As a general rule of thumb the bigger the pipe the more mid to top end power it will produce. With that being said a lot depends on engine size and build, track size and type and how many stages and overall length. There is a lot of grey area in the world of pipes.
at lower RPM's a smaller pipe may help the exhaust gases to escape quicker by an exhaust pulses' velocity pulling the next pulse.
If the smaller pipe is adding back pressure at higher rpm's, it's hurting you if you seek power in higher rpm's. You CAN go too big of an exhaust pipe where the exhaust pulses cool (expand) and lose velocity, not pulling the next exhaust pulse. Back-pressure is not good for building power. High exhaust gas velocity is good.
I'm going to do a lot of math, don't mind me
"A good section of straight pipe will flow about 115 CFM per square inch of area" and "Your intake system needs to flow 1.5 CFM per engine horsepower, and your exhaust system needs to flow 2.2 CFM per engine horsepower. " [https://m.roadkillcustoms.com/determine-proper-exhaust-tubing-size/]
.990" pipe can flow 88.5 CFM, so potential of 40 hp
1.25" pipe can flow 143 CFM, so potential of 65 hp
I think you may have better performance with the smaller pipe unless you're spending all your time in higher rpm's but that's just me
The optimum ID of the primary stage is in direct relation to how deep a breath the engine takes. The size of the ID has a lot to do with the speed of the pressure pulse, which takes us to the length of the pipe at near ideal velocities.
In my humble opinion the key to understanding exhaust system dynamics is, Degrees Per Inch (DPI). That is to say, the number of degrees the crankshaft rotates per inch the form of energy you are attempting to tune travels in the exhaust system.
The two forms of energy are, the Sound Wave and the Pressure Pulse. They travel at very different velocities.
The speed of sound in the environment of the exhaust system travels at +/- 1700 FPS and the optimum speed for the pressure pulse is 300 FPS.
Knowing the DPI for the RPM and the speed of the form of energy you are attempting to tune you can mathematically calculate the distance from the exhaust valve to the "Tuned Point" in the system. Everything starts when the exhaust valve opens.
When a full length system is designed to return the Sound Wave (now a negative sign) back to the combustion chamber, at TDC overlap, on it's second harmonic,, a speed of 300FPS will return the Pressure Pulse (now a negative sign) back to the exhaust valve on it's next opening 720 degrees later helping to draw out the next pulse. The speed of the pulse is controlled / tuned to some degree by the ID, shape and temperature of the primary pipe.
An example of calculating the DPI for the Pressure Pulse at 300FPS and let's use 6000RPM:
300FPS X12 inch's = 3600 Inch's Per Second Pulse Speed
6000RPM divided by 60 (seconds in a minute)= 100 RPS (revolutions per second)
100RPS X 360 degrees = 36,000 degrees per second, DPS, crankshaft rotation speed.
Time (seconds) is your common denominator.
So 36,000DPS crank speed divided by 3600IPS pulse speed = 10 Degrees Per Inch (DPI)
The crankshaft rotates 10 degrees per inch the pulse travels in the pipe.
The sound wave at 1700FPS X 12 inch's is 20,400IPS divided by 36,000DPS crank speed = 1.764 DPI
Let's use 75* BBDC as our exhaust opening seat timing, and the collector entrance as our tuned point.
75 + 180 = 255 degrees from exhaust opening to TDC overlap (the optimum point of sonic wave return on its second trip)
255 divided by 1.764 DPI = 144.6" total distance the Sonic Wave must travel,, divided by the number of trips in the pipe, 4 (out and back, out and back) = 36.139"
So let's see where our much slower pulse ends up at the optimum speed of 300fps.
720* (2 rotations of the crankshaft) divided by the DPI, from above, 10 = 72". Divided by the number of trips in the pipe, 2 (out and back) = 36" the tuned length for pulse return. Now since it has to travel in a pipe 36.139" long it will return "late" by .139 X 2. Actually 2.78 degrees AFTER the exhaust valve opens helping to draw out the next high pressure pulse.
In all my studies I'd never heard the term Degrees Per Inch or read the reason why 300FPS is optimum pulse speed. Guess it's true what they say, "an exhaust system is like a musical instrument".
I apologize for getting carried away, sincerely hope some of you gleam a little something from this.
^ Both posts are great and full of good information.
I'll stick by my recommendation though of the larger pipe for the OP's 390 simply based on experience.
We actually had Levi make up some pieces for us to test on the 390 platform years ago and settled on a different size(s) yet.
There are so many factors that you have to consider when choosing a commercially available pipe: c.i.d., c.r., intake flow, IN valve size, cam (lift and duration,) ignition timing, and more.
While the flow of the exhaust may be capable of some big HP numbers, it is rarely attainable considering the intake tract. Even operating at 85%, you're going to be a way underachiever to what the exhaust is capable of flowing. You have to have a super flowing intake (requiring a billet head or at the very least a welded and moved intake port) with 1.7" IN valve and the correct matched carb to get to even 50 CHP.
So, how does a 36" pipe relate to my go kart ? How can I use from this information to advance MY racing program ?
Well, first you have to put on your thinking cap.
The Loop pipes for the old FH's worked on the above principle.
But let's use the above information even better, Think Deeper and turn this info around & around to put it to good use.
The sound wave is way too fast to be of much use in a 17" open pipe or 21" if you add an RLV muffler. So forget it.
Let's TUNE our system to return the much slower Pulse to TDC overlap on it's FIRST return. Much more energy available.
While we are at it let's widen the sweet spot of our newly designed system.
For every change in inside diameter of an exhaust pipe there's always a reflection sent back to the point of origin. Smaller sends a positive reflection,, like the end cone of a 2cycle expansion chamber,, an increase in diameter sends a negative reflection back. (a good thing)
Here comes "Stepped Headers".
Using our 75* exhaust seat cam timing, from the above post, plus 180* = 255 Degrees from exhaust valve opening to TDC overlap,, our tuned point of the negative Pulse first return.
I'll make this quick,, hopefully painless,, read post above to fill in the gaps.
255* divided by 11.66 DPI for 7000RPM @ 300FPS = 21.86" divided by 2 trips in the pipe (out & back) = 10.93" First Stage.
255* / 10.0 DPI for 6000 = 25.5" / 2 = 12.75" Second Stage.
255* / 8.33 DPI for 5000 = 30.61" / 2 = 15.30" Third Stage.
255* / 7.5 DPI for 4500 = 34" / 2 = 17" End of Pipe, open or inside an RLV muffler.
Each increase in diameter, "Stage", sends a negative Pulse back up the pipe and when the RPM is "In Phase" will arrive at TDC overlap helping draw in more fresh mixture.
Ever notice on Race Car headers the first step is always at 8" to 12" from the head ? And the first Y on a Tri Y system will be approximately 18" to 24" ? This,,, is SOME of the ways a well designed exhaust system increases VE and widens the torque band.
But HEY,,,, I didn't invent this. Stepped headers have been around for a long time. But maybe now you can understand a little more of what's going on inside the simple little exhaust pipe on your go kart.
Once again,, I sincerely hope you gleamed a little something from this.
For every action, there is an an equal but opposite reaction. With a 2 cycle expansion chamber, as the positive pressure soundwave expands, it creates an equal but opposite reaction, a negative pressure wave traveling backwards towards the engine. In a straight pipe, there is no expansion (until the pressure wave gets to the end of the pipe) so there is no negative pressure wave traveling back towards the engine. At least not until that pressure wave hits the end of the pipe. You can calculate the average speed of that pressure wave, but only the average. Sound travels faster in hot air, about 1 1/2 ft./s per degree. When the exhaust wave 1st starts out, through the pipe, (I don't have any real numbers other than 2 cycle) if it's anything like the 2 cycle, the temperature is somewhere around 1000°. Maybe more. By the time it gets to the end of the pipe, it's a lot cooler. I don't know exactly how much cooler. If any of you have an infrared tire temperature gauge, you could check it. Even check the temperature of the pipe from one end to the other. It's still called "exhaust gas theory". Point is, the exhaust pulse pressure wave slows down as it travels through the pipe. More later.
"If the data does not support the theory, get a new theory" (Al Nunley)
Good deal flattop, you'll enjoy enhancing your program. The ID of the pipe is really more of a trial & error thing. We really don't know the gas/pulse speed in a given pipe unless you find the best one at a certain RPM and run the math to check the speed versus the length. Some time a smallish diameter pipe will run well for a few laps or seconds on the dyno,, then when it get's up to full temperature the performance falls off due to the hotter gasses being too fast and the pipe becoming "out of phase". Also a too short pipe may run fine with a too large ID because the slower gas speed is "in phase" with the shorter length. Or a too large ID that is rapped runs good after a few laps or 25+ second dyno pull. You may find a "slightly "smaller ID pipe will pick up TOP end power due to speeding up the gasses acting like a shorter pipe.
You'll find an engine with a short rod ratio, like the stock Briggs FH, likes a larger ID pipe than the same engine combo otherwise with a long rod ratio. The longer rod ratio affects piston dwell at TDC being more efficient. I notice the Clones run small ID pipes, my guess is they have a rod ratio in the >1.8 to 1 range and a more efficient combustion chamber. Just a guess,, haven't worked on one.
In the winter of 1993 when the WKA went to the RLV muffler on a national basis for the 1994 season, we did a lot of pipe testing. My son was running the FH .425 Purple plate JR1 class. These engines had a narrow sweet spot and just having more power was huge. After many iterations our best "home made" pipe design was tuned to the power peak,, 4800RPM. The pipe had 3 steps arranged 1/2" apart so that at 4800RPM the first step would be "in Phase" just after the intake valve opened,, the second step (a deeper step) was timed to TDC overlap,, the third step was timed just before the exhaust valve closed. 3 low pressure Pluses in the same intake stroke. No other pipe combo came close. This pipe also affected the carburetor too, big time, but that's outdated info now.
In closing, now maybe you can make an educated guess as to which pipe is the best.
I sincerely hope so, WP