The V8 Owners Forum -

Hi all,

I'm currently looking at putting a project together to aid throttle body balancing and also for checking the balance under load. My idea is to use a number of MAP sensors (2 to start off with) coupled to a micro-controller and display the readings on a small screen and also data log them to allow checking the balance with the throttle open rather than just sitting on the stop screws.

The first thing I'm trying to find is a component number for a suitable sensor, ideally reasonably cheap as I'll be needing a few of them. I'm thinking along the lines of the component used on the Megasquirt boards should do. Only for NA engines at this point so does not need to be 3 bar, anything which will measure from zero to atmospheric should do.

Any help greatly appreciated.

The MS one is an MPX4250AP and available from pretty well every good electronics supplier, as they say. About £20 plus VAT. You'll find its data sheet on the RS site.

If you're trying to use them in single runners on the engine side of the blade, they will pulse like mad and need to be heavily damped to get any sort of usable reading from them

But for a cheap 1 bar...find one off a used vehicle. Then find 8 more. Should be able to get them very cheap. ie, less than a tenner

Although avoid those from a diesel vehicle, as many do not read very well down into vacuum ranges.

stevieturbo wrote: ↑Mon Jun 13, 2022 6:15 pm If you're trying to use them in single runners on the engine side of the blade, they will pulse like mad and need to be heavily damped to get any sort of usable reading from them

But for a cheap 1 bar...find one off a used vehicle. Then find 8 more. Should be able to get them very cheap. ie, less than a tenner

Although avoid those from a diesel vehicle, as many do not read very well down into vacuum ranges.

Well I guess I should confess that I'm not actually aiming this at a V8, or a car engine at all for that matter. I'm looking at if for throttle body balancing on my bike engine. I have a manual/physical/analogue device for doing this which uses metal rods and gravity to check the vacuum. The connection is just downstream of the throttle plate. The tubes to the gauges contain a blanking plate with a small orifice which does help with damping but I'm also looking at the possibility of providing additional damping if necessary via electronic means.

I'm not really wanting to use automotive sensors due to the size, and I don't need the robustness they have. I'd much rather have something smaller which I can mount on a circuit board. Mind you the bike I have does have a couple (it's a twin cylinder) which feed it's ECU (which unfortunately I can't get into to see the readings) but I could look at salvaging some from another bike, they are only quite a small package, probably not board mountable but won't take up too much space.

Do know that MS uses a different algorithm for throttle bodies rather than a plenum. Presumably because of the less than smooth MAP signal?

I'm wondering if it would be easier to measure the average noise (SPL) from each TB? In carb days, I found a good pair of ears at least as good and any device which measured flow, for setting up twin carbs.
.

DaveEFI wrote: ↑Tue Jun 14, 2022 2:12 pm Do know that MS uses a different algorithm for throttle bodies rather than a plenum. Presumably because of the less than smooth MAP signal?

I'm wondering if it would be easier to measure the average noise (SPL) from each TB? In carb days, I found a good pair of ears at least as good and any device which measured flow, for setting up twin carbs.
.

I've heard of people doing this 'by ear' in the past, I'm sure it works well but must admit I'm more of a scientific sort and prefer things I can measure rather than 'sense'. Also I want to be able to measure this when the bike is in motion, I've never thought it was a good idea just to know the throttles are balance just when they ate on their stops, its just as important (moreso I'd argue) to know that that are balanced at all flow rates and throttle openings.
For instance, on my bike the throttle cable acts on one TB shaft and there is a linkage on the other end of that shaft to the second one. If there were any play in that linkage then the second TB may not start to open at the same time, or if it could be possible that the second opens at a different rate.

The other aspect of this is that I'm looking for a project to do using an Arduino, peripheral devices and some programming.

If you just want to ensure both are the same.....use a pressure differential sensor.

one sensor will do both.

stevieturbo wrote: ↑Wed Jun 15, 2022 9:07 pm If you just want to ensure both are the same.....use a pressure differential sensor.

one sensor will do both.

That's interesting, I was looking for a differential pressure sensor (for a different project) but struggled to find one, you don't have a suggested part type/name/number do you?

I'm sure there are others.

https://shop.vems.hu/catalog/product_in ... 0fc0537ed3

Plus some OEM use them for DPF monitoring.

https://www.delphiautoparts.com/gbr/en/ ... ure-sensor

stevieturbo wrote: ↑Fri Jun 17, 2022 6:35 pm I'm sure there are others.

https://shop.vems.hu/catalog/product_in ... 0fc0537ed3

Plus some OEM use them for DPF monitoring.

https://www.delphiautoparts.com/gbr/en/ ... ure-sensor

Cheers Steve.
The DPF one looks the closest for the purpose I'm talking about although for the purposes of my project I may prefer to monitor the individual pressures separately and perform the differential checking separately, I know that sounds like I'm just trying to make life difficult but the reasons are a little long winded to go into here.

The other sensor looks like it would be useful for something different I thought up some time ago which is a fuel pressure sensor which measures fuel pressure relative to manifold pressure which should be a constant figure. Strange thing is that advert seems to indicate this is a possible use but also indicates it is not a good idea to use with fuel pressure, which is strange. Maybe it means pump pressure rather than regulated pressure.

All good stuff though.

GDCobra wrote: ↑Sat Jun 18, 2022 6:14 pm
The other sensor looks like it would be useful for something different I thought up some time ago which is a fuel pressure sensor which measures fuel pressure relative to manifold pressure which should be a constant figure.

My ecu does this as standard. Relative FP is ideal for fuel pressure safety monitoring and trips, and far easier to see at a glance if the fuel system is behaving correctly at all times because as you say, it should be a constant value.

stevieturbo wrote: ↑Sat Jun 18, 2022 9:18 pm

GDCobra wrote: ↑Sat Jun 18, 2022 6:14 pm
The other sensor looks like it would be useful for something different I thought up some time ago which is a fuel pressure sensor which measures fuel pressure relative to manifold pressure which should be a constant figure.

My ecu does this as standard. Relative FP is ideal for fuel pressure safety monitoring and trips, and far easier to see at a glance if the fuel system is behaving correctly at all times because as you say, it should be a constant value.

That's interesting. My regulator was playing up at one point which is what gave me the idea or monitoring it, couldn't find a sensor which did it at the time so gave up on it and simply replaced the regulator but good to know I wasn't a dumb idea!

A sensor probably does exist that can do it with both liquid and air at the same time.

In the case of most ecu's or loggers, shouldn't be too hard to just use a maths channel to display relative fuel pressure via separate sensors

Although as said in the case of a low cylinder count if you're trying to use individual runners ( and even many cylinders ) there will be very erratic pressure readings in the single runner.

Thought I'd update this as I've got a little further with the project. Maybe my findings will help others.

Unfortunatly I've not sorted out any MAP sensors yet so can't currently monitor multiple channels but thought it best to do a bit of fact finding to see how plausible this may be. I realised i have a single MAP sensor available as I recently bought a 'spare' airbox for my bike and that just happened to come with a sensor attached to it so I've just been using that.

The first thing I thought I'd do was create a program for measuring and viewing the MAP signal. I was a little concerned by a couple of things posted earlier:

stevieturbo wrote: ↑Mon Jun 13, 2022 6:15 pm If you're trying to use them in single runners on the engine side of the blade, they will pulse like mad and need to be heavily damped to get any sort of usable reading from them

&

DaveEFI wrote: ↑Tue Jun 14, 2022 2:12 pm Do know that MS uses a different algorithm for throttle bodies rather than a plenum. Presumably because of the less than smooth MAP signal?

Which having seen the way my measuring gauges jump around (even though they have a small orifice in the line which should damp the signal) I tended to agree with, however this is the result I got:



It's not un-smooth at all, in fact it shows a very regular output.
I actually built in a rough and ready smoothing capability into my system but in this image it is turned off. This is why there are 2 traces, the blue one is the filtered data, the red on raw data. I've offset them here so both can be seen, infact they should overlap each other without this while there is no filtering active.

There is nothing fancy going on here. The system is simply taking readings from the MAP sensor and outputting the values, so it seems the only reason for thinking the signal would be 'rough' is that a physical gauge and the human eye can't react quick enough. Bear in mind the time period for this waveform (distance between each peak or trough) at 1500RPM is around 80 milli-seconds.

Having seen this I decided to time-stamp the signals, work out the difference between the minimum readings, calculate the time period between them and massage this into RPM, this actually works really well.



This show the RPM signal included in the trace which is hovering around 1500, exactly as per the bike's instrument.

That's pretty much it for now. I did do one other thing althogh it may not be needed in reality.
I'm only really interested in the signal when the inlet tract is active, that is the inlet valve is open and air moving through the inlet tract, this is the 'trough' area on the trace so I've added the capability to clip the higher pressures as here:



I did this as I was concerned there may be some pressure reversals in this area, which is what I'm looking for, possibly skewing my readings, doesn't look like it will be necessary but it's there if I need it, other engines may have different characteristics.



What's next?

I know it seems like I'm a long way from a working system but I feel like the hard work is done in terms of working out the unknowns. I've seen that I can read and analyse a signal from a single inlet tract, there will be no new work to monitor the other one (or others when I move on to more than 2 cylinders) I'm already detecting the lowest value (peak vacuum if you like) so all I need to do is compare the inputs and work out the difference.

Obviously I'll need to add a display for the user (I'm not expecting to run this hooked up to a laptop in practice) but I've already done that on other projects. I then need to package it so that It can sit on the bike while running. I also intend to add an SD card data logger as I don't think it's a great idea to have the rider (me) watching this while riding.

I'd also like to tap into the bikes throttle position sensor but I'm not sure how sharing sensor outputs between systems can work in practice, it may simply be a case of making the two systems share a common ground but I don't know. Bit more research needed there.

If anyone has any other ideas I'd be interested to hear.

I mentioned filtering in my previous post, thought I'd just show what I did.

This graphic shows both the filtered and unfiltered signal without any filtering applied.



They are the same apart from the offset I've applied so that both signals can be viewed. This is with the signal monitoring constant/atmospheric pressure (engine off) as hopefully it can be seen that there is a small amount of 'jitter' in the trace, the positive thing is the deviation is very small

I've then turned on the filtering and got the following result:



It's not massively differnt but hopefully it can be seen that the blue trace no longer has any 'jitter'. To do this I simply check if the new signal sample has varied by more than a threshold value and if it hasn't I don't update the value which I'm tracking. In this case I set the threshold value to 2.
I don't have any calibration data for the sensor I'm using but I do know that it's output is around 735 at atmospheric pressure and 210 at the maximum vacuum value I can apply. Given this I'm assuming a delta of 2 would equate to around 4mBar but my maths could probably stand some verification!