So...
I started thinking about how I could check the flow through each throttle body at tickover (to see if there is indeed a difference) before resorting to adjusting anything.
The right way to do this is with a carburettor flow bench which measures the air pressure drop across the slide (or butterfly in our case) or sometimes with a mass air flow instrument (in the simplest case a weighted flap which pivots open in response to increasing air flow).
Unsurprisingly, I possess neither of these but I had nothing to do yesterday and got to thinking "how hard could it be?"
The answer is - not that hard really but in trying to make a device which will compare two throttle bodies in real time, I may have run into an unsolvable problem regarding lack of suck
Behold flow bench Mk1:
In construction - as you can see its made from stuff I had available in the garage.
Basically a central vacuum plenum connected to individual measurement pots, each with a tapping for a water manometer.
With the top plates installed and a valved air bleed into the plenum to give control over the vacuum (the suck is provided by the industrial Numatic vacuum cleaner I use for work but it doesn't have speed control).
Connected up and testing using a couple of penny washers to simulate the pressure drop over the closed throttle butterfly. I still need to sort out a suitable rubber gasket and hold down arrangement for the test platform so that the TB's will seal leak free against it. It's not apparent from the picture but the water columns aren't a U-tube - they're independent (both just pull from a bottle of water on the floor).
Thus far it's a qualified success. Starting with the air bleed open, as you close the bleed, both water columns climb satisfyingly and give a nice steady pressure drop 'reading' (as I've no idea of the air flow, there will never be a quantifiable 'figure' on this reading but it doesn't matter as I only want to compare the two).
But...
(And its a big but), when one test point flow is altered (partially blank one washer hole with a fingertip), both water columns climb together in response to the increased pressure drop.
I feared this might happen and it's because both water columns are really sensing the pressure drop of the entire 'system' rather than the individual test pots. I think the way to fix this would be to introduce a pair of matched orifices into the legs feeding each pressure pot. This pressure drop before the test points might allow them to discriminate to some extent. However I'm not sure that my vacuum cleaner can provide enough flow to cope with this and it'll be a pain to dismantle it as I siliconed it all together
Or just measure one TB at a time but it'll need careful operation and checking for repeatability.
Or just give up and do a zero-zero
