saying is your not at 100% if the valve is opening an closing. You can still get some proportional control out of the valve when its not fully closing
Right?
Ian,
I can help you a little.
Without knowing the actual opening time and shutting time of the FJO valve, it is not easy to translate the flow from the duty cycle % against a 32x32 map grid.
Based on the valve is opening reliably at 10% DC at 10Hz or 100mS cycling time. I can assume the valve needs a 10mS pulse to open it fully. Shutting time will be about the same normally.
Your real control mapping range at:
- At 10Hz: it will be 10% to 90%
- At 15Hz: it wil be 15% to 85%
- At 30Hz: it will be 30% to 70%
- at 40Hz: it will be 40% to 60%
- at 50Hz: it will be no control at all.
Overflow at 40%+ DC:
This can be a CV (flow factor) mis-match between the valve and the nozzle. If the maximum flow of the valve is much greater than the nozzle flow, you will "DC% vs actual flow" linearity
Loss of flow continuity (smoothness) due to slow reacting valve:
Stretching the reasoning further, let say you are pulsing at a much lower frequency of 0.5 Hz. Duty cycle is set to 50%.
The nozzle to give full flow for 1s and nothing for the next second. Although over time (minutes), the system will give you an accurate re-production of %DC. But meanwhile you engine will hate it. The slower the valve, the more abruptness to the flow.
What to do:
Your best bet is doing what slo mentioned, offset the DC% mapping point to suit or get a faster valve. Other minor problem such as the hystersis of the magnetic circuit affecting rise and decay of magnetic flux stored in the valve core. This effect is commonly known as B-H curve. If you are operating in a small section of the curve, as suggested above somewhere between 40%-60% DC (cycle time=40Hz), you can not going to get any accuracy at all or even seeing any form of control at all - ending up to be a on/off valve.
I can help you a little.
Without knowing the actual opening time and shutting time of the FJO valve, it is not easy to translate the flow from the duty cycle % against a 32x32 map grid.
Based on the valve is opening reliably at 10% DC at 10Hz or 100mS cycling time. I can assume the valve needs a 10mS pulse to open it fully. Shutting time will be about the same normally.
Your real control mapping range at:
- At 10Hz: it will be 10% to 90%
- At 15Hz: it wil be 15% to 85%
- At 30Hz: it will be 30% to 70%
- at 40Hz: it will be 40% to 60%
- at 50Hz: it will be no control at all.
Overflow at 40%+ DC:
This can be a CV (flow factor) mis-match between the valve and the nozzle. If the maximum flow of the valve is much greater than the nozzle flow, you will "DC% vs actual flow" linearity
Loss of flow continuity (smoothness) due to slow reacting valve:
Stretching the reasoning further, let say you are pulsing at a much lower frequency of 0.5 Hz. Duty cycle is set to 50%.
The nozzle to give full flow for 1s and nothing for the next second. Although over time (minutes), the system will give you an accurate re-production of %DC. But meanwhile you engine will hate it. The slower the valve, the more abruptness to the flow.
What to do:
Your best bet is doing what slo mentioned, offset the DC% mapping point to suit or get a faster valve. Other minor problem such as the hystersis of the magnetic circuit affecting rise and decay of magnetic flux stored in the valve core. This effect is commonly known as B-H curve. If you are operating in a small section of the curve, as suggested above somewhere between 40%-60% DC (cycle time=40Hz), you can not going to get any accuracy at all or even seeing any form of control at all - ending up to be a on/off valve.
03-19-08, 04:47 PM
#12
Senior Member
Join Date: Sep 2006
Location: uk
Posts: 402
Likes: 0
Received 0 Likes
on
0 Posts
slo,
This is an interesting topic.
It appeared to me that more and more systems are advancing the electronics regardless of the mechanical performance. A fluid delievery is only as good as the limitation of the flow control valve. It is also very easy to market the "hi-tech" side of a system and people will assume the mechanical component to automatically keep up.
I have read a few posts here over the last few months, quite a few people here has fallen into this "hi-tech" trap. The end result fell way short of expectation. Putting aside the "high-defination" or "hi-resolution" controller for the time being - it is not the stumbling block.
Choosing the correct solenoid valve is the most important part of a delievery system. For example, a solenoid valve that can withhold very high pressure normally needs a very strong spring. It required a powerful coil to open the valve.
Downside is heat, powerful coil generates a great deal of heat. That is not all, it is also subject to repid valve piston-seal wear, regardless of the constuuction geometry of the moving parts - the constant hammer action during the on/off stroke.
Opening and shutting time is the next stage of consideration, based on the above requirement. selecting the valve material with the right magnetic peoperties and matching solenoid is crucial to the performance outcome of a solenoid valve.
Unlike the switching a resistor, the end result appling a electrical pulse to a coil winding + valve core is not as predictable.
In answer your question, the residual current decay after energisation, there is difference between a long pulse and a short pulse as wlll the the retracting piston inside the valve. Until you can manage the delay current properly, there is always a shift towards "slowness" as frequency increases (relatively).
New systems that comes into market seemed to be very vague on "%dc" vs "flow linearity".
As you have suggested a graduated cylinder is the way to go, trust no one.
This is an interesting topic.
It appeared to me that more and more systems are advancing the electronics regardless of the mechanical performance. A fluid delievery is only as good as the limitation of the flow control valve. It is also very easy to market the "hi-tech" side of a system and people will assume the mechanical component to automatically keep up.
I have read a few posts here over the last few months, quite a few people here has fallen into this "hi-tech" trap. The end result fell way short of expectation. Putting aside the "high-defination" or "hi-resolution" controller for the time being - it is not the stumbling block.
Choosing the correct solenoid valve is the most important part of a delievery system. For example, a solenoid valve that can withhold very high pressure normally needs a very strong spring. It required a powerful coil to open the valve.
Downside is heat, powerful coil generates a great deal of heat. That is not all, it is also subject to repid valve piston-seal wear, regardless of the constuuction geometry of the moving parts - the constant hammer action during the on/off stroke.
Opening and shutting time is the next stage of consideration, based on the above requirement. selecting the valve material with the right magnetic peoperties and matching solenoid is crucial to the performance outcome of a solenoid valve.
Unlike the switching a resistor, the end result appling a electrical pulse to a coil winding + valve core is not as predictable.
In answer your question, the residual current decay after energisation, there is difference between a long pulse and a short pulse as wlll the the retracting piston inside the valve. Until you can manage the delay current properly, there is always a shift towards "slowness" as frequency increases (relatively).
New systems that comes into market seemed to be very vague on "%dc" vs "flow linearity".
As you have suggested a graduated cylinder is the way to go, trust no one.
03-19-08, 05:05 PM
#13
registered user
iTrader: (1)
Join Date: Jan 2006
Location: San Diego
Posts: 2,469
Likes: 0
Received 0 Likes
on
0 Posts
The issue with magnetic components within any valve, is that they will rust given contact with water, mixtures of alcohol and water seem to be far less prone to rusting.
Even magnetic SS will rust in 100% water over time.
Even magnetic SS will rust in 100% water over time.
slo,
This is an interesting topic.
It appeared to me that more and more systems are advancing the electronics regardless of the mechanical performance. A fluid delievery is only as good as the limitation of the flow control valve. It is also very easy to market the "hi-tech" side of a system and people will assume the mechanical component to automatically keep up.
I have read a few posts here over the last few months, quite a few people here has fallen into this "hi-tech" trap. The end result fell way short of expectation. Putting aside the "high-defination" or "hi-resolution" controller for the time being - it is not the stumbling block.
Choosing the correct solenoid valve is the most important part of a delievery system. For example, a solenoid valve that can withhold very high pressure normally needs a very strong spring. It required a powerful coil to open the valve.
Downside is heat, powerful coil generates a great deal of heat. That is not all, it is also subject to repid valve piston-seal wear, regardless of the constuuction geometry of the moving parts - the constant hammer action during the on/off stroke.
Opening and shutting time is the next stage of consideration, based on the above requirement. selecting the valve material with the right magnetic peoperties and matching solenoid is crucial to the performance outcome of a solenoid valve.
Unlike the switching a resistor, the end result appling a electrical pulse to a coil winding + valve core is not as predictable.
In answer your question, the residual current decay after energisation, there is difference between a long pulse and a short pulse as wlll the the retracting piston inside the valve. Until you can manage the delay current properly, there is always a shift towards "slowness" as frequency increases (relatively).
New systems that comes into market seemed to be very vague on "%dc" vs "flow linearity".
As you have suggested a graduated cylinder is the way to go, trust no one.
This is an interesting topic.
It appeared to me that more and more systems are advancing the electronics regardless of the mechanical performance. A fluid delievery is only as good as the limitation of the flow control valve. It is also very easy to market the "hi-tech" side of a system and people will assume the mechanical component to automatically keep up.
I have read a few posts here over the last few months, quite a few people here has fallen into this "hi-tech" trap. The end result fell way short of expectation. Putting aside the "high-defination" or "hi-resolution" controller for the time being - it is not the stumbling block.
Choosing the correct solenoid valve is the most important part of a delievery system. For example, a solenoid valve that can withhold very high pressure normally needs a very strong spring. It required a powerful coil to open the valve.
Downside is heat, powerful coil generates a great deal of heat. That is not all, it is also subject to repid valve piston-seal wear, regardless of the constuuction geometry of the moving parts - the constant hammer action during the on/off stroke.
Opening and shutting time is the next stage of consideration, based on the above requirement. selecting the valve material with the right magnetic peoperties and matching solenoid is crucial to the performance outcome of a solenoid valve.
Unlike the switching a resistor, the end result appling a electrical pulse to a coil winding + valve core is not as predictable.
In answer your question, the residual current decay after energisation, there is difference between a long pulse and a short pulse as wlll the the retracting piston inside the valve. Until you can manage the delay current properly, there is always a shift towards "slowness" as frequency increases (relatively).
New systems that comes into market seemed to be very vague on "%dc" vs "flow linearity".
As you have suggested a graduated cylinder is the way to go, trust no one.
03-19-08, 05:22 PM
#14
Senior Member
Join Date: Sep 2006
Location: uk
Posts: 402
Likes: 0
Received 0 Likes
on
0 Posts
I agree, here is an experimnt I did last year.
I put the same clippard valve into a jar of M50/W50 and a jar of 100% water for over a month. The left valve with M50/W50 is unaffected!!!
I put the same clippard valve into a jar of M50/W50 and a jar of 100% water for over a month. The left valve with M50/W50 is unaffected!!!
03-19-08, 06:22 PM
#15
Ian,
I can help you a little.
Without knowing the actual opening time and shutting time of the FJO valve, it is not easy to translate the flow from the duty cycle % against a 32x32 map grid.
Based on the valve is opening reliably at 10% DC at 10Hz or 100mS cycling time. I can assume the valve needs a 10mS pulse to open it fully. Shutting time will be about the same normally.
Your real control mapping range at:
- At 10Hz: it will be 10% to 90%
- At 15Hz: it wil be 15% to 85%
- At 30Hz: it will be 30% to 70%
- at 40Hz: it will be 40% to 60%
- at 50Hz: it will be no control at all.
Overflow at 40%+ DC:
This can be a CV (flow factor) mis-match between the valve and the nozzle. If the maximum flow of the valve is much greater than the nozzle flow, you will "DC% vs actual flow" linearity
Loss of flow continuity (smoothness) due to slow reacting valve:
Stretching the reasoning further, let say you are pulsing at a much lower frequency of 0.5 Hz. Duty cycle is set to 50%.
The nozzle to give full flow for 1s and nothing for the next second. Although over time (minutes), the system will give you an accurate re-production of %DC. But meanwhile you engine will hate it. The slower the valve, the more abruptness to the flow.
What to do:
Your best bet is doing what slo mentioned, offset the DC% mapping point to suit or get a faster valve. Other minor problem such as the hystersis of the magnetic circuit affecting rise and decay of magnetic flux stored in the valve core. This effect is commonly known as B-H curve. If you are operating in a small section of the curve, as suggested above somewhere between 40%-60% DC (cycle time=40Hz), you can not going to get any accuracy at all or even seeing any form of control at all - ending up to be a on/off valve.
I can help you a little.
Without knowing the actual opening time and shutting time of the FJO valve, it is not easy to translate the flow from the duty cycle % against a 32x32 map grid.
Based on the valve is opening reliably at 10% DC at 10Hz or 100mS cycling time. I can assume the valve needs a 10mS pulse to open it fully. Shutting time will be about the same normally.
Your real control mapping range at:
- At 10Hz: it will be 10% to 90%
- At 15Hz: it wil be 15% to 85%
- At 30Hz: it will be 30% to 70%
- at 40Hz: it will be 40% to 60%
- at 50Hz: it will be no control at all.
Overflow at 40%+ DC:
This can be a CV (flow factor) mis-match between the valve and the nozzle. If the maximum flow of the valve is much greater than the nozzle flow, you will "DC% vs actual flow" linearity
Loss of flow continuity (smoothness) due to slow reacting valve:
Stretching the reasoning further, let say you are pulsing at a much lower frequency of 0.5 Hz. Duty cycle is set to 50%.
The nozzle to give full flow for 1s and nothing for the next second. Although over time (minutes), the system will give you an accurate re-production of %DC. But meanwhile you engine will hate it. The slower the valve, the more abruptness to the flow.
What to do:
Your best bet is doing what slo mentioned, offset the DC% mapping point to suit or get a faster valve. Other minor problem such as the hystersis of the magnetic circuit affecting rise and decay of magnetic flux stored in the valve core. This effect is commonly known as B-H curve. If you are operating in a small section of the curve, as suggested above somewhere between 40%-60% DC (cycle time=40Hz), you can not going to get any accuracy at all or even seeing any form of control at all - ending up to be a on/off valve.
It is to early to tell (for me)if the haltech fired AI is adequate enought compared to a full blown AI kit.
Good point on the high tech vs mechanical solenoids.
Thanks,
Ian
03-19-08, 07:23 PM
#16
Senior Member
Join Date: Sep 2006
Location: uk
Posts: 402
Likes: 0
Received 0 Likes
on
0 Posts
If you can measure the resistance of the valve coil. You can work out the current drain.
14.5V /coil resistance (ohm) = current (Amp). Look up teh haltech manual and see if it exceeds the output drive capability.
14.5V /coil resistance (ohm) = current (Amp). Look up teh haltech manual and see if it exceeds the output drive capability.
03-19-08, 08:09 PM
#18
registered user
iTrader: (1)
Join Date: Jan 2006
Location: San Diego
Posts: 2,469
Likes: 0
Received 0 Likes
on
0 Posts
As far as driving the valve, there isn't much the "'full blown" kits as you put it can offer ahead of the haltech
Basically all the have to offer (over what you have with the haltech) is corrections to the AI injection based on EGT or AIT or whatever.
I don't think those types of corrections are necessarily advantageous when it comes to AI on a system with a haltech where you can control fuel and ignition with all the corrections in the world, and control the AI based on load an RPM.
Basically all the have to offer (over what you have with the haltech) is corrections to the AI injection based on EGT or AIT or whatever.
I don't think those types of corrections are necessarily advantageous when it comes to AI on a system with a haltech where you can control fuel and ignition with all the corrections in the world, and control the AI based on load an RPM.
wow. Thanks for taking the time to respond. Not sure how you came up with all of this info but I am just about to read this http://forums.evolutionm.net/showthread.php?t=292953 to better understand what I am getting myself into. It]s funny that FJO suggested 30 to 50hz and I found that visually it looks like a lower frequency is best. I wonder if it has something to do with the way haltech pulses the signal.
It is to early to tell (for me)if the haltech fired AI is adequate enought compared to a full blown AI kit.
Good point on the high tech vs mechanical solenoids.
Thanks,
Ian
It is to early to tell (for me)if the haltech fired AI is adequate enought compared to a full blown AI kit.
Good point on the high tech vs mechanical solenoids.
Thanks,
Ian
03-20-08, 08:37 PM
#19
Well I did some more playing around on the AI kit today. Pretty stupid on my part the previous observations must have been with a weak battery. Since this time with a running car the nozzle and visual spray patterns changed significantly. I would have thought that a fresh charged battery would last longer the 1/2 hour.
Just goes to show you how much voltage plays a part on pump pressure and spray performance.
It even seemed to effect the frequency of the solenoid or at the very least the ability of spray at lower duty cycles.
Just goes to show you how much voltage plays a part on pump pressure and spray performance.
It even seemed to effect the frequency of the solenoid or at the very least the ability of spray at lower duty cycles.
Thread
Thread Starter
Forum
Replies
Last Post
ls1swap
3rd Generation Specific (1993-2002)
12
10-01-15 07:58 PM
