Dizzy caps and Amethyst ignition
Discussion
Any of you nice people have a knackered Dizzy cap laying around? I can see an issue putting an Amethyst on the 8 cylinder engine, as you have 8 HT lobes per revolution. This means you have only a 45' degree segment to play with before the rotor arm moves to the next cylinder, and the width of the rotor tip covers 35' of this, so the tolerances are pretty tight. So if you lock out all the mechanical timing, you need to ensure the rotor tip is just leaving the pick up point with no advance, to allow the approx 37' timing advance available over the rotor tip width. Problem is you cant see the rotor tip in relation to the pick up point with the cap on- so I need a knackered one I can cut open to see the rotor tip position at firing point. I can then trim the position of the trigger head by moving the vacuum timing plate and dizzy body so it all lines up before I lock it. It would be a pity to have to cut a good one up- obviously happy to pay postage.
Im fitting one of these to give me fully mappable ignition timing, that combined with the work that has gone on on the 14CUX mapping will give me both fuel and ignition control.
http://www.aldonauto.co.uk/shop/product.asp?P_ID=8...
Its a cheaper option than the magajolt or Megasquirt, although not quite as good as wasted spark and coil packs, it should be loads better than the clockwork dizzy weights and vacuum for not a huge investement
The instructions make it look very easy, but I dont yet have a decent base map,so Im trying to modify a Megasquirt one, but they are calibrated differently. The bit with the dizzy cap is just to make sure I dont stray out of the available period when the rotor arm is in line for any particular plug lead as the advance control is now external to to the dizzy.
http://www.aldonauto.co.uk/shop/product.asp?P_ID=8...
Its a cheaper option than the magajolt or Megasquirt, although not quite as good as wasted spark and coil packs, it should be loads better than the clockwork dizzy weights and vacuum for not a huge investement
The instructions make it look very easy, but I dont yet have a decent base map,so Im trying to modify a Megasquirt one, but they are calibrated differently. The bit with the dizzy cap is just to make sure I dont stray out of the available period when the rotor arm is in line for any particular plug lead as the advance control is now external to to the dizzy.
The timing requirements for the RV8 change when you tune the engine as you alter the engines ability to fill the cylinders at different RPM with a cam change, typically lower rpm it becomes worse when the engine is off cam and better at higher RPM. You need more timing advance to compensate for poor filling as the burn is slower to get the best power, and less when the cylinders at at full pressure. I dont know if TVR ever modified the timing over the Range Rover set up, but the Ginetta is stock timing, so Im sure I can improve it. Shunting is certainly a by product of poor cylinder fill / poor combustion
Edited by blitzracing on Thursday 2nd January 13:37
Mark .. don't forget the dizzy spins at half engine speed, so your 35degrees of rotor arm is actually 70 degrees of crank rotation. I rarely see any gains on part throttle past about 45deg of advance, so your working tolerances aren't as narrow as you might think. I've done a few ecu conversions retaining the std dizzy (locked) and never had a cross firing issue.
spitfire4v8 said:
Mark .. don't forget the dizzy spins at half engine speed, so your 35degrees of rotor arm is actually 70 degrees of crank rotation. I rarely see any gains on part throttle past about 45deg of advance, so your working tolerances aren't as narrow as you might think. I've done a few ecu conversions retaining the std dizzy (locked) and never had a cross firing issue.
Spot on, the little light went on after I posted. 
Aldon have been helpful getting the MS map transposed into their mapping system. It had me highly confused as the MS maps have wide open throttle at 100 kpa or so and drop from this point to represent a reduction in pressure below atmosphere as the timing advances. The Amethyst plots start at atmosphere, then "adds" a vacuum figure against ignition advance. Its still going to take some head scratching though..
Getting somewhere now.
Here's a screen shot of the mapping info- pretty simple really- you simply have two sets of parameters- RPM vs Advance and Vacuum vs Advance. The confusion lay in the fact that the vacuum figure is measured in inches mercury, not as a negative value below atmosphere (where 31" is about 1 bar, so I expected the value to go down towards 0 as the vacuum increased,) but as a positive value, so the inches mercury goes up as the vacuum goes up so the opposite side of the scale allows for turbo use where the figures become negative.
So Ive taken a 6 degree static timing point, that has to be added onto all of the figures under the centrifugal advance values- so it peaks at 36'. In addition to this you can add up to 12' of vacuum advance between 4 "hg (-14KPA) and 14"hg (-47KPA,) but the unit is programmed for a peak of 37'total so it wont go beyond that. I see on the mega squirt maps up to 43' peak under very light loads.
The unit does seem to be limited up to 14" hg max as this is regarded as the maximum vacuum level reached realistically, and I dont think the mapping approach allows you to put in individual timing "cells" as the MS system, but its a good start.
The system has a lovely plot facility where you simply drag the plotting points on the maps around, and it works out the values for you.
I would greatly appreciate comments on the basic timing plots from those who see RV8 maps.(Joolz?) The engine is a 3.9 with a piper 270' cam peaking at around 6200 rpm.
Here's a screen shot of the mapping info- pretty simple really- you simply have two sets of parameters- RPM vs Advance and Vacuum vs Advance. The confusion lay in the fact that the vacuum figure is measured in inches mercury, not as a negative value below atmosphere (where 31" is about 1 bar, so I expected the value to go down towards 0 as the vacuum increased,) but as a positive value, so the inches mercury goes up as the vacuum goes up so the opposite side of the scale allows for turbo use where the figures become negative.
So Ive taken a 6 degree static timing point, that has to be added onto all of the figures under the centrifugal advance values- so it peaks at 36'. In addition to this you can add up to 12' of vacuum advance between 4 "hg (-14KPA) and 14"hg (-47KPA,) but the unit is programmed for a peak of 37'total so it wont go beyond that. I see on the mega squirt maps up to 43' peak under very light loads.
The unit does seem to be limited up to 14" hg max as this is regarded as the maximum vacuum level reached realistically, and I dont think the mapping approach allows you to put in individual timing "cells" as the MS system, but its a good start.
The system has a lovely plot facility where you simply drag the plotting points on the maps around, and it works out the values for you.
I would greatly appreciate comments on the basic timing plots from those who see RV8 maps.(Joolz?) The engine is a 3.9 with a piper 270' cam peaking at around 6200 rpm.
36 total deg under load is a lot for a smallish 3.9 chamber especially if you've not got any knock detection in place, other than your ears, on the open road. I'd be taking 10 degrees out of that and creeping up on the higher numbers gradually. Trouble is, out on the road your seat of pants dyno will never tell you about the last few percent of hp left in the engine, you're just not sensitive enough to feel it. It will get you somewhere in the ball park though, and trim the last bit on the dyno.
Also it seems I was wrong about needing a separate vacuum feed from the intake.The rather unusual way the amethyst software works means it's behaviour is exactly like a distributor in that it has a basic (albeit configurable up and down at 250 rpm intervals) curve with a vacuum sensed trim applied over the top to give the 3d map.
Also it seems I was wrong about needing a separate vacuum feed from the intake.The rather unusual way the amethyst software works means it's behaviour is exactly like a distributor in that it has a basic (albeit configurable up and down at 250 rpm intervals) curve with a vacuum sensed trim applied over the top to give the 3d map.
Mark
You’re a very brave man and I’m sure lots of people are lurking with great anticipation, well I certainly am.
You’ve raised an interesting point about the width of the rotor arm because I guess the mechanical advance advances the spark and the rotor arm. After Jool’s pointing out the crank advance is halved on the rotor arm I still can’t stop thinking you need to lock the advance at the leading edge of the rotor arm so max electronic advance doesn’t go past the rotor arm.
I’m confident the rotor arm will be fine as Jools has done a few ecu conversions retaining the std dizzy (locked) and never had a cross firing issue.
Great project and well done, Steve
You’re a very brave man and I’m sure lots of people are lurking with great anticipation, well I certainly am.
You’ve raised an interesting point about the width of the rotor arm because I guess the mechanical advance advances the spark and the rotor arm. After Jool’s pointing out the crank advance is halved on the rotor arm I still can’t stop thinking you need to lock the advance at the leading edge of the rotor arm so max electronic advance doesn’t go past the rotor arm.
I’m confident the rotor arm will be fine as Jools has done a few ecu conversions retaining the std dizzy (locked) and never had a cross firing issue.
Great project and well done, Steve
37' at the dizzy is over 70' crankshaft degrees, so I think the storys about lack of advance on the dizzy is a vicious rumour is put about by the MS boys who run coil packs - 
(thats where I got the worries from anyway). I think I need to start by wiring it in and leaving the dizzy unmodified to start with, and running on map 8 that is basically all zeros, so basically you run it just as an ignition trigger and rev limiter. Once Im happy it works reliably, then lock the weights and switch across to a preconfigured map for testing. The unit can be programmed off the car with just 12 volts on it.
(thats where I got the worries from anyway). I think I need to start by wiring it in and leaving the dizzy unmodified to start with, and running on map 8 that is basically all zeros, so basically you run it just as an ignition trigger and rev limiter. Once Im happy it works reliably, then lock the weights and switch across to a preconfigured map for testing. The unit can be programmed off the car with just 12 volts on it.
Mark,
recently fitted one of these units onto my RV8 3.9l. Seemed pretty straight forward, however am having a few problems. By the way you say that total advance is 37 degrees but this is also a programmable parameter and I have set mine at 45 degrees. My problem is that when I started to compare the timing shown on the PC console with that from a timing light there was a big difference - up to 7-8 degrees(taking the 6 static into account of course!) so might be worth checking that out. As for suggested map, try this out (my static is also 6 degrees) which was taken from a heavily fiddled with conventional distributor and worked pretty well.
r.p.m advance total (with the 6 degrees static)
500 1 7 (kept this pretty low just for starting up)
1000 10 16
1500 16 22
2000 21 27
2500 25 31
3000 27 33
and above
These engines do seem to like a fairly abrupt curve, and some people even suggest all in at about 2700 rpm.
Like to hear how you are getting on with it.
recently fitted one of these units onto my RV8 3.9l. Seemed pretty straight forward, however am having a few problems. By the way you say that total advance is 37 degrees but this is also a programmable parameter and I have set mine at 45 degrees. My problem is that when I started to compare the timing shown on the PC console with that from a timing light there was a big difference - up to 7-8 degrees(taking the 6 static into account of course!) so might be worth checking that out. As for suggested map, try this out (my static is also 6 degrees) which was taken from a heavily fiddled with conventional distributor and worked pretty well.
r.p.m advance total (with the 6 degrees static)
500 1 7 (kept this pretty low just for starting up)
1000 10 16
1500 16 22
2000 21 27
2500 25 31
3000 27 33
and above
These engines do seem to like a fairly abrupt curve, and some people even suggest all in at about 2700 rpm.
Like to hear how you are getting on with it.
I was trying to work out how the best point to trigger the unit will be, as by its nature the magnetic pick up produces a wave of varying voltage depending on how fast the engine is going, so the trigger point will change against RPM to an (unknown) extent, so Ive stuck with the amp so at least you have a square(ish) wave as a consistent trigger point. I would like to trigger at TDC with a strobe, (unlike the stock set up that will be around 6-10'BTDC ) , and let the unit add the advance to this, but Ive yet to fire it up and see how the trigger rotor alignment sits with the amp trigger point. Ive removed the vacuum advance mechanism, so I not have a good range to move the trigger pick up to make sure the rotor is aligned to get the most out of the advance range. Once Im happy Ill lock the plate. Don't forget the RV8 has a ported vacuum take off that far as I can see shuts the vacuum off as the throttle plate shuts, so Ive added a new take off to remove this unknown when mapping. All theory so far until I fire it up however hopefully over the next week or so. BTW I think 45' total advance would be engine destroying under any engine condition(?).
Gassing Station | Chimaera | Top of Page | What's New | My Stuff