Custom dizzy rebuild
Discussion
I'm about to send my dizzy to the Distributor Doctor for him to remove the existing gubbins and to set it up with a specific specification as far as the weights, springs and thus advance characteristics are concerned.
Both the original (I think) dizzy on my car, and a nearly new replacement that I bought last year, have around only 12-14 degrees of mechanical advance, and talking to him, he said he could set it up any way I wanted, and also for a very reasonable cost.
Bearing in mind that I want to run the car on full manifold vacuum, I was thinking of asking for around 20 degrees mechanical advance and say 10 degrees of vacuum advance, setting it up at 8 degrees at 950 rpm with the vacuum disconnected, to give 18 degrees with vacuum connected, maxing out at 28-30 degrees at 4k rpm, without vacuum. Should it have more than 20 degrees and if so, how would you set it up so that there was a reasonable amount of safety from the risk of detonation?
Is this also an opportunity to get him to dial in a specific advance curve between these limits, ie sharp rise, linear rise, inverse rise etc. Is that possible or am I overreaching as far as the practical possibilities?
Both the original (I think) dizzy on my car, and a nearly new replacement that I bought last year, have around only 12-14 degrees of mechanical advance, and talking to him, he said he could set it up any way I wanted, and also for a very reasonable cost.
Bearing in mind that I want to run the car on full manifold vacuum, I was thinking of asking for around 20 degrees mechanical advance and say 10 degrees of vacuum advance, setting it up at 8 degrees at 950 rpm with the vacuum disconnected, to give 18 degrees with vacuum connected, maxing out at 28-30 degrees at 4k rpm, without vacuum. Should it have more than 20 degrees and if so, how would you set it up so that there was a reasonable amount of safety from the risk of detonation?
Is this also an opportunity to get him to dial in a specific advance curve between these limits, ie sharp rise, linear rise, inverse rise etc. Is that possible or am I overreaching as far as the practical possibilities?
This is a good read:
http://www.s262612653.websitehome.co.uk/DVAndrews/...
The absolute maximum is 34' at WOT on a 3.9, a little more on the bigger engines due to the slow fame from in the larger combustion chamber, but I would not go beyond 30'. You really cant work out a different advance curve without rolling road time, as you set it for peak torque at any given RPM, to the point the torque starts to drop off as you become over advanced. This will happen significantly before detonation occurs.
http://www.s262612653.websitehome.co.uk/DVAndrews/...
The absolute maximum is 34' at WOT on a 3.9, a little more on the bigger engines due to the slow fame from in the larger combustion chamber, but I would not go beyond 30'. You really cant work out a different advance curve without rolling road time, as you set it for peak torque at any given RPM, to the point the torque starts to drop off as you become over advanced. This will happen significantly before detonation occurs.
Thanks, Mark, it is interesting reading.
Martin at Distributor Doctor told me that 2 degrees is often found to be added or lost when he tests the total advance of a Lucas dizzy, from the number stamped on one of the weights. As you know, an '8' stamp signifies a maximum advance of 16 degrees but in my case, I was below a few degrees below this.
He said he has charts for the standard curves, but also said there are no standard charts for TVR's, which is not a surprise, although as he often is asked by customers to build dizzys to specific instructions for race engines, he was quite happy to talk about different characteristics and options.
It would be interesting to see what sort of chart those who have mapped their own ecus have found, and to see if there is a common theme there in the way they set up the advance after their own testing.
Martin at Distributor Doctor told me that 2 degrees is often found to be added or lost when he tests the total advance of a Lucas dizzy, from the number stamped on one of the weights. As you know, an '8' stamp signifies a maximum advance of 16 degrees but in my case, I was below a few degrees below this.
He said he has charts for the standard curves, but also said there are no standard charts for TVR's, which is not a surprise, although as he often is asked by customers to build dizzys to specific instructions for race engines, he was quite happy to talk about different characteristics and options.
It would be interesting to see what sort of chart those who have mapped their own ecus have found, and to see if there is a common theme there in the way they set up the advance after their own testing.
If you take the distributor out and remove the springs you can twist the outer central shaft in relation to the gear - so fit a protractor over the shaft and see how many degrees you get from weights fully in (no advance) to out- as you twist the shaft against the locked gear. you can see the weight stop points when you do this, so using a dremmel or like you can remove material from the back of the weight to add advance- you need to do both at same time. Dont forget 1 distributor degree is 2 crank degrees, so its not a lot of metal.
I already measured the dizzy angles in this way and again, even though it showed an 8, this method also confirmed what the adjustable strobe was telling me, that it was only giving me about 14 degrees of movement.
To dremmel off 3mm more from the weight as you said should not be to much of a problem as you say, but where Martin charges about 40 quid to recalibrate it with different weights and springs (over the phone quote, mind), I prefer to do it that way.
To dremmel off 3mm more from the weight as you said should not be to much of a problem as you say, but where Martin charges about 40 quid to recalibrate it with different weights and springs (over the phone quote, mind), I prefer to do it that way.
jazzdude said:
Bearing in mind that I want to run the car on full manifold vacuum, I was thinking of asking for around 20 degrees mechanical advance and say 10 degrees of vacuum advance, setting it up at 8 degrees at 950 rpm with the vacuum disconnected, to give 18 degrees with vacuum connected, maxing out at 28-30 degrees at 4k rpm, without vacuum. Should it have more than 20 degrees and if so, how would you set it up so that there was a reasonable amount of safety from the risk of detonation?
The above is more or less ideal with the exception of total advance which (depending on engine spec and fuel octane) can be pushed to 33 degrees from 4,000 to 4.500rpm but after that you actually want less timing, but I'll come to that later. Do also keep in mind you want a bit of adjustment in hand and the difference between idling at 15 - 20 degrees is negligible, although 18 is the ideal in my experience.You may well find going with 10 degrees at idle actually gives you 18 degrees when you connect the vacuum advance (+8), and because cruise vacuum is higher than idle vacuum you will enjoy more than 8 degrees of advance when cruising (say +10 to +12) but the thing is you really want more than that if you can get it. The engine will actually tolerate and deliver best fuel economy at 47 degrees when at the absolute optimal cruise conditions although there is not much difference 45 degrees, I'd ask the Distributor Doctor if he can give you a maximum of 15 degrees from the vac advance at peak vacuum instead of the normal 10ish.
jazzdude said:
Is this also an opportunity to get him to dial in a specific advance curve between these limits, ie sharp rise, linear rise, inverse rise etc. Is that possible or am I overreaching as far as the practical possibilities?
My recommendation, and based on my own ignition tuning work using an after market stand alone engine management system, is to build an aggressive (steep) curve that piles on the timing early in relation to engine speed. This is however also load dependent but you're very restricted in that respect with a distributor, limited to the vacuum advance in fact which only adds timing at cruise and at idle, but only at idle if you run full manifold vacuum.Proper ignition timing tuning involves a dyno and time, and with a distributor you need lots of time, this is because the distributor will require internal surgery before you can retest to see if your last combination of bob weights and springs made an improvement or a lost you torque at a certain engine speed. Even if you're only working to 1,000rpm increments, which is quite a crude way to build a curve really, the process is always going to be super time consuming.
As an absolute minimum it could easily take five runs and five distributor internal changes to find the right combination of bob weights and springs per 1,000rpm, so that's 5 x 5 = 25 changes and 25 runs on the dyno and this just to build a basic curve, so it's a little naive to assume sending you're distributor away to be modded off the car is going to give optimal results.
I'm afraid you must accept with a distributor you'll only ever end up with a compromise, which is the precise reason distributorless ignition systems were developed replacing the distributor over 25 years ago and we never went back, however aim to get as close to the following figures as you can and you should get a respectable result.
- Idle speed 1,000rpm
- Idle vacuum advance disconnected = 10 degrees
- Idle vacuum advance connected = 18 degrees
- 1,200rpm = 25 degrees
- 1,500rpm = 27 degrees
- 1,800rpm = 29 degrees
- 2,000rpm = 31 degrees
- 2,500rpm = 31 degrees
- 3,000rpm = 31 degrees
- 3,500rpm = 32 degrees
- 4,000rpm = 33 degrees
- 4,500rpm = 33 degrees
- 5,000rpm = 31 degrees
- 5,500rpm = 30 degrees
- 6,000rpm = 29 degrees
- 6,250rpm = 28 degrees
While the engine will tolerate and indeed wants 33 degrees around 4,000 to 4,500rpm (peak torque), with a distributor you're forced to set your maximum advance lower at say 30 degrees because that's the figure you'll get from just over 3,000rpm right the way to 6,250rpm and above, if you run 33 degrees of total mechanical advance you may be putting your engine at risk of detonation above 4,800rpm so best to play it safe at 30.
Of course every engine is different, the above figures are based on my 4.0 litre Rover V8 with a V8 Developments Stealth cam and a few other bits and bobs running on 95 Ron petrol, but the basic rules remain more or less the same, well on petrol anyway..... things become rather different on LPG
Thanks Dave, that's great information, exactly what I needed. I will move ahead with 20 degrees of mechanical with an aggressive curve as described.
With 15 degrees of vacuum advance, will it still idle at around 18 degrees if set at 10, vac disconnected?
I assume that raising the idle from its ECU governed 950 rpm to the 1000 rpm you recommend, is by clamping the stepper hose and using the base idle screw while setting the dizzy. Which would mean my getting it changed in the map for it to stay at 1000 rpm afterwards, which can be done and burnt back into the eprom.
On my already running nicely car on map 2 and this improved dizzy and full manifold vacuum, I'm going to also add a wideband lambda and guage, so I can also tweak the afm a little too, and hopefully bring this old 60s v8 a little bit closer to the levels those with modern Ecus are experiencing.
I resigned myself to the fact that for this car, without the services of a local experienced ecu installer and mapper, it was not worth the risk in ripping it all out and diy fitting a new system, that might have ended up with me being not that much better off if I cocked it up.
With 15 degrees of vacuum advance, will it still idle at around 18 degrees if set at 10, vac disconnected?
I assume that raising the idle from its ECU governed 950 rpm to the 1000 rpm you recommend, is by clamping the stepper hose and using the base idle screw while setting the dizzy. Which would mean my getting it changed in the map for it to stay at 1000 rpm afterwards, which can be done and burnt back into the eprom.
On my already running nicely car on map 2 and this improved dizzy and full manifold vacuum, I'm going to also add a wideband lambda and guage, so I can also tweak the afm a little too, and hopefully bring this old 60s v8 a little bit closer to the levels those with modern Ecus are experiencing.
I resigned myself to the fact that for this car, without the services of a local experienced ecu installer and mapper, it was not worth the risk in ripping it all out and diy fitting a new system, that might have ended up with me being not that much better off if I cocked it up.
jazzdude said:
Thanks Dave, that's great information, exactly what I needed. I will move ahead with 20 degrees of mechanical with an aggressive curve as described.
With 15 degrees of vacuum advance, will it still idle at around 18 degrees if set at 10, vac disconnected?
The vacuum advance unit has a curve just like a distributor, in both cases the Y axis is degrees of timing but with the X axis engine speed is replaced with vacuum measured in HG (inches of vacuum) or kPa (Kilopascals). Timing is added in direct proportion to the amount of vacuum the vac advance capsule sees just as the distributor mechanical advance mechanism adds timing in direct proportion to rising engine speed.With 15 degrees of vacuum advance, will it still idle at around 18 degrees if set at 10, vac disconnected?
A vacuum advance only has a limited range of movement so there's a mechanical limit to the amount of timing the capsule can add, you will need to talk to the Distributor Doctor about what he can give you but before worrying yourself about the amount of timing it can add far more critical is to ensure the capsule adds nothing at all when you move from cruise into even light acceleration. This transition point and understanding the vacuum figure it occurs at is essential or you'll risk detonation, the other important consideration is how rapidly it transitions from it's cruise condition to snap acceleration and how aggressively it adds timing when vacuum increases.
As we can see just tuning the vacuum advance is a science in itself, many Yank after market distributors from the likes of MSD came with an adjustable vac advance capsule but as you will not have this advantage on the Lucas distributor I would ask the Distributor Doctor to give you a capsule that adds 8 degrees of advance at idle at 45 kPa ramping rapidly to 15 degrees at 38 kPa at cruise. Above 45 kPa you need to make sure the capsule is adding less and less timing in an equally aggressive way, so by the time you're at 55 kPa and above the vacuum advance capsule adds nothing at all, like this as you accelerate you'll be falling back onto your mechanical advance curve wich should be built to give best mean torque without promoting detonation.
- 55 kPa (light acceleration) = 0 Degrees
- 50 kPa (normal driving below 2,000rpm) = 4 Degrees
- 45 kPa (idle) = 8 Degrees
- 40 kPa (entering cruise) = 12 Degrees
- 35 kPa and below (cruise) = 15 Degrees
Big changes in timing are the enemy of smooth running and distributors already tend to be jumpy especially at idle, no engine idles exactly at one fixed vacuum figure either, your camshaft profile will have a significant influence on this. The more aggressive the cam profile the more your idle vacuum will jump up and down, even if you are running a fairly mild cam vacuum at idle will likely hop between 45 kPa and 50 kPa so if you use my above figures you the run very real risk of rapidly adding and subtracting 4 degrees of timing at idle which will give you a choppy idle quality.
Some of this can be mitigated by adding timing itself, this is why raising the idle timing from the stock 8-10 degrees to the full vacuum assisted figure of 18 degrees works so well to help give you a smoother idle. The second way to improve vacuum stability is to increase the base idle speed, this is why we dial in a higher idle speed when we fit a more aggressive cam and conversely why you can achieve a lovely smooth low RPM idle on a mild cam profile. The Stealth cam I run dictates the best balance between smooth idle and lowest RPM is found running 18 degrees of timing on petrol at 1,050rpm and a whopping 28 degrees on LPG at the same target idle speed.
However I should be clear that in my case the Canems ECU controlled closed loop ignition timing based idle strategy (scatter spark) has the authority to alter the timing figures by a full 8 degrees either side these base idle timing numbers, as an example the ECU may actually drag my LPG idle timing down 6 full degrees from 28 degrees to 22 degrees to pull idle speed down enough to hit my 1,050rpm target. Equally on petrol it may add timing from 18 degrees to raise my idle speed fractionally again to hit the idle target of 1,050rpm, the truth is scatter spark is constantly but slowly adding and subtracting timing precisely on either fuel, all the driver experiences in a rock steady 1,050rpm idle... always!.
jazzdude said:
I assume that raising the idle from its ECU governed 950 rpm to the 1000 rpm you recommend, is by clamping the stepper hose and using the base idle screw while setting the dizzy. Which would mean my getting it changed in the map for it to stay at 1000 rpm afterwards, which can be done and burnt back into the eprom.
By instructing the stepper motor to bleed in more or less air engine side of the throttle butterfly the Lucas 14CUX ECU will shoot for the idle target set within the ECU, but you can squeeze things up from the 950rpm target to 1,000rpm without making changes to the chip. However you absolutely must follow the correct procedure for setting your base idle which is not as you describe above, as you are intending to run full manifold vacuum you should first disconnect your vacuum advance, you do then need to clamp the stepper hose but with the stepper motor out of the game you are looking to set the base idle screw so the engine idles low at 700 to 750rpm.Even when the engine is fully warm the stepper adds air to the value of a 250rpm increase once it's hose is unclamped, the process of hitting your target idle speed becomes a little more complex if you're running full manifold vacuum because as soon you reconnect to full manifold vacuum it's going to give you at least 8 degrees of timing which will raise your idle timing by at least 200rpm itself which in turn will add mechanical timing too. So you'll need to get back on the base idle screw winding it slowly in to calm things back down to 1,000rpm at 18 degrees of full manifold vacuum supported idle timing.
Once the stepper feed hose is unclamped there's still work to be done, to be honest it's all a bit of a balancing act between the base idle screw air bleed, your target idle speed, and hitting the perfect 18 degrees of idle timing. This is difficult enough in itself but remember the ECU will also likely be constantly be moving the stepper motor in and out as it trys to hit the target idle speed, once the hose is unclamped this can make finding the final settings virtually impossible. Also when fitted with the overly small plenum arrangement we have the Rover V8 becomes an engine that's very sensitive to the volume of air bleed especially at idle, a stable fixed volume of air is what you want to tune a nice stable idle and good low speed drivability, but sadly unless everything is perfect including the health of the ignition system, the idle ignition timing chosen, the base idle setting and ultimately how well the engine block is electrically earthed the stepper motor can conspire against you in this respect.
On my Chimaera I have a more sophisticated Bosch PWM Canems ECU rotary idle air control valve, well it's actually a cheap pattern part fake Bosch idle valve that came with my Canems inhalation, rather than moving in steps as is the case with the Lucas stepper motor it offers and infinitely variable air bleed. Sounds great right? well not really as I have recently suffered an earth loop where the idle valve was causing the ECU to loop the priming pulse on pre-cranking key on conditions, so as an experiment I removed my idle valve to explore just how much the engine actually needs it?
This was quite revealing, my tests proved just how little a correctly setup Chimaera actually needs the involvement of an ECU managed extra air valve. OK so I do have the additional advantage of scatter spark which TBH is truly game changing, if configured correctly scatter spark actually completely removes the need for an extra air valve in 99% of all situations which is what I suspect Powers discovered when they designed the MBE setup for Chims and Griffs and the reason they chose not fit an idle valve at all. With the air bleed no longer constantly going up and down a far more stable idle can be achieved and low speed drivability is enhanced too.
I suspect the problems I've encountered with my Canems controlled Bosch idle valve are more down to earthing issues that are so common on our cars so I'll be addressing this in the spring myself as the ECU installers themselves failed to correct the problem. It's also important to point out there is one condition where the engine does need extra air, cold start, however its surprising how little extra air it actually needs even under this condition. Under cold start cranking conditions and with coolant temps between 0-40 degrees you do need a declining taper of extra air but that's it, this is of course aided by my Canems coolant temp ignition correction and scatter spark features so it may be a 14CUX and distributor Chimaera engine needs a little more help in the form of extra air?
To solve my looping priming pulse issues and until I'm able to properly turn on the earthing issues I simply went back to my old school carburetor tuning experience by fitting a manually controlled cable operated fast idle system. Basically I built a simple manual extra air valve system to give me the additional air required during cold start cranking and the 0-40 degrees coolant temp warm up phase. As soon as coolant temp passes above 40 degrees I can push the knob fully home and for the rest of the day it is not required at all.
With my manual extra air valve knob pushed fully home the engine falls back on the tiny fixed air bleed from the base idle screw, a bit of a dirty air bleed from the crankcase breather setup that in my case is now properly valved, and of course ignition timing which in turn is controlled by the Camnems ECU through closed loop idle using scatter spark to precisely maintain my 1,050rpm target. The results are very noticeable improvements in idle stability and low speed drivability, clearly my pattern part Chinese fake Bosch idle valve was doing things the duty cycle figure on my laptop was not revealing. My manual idle valve is a simple old school solution to the requirement for extra air that designers solved way back in vintage car days by providing a hand throttle.
Based on the clear inarguable results of my experiments it's hard not to come to the conclusion that 99% of the time a Chimaera simply doesn't need the sophistication of an ECU controlled idle valve at all, be that a Bosch PWM valve or the old Lucas stepper motor. In my case and for a number of reasons the idle valve simply gave more trouble than it was worth, this may however change when I improve my engine earthing. But I still subscribe to the lesson drummed into me when I trained as a mechanic ie "all vacuum leaks are bad" and "you can't expect to tune an engine effectively when vacuum leaks are present", Ok so this training related to tuning carb fed engines but an internal combustion engine is still an internal combustion engine despite if it is on injection or a carb.... so as far as I'm concerned these fundamental rules still stand.
It's worth remembering all extra air valves (idle valves) do is introduce an managed vacuum leak so if there are earthing issues, the ECU itself has issues, or if the valve is a poor quality pattern part it's going to be beneficial to replace the system with a simple cable operated fast idle valve as I have successfully implemented myself. The complex PWM electric idle valve and ECU are simply replace with a manual valve controlled by my brain and hand, no annoying earth loops and when I push the knob in I know the valve is shut so I know there's no air bleed or what the old mechanics I worked with would have called a vacuum leak
jazzdude said:
I resigned myself to the fact that for this car, without the services of a local experienced ecu installer and mapper, it was not worth the risk in ripping it all out and diy fitting a new system, that might have ended up with me being not that much better off if I cocked it up.
Absolutely the right approach, with a proper tune/setup and few tricks like running full manifold vacuum you can expect the engine to be super smooth, start with the basics like HT leads, setting your timing, cleaning the air filter, properly setting up the base idle, TPS and throttle butterfly rest position and you'll be making progress for sure.Further improvements come from replacing the dreadful the inappropriate NGK B7ECS with a set of BPR6ES or better still BPR6EIX and deleting the failure prone plug extenders. improve the engine earthing and maybe even delete the stepper motor altogether replacing it with a manually controlled fast idle valve as I have done can make a massive improvement too.
At the end of the day it's just an old Rover V8, it's an engine that ran beautifully smoothly in old Rover P5s & 6s and early Range Rovers on just a par of SUs and a distributor, so there's really no reasons why it shouldn't run just as well if not better on the Lucas 14CUX injection system which indeed it did on the Range Rovers fitted with the exact same setup.
Saying that the more emissions based add-ons, ignition timing and fueling strategies Land Rover threw at the venerable but polluting old Rover V8 to make it meet ever stricter emissions standards the more it becomes difficult to tune for a nice idle and smooth running. This is not an engine that likes to run an emissions friendly retarded 8-10 degrees idle and it's not keen of running lean at idle either, while 14.7 parts air to 1 part fuel by mass is indeed chemically perfect combustion for petrol it is in fact for old Rover an inappropriately lean diet.
I would recommend to everyone to remove as much emissions junk as possible and return the engine to a fuel and ignition strategy it's designers intended when they created the power plant back in the late 1950's, the TVR exhaust manifolds are also a very poor design too with very unequal length headers. The front cylinders get a super short 5" run while the rears get a far more appropriate longer run that's at least three to four times longer than the font cylinders, so even if you cant afford a better equal length setup or the 180 degree masterpiece from Clive Ford you should at least ditch the nasty pre-cats and even more importantly delete the very restrictive dustbin sized main cat.
As we can see there's a lot of things that can be done to make a standard 14CUX and distributor equipped TVR Chimaera run a lot smoother and for very little money, and all this before you tackle the fueling side by changing the chip. The engine just needs some sympathetic and logical finessing at the hand of someone who knows what they are doing the results being near after market ECU levels of drivability are really within the grasp of all for a very small investment.
Finding a right minded MoT tester who is happy to overlook the emissions test is key, with this hurdle overcome you're set to progress to unstrangling the engine and giving it what it really wants.... at which point it will thank you with a super smooth idle and great drivabiliy
See, no after market ECU required
Gassing Station | Chimaera | Top of Page | What's New | My Stuff