Hi, I'm interested in measuring each cylinder's vacuum independently rather than aggregating them into a single measured value via a reservoir.
I've just started to look into this, but I'm considering using some logic, outside of the MJLJ, to multiplex 4 sensors which would be controlled by two of the User Outs (0-3). This would obviously require some firmware and application sw modifications do cycle through the 4 states, and create 3 more buckets for the readings to be saved in. I've just ordered my MJLJ and have read somewhere on the many forums that the software is open, but I'm not sure what exactly that means?
Does anyone know if this has been done or whether this approach is overkill, lunacy, and/or brilliant (not)?
Thanks for any help with this!
Best,
Jeff
4 x map sensor mods to monitor each cylinder independently
Moderators: JeffC, rdoherty, stieg, brentp
-
alpine-jeff
- Posts: 8
- Joined: Wed Feb 11, 2009 3:58 am
- Location: Northern California
Hi Jeff
I posted a similar suggestion last year. I think the use of a multiplexer and the user outputs to select the correct channel would give greater logging capabilities by allowing 4 or 8 analog inputs for things like water temps, oil temps, oil pressure, EGT, lamda and many more.
The external software shouldn't be a problem but I'm not sure if the MJ has enough internal memory? I've got 2 V3s on my cars and would love to help develop this type of feature. If the MJ hasn't enough memory could we loose some other feature like dual tables? I don't really see a point in this unless you run different fuel types/octanes.
Is this possible Brent and would you be willing to develop it?
Regards
Dave
I posted a similar suggestion last year. I think the use of a multiplexer and the user outputs to select the correct channel would give greater logging capabilities by allowing 4 or 8 analog inputs for things like water temps, oil temps, oil pressure, EGT, lamda and many more.
The external software shouldn't be a problem but I'm not sure if the MJ has enough internal memory? I've got 2 V3s on my cars and would love to help develop this type of feature. If the MJ hasn't enough memory could we loose some other feature like dual tables? I don't really see a point in this unless you run different fuel types/octanes.
Is this possible Brent and would you be willing to develop it?
Regards
Dave
Have you some info that suggest this is worth doing and a system model for all the components?:
- The MAP for any one cylinder varies quite violently through the 4 strokes - so when will you measure it?...you presumably don't want to average it, as that would be done by a reservoir for each cylinder and then I doubt that there's much difference between an averaged sibgle cylinder and an average of all cylinders...
- MAP sensors have a response time (not sure what it is) but it's quite quick, so that needs to be taken into account
- any vacuum pipe between the manifold acts as a low pass filter (aka reservoir) so needs to be taken into account
- etc
If you really want to do it, probably cheaper, simpler and quicker to use 4 EDIS modules driven off one VR sensor, 4 MJLJs and 4 coilpacks
Wiring might be a bit spaghetti-like....
The firmware/application sw is the same thing and it's no longer "open" as in the source code being publicly available. The Configurator sw has, as far as I know, never been publicly available.
The User Outputs are only triggered on load/rpm/aux values and I can't see how they could be used to generate rapid, and co-ordinated, multiplexing signals - oh, I see, you're going to completely rewrite the firmware...
....
Why do you think you need this? I there any OEM/extreme operating condition solution that already does this?
But presumably I've missed the point completely?
- The MAP for any one cylinder varies quite violently through the 4 strokes - so when will you measure it?...you presumably don't want to average it, as that would be done by a reservoir for each cylinder and then I doubt that there's much difference between an averaged sibgle cylinder and an average of all cylinders...
- MAP sensors have a response time (not sure what it is) but it's quite quick, so that needs to be taken into account
- any vacuum pipe between the manifold acts as a low pass filter (aka reservoir) so needs to be taken into account
- etc
If you really want to do it, probably cheaper, simpler and quicker to use 4 EDIS modules driven off one VR sensor, 4 MJLJs and 4 coilpacks
Wiring might be a bit spaghetti-like....The firmware/application sw is the same thing and it's no longer "open" as in the source code being publicly available. The Configurator sw has, as far as I know, never been publicly available.
The User Outputs are only triggered on load/rpm/aux values and I can't see how they could be used to generate rapid, and co-ordinated, multiplexing signals - oh, I see, you're going to completely rewrite the firmware...
....Why do you think you need this? I there any OEM/extreme operating condition solution that already does this?
But presumably I've missed the point completely?
Yep, it's called RaceCapture http://www.autosportlabs.net/Main_Page !dave9469 wrote:I think the use of a multiplexer and the user outputs to select the correct channel would give greater logging capabilities by allowing 4 or 8 analog inputs for things like water temps, oil temps, oil pressure, EGT, lamda and many more.
-
alpine-jeff
- Posts: 8
- Joined: Wed Feb 11, 2009 3:58 am
- Location: Northern California
Hi Dave, Martin,
Thank you both for your comments & suggestions.
Martin, I would expect to use 4 x MPX4250A on a separate board and select with map device I use to feed the analog input on the 68HC908. The challenge will be, as Dave indicated, available memory on the MJ, and modifying the s/w to manage two of the user outputs to drive the map selection and receive, store and average the various input values over some time period. Certainly not rocket science compared to the engineering (s/w and h/w) for rest of this nice device.
Thanks also for pointing me toward the RaceCapture. Unfortunately all I've ever been able to experience are some nifty images and a rejection email to my request to be a beta tester. I'm not a race car driver, I'm simply an enthusiast that would like to profile my old Sunbeam (which has had a frame up restoration with some new features) a little tighter than the old Lucas distributor can offer. I suspect that the majority of Mega* users were simple folks like me. But the market, at least where the real $$ can be had these days, is in the professional data acquisition market. Oh well, gotta love free markets (well... maybe not these days.....?).
Thanks! Jeff
Thank you both for your comments & suggestions.
Martin, I would expect to use 4 x MPX4250A on a separate board and select with map device I use to feed the analog input on the 68HC908. The challenge will be, as Dave indicated, available memory on the MJ, and modifying the s/w to manage two of the user outputs to drive the map selection and receive, store and average the various input values over some time period. Certainly not rocket science compared to the engineering (s/w and h/w) for rest of this nice device.
Thanks also for pointing me toward the RaceCapture. Unfortunately all I've ever been able to experience are some nifty images and a rejection email to my request to be a beta tester. I'm not a race car driver, I'm simply an enthusiast that would like to profile my old Sunbeam (which has had a frame up restoration with some new features) a little tighter than the old Lucas distributor can offer. I suspect that the majority of Mega* users were simple folks like me. But the market, at least where the real $$ can be had these days, is in the professional data acquisition market. Oh well, gotta love free markets (well... maybe not these days.....?).
Thanks! Jeff
OK - as simple question or five:
- how much variance in timing would you expect, across the cylinders, if there was a "black box" that could do individual cylinder timing by measuring individual intake manifold pressures?
- if it's small, is it smaller than the resolution of the EDIS module? (what is the resolution of the EDIS module?)
- would this variance be constant, or change with rpm?
- how on earth would you get the best map, even on a rolling road?
- how many cars have genuinely independent intake manifolds (like twin choke twin carb setups - Weber 40s et al) rather than slightly upstream (towards the carb) splits in the manifold - and how does the pressure vary an inch or two downstream of the split, compared with at the split?
Genuine questions, honest....
- how much variance in timing would you expect, across the cylinders, if there was a "black box" that could do individual cylinder timing by measuring individual intake manifold pressures?
- if it's small, is it smaller than the resolution of the EDIS module? (what is the resolution of the EDIS module?)
- would this variance be constant, or change with rpm?
- how on earth would you get the best map, even on a rolling road?
- how many cars have genuinely independent intake manifolds (like twin choke twin carb setups - Weber 40s et al) rather than slightly upstream (towards the carb) splits in the manifold - and how does the pressure vary an inch or two downstream of the split, compared with at the split?
Genuine questions, honest....
-
alpine-jeff
- Posts: 8
- Joined: Wed Feb 11, 2009 3:58 am
- Location: Northern California
Hey Martin,
I'm afraid I may not have asked my question correctly or you misunderstood. I don't want to control each cylinder's timing individually, I merely want to monitor the vacuum, from the intake manifold, and switch based particular cylinder intake cycle. The individual measurements would not effect the any advance, however the aggregate of the 4 would. I'm more interested in the relative percentage differences across the cylinders, which should be the same regardless of the rpm. I have a pair of dcoe40's and would like to monitor each separately mainly for tuning and preventative problem detection. This little 4 cylinder is a bitch to get / keep tuned.
Thanks for your willingness to jump in on this discussion.
Best,
Jeff
I'm afraid I may not have asked my question correctly or you misunderstood. I don't want to control each cylinder's timing individually, I merely want to monitor the vacuum, from the intake manifold, and switch based particular cylinder intake cycle. The individual measurements would not effect the any advance, however the aggregate of the 4 would. I'm more interested in the relative percentage differences across the cylinders, which should be the same regardless of the rpm. I have a pair of dcoe40's and would like to monitor each separately mainly for tuning and preventative problem detection. This little 4 cylinder is a bitch to get / keep tuned.
Thanks for your willingness to jump in on this discussion.
Best,
Jeff
-
alpine-jeff
- Posts: 8
- Joined: Wed Feb 11, 2009 3:58 am
- Location: Northern California
That's half of my overall goal. This half also maintains the current single map advance scheme by averaging the 4 maps together.
But the second half need to be able to cycle through the N-cylinders and maintain some buffer depth of each cylinder for either real time or post processing and analysis.
So my original thinking, as I laid out in the first post, is to use two of the User Outs to, in my case, cycle from 0 - 3 (four cylinders total). A 6 and 8 cylinder would need three bits.
Dave brought up an interesting issue: memory limitations. I haven't even looked at what's available and what's used. It may be that the only processing possible would be real time, small sample and display.
So Martin, are you affiliated in anyway with AutoSportLabs? What I'd like to look at would be what's involved in modifying the existing firmware and app's ui and display.
Many thanks,
Jeff
But the second half need to be able to cycle through the N-cylinders and maintain some buffer depth of each cylinder for either real time or post processing and analysis.
So my original thinking, as I laid out in the first post, is to use two of the User Outs to, in my case, cycle from 0 - 3 (four cylinders total). A 6 and 8 cylinder would need three bits.
Dave brought up an interesting issue: memory limitations. I haven't even looked at what's available and what's used. It may be that the only processing possible would be real time, small sample and display.
So Martin, are you affiliated in anyway with AutoSportLabs? What I'd like to look at would be what's involved in modifying the existing firmware and app's ui and display.
Many thanks,
Jeff
Jeff, more questions for you.
Are you having trouble synchronising the carbs?
Are you going to design a new intake/plenum to compensate for any variances in the one you have?
Are your combustion chambers cc'ed and are all your ports the same, as in cnc or hand ported?
The reason I ask is the trouble you will need to go to and the questionable, if any, benefit of 4 map sensors and a multiplexed datalog of such sensors.
Are you having trouble synchronising the carbs?
Are you going to design a new intake/plenum to compensate for any variances in the one you have?
Are your combustion chambers cc'ed and are all your ports the same, as in cnc or hand ported?
The reason I ask is the trouble you will need to go to and the questionable, if any, benefit of 4 map sensors and a multiplexed datalog of such sensors.