FE Power Forums
FE Power Forums => The Road to Drag Week 2014 => Topic started by: jayb on June 23, 2014, 12:37:55 AM
-
Dyno weekends tend to be exhilarating and frustrating all at the same time, and this one was surely that! When I wrote my last update I figured I'd be spending all weekend wiring the EFI system onto this engine, and would be lucky to get it running. That all went out the window on Monday when I emailed Scott Clark and told him that I probably wouldn't be ready to run this weekend, because I didn't have the engine wired and I wanted to have it up and running before he drove up with his double throwdown, triple whammy 8 oxygen sensor tuning setup. But in his return email Scott pointed out that he was not tied up on Friday, and if I could take Friday off he would come up and help me wire the engine, and we could get it running and tuned over the weekend.
Well, I had another errand that I had to run on Friday morning anyway, so I had already taken the day off. Scott had to drive up from Omaha to my place, and I figured with both of us working on the wiring we might be able to knock it out by the end of the night on Friday. So we made the arrangements, and I spent some time during the week getting all the mechanical stuff done for getting the engine running on the dyno. On Thursday night I was clipping part of the wiring harness out of my Shelby clone and putting it in place on the engine, and at the end of the night on Thursday, for once it appeared that I had overestimated the time it was going to take to get the project done; it seemed like between the two of us Scott and I could indeed get the wiring done on Friday.
Friday I got back from my morning errand by 11:00 or so and started working on the wiring shortly thereafter. Scott showed up around 2:00 and between the two of us we had it finished up and were ready to start the engine by 6:00 PM. About that time another guy that Scott knows, Mark Dahlquist, also showed up. He drove down from Fargo ND to meet up with Scott and drop something off with him, and he hung around to help out with the engine. I was really lucky to have both of those guys here to help out. Some of you may know that Scott is kind of the EFI tuning guru for a lot of motorsport teams; he tunes for Bischoff and Ray Barton at Engine Masters, and of course those teams always do well. He also tunes for some of the Bonneville teams, and he is actually making his living at this point tuning engines. I hadn't met Mark before this weekend, but he is also an Engine Masters competitor; his first year was last year, and he finished 9th (I think) with a Pontiac engine, with Scott tuning the engine for him. The most recent issue of Popular Hot Rodding has a writeup on Mark's EM engine. Talk about qualified assistance! I'm practically a newb compared to these guys.
Anyway, after Mark showed up Scott spent some time setting up the Megasquirt MS3X EFI system, and then we tried to start the engine. First time out on this stuff you are always concerned about getting a good signal from the crank sensor. I was using a Ford VR sensor on the crank target wheel, and a Cherry digital sensor for the cam position sensor. After several tries we got the engine started, but I shut it off right away because we hadn't put coolant in the engine yet. When assembling this engine I had been concerned about internal water leaks, so I had decided that the first thing I was going to do was to run the engine on the dyno with a radiator, so that after the water temp came up I could dump in some Moroso ceramic sealer and circulate that through the engine for a while to seal up any seepage that might be present. After we knew the engine would start, I got to work setting up the radiator and the cooling hoses. Mark had the idea to put some hooks in the ceiling and hang the radiator from a couple of tie down straps; that worked out very well, and shortly thereafter I had the cooling hoses attached to the water pump and the thermostat housing, and the electric fans wired up. I filled the engine with distilled water at that point, checking carefully for any leaks, but there weren't any, so that was a good sign.
Next we fired up the engine, but Scott was concerned about the crank signal because the engine wasn't running that well. The MS3X system has a diagnostic mode where you can watch the signal from the crank sensor, after conversion to a digital signal (the Ford VR sensor puts out an analog or sine wave signal). This is just like having an oscilloscope screen on your computer so you can watch the crank and cam signals. This diagnostic mode was showing errors with some of the teeth on the target wheel; the sensor was losing teeth, and so the missing tooth on the wheel was not being detected properly. We messed around with the sensor airgap to try to resolve this problem, and it seemed like it mostly went away, but still would give us some intermittent errors when the engine was running. In any case though, we could get the engine started, and it sounded pretty good. Finally around 10:00 PM we just decided it was getting too late and that we needed to run the engine to get the sealer circulating; I didn't want to leave it overnight with water in it and no sealer, and also I wanted to circulate the sealer and then later drain the coolant out of the engine, per the sealer's instructions. So, we fired it up with a less than ideal crank sensor signal and started warming up the engine. It wouldn't idle much below 1500 RPM at this point, but that was OK for Friday night's purposes. After the engine was warmed I dumped in the bottle of sealer, put the cap on the radiator, turned on the electric fans, and let the engine run. The CVR water pump seemed to keep right up with the cooling system; it did gradually warm over 30 minutes from 180 degrees to 210 degrees, but by that time the sealer was well circulated and we had run the engine twice as long as was recommended by the sealer instructions, so we shut it down and called it a night. Scott and Mark took off, and around midnight I came back out to the shop, drained the water out of the engine, and then went to bed.
Saturday morning I was up early, taking the radiator off the engine and hooking up the dyno's cooling system. After Scott and Mark arrived back we got to work on the startup procedure for the engine again. We continued to have trouble though with the crank signal. For a while it would work just fine, with no errors, then suddenly we would get bunches of errors all at once, and the engine would sound like crap. One thing we tried was changing the lead on missing tooth of the crankshaft target. My target wheel, like the Ford ones, has a tooth every 10 degrees around the wheel, with one tooth missing to allow the EFI unit to sync up to the wheel, and know where top dead center is. I had the tooth #1, which is the first tooth after the missing tooth, set for around 65 degrees BTDC. My target wheel is drilled for multiple mounting positions, in 20 degree increments, so we tried advancing the target wheel so that tooth 1 was at 45 degrees, and retarding it so that tooth 1 was at 85 degrees, but still couldn't get a good consistent crank signal. This was the same issue that was dogging me at Drag Week in 2011, so we really needed to get this problem solved. Scott began promoting use of another Cherry digital sensor, like the one we were using for the cam sensor, on the crank; he had recently had good luck with those. I did have one spare Cherry sensor, so finally I installed one in place of the Ford VR sensor. There was an immediate improvement in the cranking signal from the sensor, and the engine began starting much more easily. We had it up and running for several minutes at a time while Scott tuned the A/F ratio and logged the sensor diagnostic data. We were still seeing some drop out of the crank signal, though, so Scott thought we should try moving the target wheel again. After I did that, suddenly we had nothing for signal, and the engine wouldn't start. Mark went into the dyno room and wiggled the sensor, and found that it was loose! This was my fault; I had only put the nuts that held the sensor in place finger tight, and they had come loose and the sensor had started to wobble around. I tried to put the sensor back in its original position, but it looked like the sensor body had been dinged by the target wheel, and it just wouldn't work anymore. So, our crank sensor was no longer working.
We thought about going back to the Ford VR sensor, but I did have a couple of other digital output sensors made by Hamlin that would also work. Unfortunately the Hamlin sensors wouldn't physically fit in the crank sensor mounting bracket that I had machined for this engine. So, in the end what we decided to do was take the Cherry sensor out of the left valve cover, use that one as the crank sensor, and put one of the Hamlin sensors into the valve cover to act as the cam sensor.
All this screwing around with the sensor took the whole morning and most of the afternoon, but when we got the new sensor configuration set up it seemed to work really well. Finally we were ready to make some dyno pulls. We started with pulls from 3000-5000 RPM, and gradually worked our way up to 5000-7200. The engine sounded really good, but was down just a little bit on power compared to what I was expecting. (By the way, this will be disappointing for some of you guys but I won't be sharing any of the dyno results from this engine just yet. There are Drag Week competitors watching this web site, and I don't want to tip my hand before the event. I will post the dyno results on this blog in September, after Drag Week has started.) While we were doing these pulls Scott was messing around with a software feature in the MS3X called VVT. VVT stands for variable valve timing, and it allows the MS3X to control the variable valve timing actuators found on some modern engines. What he was discovering, though, was that despite the fact that this engine doesn't have VVT actuators, the MS3X can log the position of the cam sensor during a pull. This would allow us to see the cam with the cam sensor on it advancing or retarding during the pull!
Back in 2006, on the second SOHC I ever built, I added proximity sensors and targets to the crank and both cams, and ran multiple experiments over a 4 week period trying to determine exactly how much the cams were moving with engine speed. The datalogging capability I had back then was SLOW, and so I ran the pulls really slow, and tried to get data every 500 RPM or so. It was quite the torture test for the engine, and gave me some surprising results. What I found was that while the right cam retarded with RPM (as everybody always said they did), the left cam actually advanced with RPM. This led to about a 3-4 degree variation in cam timing between the two cams, so ever since this I've been setting up my SOHC engines with the right cam advanced 3-4 degrees compared to the left cam. The two graphs reprinted below summarize all this data:
(http://fepower.net/Photos/Road to Drag Week 2014/origcamphasedata1.jpg)
(http://fepower.net/Photos/Road to Drag Week 2014/origcamphasedata2.jpg)
The "calculated" vs. "measured" data in the second chart refers to calculating the difference between the two cam signals when they are compared to the crank signal, and then a directly measured difference between the two cams. The data was taken on different days, so there are some minor differences, but the general trend is the same. Now that we could see this same data logged directly from the EFI system, I was anxious to get some confirming data.
The cam sensor was on the left cam, so we were expecting it to advance. After the next pull, the data seemed to show that the left cam was retarding, and by quite a significant amount! However, there was some question about the data itself; could the direction be wrong? We weren't sure, but in any case we were seeing something like 8 degrees of change in the cam phase. Yikes!
While we were thinking about this, we decided to go ahead and put Scott's 8 O2 sensor setup on the engine. Here's a picture of the engine with this setup installed:
(http://fepower.net/Photos/Road to Drag Week 2014/8o2s.jpg)
This setup uses the closed loop corrections available in the MS3X to monitor the O2 sensors in each primary pipe, and then hold the injectors open a shorter or longer period of time, to try to hit the targeted A/F ratio in each cylinder. After getting all the O2 sensors installed Scott configured the CAN communications in the software, and we were ready to run. We targeted the A/F at 12.6:1, and ran the pull hoping for a significant power increase. We did see a bump of about 10 HP, but it wasn't a real big improvement. This basically meant that the sheet metal intake I had built for this engine, previously known as "the steaming pile" because of the problems it seemed to give us in 2011, was actually pretty good. In fact, Scott was praising the steaming pile by the end of the day Sunday as one of the better sheet metal intakes he's seen with respect to fuel distribution.
Last time I ran this engine on the dyno, in its 585" form, it had wanted to run a little lean. So we ran another pull, changing the targeted A/F ratio to 13:1, and the engine really picked up, to the tune of about 18 HP. That was good news. We continued to log the cam data, and we continued to see what appeared to be a big retard in the left cam during the pull. It was about the end of the night on Saturday, so we made plans for the next day and called it a night.
This morning (Sunday), the first thing I did when getting out to the shop was to take my second Hamlin sensor, and install it as an auxiliary cam sensor in the right valve cover. I was certain that the right cam would retard with engine RPM, so if the right and left cams were moving in the same direction, we would have confirmation that both cams were retarding. Thinking about this a little more, I began to suspect that the long, slow pulls I had done on the test engine back in 2006 were not really reflective of what would actually be happening at the track, and that maybe with the rapid logging available using the MS3X, and a normal dyno pull rate of 300 RPM/sec (rather than the 25 or 50 RPM/sec I was using in 2006) might make the cams behave differently. When Scott arrived the first thing we did was run the test to look at the left and right cams together. Sure enough, they were both moving in the same direction, and both retarding a significant amount. Here are four screen shots from the MS3X datalogger that show this effect:
(http://fepower.net/Photos/Road to Drag Week 2014/Camsensor1.JPG)
The vertical blue line running through the graph has numbers next to it, and these are the ones we want to pay attention to. The top graph shows engine RPM, which is about 1200 at this point in the log. The bottom graph shows the cam sensor angles. These are phased arbitrarily, so what we are looking for is changes in these numbers, not absolute values. The red line is the right cam sensor, and the white line is the left cam sensor. At this point in the log the right cam is at 127.6 degrees, and the left cam is at 156.9 degrees. Off to the right of the vertical blue line, you can see the engine RPM during the dyno pull. Also, the left cam line goes vertical in the middle of the pull, but that's not real data, that is just a noise spike. Here are the other graphs with the blue line positioned at various points during the pull:
(http://fepower.net/Photos/Road to Drag Week 2014/Camsensor2.JPG)
In this graph we are at the start of the pull, 5000 RPM, and the right cam has retarded to 116.7 degrees, while the left cam has retarded to 153.7 degrees. Here's the next graph:
(http://fepower.net/Photos/Road to Drag Week 2014/Camsensor3.JPG)
Here we are at 6360 RPM, and the right cam has retarded to 115.3 degees, while the left cam has retarded to 149.7. Here's the next graph:
(http://fepower.net/Photos/Road to Drag Week 2014/Camsensor4.JPG)
In this case we are almost at the end of the pull, at 7180 RPM, and the right cam has retarded to 113.7 degrees, while the left cam has retarded to 147.0.
To be honest I'm not really sure I believe these numbers; that is a tremendous amount of chain stretch, and it just seems unreasonable to me. The difference between the right cam at idle (127.6 degrees) and the right cam at 5000 RPM (116.7 degrees) is just not believable. Also, the left cam is retarding more during the pull (6.7 degrees) than the right cam (3.0 degrees). This is in spite of the fact that as the engine spins the length of the chain between the drive sprocket and the left cam is shorter than the length of the chain between the drive sprocket and the right cam, in the direction of engine rotation. If anything, the right cam should be retarding more than the left cam, if indeed they are both retarding as this data seems to suggest. I can see several potential sources of error in this data, including variability in the targets (which are just bolt heads), and frequency response of the Hamlin sensors. Just doing a quick calculation, at 7000 engine RPM the cams are spinning at 3500 RPM. This is 58.33 rotations per second. To discriminate 0.1 degrees of accuracy the update frequency of the sensors must be 1/(58.33 * 3600), or about 210 KHz, which is pretty fast for a magnetic sensor. Even to discriminate 1.0 degrees the frequency response still has to be 21 KHz. I need to do some research on the sensors to see if they are up to the drill here.
In any case more happened on Sunday than I have reported so far; I will update this post tomorrow with more information on Sunday's tests - Jay
-
Very interesting! I'm sure you have thought of it and maybe now isn't the time with Drag Week coming up but....
Can you rig up some kind of crude plexiglass or Lexan covered front cover and video the action of the chains and adjuster to determine if they really do retard that much? Hate to think a good cover may have to be sacrificed to do.
-
According to legend Sneaky Pete Robinson did exactly that when developing his fuel motor in the 1960s. He put a degree wheel on each cam, and stood in the dyno room with a timing light and oil spraying everywhere, watching the cam timing! I won't be doing anything like that LOL...
-
Glad to see it up and running, hope it will stay leak free for you! How is the torque with that intake set-up? Good enough to take some gear out of the car like you did with the Galaxy? Hopefully the power will come around with some run time, and additional tweaking. Joe-JDC
-
Jay, what an incredibly detailed account of this weekend's events! :o I was very grateful to be included this weekend and even though I didn't do much I sure had a great time! I hope to visit you again sometime. I also wanted to point out that it's still possible you are looking at crankshaft degrees which would make the cam degree error reading 1/2 (unless you verified that aspect after I left).
-- Mark Dalquist
-
Jay
Thanks for taking the time to write all this down for us.
It's fascinating reading.
-
Jay, I am no electrical engineer, but any chance you could be seeing any kind of interference that would change the readings of the sensors or inductance along the route of the wire?
Even with tha being said, looks like the numbers could be a harmonic causing some chain "waggle" at different rpm
Does anyone make a gear drive or a belt drive?
-
Does anyone make a gear drive or a belt drive?
FPP made a belt drive back in the early '80s, but I only ever saw one on Rons engine. And I thought I remembered Sneaky Pete ran a gear drive at one time, or somebody did, but they were all one-off designs and never really sold.
Wait till September? What is this, a season show ender? Another "Who shot J.R.?" ;D ;D
Jay, you may not want to totally discount the less retarding on the right cam. Even though the chain from the drive sprocket to the right cam is longer, there may be something other than chain stretch going on. Momentum has a way of making chains follow an arc pattern, even on drive sides. If the chain were doing something to that effect, it could be "pulling" the right cam ahead, or advancing it, relative to the left cam. The open chain area between the left and right cams is fairly significant and even though there is a short bottom guide, it may not be enough. That was the reason for the discussion earlier about modern engines and their chain controlling devices. It's very important in high revving engines.
There is a top guide, but the chain arc may not simply be "up and out", it could also be "down and in". You can see this on blower drives and just about any belt/chain drive system, and it all depends on length, speed, applied torque etc. as to whether its "up/out" or "down/in". Something that only sophisticated strobe or very high speed cameras could detect....the stuff that manufacturers have access to. ::)
Just a thought.
Boy do I love carburetors and simple electronic ignitions...lol
-
This week's update continues here. After running the pulls with both cam sensors installed, and seeing that they were more or less repeatable for the cam phasing shift, I decided that the next logical course was to change the cam timing. I had the right cam degreed at 105 and the left cam at 108, but on the engine stand I had checked piston to valve clearance all the way up to 99 and still had about .050" on the intake valve, so I had some room to advance the cams to compensate for the chain stretch that the sensors seemed to be indicating was happening. About the time I got started on the cam timing change Mark had to leave, and then my friend Kevin R showed up to help out. Always nice to have some wrenching help on the dyno.
I got started on the cam timing change by pulling the right side inspection cover so I had the opportunity to look at the witness marks from the chain hitting the tensioner arm. What I saw indicated to me that the chain was still too loose; notice the long stripe in the machinist's blue dye on the top of the tensioner arm:
(http://fepower.net/Photos/Road to Drag Week 2014/cammark2.jpg)
I think that's too much contact, even after some dyno pulls. I like to see just a little nick on that tensioner in the dye; after the engine went back together and I had tightened up the chain bolt an eighth of a turn, I made a few more pulls, this was what the mark looked like:
(http://fepower.net/Photos/Road to Drag Week 2014/cammark1.jpg)
So, next it was on to the timing change. Now if this had been last year or the year before, my friend Hemi Joel (aka Captain Stabbin') would have been here with his video camera, and he would have recorded the next events for posterity. Scott asked me how long it would take to change the cam timing. I said, "Oh, about 20 minutes per side." When Joel finds out he missed the opportunity to get that on tape, and then the following 5 hour debacle to get the cam timing changed, he will be very disappointed LOL!. Advancing the right cam was complicated by the fact that I was out of adjustment on the alignment pins that move the cam gear with respect to the camshaft itself. This meant that I had to move the sprocket on the timing chain by one tooth, and then go back to the most retarded position for the alignment pin. After pulling the valve cover and loosening the chain, I removed the sprocket and rotated it one tooth. About an hour later I figured out that I had moved it the wrong direction. The first thing that happened after I got the sprocket back on was I rotated the engine over by hand to check the cam timing, but hit a dead stop. Uh-oh, that was a piston running into a valve! Thank goodness I didn't try to rotate the engine over with the starter, or I would have bent a valve for sure.
I scratched my head for a while trying to figure this out, and eventually decided to pull all the rocker shaft clips on the right head so that there was no way the valves on cylinders 1-3 would open, and then back the adjusters off the rockers on cylinder 4 to minimize any potential piston to valve contact. Then, I was going to try to rotate the cam to check the timing on #1 so I could figure out where I went wrong. Along the way I loosened up the timing chain in case I had to pull the sprocket off again. And then I made another cardinal error; I rotated the crank with the timing chain loose, and all the rockers on the left head still on the valves. After rotating the crank a few degrees I heard a pop, and then another couple of pops. On the second set of pops I notice the timing chain jumping, and it dawned on me that with the chain loose, and the left cam hard to rotate because of the valve spring pressure on the cam, that I had just jumped a few links on the left side sprocket.
At this point I was basically back to square one on the cam timing for this engine. I had to start completely from scratch, which is not easy with the front cover on the engine. With the cover off, you can line up the colored chain links with the dots on the timing gears, and know that you are very close right from the start. But with the cover on there is no way to see the chain and center timing gear, and you can't move the chain with respect to the center timing gear anyway. I ended up taking the bolts out of the left cam sprocket so that the left cam wouldn't move, going back to the right cam and getting it timed correctly, and the pulling the rocker arms off the valves on the left cam, getting that cam lined up, and then degreeing that one. It is a lot easier to write this than to actually do it LOL! It took me until 5:00 on Sunday to get this done, and get the engine back together.
When I did have the engine back together, though, I had changed the left cam to 103 degrees (from 108), and the right cam to 102 degrees (from 105). If the retard numbers we were seeing were accurate, this would put the cams right around straight up in the RPM range of 5000-7000. I started the engine with some trepidation, fearing I may have done some damage to the engine during the whole cam timing debacle, but it lit right off and sounded fine. After warm up we ran the dyno pull AND..... the engine was down 20 HP. Sounded good during the pull, though.
After that pull we pretty much called it a day. I had a whole list of things I was going to try on Sunday, but due to the cam timing fiasco I didn't get to try anything but that. After this I was even more suspicious of the cam timing that was being reported by the MS3X, because if those numbers were right the engine should have picked up with the cam timing change. Also, today I got some more information on the Hamlin sensors I'm using for cam timing, and their specs are really not ideal for this application, so I'm going to get some of the sensors that I used with my original cam timing experiments several years ago, which are very fast and accurate, and plug those in to see what numbers they give. Scott also told me he is going to check with the Megasquirt designers to see if what we are trying to do is even possible, or if there is some kind of electronic signal processing going on that is preventing us from getting accurate cam timing data. We will see how all this works out in the next week.
Other experiments I want to run are to move the throttle bodies to the top of the intake manifold, rather than the front, to see if that improves performance, and to turn down the dry sump oil pressure a little (currently running at 80 psi) to save power. I'm also not getting very good crankcase vacuum from the dry sump, so I may have a leak to track down there. I fully expect to spend another full weekend on the dyno with this engine next weekend, to try to iron out these little bugs and really dial this engine in. But I'll tell you guys something, this thing sounds like a million bucks at speed. Next weekend I'll run a video of a pull so you can hear it.
I'll post another update next Sunday night - Jay
-
Glad to see it up and running, hope it will stay leak free for you! How is the torque with that intake set-up? Good enough to take some gear out of the car like you did with the Galaxy? Hopefully the power will come around with some run time, and additional tweaking. Joe-JDC
Joe, the torque curve is very broad and flat, but it is down a little from where I expected it to be. According to Pipemax my intake runners are a good 3" short, and that is one good reason for the torque being a little lower. But its not bad, really...
-
Jay, what an incredibly detailed account of this weekend's events! :o I was very grateful to be included this weekend and even though I didn't do much I sure had a great time! I hope to visit you again sometime. I also wanted to point out that it's still possible you are looking at crankshaft degrees which would make the cam degree error reading 1/2 (unless you verified that aspect after I left).
-- Mark Dalquist
Mark, for sure we are looking at crankshaft degrees on the MS3X, but still that is a lot of error, more than I saw the first time I ran these tests back in 2006 or so. In any case it was great having you here, and you are welcome back at any time - Jay
-
Jay, I am no electrical engineer, but any chance you could be seeing any kind of interference that would change the readings of the sensors or inductance along the route of the wire?
Even with tha being said, looks like the numbers could be a harmonic causing some chain "waggle" at different rpm
Does anyone make a gear drive or a belt drive?
Ross, electrical interference is certainly a possibility, but these are digital sensor outputs so I think that explanation is not all that likely. And I think for sure there is a harmonic in the chain causing the timing to move around; I saw this with the first data I collected on the phenomenon back in 2006.
Here are a couple of photos of gear drives for the SOHC. The first one is a custom drive, I think, and the second one is the one that Sneaky Pete Robinson developed. I think that my friend Brutal Bob may have one of the Sneaky Pete drives:
(http://fepower.net/Photos/Road to Drag Week 2014/Gear Drive SOHC.JPG)
(http://fepower.net/Photos/Road to Drag Week 2014/Sneaky Pete Robinson Gear Drive.jpg)
Needless to say, I don't have access to anything like that at present. I'm thinking I need to really figure out if the data I'm collecting on the cam timing change is accurate, and then make any changes from there...
-
Wait till September? What is this, a season show ender? Another "Who shot J.R.?" ;D ;D
Who shot J.R. LMAO!! ;D ;D
-
I would have thought that the larger pin double roller chains from Munro wouldn't stretch AS MUCH as the stock chains? If anything, wouldn't the cam timing reflect that? At 103 and 102, does that mean your experiencing Less stretch or more stretch than normal (and I use normal sparingly with SOHC's)? LOL
-
That was also my thought, but I don't think we've proven anything yet. For one thing, I'm not confident that the data I'm getting from the sensors is accurate. I need to work on that issue this week. For another, the valve springs on this engine are quite a bit stiffer than the springs on the engine I had tested previously, so comparing this chain with the .250" pins and these springs is different than comparing the stock type chain with much lighter springs. So, the jury is still out...
-
I can't imagine the badass noise that would come from such a group of timing gears for a SOHC. :o
All this technobabble about Megasquirt,Cherry,Hamlin,digital,analog,magnetic datalogging gizmos has me lost. I'm like a penguin cruising thru the desert....really lost! But I do enjoy the posts of your weekend efforts Jay. 1000hp or bust! Go Jay go!
-
Yes, the noise from those gears would be positively frightening. Think of the noise from a normal gear drive setup, with three gears only, and then add another 5 or 6 gears!
I don't know if anybody has noticed this, but looking at Sneaky Pete's gear drive, he is turning the right cam backwards! What the heck is that all about? Maybe he is trying to compensate for the side to side differences in the SOHC valvetrain. As set up stock, the right cam pushes on the rocker arm roller wheel to start the intake valve opening, and the left cam pulls on the rocker arm roller wheel to start the intake valve opening. This leads to differences in how the valve acts from side to side, even if the cam lobes are identical. Comp addressed this with their most recent cam grinds for the SOHC, by making slightly different grinds for the left and right cams. Maybe Sneaky Pete was trying to address it with reverse cam rotation back in the '60s. I wonder if you have to use two left cams with that setup, or a custom grind on the right, or what...
Edit: Come to think of it, two left cams wouldn't work because of the bearing spacing. Must have required a custom cam...
-
Thanks Jay for having us up this weekend. You always are tackling the odd projects and I always learn a thing or two.
I believe the sensors might wander as much as the 1.4* Jay calculated, but my money is on the megasquirt recording the falling edges accurately. Another issue might be that we used a simple bolt head for the sensor targets. But - the data we logged still screams "your cams are moving at least 9* from idle to 7000rpm" even if the absolute measurements might not be .01* accurate.
I submitted a question and our data to the Megasquirt gurus, told them what we're trying to do and asked what the best way is to turn Jay's MS3 into a highly accurate logger for the cams. With the crankshaft trigger wheel and ignition coils we use, the ignition timing light makes the crankshaft look like it is not moving, so the basic timing functions of this ECU are definitely working. But I told them "we highly doubt those cams are moving as much as the logs show" so let's see what they say! they have always been helpful to me in the past.
This is a fun project for sure!
-Scott
-
Thanks for checking with the Megasquirt guys, Scott, I'm very interested in what they have to say. Here's another odd thing about those logs. If you look at the log for the cam that is mapped in red, it is changing fairly dramatically before the pull, without any change in engine RPM! At a steady 1200 RPM, it looks like that cam retards 6-7 degrees! I can't imagine that's right, but you never know...
-
but my money is on the megasquirt recording the falling edges accurately.
-Scott
Would that be a trigger positioning thing, or a trigger type problem.
-
but my money is on the megasquirt recording the falling edges accurately.
-Scott
Would that be a trigger positioning thing, or a trigger type problem.
I'd believe a possible triggering issue (sensors, targets, or circuit conditioning on the Megasquirt inputs we are using) causing error... but I am confident in the processor and code accurately recording the position of the falling edge signals it's seeing. Otherwise we'd have to question the ignition timing, and anything else referencing crankshaft position.
-
Jay - I recognize that fist timing gear set picture. I bought a 2x4 SOHC intake and a bunch of timing stuff from an older gentleman named Nat in Lake Havasu, AZ. I want to say it was around 1999 or 2000. That was his engine, and he had those gears custom machined. It looked like the machinist knew what he was doing.
If you think timing a SOHC from scratch with the front cover on is hard, you should try it with the engine installed in a 427 Cobra! No friggin' room, and it was a bear to even see the timing marks on the damper.
Your timing adventure makes sense. Advance tends to give you more down low. You must not be losing as much timing as those sensors indicate...
- Bill
-
Yes, Pete's drive does spin that cam backwards yet.....
-I wonder if he did this solely to simplify the gear drive with one less gear, keep it compact and under a cover (his was a custom one of course) that was somewhat like the OEM unit?
Or......
-was the backward spinning cam also a part of improving the longevity and/or accuracy of valve timing events in addition to the gear drive idea?
You seem Jay to be in touch with other cammer builders who may know. Earl Wade if he is still with us? Maybe Jon Kaase.
Yes, the noise from those gears would be positively frightening. Think of the noise from a normal gear drive setup, with three gears only, and then add another 5 or 6 gears!
I don't know if anybody has noticed this, but looking at Sneaky Pete's gear drive, he is turning the right cam backwards! What the heck is that all about? Maybe he is trying to compensate for the side to side differences in the SOHC valvetrain. As set up stock, the right cam pushes on the rocker arm roller wheel to start the intake valve opening, and the left cam pulls on the rocker arm roller wheel to start the intake valve opening. This leads to differences in how the valve acts from side to side, even if the cam lobes are identical. Comp addressed this with their most recent cam grinds for the SOHC, by making slightly different grinds for the left and right cams. Maybe Sneaky Pete was trying to address it with reverse cam rotation back in the '60s. I wonder if you have to use two left cams with that setup, or a custom grind on the right, or what...
Edit: Come to think of it, two left cams wouldn't work because of the bearing spacing. Must have required a custom cam...
-
More here Jay:
http://www.network54.com/Forum/76346/thread/1292363027/Gotta+love+it-+Harvey+Crane's+take+on+the+SOHC.....
Both the link I posted above and the text below came from Harvey Crane's long dissertation on Pete, cammers, early Crane history etc. Funny, I had run across his text long ago and posted same in the old FE Forum.....which got me to thinking to go looking for it! You refreshed my memory Jay when you said Pete had stood in the dyno room...etc. Your own 2006 SOHC chain comments in the link are also a good refresher for our posters here.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
After NASCAR banned this engine in 1964,Ford went drag racing with it.In Top Fuel and AFX,the fore-runner of Funny Car,the engines gave problems with consistent performance."Sneaky Pete"Robinson tested the engine and found the cam drive chain stretched so much,the cams went from 4 degrees advanced to 4 degrees retarded at 6,000 RPM and even more at 8,000+ RPM.The engine also had one bank of cylinders that ran hotter than the other,this was due to both cams turning in the same direction.Since the profile of the cam is different for lift and close,one cam had a different lift profile than the other,resulting in a different air / fuel charge in left and right cylinders.Pete's gear drive turns the cams in opposite directions so the intake profile on both banks is the same.Pete won his first major event after he installed his gear drive,but sadly was killed shortly after and no one picked-up where he left off.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
So, although you have seen all this before Jay, your current data methinks begs another serious look at whether the data is correct and then what fix, maybe even short of a 100% gear drive, could be designed. JMO!
-
More here Jay:
http://www.network54.com/Forum/76346/thread/1292363027/Gotta+love+it-+Harvey+Crane's+take+on+the+SOHC.....
Both the link I posted above and the text below came from Harvey Crane's long dissertation on Pete, cammers, early Crane history etc. Funny, I had run across his text long ago and posted same in the old FE Forum.....which got me to thinking to go looking for it! You refreshed my memory Jay when you said Pete had stood in the dyno room...etc. Your own 2006 SOHC chain comments in the link are also a good refresher for our posters here.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
After NASCAR banned this engine in 1964,Ford went drag racing with it.In Top Fuel and AFX,the fore-runner of Funny Car,the engines gave problems with consistent performance."Sneaky Pete"Robinson tested the engine and found the cam drive chain stretched so much,the cams went from 4 degrees advanced to 4 degrees retarded at 6,000 RPM and even more at 8,000+ RPM.The engine also had one bank of cylinders that ran hotter than the other,this was due to both cams turning in the same direction.Since the profile of the cam is different for lift and close,one cam had a different lift profile than the other,resulting in a different air / fuel charge in left and right cylinders.Pete's gear drive turns the cams in opposite directions so the intake profile on both banks is the same.Pete won his first major event after he installed his gear drive,but sadly was killed shortly after and no one picked-up where he left off.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
So, although you have seen all this before Jay, your current data methinks begs another serious look at whether the data is correct and then what fix, maybe even short of a 100% gear drive, could be designed. JMO!
Interesting! so they found 8* of movement - similar to what we might have seen.
-
Funny I didn't see that much timing variation in 2006 during my testing. The original stock chains were not as good as the chains I was running in 2006, based on the chain pull-test data, so maybe they stretched more than the ones I was using. I still think the data is suspicious, especially the data for the right cam, which is the red line in the graphs on the previous page. With no change in engine speed, that right cam is retarding 3-4 degrees in that graph. I don't believe that...
In any case, my precision sensors are on the way and I will be able to test with them installed this weekend. Should be an interesting comparision...
-
Yes, Pete's drive does spin that cam backwards yet.....
-I wonder if he did this solely to simplify the gear drive with one less gear, keep it compact and under a cover (his was a custom one of course) that was somewhat like the OEM unit?
Or......
I have an old magazine article....somewhere ::) that spoke about the Sneaky Pete drive system and he spoke about the reverse cam being needed because the gears had to be made to fit the space and engine design under the cover, and there was just no way to add another gear. I remember it talking about all the work and effort that went into the system.
-
How will you test the new sensors?
-
The new Honeywell sensors respond at a higher frequency (100 KHz) than the Hamlin sensors (17 KHz), and also use simpler electronics. The signal processing electronics used in the Hamlin sensors and the Cherry sensors is very sophisticated, but not designed as much for accuracy and repeatability as they are for detecting a wide variety of ferrous targets. The Honeywell sensors require a magnetic field applied to the sensor, not just a ferrous target like a gear tooth. So, instead of using a bolt head as the target like I have in the tests so far, I've acquired some small Honeywell magnets that are delivered in aluminum holders with an 8-32 screw on the back of the holder. So what I'm going to do (which is the same thing I did in 2006) is center drill the bolt head that I was using for a target, and install one of these magnets on each cam. With the 100 KHz frequency response of the sensors and a repeatable magnetic field I should be able to discriminate down to 0.2 degrees of crank timing at the cam.
If I get the same results as last weekend then one more thing I'm going to try is a suggestion from the Megasquirt guys that Scott got for me, to remove some of the filtering from the inputs to the ECU to make sure that the filters are not skewing the data somehow. And after that, I think I will have to accept that the data is the data LOL!
-
Jay, I love it when you talk in Greek! Haha!
Sounds like your persistence (as usual) will result in good and accurate data from which a decision can be made. Best of luck!
-
Funny I didn't see that much timing variation in 2006 during my testing. The original stock chains were not as good as the chains I was running in 2006, based on the chain pull-test data, so maybe they stretched more than the ones I was using. I still think the data is suspicious, especially the data for the right cam, which is the red line in the graphs on the previous page. With no change in engine speed, that right cam is retarding 3-4 degrees in that graph. I don't believe that...
In any case, my precision sensors are on the way and I will be able to test with them installed this weekend. Should be an interesting comparision...
Is it possible the chain is stretching more due to higher valve spring tension?
And I am not sure the gear drive would be intentionally that noisy. The high pressure oil pump and camshaft in my 6.0 Diesel are gear driven (as are 7.3 Powerstrokes and 5.9 Cummins) and do not exhibit that old school gear drive sound.
Josh
-
More chain stretch due to higher valve spring pressure is certainly possible; what's confusing to me is the behavior of the left cam, which advanced a few degrees at higher engine speeds when I tested the other SOHC in 2006, and now appears to be retarding, like the right cam. I don't think I can chalk that behavior up to valve spring pressure.
On the gears, I'll bet your diesel gears aren't straight cut. Straight cut gears like the ones in the pictures earlier in the thread are really noisy, but stronger than angle cut gears.
-
You ready Jay, you had a long recovery from the last engine LOL. I would like to see the results of your new and improved engine.
-
More chain stretch due to higher valve spring pressure is certainly possible; what's confusing to me is the behavior of the left cam, which advanced a few degrees at higher engine speeds when I tested the other SOHC in 2006, and now appears to be retarding, like the right cam. I don't think I can chalk that behavior up to valve spring pressure.
On the gears, I'll bet your diesel gears aren't straight cut. Straight cut gears like the ones in the pictures earlier in the thread are really noisy, but stronger than angle cut gears.
That's true, they are angle cut on the diesel.
I got to thinking about the Ford Indy 255. I have been around a couple of those as they were started and ran, and did not exhibit the infamous gear drive noise.
(http://www.wrljet.com/fordv8/images/64-indy-gears.jpg)
Josh
-
FYI friend of mine built a NASCAR cup motor by collecting parts and pieces and of course buying a bunch of new stuff. Most amazing was what looked like a stock SBF timing cover with of course the large bolt hole that gets a dry-sump stud, but the interesting thing was the long Teflon guide that required significant filling/build up of aluminum inside the cover to support the Teflon strip. It was literally kissing the chain. When you think of NASCAR teams addressing chain stabilization on a simple little crank:cam drive chain ........and that monster chain on the SOHC....................
-
Nascar uses timing belts and has for nearly 2 decades.
Josh
-
The one in my friends less then 10 year old Cup motor (previous Gen) is solid aluminum (not bolted in) with the Teflon on the face. His is much more parallel with the chain then the one shown FWIW.
http://www.circletrack.com/enginetech/ctrp_1003_nascar_sprint_cup_engine/photo_12.html
-
More chain stretch due to higher valve spring pressure is certainly possible; what's confusing to me is the behavior of the left cam, which advanced a few degrees at higher engine speeds when I tested the other SOHC in 2006, and now appears to be retarding, like the right cam. I don't think I can chalk that behavior up to valve spring pressure.
On the gears, I'll bet your diesel gears aren't straight cut. Straight cut gears like the ones in the pictures earlier in the thread are really noisy, but stronger than angle cut gears.
just put a straight cut dog box behind it... you'll never hear the cam gears again... (well i guess when the clutch is in.... or you're in neutral)
-
The one in my friends less then 10 year old Cup motor (previous Gen) is solid aluminum (not bolted in) with the Teflon on the face. His is much more parallel with the chain then the one shown FWIW.
http://www.circletrack.com/enginetech/ctrp_1003_nascar_sprint_cup_engine/photo_12.html
You're link leads to an open wheel modified Outlaw style engine and not a full on Nascar Winston Cup/Nextel/Sprint Cup engine.
The big boys use belts, the lesser classes use whatever.
Josh
-
Not sure on that Josh. Read all the other picture captions. This one is interesting.
http://www.circletrack.com/enginetech/ctrp_1003_nascar_sprint_cup_engine/photo_14.html
Has adjustable rockers UNLIKE current Cup motors. So they must use lash caps like SOHC's used to? Since SOHC adjustable rockers seem to be a problem maybe non adjustable are the better way.
Perhaps they found the chain is better than a belt, that billet timing cover doesn't appear to have been made for a lesser series.
-
Not sure on that Josh. Read all the other picture captions. This one is interesting.
http://www.circletrack.com/enginetech/ctrp_1003_nascar_sprint_cup_engine/photo_14.html
Has adjustable rockers UNLIKE current Cup motors. So they must use lash caps like SOHC's used to? Since SOHC adjustable rockers seem to be a problem maybe non adjustable are the better way.
Perhaps they found the chain is better than a belt, that billet timing cover doesn't appear to have been made for a lesser series.
Once again, BB's link is to an Open Wheel World of Outlaw's engine.
Notice the external timing belt of the Ford FR9 Sprint Cup Nascar Engine
(http://www.cpgnation.com/filehost/files/8/Ford%20NASCAR%20Small%20Block/P1040475-1.jpg)
The Yates/Ford 1991+ Nascar Timing Belt set-up
(http://cdn1.bigcommerce.com/server1900/e457d/products/245/images/739/RYR-BD-107__59254.1291168767.1280.1280.jpg?c=2)
Josh