I could just ring Pete, but....

[Huzo]

The amount of times I've read that butchers have buggerised around with Sacred Screws and basically rooted sets of TB's cannot be overestimated.
I'm guilty of it myself, so here's the thing...
If one was to take a known good TB and hook it up to a flow source of nominal but not necessarily known value and supply air, there would be a certain pressure in the duct.
Now, if you had a U tube manometer plumbed into that duct, a pressure reading would show and that would be the end of it...
Almost.
So now if you got a "Y" shaped duct and put the good calibrated TB on one outlet and the shagged one on the other, you would have a different reading on each U tube since there is a different flow through the buggered one than the known good one.
If you were to close the bleeds and manipulate the SS on the suspect one so as to achieve an equal reading on both manometers, would that not indicate that the flow rate was now equal on both and therefore the adjusted one was flowing at the same rate as the good one given that the air supply was identical to both ?
This not an attempt to balance two TB's, it's trying to set a bad one to the value of a known good one.
The flow rate does not need to be known per se, just knowing that it's now flowing at an equal rate to the factory one and that factory one is at the correct spec.
I thought a standard Norge airbox with the good TB on one duct and the dodgy one on the other, with flow supplied through the snorkel, might be a place to start...?

[Zoom Zoom]

I'm far from being an expert, but it sounds logical to me. You are, after all, attempting to equalize air flow. 

John Henry

[lucian]

Huzo, here is how I was able to match my buggered LH tb to my un buggered right side. Only need to pull the  intake rubber couplings, and alum. injector manifolds. Trim and fit the gauge until a precision fit is achieved on your right un tampered side and then check angle with protractor. should be 4.5 to 4.8 degrees. you can than use it to adjust the LH side. I would then do the usual valve adjust and  sync tb's and reset tps and self learners .

http://wildguzzi.com/forum/index.php?topic=86663.0

[Huzo]

Huzo, here is how I was able to match my buggered LH tb to my un buggered right side. Only need to pull the  intake rubber couplings, and alum. injector manifolds. Trim and fit the gauge until a precision fit is achieved on your right un tampered side and then check angle with protractor. should be 4.5 to 4.8 degrees. you can than use it to adjust the LH side. I would then do the usual valve adjust and  sync tb's and reset tps and self learners .

http://wildguzzi.com/forum/index.php?topic=86663.0

This all adds to to the combined pool of knowlege Lucian.
I have resolved to not delve into this area at all since I'm fortunate to be able to access help if/when needed, but I do admit to pondering from outside the circle..

[Kiwi_Roy]

I'm an instrumentation technician, I like your thinking. 
 

[lucian]

Huzo, I think your airbox , manometer idea is a brilliant one, short of trying to build your own flow bench. The beauty of it is that you would actually be measuring air flow and not just mechanical throttle plate angles as with my gauge. I am sure that even the slightest differences in manufacture between the left and right sides could have an adverse effect on balance despite mechanically being set to exactly the same physical angle. In my case the gauge method achieved acceptable results and my griso went from running like a toad with the hiccups at low throttle openings, to a smooth running machine at  idle and low rev's and that's close enough for me. 

[Murray]

No idea if it works but you could go really old school assuming you are not deaf as a post. Pice of garden hose or something similar down the good throttle body, listen to the noise  move hose to the bad one and tune till the noise matches it. Something about balancing triple and twin SU setups on Holden sixes and Mini's, I assume it would also work well on webbers.

[Old Jock]

Good idea Peter but an awful lot of variables could affect it. Including the accuracy of the construction of your duct (big time)

I have no idea what accuracy is required to set these things up properly, but I'd be tempted to use the same duct and then swap the bodies over

If the accuracy requires a good flow measuring rig I would very much doubt that any of these idea would work in practice, again depends on the accuracy you are trying to achieve

Back in my turbine days, I did a bit on gas measurement. When calibrating orifice plates the whole section has to be taken to the lab and if the plate is subsequently disturbed the cal is toast.

It's an amazingly finicky affair

John

[Huzo]

Good idea Peter but an awful lot of variables could affect it. Including the accuracy of the construction of your duct (big time)

I have no idea what accuracy is required to set these things up properly, but I'd be tempted to use the same duct and then swap the bodies over

If the accuracy requires a good flow measuring rig I would very much doubt that any of these idea would work in practice, again depends on the accuracy you are trying to achieve

Back in my turbine days, I did a bit on gas measurement. When calibrating orifice plates the whole section has to be taken to the lab and if the plate is subsequently disturbed the cal is toast.

It's an amazingly finicky affair

John

Hmmmm yes Jock.
I started to wonder myself about the efficacy of the whole she-bang.
If the good TB is calibrated to flow x amount, introducing the dud into the plumbing may contaminate the purity of the measurement on the good one and a definitive result may be unobtainable, I'm very hazy on this..
Even if you flowed the good one in isolation and got a manometer reading then attached the dud, you could adjust the screw to achieve the same manometer indication, but you could not be sure that the supply rate was constant. I do believe you would have to have them manifolded, but as you say John, there'll be more bugs than a boarding house bed to be ironed out.
Obtaining data on the pressure indication within the duct for a good TB @ the designated flow rate may be advantageous because even if the dud TB contaminated the reading, once you had achieved that steady state reading with both TB's slurping, you could be relatively sure that the newly adjusted one was performing exactly the same as the un molested one, even though you had no numerical data, you are only searching for balance.
I am tempted to trust the airbox method because the box itself is supplying air to both TB's and the flow rate is relatively low I'd imagine, so the pressure would have good opportunity to equalise bi laterally.
Of course this is conventional wisdom and the whole shooting match might not be worth a pinch of piss..
Another thing that has me worried is the strong sound of crickets coming from up around  the Bungendore/Canberra region, that usually means the theory is full of holes..

[Kiwi_Roy]

So the Left Hand throttle body is calibrated on a flow bench and the idle stop is daubed in yellow paint.
They also set the link rod and paint it.

There is no magic paint on the Right Throttle body, how is that calibrated, its effected by both the Bell Crank and the Link rod length.

There's no paint on the Bell Crank, as soon as you touch that the Right Hand throttle moves.

The link rod in a conventional Link/lever is used to adjust Linearity
I believe the Bell Crank is more like a Zero adjustment, it seems to effect the idle as much as it effects the balance at speed.

It would be interesting to read a write-up on setting them on a flow bench.

I don't think pressure would be sensitive enough, a flow meter would be better.

Have you thought about just measuring the settings with feelers etc. compared to a good set, the idle stop for instance should be fairly easy to replicate
(for example prop the butterfly open with a thin rod then measure the gap between butterfly and stop with feelers)


Why even use a Flow Bench?
In theory every bike would be fuelled exactly the same,
No need to do a balance
No need for tweaking the map
That's my theory anyway



 

[pete roper]

The stop screws on both TB's are sealed. I imagine they are calibrated without the linkage rod installed. The bell crank will be set in the middle position and the linkage rod installed and locked. Both air bleeds will be closed when this procedure is carried out and then inblances caused by differential performance of the two cylinders will be compensated for by bell crank and air bleed.

My guess

[Huzo]

I think it's gunna' be the same as what the Doctor said the day I was born....
"This is going to get ugly..."
Just for the sake of a good chat, purely measuring the gaps and plate angles etc... will not allow for other anomalies in manufacture.
I guess we all know that a flow bench is what's required, but I cannot resist trying to escape through the side door..

[Kiwi_Roy]

The stop screws on both TB's are sealed. I imagine they are calibrated without the linkage rod installed. The bell crank will be set in the middle position and the linkage rod installed and locked. Both air bleeds will be closed when this procedure is carried out and then inblances caused by differential performance of the two cylinders will be compensated for by bell crank and air bleed.

My guess

My Griso only has a stop on the Left, it's not even drilled on the right.
I would like to see a RH stop, I used it to good effect on my high mileage VII Sport

[Moto]

Huzo, I think your idea makes sense.  I would second the idea of switching the two throttle bodies from side to side after you believe the calibration is done, as a check.

I also wonder about the appropriate air flow rate for this kind of check. Do we know what rates are used in flow benches? It would not be hard to calculate an approximate flow rate in the motor when it is operating at cruising speed,* but I don't know if this is appropriate for your test setup. It could be that the sensitivity (and damping characteristics) of your measuring device change at different rates, or that the characteristics of the throttle bodies themselves are more critical at some flow rates than others.

Good topic for speculation!

Moto

* I might calculate the volume of air containing a stoichiometrically correct amount of oxygen to combine with the fuel consumed at given mpg and speed, converted to units per second.

[Huzo]

Huzo, I think your idea makes sense.  I would second the idea of switching the two throttle bodies from side to side after you believe the calibration is done, as a check.

I also wonder about the appropriate air flow rate for this kind of check. Do we know what rates are used in flow benches? It would not be hard to calculate an approximate flow rate in the motor when it is operating at cruising speed,* but I don't know if this is appropriate for your test setup. It could be that the sensitivity (and damping characteristics) of your measuring device change at different rates, or that the characteristics of the throttle bodies themselves are more critical at some flow rates than others.

Good topic for speculation!

Moto

* I might calculate the volume of air containing a stoichiometrically correct amount of oxygen to combine with the fuel consumed at given mpg and speed, converted to units per second.

The stoichometric approach is way above my head these days and probably always was.
I guess if you take a 600 ml cylinder @ 3,000 RPM that's 50 strokes per second, divided by 2 for a four stroke= 25 per second x 0.6 litres= 15 litres/sec if you had 1 atmosphere in the TB.
But you don't...
It must be less than 1 or ambient pressure would not charge the cylinder, so I'm losing track of how that would be derived.
I started to wonder if you could get a suitable fluid like methylated spirits or petrol etc... and arrange a vertical tank of 5 litres or so and attach a good calibrated TB to the bottom via a coupling similar to the intake tract on the engine.
Close the bleeds.
Have a good fitting plug on the outlet end of the TB.
Fill the tube up to a pre bored overflow hole near the top of the tube and allow air to escape upwards so as to purge the system.
So you've now got a Tube 50 mm or so diameter and a couple of metres high full of liquid.
Remove the plug and do a series of cycles to check the time taken to empty the tube.
If you measured the time taken to empty the tube, you would have the flow rate of a known good TB, then attaching the buggered one and filling the tube, you could manipulate the SS (sacred screw) over a series of tests to achieve the same flow rate.
Seemingly, this would the translate back to airflow...?
As I see it at this stage, any anomalies in machining etc.. would be negated because the flow of fluid through the device, is mimicking the airflow in general use so any leakage can be adjusted out by alteration of the SS.
The rate would be different for air, but they would be the same as each other after adjustment.

[Moto]

Huzo, your approach again makes sense.

As for stoichiometry, I was thinking of calculating the gasoline consumed per unit of time like this, using some round numbers for an example. Suppose the bike gets 50 mpg at 60 mph, and we want to know what the rate of air flow must be.

In one hour the motorcycle goes 60 miles, using 60/50 = 1.2 gallons of fuel. At 6.3 pounds per gallon, that's 7.56 pounds of gas per hour, or 3432 grams. Dividing by 3600 seconds per hour, the rate is about 0.95 grams of gasoline per second.

Multiplying 0.95 times the stoichiometric ratio of 14.7 gives about 14.0 grams of air per second as the requirement. This calculation doesn't require an assumption about the efficiency of the cylinder filling.

How you measure the flow rate of air in grams per second is another question. I don't know the answer to that, but do note that mass airflow sensors, as used in fuel injection, apparently do it. Maybe one of those could enter into a setup.

Moto

[Huzo]

Huzo, your approach again makes sense.

As for stoichiometry, I was thinking of calculating the gasoline consumed per unit of time like this, using some round numbers for an example. Suppose the bike gets 50 mpg at 60 mph, and we want to know what the rate of air flow must be.

In one hour the motorcycle goes 60 miles, using 60/50 = 1.2 gallons of fuel. At 6.3 pounds per gallon, that's 7.56 pounds of gas per hour, or 3432 grams. Dividing by 3600 seconds per hour, the rate is about 0.95 grams of gasoline per second.

Multiplying 0.95 times the stoichiometric ratio of 14.7 gives about 14.0 grams of air per second as the requirement. This calculation doesn't require an assumption about the efficiency of the cylinder filling.

How you measure the flow rate of air in grams per second is another question. I don't know the answer to that, but do note that mass airflow sensors, as used in fuel injection, apparently do it. Maybe one of those could enter into a setup.

Moto

Yep, you shone a little bit of light on a dusty old corner of my 1976 chemistry class, but alas, I think that may have been the one I skipped out early on to head off to the beach on the T 500.
Wish I'd stayed in the class and kept the Suzuki...

[Kiwi_Roy]

Air flow is tricky, it's too hard to measure
I was looking at my throttle bodies on the bench, I wanted to check if they were set roughly the same so i wandered out to the garden tap and filled both inlets with water
they took about the same time to drain through the butterflies, close enough, water won't harm them.

Use water flow under standard conditions, you can verify that by weight or volume

It would be easy to make a simple test stand using a column of water that could be easily replicated across the globe.

Huzo, I sent you a PM

[Huzo]

Air flow is tricky, it's too hard to measure
I was looking at my throttle bodies on the bench, I wanted to check if they were set roughly the same so i wandered out to the garden tap and filled both inlets with water
they took about the same time to drain through the butterflies, close enough, water won't harm them.

Use water flow under standard conditions, you can verify that by weight or volume

It would be easy to make a simple test stand using a column of water that could be easily replicated across the globe.

Huzo, I sent you a PM

Is this in response to post #14 KR..?

[Kiwi_Roy]

Is this in response to post #14 KR..?

No, I read "As for stoichiometry" and my eyes glazed over that's as far as I got

I just re-read post #14, exactly that
I came up with the same idea only I would use water, make a simple column out of plumbing fittings with an overflow to give you a constant head then catch the water coming through the butterfly and either weigh it or measure it in a graduated beaker. ( I think it would be in the order of one or two liters per minute)
If you publish the dimensions guys anywhere would be able to replicate the column and reproduce the results.
The throttle bodies should fit to the column using the same rubber sleeve used on the bike.
Don't separate Left from Right, leave it intact.
Fit the buggered up LH body, flow it and adjust the stop until you get the right quantity over time either from measuring a good one or just from the Liters / minute specs from other owners (experts)
then flip it around and do the RH body adjusting the bell crank to get the same results.

No need for a plug, mount it over the sink or outside, set it up so the overflow is running nicely then just grab the flow for a set time and measure it. 

Yes, when we specify control valves we use Cv, even for gas it's the way of comparing one valve to another
1 Cv is the size that will flow I USG per minute at 1 psid
The nice thing about water, it's pretty stable unlike air that changes with temperature or pressure, the calculations are much simpler than trying to do it as gas.

[Huzo]

I'm more along the lines of..
Fill the tank to the overflow and see how long it takes to empty.
Not..
Fill the tank to the overflow and see how much you get in a unit time.
It's easier to measure time accurately than volume.

[Kiwi_Roy]

I'm more along the lines of..
Fill the tank to the overflow and see how long it takes to empty.
Not..
Fill the tank to the overflow and see how much you get in a unit time.
It's easier to measure time accurately than volume.

Ok, but I sent you a sketch.
If you use a tank how do you guarantee Bill in the back of beyond uses the same size tank for his?
Volume x time is easy, leave it running longer to get better accuracy it can also be weighed as an alternative to measuring the volume, using a constant head the size of the pipe or tank doesn't matter.

I think your idea of a tank is ok for a one off.

Actually I don't think you will need a tank, just use a pipe similar in diameter to the throttle body, I think you will find it gives you sufficient time to use a stopwatch.

Try it both ways

If you don't like measuring by Volume, weigh it, 1 liter of water weighs 1 Kg

[lucian]

You guy's got me thinking , there has to be a way to accurately compare flow rates on these tb's while still on the bike. Here's my thought. How about a spirometer you could hook on to the outlet side of the tb after removing only the rubber couplings and injector manifolds?  I'm thinking a 4 foot long clear plastic tube with an  ID slightly larger than a weighted ping pong ball.  Connect top of this tube to a tee and connect one side of tee to a shop vac and the other side of tee to the TB .  Graduate the tube with a tape measure and adjust the weight of the ball until you get it to float under  the vacuum inside the tube, within the range of the graduations.  If this could be accomplished, I would think  it would be very sensitive to any adjustment of the throttle plates.  Sketch is in the works.  Somehow this needs to be put to bed.

[Kiwi_Roy]

We in the instrument trade call those Rotameters or Variable Area Rotameters
The tube is tapered, as the ball rises up the tube more and more area is between the ball and the wall (the variable area) until it balances. Actually the calculations for a large gas one are very complex, they can get very expensive


Sent from my iPhone using Tapatalk

[lucian]

Expensive?  Guess it's back to the drawing board.   

[Kiwi_Roy]

Dont discount the idea though, the point where the throttle stop makes contact is very small, with the inlet valve closed you might be able to suck the air back out through either one of the stepper ports or even the connection where it connects to the cannister and them through a half inch rotameter

I still like your liquid idea better, easier to replicate.

And cheap - Guzzi content

The shop vac would be a variable, no two the same

Really though why do you need to do it in-situ, it only needs to be done once and then you fill the grub screw with the sacred yellow paint,
when you do a throttle balance after that it doesn't get touched.  IMHO you will get better results with your water flow test, it's brilliant in its simplicity.

[lucian]

Gotta give credit where credit is due,  Huzo's  the one with the waterometer , I agree excellent Idea 

[Huzo]

Gotta give credit where credit is due,  Huzo's  the one with the waterometer , I agree excellent Idea 

Is it possible I have damaged something by pouring the water in while they are on the bike..?
The engine won't go over TDC even with two people on it towed behind the car..!
The back wheel just skids, should I force it..?

[Paul Brooking]

Is it possible I have damaged something by pouring the water in while they are on the bike..?
The engine won't go over TDC even with two people on it towed behind the car..!
The back wheel just skids, should I force it..?

This will fix it

[Huzo]

This will fix it

currently on the M2 under the honeypot road overpass waiting for my log book to catch up.
Adelaide is bigger that I thought..!