Tweet
about the only performance application i know of where low impendance are really common is with the Sy/Ty crowd...
-
Rick...Originally posted by ollopa View PostWell there hasn't been any input on my end of the project, so I'll just check in with where I'm at and what I'm thinking. I just got a wideband O2 sensor and am in the midst of fine tuning a heavily modified engine for a 302->351 swap, so I have been a little distracted after all. I grabbed an injector and threw together a MOSFET driver for it. It works great, so I've been thinking about the digital drive. I know I can get 1ms resolution without any trouble, but I'd like to try and get 0.1ms or 0.5ms resolution for some rudimentary injector characterization.
Driving all the injectors simultaneously would be really harsh on the power supply, so I'd like to have some options for driving patterns: simultaneous, sequential, and bank (batch). Sounds a bit like EFI options
The trade off here is that the more injectors you drive at one time, the sturdier the power supply you need, but the fewer you drive at one time, the longer it will take overall to fire all the injectors. It's only software, so why not make all the options available and let the user decide, right?
So we'll be able to select the number of injectors connected from 1 to 8, and the drive pattern of sequential, simultaneous, and bank. Bank drive will have a couple different configurations depending on the number of injectors connected. For example, if you have injectors then you could fire four banks of two or two banks of four (2/2/2/2 or 4/4). If you have 6 injectors, you could configure banks of 2/2/2 or 3/3. Whatever your power supply can handle and whatever you personally like.
Duration will be programmable and will have a couple modes. You can either specify the number of pulses to send or go into continuous mode where you push a button to start sending pulses and it will keep sending until you push the button again to stop.
I'll post a video of the prototype this weekend. I haven't built a FrankInjector yet so I can't show injector spraying, but I'll be able to demo the user interface and let you listen to injectors clicking.
-Rick
Not being a Wiz with electrical devices... and in the wake of reading all your intuitive technical ideas on how to make your machine... I am having trouble visualizing what it will look like... But FWIW... I can help cut down on the amount of time to completion needed to get yours up and running by suggesting that you send me a working model...noting that there is no reason fro you to have to get the harness wiring or other support equipment (on a fully assembled FrankInjectror Machine...for that matter...) when I am very willing to act as your Guinea Pig to do some preliminary step testing. If you have scoured my online test videos...you will find that I always try to follow an objective, scientific approach when doing all of the trial and error testing that lead to the completion of two working models out of three tries (the first and the last both work...with the last more closely adhering to the principles proscribed by Mikhail Kalashnikov when he designed his AK47 to have " rugged simplicity and reliability). So...if you don't mind the lag time of sending and receiving your test prototypes... I am ready, willing and able to assist you with things. And with no intention to stifle any of your creativity...the only other question I have is...is it really necessary to build in so much performance variability into your device...when actually...even if the pintles were being opened and closed at just a few basic pulse rates...these would suffice for the bottom-line purpose of getting the EFI cleaning fluid cycling through the injectors for as long as necessary to clean them?
About the power supply demands... unless you intend upon building your own stable power source...just know that way back when I was struggling with this issue, I followed the suggestions of Superdave and Robertisaar to use a switchable ATX power supply (12 Volts @ 5 Amps) and it never once burped while actuating two OTC devices to handle all six injectors for extended cycling periods off of a single (+) and (-) connection... so if the idea of your device is to build and offer sort of a "HeathKit" design... the ATX PS are ubiquitous and fairly cheap for the end users (The FrankenBuilders) to get their hands on...and would be one less thing for you to have to worry about... and avoid any "Mission Creep" problems... but I suppose I am just a little too antsy for the final parts and pieces to be built... Sorry about that...Last edited by 60dgrzbelow0; 06-27-2010, 12:12 PM.Comment
-
Brad.... Just an Idea ...
I don't know what is in the air or the drinking water around the nation right now... but I am getting an exponential increase in the number "hits" on my photobucket within the last week or so...specifically on "The FrankInjector Machine" stuff. With the approval of site Owners, Admins and Mods... I would like to take advantage of this flurry of activity and place the a HQ image of the LOGO of www.60degreev6.com on my opening page with a short video touting the sites high points and qualities here and offer a standing invitation for more people to join. I will follow your lead (and message scripting if required) to get the message out and help to support increasing our membership role...
Last edited by 60dgrzbelow0; 06-27-2010, 12:44 PM.Comment
-
Robert and Rick...
JICYDK...
The latest version of Fedora Linux (13) includes a "Science Package" that includes a new open source PCB design app:
Comment
-
Well until I make a PCB and put it in a box, it looks like a breadboard attached to an STK500
It's going to look like a small box with an LCD an a knob on it, some wires coming out of it.
I know you're eager to test something, but give me a chance to build it first. Before I can take it off my desk and send it to you, I need to lay out a board and stuff it into a box. Not hard, but it takes time.
For the sake of cleaning, you're right--you could just connect a 12VAC power supply directly to the injectors and get the pintles to cycle. It's important to compare injectors to each other at short as well as long pulsewidths, however. Some injectors may spray fine when wide open but flow the wrong volume or have a bad pattern at idle pulsewidths. I could just hard-code a few pulsewidths that I deem to be important and useful, but why? It isn't really any more effort to make that a fully programmable parameter.
I run a ham radio off a computer power supply, so I'm already well aware of their usefulness outside of computer cases. I'm not designing or having anything to do with 12V power supplies right now, so don't worry. I'm actually already done with the batch/sequential/simultaneous code, so that's not a distraction either. I'm just working on the user interface and considering enclosure options at this point. Speaking of which, I'm gonna return to that user interface code right now....Comment
-
Well I did some testing and bugfixing today. I had a stuck injector that had already been through the ultrasonic cleaner and a heated soak in degreaser. I was able to get it unstuck by connecting it to my prototype and taping it on my desk upside-down. After a few taps it started to open a little, and then slowly the buzzing became more pronounced and it opened fully. I drove a couple other injectors for several minutes with good results. I made a little error when I did the design and wound up designing it for 0.05ms resolution instead of 0.1ms resolution. It worked fine with 0.05ms resolution but that's just ridiculous, so I scaled it back to 0.1ms.
Everything is looking good right now. I just need to do a little more coding on the user interface, maybe add a bootloader, and put a board together. I'm going to be really busy the next couple of weeks, so I will work on this a little here and there and probably only check in on the weekends. I think I can have something ready to ship out in about 4 weeks. I'll need to put in an order for some more parts this week.Comment
-
Sure, go for it.Originally posted by 60dgrzbelow0 View Post-Brad-
89 Mustang : Future 60V6 Power
sigpic
Follow the build -> http://www.3x00swap.com/index.php?page=mustang-blogComment
-
Originally posted by ollopa View Post
Rick... That is some very good news... I think you should only work on this project when time and components are handy and whenever you are inspired and inclined to do so. I would be curious to see if you have the same (+) results I got when actuating the ones I was cleaning using the silicone sealed OEM GM Harness... and in those tests...they neither leaked nor shorted during any circumstances. Optimally, the silicone seal compresses tightly, courtesy the metal clips on the harnesses and does a great job of keeping the electrical contacts inside bone dry. I have observed that some of the used/older EFIs I've examined often exhibited rust and corrosion on the two metal tine connectors that seemed to have happened due to long use and/or suffered enough vibration to remove either the cadmium or galvanic coatings on them and expose them enough to cause corrosion. I was just wondering whether or not this would be a side issue to refurbishment that would reduce or change the electrical properties enough to cause a problem...and if so... as an adjunct to a more complete repair of damaged or worn EFIs... I think I might have a fix that would allow for light nickel plating of the tines in a simple manner that would not damage the EFIs and refresh their ability to maintain a good circuit . Many months back...I did some experimenting with electroplating and memorialized the results with images and videos here:
The only concern I wanted to mention was that the injectors might be prone to wear if activated repeated when completely dry... On my end....whenever I had to work with them "dry" I gave them a quick shot of WD-40 when they are first fired up to allow the lubricant the chance to work its way inside and prevent any accidental wear or damage.Last edited by 60dgrzbelow0; 06-28-2010, 06:47 PM.Comment
-
Hmmm. I've never had injectors with corroded spades. I've never even noticed if they were plated or not, actually. Mine look like they might be plated or they might just be a bare alloy. I didn't spend much time looking. At 12V, you don't have to worry about shorting when submerged unless you're dunking them in salt water. I'm guessing the conductivity of purple power is pretty low. You did make me wonder if they are coated and if so, would the ultrasonic action degrade the coating. It knocked some paint off my injectors, but the electrical contacts still work and look fine. Your idea about re-plating the terminals is not a bad one, but I doubt if it would be worth the cost, effort, and toxicity/cleanup. I would just use Caig's Deoxit + power booster to clean up the corrosion and leave a protective coating. If I came across a harness with bad seals, I'd probably clean up the contacts and just squirt a little dielectric grease in there to prevent further corrosion.
Thanks for the tips about injector care. Thankfully I'm testing with a couple sets of injectors that have almost no value to me. I have some dime a dozen 19lb injectors and a set of 36lbers that have issues. I also have some 30lb guys I actually want to clean and use, and then there are the 3 or 4 sets of TBI injectors...
But I won't be dry-firing for long... My busy schedule has been rearranged a little and so I took some time the last couple of days to work on my own Frankinjector... Behold my uh... "Rickinjector Rail!" Yeah.
Any resemblance this bears to your copper pipe manifold is purely coincidence. I didn't even see your design until after I had put this together. Let me tell you the one critical difference, though: I'm using 0.014 wall thickness 9/16" OD brass tubing to compress the O rings instead of 1/2" OD copper pipe. If you do the math, the ID for the brass tubing comes out to 0.5345". I can tell you from direct comparison that this tubing is much tighter than copper pipe. According to my research, copper pipe comes in K, L, and M grades. The inner diameter of 1/2" K is 0.527", 1/2" L (most common in my stores) is 0.545, and 1/2" m is 0.569. If one could find the thick walled K copper pipe, it might work even better than the brass tubing. But I think the brass tubing will work fine.
Many more pics Here
I made two attempts, the first using 5/8" brass tubing for the main rail and a standard piece of stamped steel for the bottom retainer. This was OK but I ruined the thin walled 5/8" tube when doing some drilling. That's when I decided to keep the 9/16" pipes and put them into a larger and sturdier 3/4" copper pipe. This worked out great. There's one trick I didn't take a picture of that anyone should know if they want to try to recreate this. To keep the brass tubes square and in alignment with each other, I places a 1/2" square stick of poplar in the copper tube and let the brass tubes rest on one of the flats of the stick.
I silver soldered everything with a propane torch. That was a new experience. The silver solder is very strong. It worked out well. My technique sucked until the very end of the project, though. Drilling 9/16" holes in the copper tubing was also a pain. Everything up to 3/8" or so was no problem, but 1/2" and 9/16" tended to tear up the pipe. I'm sure that using the right $40 bit would have made that a cakewalk, but I'm cheap. I found that going slow on the drill press up until it was just about to punch through, and then quickly jamming the press down made acceptable holes.
The bottom piece is something a little bit different. I drilled 3/8" holes which are a perfect fit for pintle caps but also work when the caps are removed. Then I made little rings out of 1/2" copper and placed them around the holes. These rings or cups are to allow the injectors to be flipped upside-down for backflushing. The O rings don't need to seat or seal in them. They just keep the injector pointed over the hole. I soldered a 1/4" brass strip on edge underneath the bottom brass plate to stiffen it and prevent flex under load. It's very rigid.
The final note is that I flared the ends of the 9/16" tubes and copper rings to make it easy for the O-rings to be inserted. Insertion probably would have been impossible without cutting the O-rings otherwise. Flaring tools are like $25 so I used a hammer and a $6 swaging tool instead
I gave it a leak and pressure test in a sink full of water and it is doing fine. My digital pressure gauge (also from the junk drawer) is acting up, so I am not sure what pressure I got it up to. Enough to get a stream of bubbles coming out of the regulator return line. The regulator was last set to about 40 PSI.
Now I'm a little more motivated to layout a board for the electronics.
Last edited by ollopa; 07-01-2010, 05:27 AM.Comment
-
Nice looking manifold Rick. It should work great.MinusOne - 3100 - 4T60E
'79 MGB - LZ9 - T5
http://www.tcemotorsports.com
http://www.britishcarconversions.com/lx9-conversionComment
-
Rick... All I can say is...WOW!
At last... Everyone will be able to see what is possible from the imagination and hands of a Real Design Engineer! What you have made is so much more elegant and build-able, too. I'm certain everyone is as curious as I am to see how you set up the arrangement of your graduated cylinders for the Flow Matching portions of your test(s). If you design and create everything else as clean and compact as your "RiXInjector Rail"...I think it might induce more folks to try and build their own versions using design plans that will be much easier than mine to make and use. I am thrilled with your participation and I know that yours will make for a better success. If you can...it would be nice to see your present and future build images via a sort of permanent link to your photobucket posted at either the top and/or bottom of your threads. I would like your permission to "leech" some of your images to my own photobucket that will serve to give you credit and lead those other viewers back to the forum and read more on how you are doing things. Just let me know if that is Okay... Oh..the only reason I call my machine "The FrankInjector" is in honor of James Whales' 1931 movie "Frankenstein"...
Comment
-
I'm curious, too. A little bit less planning and designing went into this than you might imagine. I measured the ID of the holes of some rails I have, then hit the hardware store to see what was available to work with. Then it's back home to arrange the bits and pieces and do the best with what I've got. It's an iterative processes as you can see in the photos from my first attempt. Kind of reminds me of playing with Legos and Construx...Originally posted by 60dgrzbelow0 View Post
Obviously I need to build a stand and find a better way of clamping the rail down than zip ties. Actually, some zip ties have release levers so they are reusable and that might be an option for the truly lazy... I've got a different idea, though. I found some electrical grounding clamps that look like this:
The ground wire screw is 1/4-20. I'm thinking about drilling two holes at opposite ends of the base plate, clamping the grounding clamps on the fuel rail above, and screwing a threaded rod into the clamp. The base plate can then be held in place with two wing nuts. I'm just hoping that the wing nuts will have a low enough profile not to interfere with the spray pattern.
To suspend the whole thing over the graduated cylinders, I'm imagining a wooden A frame at either end of the fuel rail with pipe hangers attached at the apex. The hangers I have will fit over the end caps of the pipe and have a screw clamp, so I'll be able to rotate the assembly into position and then lock it down.
"RiXInjector Rail" Thanks for protecting my anonymity
it would be nice to see your present and future build images via a sort of permanent link to your photobucket posted at either the top and/or bottom of your threads.
You mean in my signature? I did bury a link in my post somewhere, but I can see how it would be hard for anyone to notice once this thread goes to another page.
Granted! Do whatever you want with those photos. Go easy on the credits. Just a link back to the forum post or something is enough for meI would like your permission to "leech" some of your images to my own photobucket
Now you didn't really think I didn't catch that reference, did you?Oh..the only reason I call my machine "The FrankInjector" is in honor of James Whales' 1931 movie "Frankenstein"...
We used to have a server in the office cobbled together from scrap parts that we referred to as "Frankenstein." I've also seen Tron while it was still fairly fresh and I'm no stranger to the saga of Dune
Comment
-
had an interesting idea...
how easy would it be to integrate a standard audio port into a device? thinking being that maybe it could be incredibly easy to literally plug a computer in, play a "sound" file in something(hell, maybe even repeat it) and have it trigger the injectors according to wether or not there is any sound... running them in specific duty cycles would be easy, i like messing with Audacity and creating tone patterns, makes setting up equalizers much easier... just wondering how much effort it would be for something like this.Comment
-
Well, a signal is a signal is a signal, ain't it? I mean from a controller input standpoint, of course. As to the corroded injector pins, I don't recall seeing that many on GM's, mostly on Nissans and Volvos.
If you ain't rock and roll, you must be driving a HondaComment
-
It could be done without too much effort, but it doesn't make much sense (to me, anyway) to do it that way unless you had no other choice. The injectors should be fired with fast digital signals, but audio is analog, so you'd need to have a clipper or trigger to convert the analog to digital. The reliability of the coupling would depend on the amplitude of the sound source. It seems kludgy to me. Is it just that you're worried about the user interface being awkward or confusing?Originally posted by robertisaar View PostComment
Comment