Notices
944 Turbo and Turbo-S Forum 1982-1991
Sponsored by:
Sponsored by: Clore Automotive

AEM EMS in a 944/951/952

Thread Tools
 
Search this Thread
 
Old 06-09-2011, 05:27 PM
  #76  
reno808
Rennlist Member
 
reno808's Avatar
 
Join Date: Mar 2005
Location: In the garage trying to keep boost down
Posts: 8,809
Likes: 0
Received 0 Likes on 0 Posts
Default

Originally Posted by Chris White
Haltech will too....but I refuse to even entertain the idea!
how come??
Old 06-09-2011, 05:28 PM
  #77  
333pg333
Rennlist Member
 
333pg333's Avatar
 
Join Date: Feb 2006
Location: Australia
Posts: 18,926
Received 98 Likes on 81 Posts
Default

PM sent Thom.
Old 06-09-2011, 06:26 PM
  #78  
A.Wayne
Formula One Spin Doctor
Rennlist Member
 
A.Wayne's Avatar
 
Join Date: Jan 2005
Location: RPM Central
Posts: 20,448
Likes: 0
Received 0 Likes on 0 Posts
Default

Originally Posted by Chris White
Haltech will too....but I refuse to even entertain the idea!


Originally Posted by Thom
I run full sequential (fuel + spark) with my Vi-PEC, stock crank trigger + a cam angle sensor.
I have compared both Vi-PEC/Link softwares and they look 99% identical.

Not sure I understand the benefits of running a 60-2 tooth trigger, as the stock 132 tooth trigger set up should offer better resolution...?



Not sure they went this direction to increase performance but rather to reduce scale manufacturing costs within their partnership with Bosch who were certainly beginning to make set ups for other cars also using a 60-2 tooth trigger.
They introduced the 60-2 set up with the 944S in '87 yet they didn't convert the 8V cars to it.
The older Bosch system is specific for that setup the 60-2 done on the newer cars are for more resolution not less. Many say they can work with anything , stick a digital timing light on them and watch them run ...


Old 06-09-2011, 09:02 PM
  #79  
Chris White
Addict
Rennlist Member

Rennlist Small
Business Sponsor

 
Chris White's Avatar
 
Join Date: Jul 2001
Location: Marietta, NY
Posts: 7,505
Likes: 0
Received 36 Likes on 27 Posts
Default

Originally Posted by schip43
Yes it is under a grand but it leaves you to do everything! I'm sure there are those on here that can deal with it but many cannot.
If you are just looking at base costs – the Haltech Sport 1000 ECU is just over $1300 and the Electromotive Tec-S is $1249 for just the ECU.

Originally Posted by Thom
Not sure I understand the benefits of running a 60-2 tooth trigger, as the stock 132 tooth trigger set up should offer better resolution...?
The stock system is not a trigger wheel, is a starter ring gear! A proper trigger wheel, regardless of tooth count, has square teeth to give the best trigger signal. Some of the systems I have looked at that use the stock ring gear do not calculate engine rpm in real time, they calculate it once a revolution. On a fast accelerating engine this will cause timing errors. Real high performance ECUs calculate the crank position every tooth and as a result can give ¼ degree accuracy in the ignition timing. As A Wayne mentioned – put a timing light on it and see how steady the system is at idle. If it moves at all at idle you can only imagine how sketchy it will get at speed in dynamic conditions.

Originally Posted by reno808
how come??
See above…plus I have no wish to reuse old sensors in mission critical areas. The only old sensors that I ever reuse and coolant temp sensors – and if it’s a race engine I replace those too. I also want the ability to set the trigger up the way I want to – set the lead time of the home signal (missing tooth on a 60-2 system or a cam position sensor). The position of the home signal is dependent on the max ignition advance you want to use. Some systems are locked into a specific lead timing, some are not. Another reason for preferring not to use the stock trigger is accessibility. There is a definite benefit to having quick and easy access to any vital sensor on a track car. You never know when you may have to change something!

I am sure people can get het stock trigger system to work, and maybe even work well – I just don’t think its optimal and I am not into compromises when it comes to management systems – if you want to install and set one up then do it the best way possible.
Old 06-10-2011, 01:37 AM
  #80  
333pg333
Rennlist Member
 
333pg333's Avatar
 
Join Date: Feb 2006
Location: Australia
Posts: 18,926
Received 98 Likes on 81 Posts
Default

Pretty much what I've found. Seems as if Electromotive, Wolf, Motec, and also LINK have discovered this.
Many are in fact offering a front wheel kit for this reason. This is why I am interested in Thom's experiences?
Old 06-10-2011, 05:08 AM
  #81  
Thom
Race Car
 
Thom's Avatar
 
Join Date: Mar 2002
Posts: 4,329
Received 41 Likes on 31 Posts
Default

My experience with adjusting the base timing on my set up with a timing light is that it did not move one bit, and that was with watching the OT mark for a good 10 minutes to check how stable the system was (serious headache following...!). This tells me the CAS used on my set up is precise enough for this application.

As for tooth squareness :

a/ I'm not convinced that the starter ring gear is any less square than the tooth on a S/S2/968 flywheel - did a clutch job a 944S last weekend and the tooth on the flywheel didn't seem particulary "more square" than those on the ring gear.

b/ How could "tooth squareness" matter when used only for a reference sensor dedicated to measure engine speed only? Can the version of the factory Bosch ECU used with the 60-2 set-up (16V engines) compensate for the varying distance between the position sensor and the tooth during the arch of circle of the period of the tooth? I doubt it since the same position sensor is used on both 132 and 60-2 set ups.

c/ If the ECU cannot compensate for the varying distance between the sensor and the tooth, which will be varying greater the wider the tooth, the maths just say that for 132 counts for a revolution the sensor is triggered more than twice than with a 60-2 set up, hence better resolution at measuring engine speed.

Last edited by Thom; 06-10-2011 at 06:35 AM.
Old 06-10-2011, 06:19 AM
  #82  
Raceboy
Three Wheelin'
 
Raceboy's Avatar
 
Join Date: Jul 2005
Location: Estonia
Posts: 1,631
Received 17 Likes on 15 Posts
Default

On good systems, reference sensor is only used for determining TDC position and engine speed is calculated by teeth. Cam sensor indicates which TDC (1st or 4th cylinder) it is.
Old 06-10-2011, 08:48 AM
  #83  
Chris White
Addict
Rennlist Member

Rennlist Small
Business Sponsor

 
Chris White's Avatar
 
Join Date: Jul 2001
Location: Marietta, NY
Posts: 7,505
Likes: 0
Received 36 Likes on 27 Posts
Default

Originally Posted by Thom
My experience with adjusting the base timing on my set up with a timing light is that it did not move one bit, and that was with watching the OT mark for a good 10 minutes to check how stable the system was (serious headache following...!). This tells me the CAS used on my set up is precise enough for this application.

As for tooth squareness :

a/ I'm not convinced that the starter ring gear is any less square than the tooth on a S/S2/968 flywheel - did a clutch job a 944S last weekend and the tooth on the flywheel didn't seem particulary "more square" than those on the ring gear.

b/ How could "tooth squareness" matter when used only for a reference sensor dedicated to measure engine speed only? Can the version of the factory Bosch ECU used with the 60-2 set-up (16V engines) compensate for the varying distance between the position sensor and the tooth during the arch of circle of the period of the tooth? I doubt it since the same position sensor is used on both 132 and 60-2 set ups.

c/ If the ECU cannot compensate for the varying distance between the sensor and the tooth, which will be varying greater the wider the tooth, the maths just say that for 132 counts for a revolution the sensor is triggered more than twice than with a 60-2 set up, hence better resolution at measuring engine speed.
Looks like a little basic info is in order - a magnetic (regular or Hall Effect) sensor does not count ‘teeth’, it measures the ‘change of state’ of the trigger wheel. An electrical charge is created when the leading edge of the tooth passes by the sensor and another charge is created when the trailing edge passes the sensor. The amplitude of the signal is based on proximity. If the trigger tooth has any ‘ramp’ to it (less than completely square) it will create a signal with some analog function and the ECU will have to ‘guess’ at what voltage level represents the actual triggering event. Starter gears tend to have less that square teeth – they need to in order to mesh with the starter without causing undue wear.

All that being said – the 944 starter ring gear does have a profile that is ‘squarer’ than most.

As for the compensation for varying distance – its not an issue since the signal voltage generated by the small change is distance is minimal – it is also important to note that the signal delta will be opposite to the rising leading edge. Most good stand alones allow you to choose between a rising or falling trigger. If you chose the rising signal then the slight falling due to the increase in gap distance during the tooth passage is truly irrelevant. I will give you credit for a good try on that point!

In truth the stock ring gear should do an acceptable job in creating a signal for an ECU. My main objection to it is the idea that you can set up your new ECU using the old stock sensors. If you replace the stock sensor with new ones and make up your own wiring harness from the new sensors to your new ECU that will greatly reduce my objection to the implementation of a stock type set up. I still find the position of the sensors and the gap adjustability a maintenance annoyance! I still have a concern about the use of the trigger signal. Some of the systems that I have looked into that have the ability to use the stock trigger system do not calculate the engine speed at the passing of every tooth. Some just use a software based system that calculates the engine speed once a revolution. So you have more trigger events but a lot less final resolution due to the software algorithm.



Originally Posted by 333pg333
Then there's the added exercise of squeezing it in when running dry sump. Bit less room.
Not a problem……even with the auto tensioner in place!
Old 06-10-2011, 09:16 AM
  #84  
Thom
Race Car
 
Thom's Avatar
 
Join Date: Mar 2002
Posts: 4,329
Received 41 Likes on 31 Posts
Default

Thanks for the clarification.

It goes without saying that installing a modern ECU/wiring harness without all new sensors (OEM or not) makes little sense.

Yes, the factory location of the crank sensor is not the most convenient, but to be honest it should not need to be adjusted every other day and it still probably requires less work than installing a replacement sensor elsewhere.
Old 06-10-2011, 09:27 AM
  #85  
333pg333
Rennlist Member
 
333pg333's Avatar
 
Join Date: Feb 2006
Location: Australia
Posts: 18,926
Received 98 Likes on 81 Posts
Default

Originally Posted by Chris White
Looks like a little basic info is in order - a magnetic (regular or Hall Effect) sensor does not count ‘teeth’, it measures the ‘change of state’ of the trigger wheel. An electrical charge is created when the leading edge of the tooth passes by the sensor and another charge is created when the trailing edge passes the sensor. The amplitude of the signal is based on proximity. If the trigger tooth has any ‘ramp’ to it (less than completely square) it will create a signal with some analog function and the ECU will have to ‘guess’ at what voltage level represents the actual triggering event. Starter gears tend to have less that square teeth – they need to in order to mesh with the starter without causing undue wear.

All that being said – the 944 starter ring gear does have a profile that is ‘squarer’ than most.

As for the compensation for varying distance – its not an issue since the signal voltage generated by the small change is distance is minimal – it is also important to note that the signal delta will be opposite to the rising leading edge. Most good stand alones allow you to choose between a rising or falling trigger. If you chose the rising signal then the slight falling due to the increase in gap distance during the tooth passage is truly irrelevant. I will give you credit for a good try on that point!

In truth the stock ring gear should do an acceptable job in creating a signal for an ECU. My main objection to it is the idea that you can set up your new ECU using the old stock sensors. If you replace the stock sensor with new ones and make up your own wiring harness from the new sensors to your new ECU that will greatly reduce my objection to the implementation of a stock type set up. I still find the position of the sensors and the gap adjustability a maintenance annoyance! I still have a concern about the use of the trigger signal. Some of the systems that I have looked into that have the ability to use the stock trigger system do not calculate the engine speed at the passing of every tooth. Some just use a software based system that calculates the engine speed once a revolution. So you have more trigger events but a lot less final resolution due to the software algorithm.





Not a problem……even with the auto tensioner in place!
Hehe, I was hoping you'd post that. I couldn't find it. Thanks.
Old 06-11-2011, 01:00 PM
  #86  
Thom
Race Car
 
Thom's Avatar
 
Join Date: Mar 2002
Posts: 4,329
Received 41 Likes on 31 Posts
Default

This is what Ray Hall from Vi-PEC has to say about the triggering :

Originally Posted by Ray Hall
The [Vi-PEC] ECU does not use RPM. It reads every tooth. The time between the current tooth and the previous tooth is used to calculate when the next tooth should occur. It then looks to see if any ignition or injector events that need to be processed before the next tooth, and deals with them. This gives very accurate ignition and injector timing.
I suppose this means that the ring gear on our cars performs well enough when used with an ECU that can count each of the 132 teeth of the ring gear, such as the Vi-PEC.

I don't have any experience with the Link ECU but if it's as similar to the Vi-PEC as the user interface suggests, I suppose the Link will perform just as well.

From that point I'm not sure how a 60-2 set up could offer as much resolution, unless perhaps if using some compensating algorithm.

Last edited by Thom; 06-11-2011 at 01:20 PM.
Old 06-11-2011, 01:24 PM
  #87  
theedge
Addict
Rennlist Member
 
theedge's Avatar
 
Join Date: Aug 2003
Location: Canada, Eh?
Posts: 14,242
Likes: 0
Received 3 Likes on 3 Posts
Default

Originally Posted by Thom
This is what Ray Hall from Vi-PEC has to say about the triggering :



I suppose this means that the ring gear on our cars performs well enough when used with an ECU that can count each of the 132 teeth of the ring gear, such as the Vi-PEC.

I don't have any experience with the Link ECU but if it's as similar to the Vi-PEC as the user interface suggests, I suppose the Link will perform just as well.

From that point I'm not sure how a 60-2 set up could offer as much resolution, unless perhaps if using some compensating algorithm.
Vipec is similar to the Link you mean Link makes the Vipec.
Old 06-11-2011, 02:07 PM
  #88  
Chris White
Addict
Rennlist Member

Rennlist Small
Business Sponsor

 
Chris White's Avatar
 
Join Date: Jul 2001
Location: Marietta, NY
Posts: 7,505
Likes: 0
Received 36 Likes on 27 Posts
Default

Originally Posted by Thom
I suppose this means that the ring gear on our cars performs well enough when used with an ECU that can count each of the 132 teeth of the ring gear, such as the Vi-PEC.

I don't have any experience with the Link ECU but if it's as similar to the Vi-PEC as the user interface suggests, I suppose the Link will perform just as well.

From that point I'm not sure how a 60-2 set up could offer as much resolution, unless perhaps if using some compensating algorithm.
You identified the interesting part in your last statement. I can speak for the Electromotive system – which by the way developed, patented and licensed the 60-2 tooth trigger wheel system). The electromotive ECUs use a complex modeling system to use the trigger impulse, the current rotational speed and the rate of change to identify the current crank angle accurately. If you just look at the teeth and you are close, but not perfect due to the dynamic nature of an engine. In reality most of the current crop of EMS do a pretty good job, in the past there were some that had pretty significant errors in timing.

My main objection still stands – old sensors and wiring are not a good starting point for a good management system and the original set up is not easy to work on if track side tweaks or repairs are needed.
Old 06-11-2011, 02:08 PM
  #89  
Thom
Race Car
 
Thom's Avatar
 
Join Date: Mar 2002
Posts: 4,329
Received 41 Likes on 31 Posts
Wink

Hey Brian, does it really matter which of the chicken or the egg came first...
Old 06-11-2011, 04:34 PM
  #90  
Chris White
Addict
Rennlist Member

Rennlist Small
Business Sponsor

 
Chris White's Avatar
 
Join Date: Jul 2001
Location: Marietta, NY
Posts: 7,505
Likes: 0
Received 36 Likes on 27 Posts
Default

Originally Posted by Thom
Hey Brian, does it really matter which of the chicken or the egg came first...
Depends if you want an Omlet or a Chicken Sandwich!


Quick Reply: AEM EMS in a 944/951/952



All times are GMT -3. The time now is 03:57 PM.