vintageracer
02-08-2015, 04:13 PM
From Fullsizechevy.com. A post about the old TPI system I find interesting!
This response was to a post where the OP wanted to use a late model 880 truck block, remove the Vortec heads, use TPI heads, TPI Injection with a boosted power adder. Many responses as to why this was not a good idea. There were also discussions about the differences in SFI, Multi-port, TBI ect.
This particular response appeared to me to be a great explanation about why the old TPI was so limited compared to today's technology but was considered then the "Cutting Edge" technology behind TPI.
I hope you enjoy this explanation as much as I did.
As is anything on the internet it could all be BS but it sure sounds good!!!
"TPI doesn't particularly generate "low RPM" torque. Common misconception. TPI generates "3600 RPM torque".
The way it works is, it uses the compression-rarefaction pulses, aka SOUND, created in the runners by the air rushing down them at high speed and slamming into the back of the intake valve as it closes.
The speed of sound is about 1100 feet per second, + - the effect of temperature and pressure; just to keep the numbers round and handy, let's call it 1000 feet per second. That means sound travels 1 foot in one millisecond. Keep that constant firmly in your mind as we go on.
Imagine what's going on in the runner during an intake event... the valve is open, the pressure in the cyl is lower than atmospheric because the piston has been going down, and the atmosphere is pushing air down the runner. This tall column of air is now rushing down the runner toward the cyl. It's heavy, and has inertia; i.e. once it's moving, it wants to STAY moving. Presently the intake valve closes. The air is still moving, trying to rush on down there; it smashes into the back of the freshly closed valve, and bounces off, creating a pulse of SOUND having positive pressure. This pulse travels back UP the runner at the speed of SOUND (remember the 1 foot per millisecond #) and reaches the plenum. It bounces around in there a bit, and some of it finds its way into THE NEXT runner in the firing order, which is now experiencing ITS intake event. This pulse from thes PRIOR event ADDS to the pressure in the runner of this NEXT event, packing the cyl a little bit more full than the atmosphere alone. The process repeats itself at each firing; each cyl's intake fill event is reinforced by the pulse left over from the PRIOR cyl's intake valve closing event.
OK now that you understand HOW IT WORKS, here's the numbers. The sum total of the length of the intake tract from the plenum to the valve is about 22"... just shy of 2 feet. So, the SOUND wave travels UP the runner at 1 foot per msec, passes a few inches through the plenum, and down the NEXT runner at 1 foot per msec, to reach the cyl. This means the sound wave takes 4 msec to get from one cyl to the next one in the firing order, right? So when the time between cyl firings is 4 msec, the "tuned" aspect of "TPI" is in effect, right? All you gotta do to now is figure out what RPM corresponds to 4 msec between cyl firings and you know what THE ONLY RPM that TPI reinforces is.
Well, 4 msec is 1/250 of a sec, meaning this is when there are 250 firings per sec. Again, permit me to round off, just to keep the numbers easy; call it 240 which is plenty close enough. Since 4 cyls fire per crank revolution, then that means this occurs when there are 240/4 crank revolutions per sec, which is 60. 60 revs/sec = 3600 RPM.
Eh??
So you look at a dyno run for a TPI motor, and you will see this "Mount Everest" looking feature on the torque curve, of this GIANT peak of torque at 3600 RPM.
Sounds great in principle.
But there are 2 problems. The first is, in order to get the pulse that makes TPI work to occur AT ALL, the intake valve has to close VERY early in the cycle. Cams that close VERY early are by definition VERY small. THis means, in order to get TPI to work AT ALL, you MUST use a cam that has NO UPPER RPM CAPABILITY WHATSOEVER, which is where "low RPM torque" comes from. It's not that TPI as such makes "low RPM torque", it's that it FORCES the cam to only be capable of "low RPM torque". The second is, there's a negative pulse of sound right behind that positive one (sound being what it is), and that NEGATIVE pulse has THE EXACT OPPOSITE effect on cyl fill if it reaches a filling cyl. Well it works out that exactly that occurs; at 5400 RPM, WHATEVER benefit you got fro the "T" effect at 3600 RPM, you get AN EQUAL AND OPPOSITE effect (destructive resonance) at 5400 RPM. TPI's torque curve therefore, instead of looking like a gentle rising curve up toward some RPM and then a gentle slope off, looks like a relatively flat line until 3000 RPM or so, then steeply rising to a peak at 3600 RPM, then STEEEEEEPLY falling off from there. The HP curve, instead of rising at a 45° angle to the peak HP RPM (because HP is the product of torque x RPM x a constant), looks like a 45° curve up to 3000 RPM, a steeper rise to 3600 RPM, then a flat line, then it craters at around 4800 RPM. Since HP is torque x RPM x a constant, then if torque falls off as RPM increases, then HP never builds up either.
It's WEAK. No matter what you do, those long runners force it to be WEAK. The only way around it is with boost. Since of course with boost, all of the atmospheric pressure related behaviors are overwhelmed by the artificially raised pressure in the intake tract.
THAT'S why you don't want TPI. Has nothing to do with whether it's SFI or not. At the time it was introduced, it was "state of the art", when compared to 70s smogger motors; now, it's hopelessly obsolete and underperforming compared to what has come after. It's why, for example, a ZZ4 is "rated" at 345 HP gross with a carb (probably 275 or 280 net HP), but the IDENTICAL SAME LONG BLOCK except with a weeenier cam is "rated" at 230 net HP when it has TPI on top."
This response was to a post where the OP wanted to use a late model 880 truck block, remove the Vortec heads, use TPI heads, TPI Injection with a boosted power adder. Many responses as to why this was not a good idea. There were also discussions about the differences in SFI, Multi-port, TBI ect.
This particular response appeared to me to be a great explanation about why the old TPI was so limited compared to today's technology but was considered then the "Cutting Edge" technology behind TPI.
I hope you enjoy this explanation as much as I did.
As is anything on the internet it could all be BS but it sure sounds good!!!
"TPI doesn't particularly generate "low RPM" torque. Common misconception. TPI generates "3600 RPM torque".
The way it works is, it uses the compression-rarefaction pulses, aka SOUND, created in the runners by the air rushing down them at high speed and slamming into the back of the intake valve as it closes.
The speed of sound is about 1100 feet per second, + - the effect of temperature and pressure; just to keep the numbers round and handy, let's call it 1000 feet per second. That means sound travels 1 foot in one millisecond. Keep that constant firmly in your mind as we go on.
Imagine what's going on in the runner during an intake event... the valve is open, the pressure in the cyl is lower than atmospheric because the piston has been going down, and the atmosphere is pushing air down the runner. This tall column of air is now rushing down the runner toward the cyl. It's heavy, and has inertia; i.e. once it's moving, it wants to STAY moving. Presently the intake valve closes. The air is still moving, trying to rush on down there; it smashes into the back of the freshly closed valve, and bounces off, creating a pulse of SOUND having positive pressure. This pulse travels back UP the runner at the speed of SOUND (remember the 1 foot per millisecond #) and reaches the plenum. It bounces around in there a bit, and some of it finds its way into THE NEXT runner in the firing order, which is now experiencing ITS intake event. This pulse from thes PRIOR event ADDS to the pressure in the runner of this NEXT event, packing the cyl a little bit more full than the atmosphere alone. The process repeats itself at each firing; each cyl's intake fill event is reinforced by the pulse left over from the PRIOR cyl's intake valve closing event.
OK now that you understand HOW IT WORKS, here's the numbers. The sum total of the length of the intake tract from the plenum to the valve is about 22"... just shy of 2 feet. So, the SOUND wave travels UP the runner at 1 foot per msec, passes a few inches through the plenum, and down the NEXT runner at 1 foot per msec, to reach the cyl. This means the sound wave takes 4 msec to get from one cyl to the next one in the firing order, right? So when the time between cyl firings is 4 msec, the "tuned" aspect of "TPI" is in effect, right? All you gotta do to now is figure out what RPM corresponds to 4 msec between cyl firings and you know what THE ONLY RPM that TPI reinforces is.
Well, 4 msec is 1/250 of a sec, meaning this is when there are 250 firings per sec. Again, permit me to round off, just to keep the numbers easy; call it 240 which is plenty close enough. Since 4 cyls fire per crank revolution, then that means this occurs when there are 240/4 crank revolutions per sec, which is 60. 60 revs/sec = 3600 RPM.
Eh??
So you look at a dyno run for a TPI motor, and you will see this "Mount Everest" looking feature on the torque curve, of this GIANT peak of torque at 3600 RPM.
Sounds great in principle.
But there are 2 problems. The first is, in order to get the pulse that makes TPI work to occur AT ALL, the intake valve has to close VERY early in the cycle. Cams that close VERY early are by definition VERY small. THis means, in order to get TPI to work AT ALL, you MUST use a cam that has NO UPPER RPM CAPABILITY WHATSOEVER, which is where "low RPM torque" comes from. It's not that TPI as such makes "low RPM torque", it's that it FORCES the cam to only be capable of "low RPM torque". The second is, there's a negative pulse of sound right behind that positive one (sound being what it is), and that NEGATIVE pulse has THE EXACT OPPOSITE effect on cyl fill if it reaches a filling cyl. Well it works out that exactly that occurs; at 5400 RPM, WHATEVER benefit you got fro the "T" effect at 3600 RPM, you get AN EQUAL AND OPPOSITE effect (destructive resonance) at 5400 RPM. TPI's torque curve therefore, instead of looking like a gentle rising curve up toward some RPM and then a gentle slope off, looks like a relatively flat line until 3000 RPM or so, then steeply rising to a peak at 3600 RPM, then STEEEEEEPLY falling off from there. The HP curve, instead of rising at a 45° angle to the peak HP RPM (because HP is the product of torque x RPM x a constant), looks like a 45° curve up to 3000 RPM, a steeper rise to 3600 RPM, then a flat line, then it craters at around 4800 RPM. Since HP is torque x RPM x a constant, then if torque falls off as RPM increases, then HP never builds up either.
It's WEAK. No matter what you do, those long runners force it to be WEAK. The only way around it is with boost. Since of course with boost, all of the atmospheric pressure related behaviors are overwhelmed by the artificially raised pressure in the intake tract.
THAT'S why you don't want TPI. Has nothing to do with whether it's SFI or not. At the time it was introduced, it was "state of the art", when compared to 70s smogger motors; now, it's hopelessly obsolete and underperforming compared to what has come after. It's why, for example, a ZZ4 is "rated" at 345 HP gross with a carb (probably 275 or 280 net HP), but the IDENTICAL SAME LONG BLOCK except with a weeenier cam is "rated" at 230 net HP when it has TPI on top."