Variable Valve Timing – and why its important. A few years back Motorsport Electronics took it upon our selves to reverse engineer Ford’s elusive VVT as used on the ST170, and create initially a stand-alone VVT Controller (the VVT Pro) and have since we have integrated it into all of our ECUs. We document the process, the discoveries, and explain why if your engine has VVT – you would SILLY to not make the most of it!
This project came about when looking at torque curves of the ST170 – when using aftermarket engine management, the normal option was to delete the ST170s VVT, locking the inlet cam at a certain position – and there was a mass of misinformation and just plain misunderstanding flying around the internet. You could choose a torquey bottom end, or a peaky top end – but you sacrificed power somewhere. VVT simply allows the cam timing to change based on things such as load and RPM – resulting in optimisation of power throughout the rev range – why give that up?
The reason: Mainly due to the expense. Very few ECUs at the time offered what we call “closed loop” VVT – that is, reading the cam position from the camshaft sensor, and comparing it to the crank position to asses the cams relative angle, then changing oil flow through an electrically operated solenoid to move the cam. Most ECUs (and many still do) only allow the user to turn the oil flow “full-on” or “all-off” at a certain RPM setpoint – never a good idea, and never going to get the best from the engine!
Even if you modulated the valve in an Open-Loop method and said ‘30% on’ at x-rpm and ‘75% on ay Y RPM’, it still isn’t going to be right – that valve will flow different amounts of dependant on lots of factors, oil temp, viscosity, engine speed, resulting in the cam being potentially at any angle – most likely upsetting fuel tables as well!
So – the first step – measuring how the OEM ECU does it…
First we bought a stock ST170 to work on – unmodified, making around the normal 150ish BHP (not 170 as some would like to believe!)
Using an oscilloscope measuring various things such as crank/cam patterns and voltages we made the following observations:
- The VVT solenoid is driven by 12V, switched to ground by the OEM ECU
- The cam swings approximately 45 degrees from rest to fully advanced – the valve “failed” it sits at rest.
- It is controlled by the OEM ECU in a closed loop manner.
- Simply driving it with a PWM freq. will not work as oil pressure/temp/wear etc will effect the VVT valves flow, and hence cam position.
- The Ford ECU monitors the cam position in relation to the crank position and then uses a PID control loop to Adjust the PWM duty to reach the desired cam angle, looked up in a map that is defined AT LEAST by RPMxLOAD.
- VVT actuation does not begin until at least 1300 RPM.
With the stock crank pattern (36-1) and the now discovered cam trigger pattern of “8-3” found, we set about writing a simple decoder that would measure the angle of the missing tooth of the crank in relation to the extra tooth of the cam. We then did some WOT throttle runs, allowing the stock ECU to control everything, while we “snooped” this data. Plotting this on a graph proved very interesting.
What is worth noting is that control strategies such as ‘turning the VVT solenoid fully on at 1500rpm’ or other such tricks are NOT going to give the results that are needed. The cam is clearly shown to fully advance and then begin to taper back as RPM increases – resting around “mid advanced” from around 5k RPM to the redline. One thing many find if they run an ST170 without cam control, is the engine becomes VERY flat above 5,000RPM, and this may indeed be the reason…
Talking to a few of our dealers – it became apparent (and this was before the launch of our ME221 ECU), that a nice ‘add-on’ option of a little box of tricks that allowed the end user to tune their VVT in conjunction with whichever ECU they choose to use (or even carbs) would be an excellent option for a product to design. With the new data in hand, and wanting to carry out more experimentation, the VVTPro was designed, software written and a tuning interface devised. Still for sale now, the VVTPro gives great results for those with lower-tech ECUs to get the best from their engines – an excellent stop gap before moving to an advanced ECU like our ME221 for the ST170.
The time had come to hit the dyno. With our initial software we were able to lock the cam (well, make the cam angle stay fixed by automatically targeting a desired angle) allowing us to research different outcomes at different fixed positions – which in turn allows us to optimise what angle we should use at what RPM to give the best torque. Lets have a a look at the series of dyno graphs to see how this evolved…
In each plot – green is the OEM ECU in control (closed loop) verses Red with our VVTPro controlling the cam and holding a certain angle for the entire rev range. (What a VVT Delete kit would do).
You can see how 10* locked really suffers throughout the rev range, with 20* giving the best peak power with loss of bottom end torque (almost 40%!), and onwards to 50* matching OEM bottom end torque in exchange for less torque further up the rev range.
Okay so looking at this we can see that if you are going to lock your cam into a position, aim for approx 40* advance. That’s the ‘average’ best, and I say best if you are willing to live with a lot less lower bottom end torque, poor idle/emissions, and a fluffy top end/peak power. Note that at 30* advanced we actually IMPROVE on the OEM power output at the RPM range 5500 onwards…
Of course this is going to cost something like £100+ to get the VVT delete/lockout hardware, then the time to fit the hardware, change the belt, DTI the cam angle (and the tools and know how to do so)…. the list goes on…
What we really need to do is change the cam angle based on in this case RPM alone to give us the best of ‘all’ worlds. Lets go along the highest torque line and pick the curve that best represents the most torque at that point/range and punch those numbers into our VVT controller, and doing so, we can see we need to run like the table below, and, interpolate smoothly between this angles as RPM climbs.
(RPM -> CAM ADVANCE (degress))
0-1000 -> 0*
1000-2000 -> 50*
2000-3000 -> 40*
3000-4000 -> 50*
4000-5000 -> 30*
5000-5500 -> 20*
6000+ _> 10*
In conclusion, and a few years on from when we first did these experiments, and designed the VVT Pro, we can show just how important variable Valve Timing control is. One reason we made sure even our affordable ME221 Engine Management System would feature this control strategy – also as used on our Plug-n-Play model for the 2001-2005 Mazda MX-5 VVT Engine. If you choose not to make use of VVT, its simply throwing away free horsepower the original engine designers spend millions implementing!
Here’s a final sign off image of an Mk1 Escort that featured an ST170, bike carbs and our Nodiz Pro ignition system – both before VVT control and using a VVT Pro – the results speak for themselves!
TL:DR – There’s a video describing the system as well!