Greetings: Steve A, Leatherneck, Rockhead(???), Louie, et al!
Think I'll join the "fray".
I'm a "downsized" HVAC engineer (AHH: capitalism!... but it's GREAT!!) from one of the big 5 mfcts; now in private practice for a small firm. And by the way Louie: Hear! hear! on what you say about the need for educated designers in our field ... there is a definite dearth!!! ... which is why I was snapped up and allowed the priviledge of working from my home. You're right: this forum is a GREAT TOOL for education and for "chats by the coffee machine" which I can no longer have since the end of my "corporate" life.
I MADE those "mfcts fan perf tables" (ESP) you all look at for about 12 years of my life. The reason so many opinions abound is, as Louie says, everyone thinks "It's not rocket science" (which it ain't); BUT it does take day in and day out exposure to be comfortable with the terminology.
As Leatherneck pointed out: ESP is the STATIC pressure across the unit ... inlet to outlet. And what is catalogued in the lab is per ASHRAE and ASME test procedures: a static pressure "rake" consisting of several STATIC pressure measurement points (various areas of the duct in a single plane ... to get an average) is positioned 3 to 5 "duct diameters" downstrean and upstream of the AHU and the delta-P or ESP is measured at various airflows; as determined by a calibrated ASME nozzle. The reason for more than one point of measurement is to obtain an "average"; anf the reason for the "3 to 5 duct diameters" is to eleiminate turbulent inlet and outlet effects of the unit; without inadvertently including duct friction loss ... if you get too many diameters downstream, there will be friction losses which will unfairly penalize the AHU's apparent performance. Leathernecj is right: You don't care about what's happening IN the AHU: it's a black box that can produce so many CFM against so much resistance: Think of a PUMP: suction as well as discharge.
Anyway, I don't want to "ramble" too much.
Getting back to Steve A's problem: The difference between lab catalogued performance and field measured performance CAN BE (but not always) like night and day... but what else can mfcts do?? Think of it: X number of static P points in the same plane vs ONE in the field. And how close to "90 degrees to the flow" is that probe you're sticking in the duct vs a lab setup?? The slightest deviation from 90 degrees and you're picking up some velocity pressure. And what of turbulence effects??? How many field installations have 3 to 5 straight runs of duct into and out of the AHU?? I think you may be starting to see the limitations of field tests vs thew standardized lab conditions under which data is catalogued.
Mfcts will often publish "base" unit performance with deducts for options like 1" filters or 2" filters, X kW electric heat, hi-eff coils, economizer dampers, wet coil (for cooling, dry coil (for heating), etc. Then "ESP" capability of a particular AHU is the "delta-P" measured from inlet to outlet LESS the above mentioned options ... in other words: what's left for the distribution system: the ducts, elbows, grilles, etc. We designers typically design to 0.08"/100ft for supply and 0.1"/100 ft for return of STRAIGHT pipe .. and "hope" everything else falls out in the "cushion". While this may work for residences and light commercial jobs, the analysis for extensive, high-pressure syatems like say the Wold Trade Center need to be more astute or exacting ... but still, there's "pad".
Often manufacturers will offer oversize drive kits (bigger motors and/or sheaves.. for more RPM) if the distribution system losses are really big.
Back to Steve's question: Max ESP published on the nameplate doesn't mean the unit will "blowup" if that value is exceeded ... it simply means that the standard drive can only do so much... beyond which you will need an "oversize" kit.
But more specifically: Using the fabulous Ductulator (which I hope Louie now love) tells me that the sound problem lies in excessive return velocity to the units. Assuming a general avg of 400 cfm/ton, the 3 ton should be moving 1,200 cfm. The ductulator tells me that with a 12" dia return duct, the velocity will be 1,500 fpm. "Industry" guidelines to avoid "sound" problems in residences (assuming you have reasonably unimpaired hearing)are: Main Ducts: 700 - 900 fpm; Branch Ducts: 600 fpm; Branch Risers: 500 fpm. So the velocity is too high .. no doubt thru the installer's unawareness of these guidelines or efforts to cut costs ... or physical constraints of your residence's construction. For a single return duct, the 3 ton would need a 15 1/2 to 17 1/2 inch diameter duct to satisfy the 700 to 900 fpm guideline. The 5 ton, with 3 returns of 10 inches (2,000/3 = 667 cfm/duct has a high velocity from a sound perspective as well; at 1,200 fpm per my ductulator. Those returns should be 12" to 14"....But would they fit? Or would you, the owner have objected to the higher cost??
You didn't say what your main supply duct sizes were ... and if they were round or rectangular there could be "challenges" here as well! All is not lost, though, as duct liners and stiffeners might yet be an option ... even "white noise" backgrounds or there are products on the commercial market that "cancel" air noise by exotic accoustic wave collision techniques.
It's a constant tug of war: keeping the designer, contractor and owner, occupant satisfied ... which makes this job a satisfying challege ... for those who like challenge!!
I hope this helps (some??)
