CSX deturbocharged some GP40-2 locos and now call them GP38-2S's. 2,200 HP instead of the standard 3,000 HP turboe'd version.
Externally, there is no difference. Unlike some other deturboing programs made by other railroads, CSX made no distinctive changes to the outside of the loco (Many times, there would be removal of the single turbo exhaust stack and the addition of two or four normal exhaust stacks, maybe even the addition of a paper air filter box.)
A pic of such a derated loco here, #4422 from RR Pictures Archives:
My question is for those who may have heard one.
What do these locomotives sound like? A normal GP38-2?
Interesting factoid: Apparently the turbos were not removed. They were somehow made to not spin, or made to mechanically spin like a Roots blower.
Also interesting, Athearn recently released a model of #4422 as a GP40-2, so setting the sound decoder to the proper prime mover sound is a goal of mine.
- Douglas
DoughlessInteresting factoid: Apparently the turbos were not removed. They were somehow made to not spin, or made to mechanically spin like a Roots blower.
A derated 645 will just keep the mechanical drive to the blower and have exhaust ducted to the stack rather than through a turbine.
Overmod Doughless Interesting factoid: Apparently the turbos were not removed. They were somehow made to not spin, or made to mechanically spin like a Roots blower. So you know how these worked: instead of expensively machined and geared Roots blowers, they use the turbo compressor wheel as a non-positive-displacement blower for scavenge air at lower engine notches. This is driven by a gear and clutch, mechanically, at lower speeds where the engine makes inadequate exhaust mass flow for turbocharging -- in higher notches when the turbine develops enough torque to start 'taking over' there is an overrunning clutch that unloads the gear drive. A derated 645 will just keep the mechanical drive to the blower and have exhaust ducted to the stack rather than through a turbine.
Doughless Interesting factoid: Apparently the turbos were not removed. They were somehow made to not spin, or made to mechanically spin like a Roots blower.
So you know how these worked: instead of expensively machined and geared Roots blowers, they use the turbo compressor wheel as a non-positive-displacement blower for scavenge air at lower engine notches. This is driven by a gear and clutch, mechanically, at lower speeds where the engine makes inadequate exhaust mass flow for turbocharging -- in higher notches when the turbine develops enough torque to start 'taking over' there is an overrunning clutch that unloads the gear drive.
So do they sound like a GP38 at lower revs, then a GP40 as the revs build?
Before I started the thread, I assumed that a GP38-2S would sound like an unturboed 645 powered GP38-2, albeit with slightly different note having the exhaust going through one large stack instead of two smaller, or in some cases four, stacks.
I wanted confirmation from folks who may have heard an actual -2S, and provided other information to maybe discuss the program in general.
I'd expect them not to sound like a GP38-2 at all; that's a Roots-blowered engine with distinctive gear sound.
At low notch it should sound like a GP40, up to the point the turbo boost starts to build up in the exhaust plenum. I have seen some threads about what an 'overdriven' mechanical turbo drive sounds like (e.g. when the overrunning clutch fails to disengage and the gearing takes load at over 900 engine rpm) and that might be a guide to thd sound if extensive beefing-up of the stock mechanical components has been made in the conversion.
I'd ask this over on RyPN where people can give you firsthand descriptions of what was changed and why -- they would also likely know where to hear a sound clip of one at different throttle positions.
Overmod I'd expect them not to sound like a GP38-2 at all; that's a Roots-blowered engine with distinctive gear sound. At low notch it should sound like a GP40, up to the point the turbo boost starts to build up in the exhaust plenum. I have seen some threads about what an 'overdriven' mechanical turbo drive sounds like (e.g. when the overrunning clutch fails to disengage and the gearing takes load at over 900 engine rpm) and that might be a guide to thd sound if extensive beefing-up of the stock mechanical components has been made in the conversion. I'd ask this over on RyPN where people can give you firsthand descriptions of what was changed and why -- they would also likely know where to hear a sound clip of one at different throttle positions.
The GP40's turbo whine is evident at start up and through out the rev/power band. Its constant.
If the 38-2Ss do not have active turbos, (at least most of the time) then there is no whine, correct?
Maybe not sounding like a pure roots blown PM, but the turbo whine, which is a pretty dominating sound, would be eliminated.
I think the point of the program was to facilitate more a responsive throttle when switching cars, as there is a delay when the turbo spools. A turbo spool is a PITA if the engineer is trying to throttle up and down in short spurts when shunting cars around. So the turbo is "shut off" in these derated GP40-2s. "shut off" as opposed to totally removed. Either way, it seems to me the whine would be eliminated if the turbo does not spin, and I think that's the point of CSX derating these particular GP40-2s.
What is RyPN?
The EMD "Roots blower" sound is very distinctive across the rpm range -- these are three-lobe carefully-machined rotors that run with only a few thou of clearance, with very precise gears to keep them out of interference. Their sound is a major part of the "567 experience", less so in the 645s. I don't believe there is a 'factory' 710 with positive-displacement blowers -- among other things it would be a relative fuel hog at higher notch -- but there have been discussions on how it could be achieved.
Much of the turbo 'sound' at low notch would likely be compressor noise, not turbine acceleration. How much of that is compressor-wheel aerodynamics and how much of it is gear noise, I don't know. I'd expect the drive noise to change as the turbine accelerates the rotor past the gear-drive speed and unloads it; you'd still get gear noise but it wouldn't be stressed power transmission noise.
The whine of the gear noise would be preserved up to at least the fifth notch in a 'converted' turbo engine, so any noise you hear down low is likely from that source, and you may be able to come to discriminate the turbine whine from the loaded and then unloaded gear-drive whine over time now that you know how the sound is being produced.
While there is no doubt whatsoever that the large EMD turbo is a pain in any kind of flat switching or extensive speed/load changes where high horsepower is a factor, I'm not sure the turbo-lag issue at lower notches is as much of a concern. Operationally the boost ought to come up fairly briskly in changes between the lower notches, as the device is functioning as a supercharger in that range, speed proportional to crank rpm, so there is no 'turbo lag' effect at all; the catch is that all the required horsepower to run the blower is directly off the crank, at fixed ratio and no multiple-stage speed adjustment, so there's no 'big saving' from exhaust turbocompounding. Once you get the engine accelerated and the fuel feed balanced, the exhaust volume begins to appreciate to the point it drives the turbine enough to start unloading the gear drive -- but there is no formal disconnecting clutch on that drive, so it continues to make engaged-gear noises, just not fully loaded/crowning-tooth ones.
"RyPN" is the forum part of the old "Eleanor P" magazine site, Railway Preservation News. That's probably the best site for finding serious preservationists with their serious-preservation hats on, ready to talk business in a technical context. The "O gauge forums" have a wealth of people who are very knowledgeable 'in the real world' but they go there more to play than to talk technical shop...
Your opinion is that it would still sound relatively indistinguishable from a GP40-2, and not totally different like a GP38-2. I only have those two choices on the sound decoder, with minor variations.
BTW, I found this excerpt of a thread. The thread is on this site.
EMD's that are turbocharged still have the roots blower that is crankshaft driven, but have an overrunning clutch that also allows them to be driven at higher speeds by an exhaust turbine as a turbo. At low RPM/loads and when starting, the roots blower is driven by the crankshaft to supply scavenging air just like all 2 cycle EMD's. But at higher throttle settings when there is enough exhaust flow, the turbine in the exhaust will produce enough power to spin the blower faster than it normally would spin if driven by the crankshaft. The overrunning clutch will let the blower run faster, supplying more air and thus making more horsepower.Most engines that have a turbocharger or supercharger have just one or the other; EMD is somewhat unique in having a combination of the two on their "Turbo" models.
So if CSX deactivates the turbo, the PM still has a roots blower. Maybe a GP38-2 sound file is the better choice.
I would be very careful believing that, because in my opinion it is not so.
Roots blowers are geared directly without a clutch, so they would continue to turn and draw a high percentage of crank HP at high notch... the reason a 38 is a fuel hog if you run it at higher governor speed.
On the other hand, once the turbo kicks in, the Roots blower would see 'turbo boost' back pressure and take even more horsepower even though its positive displacement from the intake air tract remains fixed.
The engines I'm familiar with use the turbocharger 'compressor' as the scavenge-air blower, in place of Roots blowers, by means of the geartrain mentioned earlier. Think of it as being like a Paxton supercharger instead of a 6-71 Jimmy blower on an automobile engine. The gearing also turns the turbine along with the compressor wheel, but until there is sizable exhaust-gas mass flow (from burning the greater mass of fuel that the Woodward feeds with advancing notch to hold commanded rpm) the turbine will not develop enough torque to take over from the now-faster-turning gear drive. Note that you may hear turbo noise as the exhaust goes through the blading, akin to how a siren makes noise, but you will still also hear the direct-drive gearing.
Now on a GP40 there is nothing like a Bendix drive (to my knowledge): when the turbine overruns the clutch the gear drive remains in mesh and keeps making noise, but very little power now goes through the gears and teeth so any 'whine' is reduced. The sound of a derated turbo would be this gear noise if made to produce the full volume of scavenge air for (reasonably) full combustion with the engine turning at notch 8 speed under whatever load.
Your problem remains that none of the Roots noise is correct (although many railfans won't recognize that) BUT any actual turbine noise, or compressor noises at higher speed or boost than around stoich, wouldn't be correct either -- that being the functionality that the rebuilt gear drive to the turbo wheel is now fully providing.
To modrl this correctly you'd need to record your own sound files of a de-turboed engine (vs. one that has had a blower engine swapped in) and use them in some kind of 'project'. Whether that level of mechanical nerd correctness is worth it is your choice.
OvermodTo modrl this correctly you'd need to record your own sound files of a de-turboed engine (vs. one that has had a blower engine swapped in) and use them in some kind of 'project'. Whether that level of mechanical nerd correctness is worth it is your choice.
That's not going to happen. I'm not trying to nail down exactly what a derated PM would sound like, I'm trying to ball park this by understanding if I'm going to hear that turbo whine or not. That is the big difference in sound between the turbo and roots versions.
There is a distinct difference in the decibal level of "high pitched whine" coming from a turboed 645 than a non turboed 645.
Based upon your technical explanation and from that other thread:
If I understand correctly, the 645 has only one "finned" wheel apparatus that acts like a belt driven supercharger at low rpm or notch, then acts as an exhaust driven turbocharger at high RPM. A clutch of sorts disengages the belt drive to allow the exhaust drive to take over. I would think, that the basic design of the 645 nonturbo and 645 turbo is substantially the same, with only the "turbo" portion attribtable to the actual clutch assembly that alters the roots blower to a turbo blower. There would not be two separate "finned" apparatusses (apparati?)
So if that clutch assembly is disengaged (or completley removed as in other derating programs) we're left with only the sounds the roots emits, from start up to last notch.
I'm having a hard time understanding how a derated turbo blown prime mover that is no longer turbo blown (has its clutch deactivated) would still emit the same high pitched whine. The overall sound must be bent more towards the pure roots blown sound.
And yes, the 567 has more of a whine (different sounding whine than a turbo) than a 645. I hear that.
You mistook my analogy. The Paxton is belt-driven; the 645 is gear-driven, with an overrunning clutch (think 'sprag clutch' unlocking to let the turbine spin the centrifugal blower faster than the gear drive, but locking up if the turbine 'contribution' falls below the gear speed.
THERE IS NO ROOTS BLOWER OR DRIVE ANYWHERE ON A TURBO 645. There is no Roots blower on a 645 turbo that has been converted to use turbo parts as a centrifugal blower. If you have a 645 engine with Roots blowers, you could not use a turbo on it, gear-drive-only included, without removing the Roots blowers and their drive, and substituting the gear drive as used on a turbo engine.
Yes, there are different sizes of turbo assembly, and yes, there might be a special or different impeller size, or better parts in the gear drive, on an engine without exhaust-turbine boosting.
I do not think you could separate the impeller whine from the combined impeller and turbine whine in the sound file of different notches and loads of a turbo 645. Nor do I think you can predict the sound of a 'deturboed' engine in high notch that does not have torque on the geartrain relieved by the overrunning clutch releasing.
Certainly there is no way any Roots sound file will be any better than an abstraction: even the way the blower loads with acceleration is different.
OvermodTHERE IS NO ROOTS BLOWER OR DRIVE ANYWHERE ON A TURBO 645.
OK, I didn't necessarily misread what you said, I was also thinking about what this guy said:
EMD's that are turbocharged still have the roots blower that is crankshaft driven, but have an overrunning clutch that also allows them to be driven at higher speeds by an exhaust turbine as a turbo. Most engines that have a turbocharger or supercharger have just one or the other; EMD is somewhat unique in having a combination of the two on their "Turbo" models.
And this is what you said initiailly:
This is driven by a gear and clutch, mechanically, at lower speeds where the engine makes inadequate exhaust mass flow for turbocharging -- in higher notches when the turbine develops enough torque to start 'taking over' there is an overrunning clutch that unloads the gear drive.
I'm in no position to argue. Its just that in my position of ignorance, both statements tend to say the same thing.
Edit: Perhaps I'm using the word "turbo", when I should be using the word "exhaust turbine". I think of a turbo as being a turbine that is driven by exhaust, but that may not be the technically correct wording in this case.
I'm assuming that the turbine spinning at high RPM is what produces the whine (the whine in question, not the roots whine), and that rerouting the exhaust so that the turbine never spins with the speed of exhaust motivation would cause the loco to emit a different sound than if the turbine was spinning vigorously.
And I'm thinking what we're left with is the sound of the turbine being driven by crankshaft..or gear driven....at slower RPMs producing less hp.
I'm thinking that's what happens when CSX "pins the turbo" in the GP38-2S program.
Still having difficulty getting past that.
(to be clear about what specific whine I'm talking about, go to the soundtraxx website and listen to the turbo and non turbo 645)
It has the turbo whine, but not quite the same as a 40-2.
https://www.youtube.com/watch?v=kbCRozT8sjg
https://www.youtube.com/watch?v=Fe57q4BmVfg&t=321s
An "expensive model collector"
n012944 It has the turbo whine, but not quite the same as a 40-2. https://www.youtube.com/watch?v=kbCRozT8sjg https://www.youtube.com/watch?v=Fe57q4BmVfg&t=321s
Thanks. Exactly what I was looking for.
The high pitched whine is not as pronounced as a typical 40-2, IMO. As it passes off screen at the very end after the notching settles down, the turbo whine is nearly overtaken by the typical 645 or even 567 whine.
The sound decoder has 710 files too, which seem to have a deeper toned and less pronounced turbo whine than the 645.
I'll play around with the 710 files as maybe one of those would be the closest facsimile.
OK, let's try this from a different angle.
The EMD two-stroke engines need two kinds of air to run. The first kind is scavenge air, which sweeps out the exhaust before compression starts. The other kind is supercharging air, which adds pressure to the necessary volume of scavenge air to increase charge density, which then lets the injector use more fuel to make more engine power.
The first kind of air is only incidentally 'supercharging' in the sense that word is used for automobile engineering or drag racing. That is not helped by the happy use of GM two-stroke Diesel engine scavenge blowers as automobile superchargers!
A 'turbocharger' (more properly originally called a turbosupercharger) is an air compressor, usually a centrifugal compressor, driven by a turbine. That turbine is usually spun by exhaust gas, because that allows heat and energy recovery from the exhaust downstream from its doing useful work in a piston engine, but you could use any other pressure source to spin the turbine... compressed air, or a separate burner.
To digress just a second: you can use an exhaust turbine for mechanical engine torque instead of air supercharging -- the Wright turbo compound airplane engines on '50s piston airliners added a few hundred 'free' horsepower that way, and Volvo among others has experimented with the idea on diesel trucks. That is not useful for locomotive service using 8-notch governor MU control for various boring reasons, so we ignore it for now.
The compression of the supercharging air is going to make noise no matter how the compressor is driven -- you may be familiar with the noise a belt-driven Paxton makes. That is the kind of whining noise the compressor 'side' of a turbocharger makes.
Meanwhile, the exhaust gas flow through the turbine will also make a noise going through the blading. Most people have no idea what this sounds like 'independently' as without a load an exhaust turbine can rapidly spin to self-destructive speed.
Note that the compressor of a jet engine is substantially a turbocharger, in fact using a very substantial percentage of the actual fuel energy released in a Brayton-cycle engine, so you should not be surprised at the character of the noise the device produces. The important thing to recognize is that part of this noise is compressor noise, and some of it is turbine noise.
In the conversion of the 'deturboed' locomotives, the turbine is no longer used as a source of boost. Ideally it would be removed, but for structural and cost reasons it may be necessary, or desirable, to keep something on the shaft in its place for balance... and just leaving the turbine in its housing, plated off from the combustion exhaust, might be a cost-effective approach in this case. This might be done in a way that produces noise when the compressor at the other end of the shaft is being mechanically spun to generate proportional scavenge air.
The compressor noise would likely be the same proportional to speed whether turbine or gear-driven. If a turbine wheel were retained for balance its noise would be different than if exhaust were being blown through it; we could probably do some comparative spectral analysis to figure out what parts of the deturboed whine came from an isolated exhaust turbine BUT remember the engine would not run without scavenge air, so the compressor and its possibly beefed-up gear drive would need to be running at full output, with corresponding noise, unless you had a separate coordinated air source... I find this unlikely and difficult to provide for the 'value' gained about what makes what part of the whine.
I don't find the 645 engine exhaust sound to be the same as a 567's, and I find a 710's more different still. While it is technically possible to make a Roots-blown 710, there is no earthly reason to do so and to my knowledge no one in North America has tried... any deturboed 710 would use a rebuilt EMD-turbocharger-based approach (or if the market permitted, a gear-driven solution more efficient at moving the proportion of scavenge air than a Roots blower). Any real 710 sound file would be 'further out' than the comparable 645 file.
Overmod OK, let's try this from a different angle. The EMD two-stroke engines need two kinds of air to run. The first kind is scavenge air, which sweeps out the exhaust before compression starts. The other kind is supercharging air, which adds pressure to the necessary volume of scavenge air to increase charge density, which then lets the injector use more fuel to make more engine power. The first kind of air is only incidentally 'supercharging' in the sense that word is used for automobile engineering or drag racing. That is not helped by the happy use of GM two-stroke Diesel engine scavenge blowers as automobile superchargers! A 'turbocharger' (more properly originally called a turbosupercharger) is an air compressor, usually a centrifugal compressor, driven by a turbine. That turbine is usually spun by exhaust gas, because that allows heat and energy recovery from the exhaust downstream from its doing useful work in a piston engine, but you could use any other pressure source to spin the turbine... compressed air, or a separate burner. To digress just a second: you can use an exhaust turbine for mechanical engine torque instead of air supercharging -- the Wright turbo compound airplane engines on '50s piston airliners added a few hundred 'free' horsepower that way, and Volvo among others has experimented with the idea on diesel trucks. That is not useful for locomotive service using 8-notch governor MU control for various boring reasons, so we ignore it for now. The compression of the supercharging air is going to make noise no matter how the compressor is driven -- you may be familiar with the noise a belt-driven Paxton makes. That is the kind of whining noise the compressor 'side' of a turbocharger makes. Meanwhile, the exhaust gas flow through the turbine will also make a noise going through the blading. Most people have no idea what this sounds like 'independently' as without a load an exhaust turbine can rapidly spin to self-destructive speed. Note that the compressor of a jet engine is substantially a turbocharger, in fact using a very substantial percentage of the actual fuel energy released in a Brayton-cycle engine, so you should not be surprised at the character of the noise the device produces. The important thing to recognize is that part of this noise is compressor noise, and some of it is turbine noise. In the conversion of the 'deturboed' locomotives, the turbine is no longer used as a source of boost. Ideally it would be removed, but for structural and cost reasons it may be necessary, or desirable, to keep something on the shaft in its place for balance... and just leaving the turbine in its housing, plated off from the combustion exhaust, might be a cost-effective approach in this case. This might be done in a way that produces noise when the compressor at the other end of the shaft is being mechanically spun to generate proportional scavenge air. The compressor noise would likely be the same proportional to speed whether turbine or gear-driven. If a turbine wheel were retained for balance its noise would be different than if exhaust were being blown through it; we could probably do some comparative spectral analysis to figure out what parts of the deturboed whine came from an isolated exhaust turbine BUT remember the engine would not run without scavenge air, so the compressor and its possibly beefed-up gear drive would need to be running at full output, with corresponding noise, unless you had a separate coordinated air source... I find this unlikely and difficult to provide for the 'value' gained about what makes what part of the whine. I don't find the 645 engine exhaust sound to be the same as a 567's, and I find a 710's more different still. While it is technically possible to make a Roots-blown 710, there is no earthly reason to do so and to my knowledge no one in North America has tried... any deturboed 710 would use a rebuilt EMD-turbocharger-based approach (or if the market permitted, a gear-driven solution more efficient at moving the proportion of scavenge air than a Roots blower). Any real 710 sound file would be 'further out' than the comparable 645 file.
The compressor noise would likely be the same proportional to speed whether turbine or gear-driven.
Yes, the point of the exhaust driving the turbine is to increase the speed and provide more hp compared to gear driving the turbine (or its it compressor, or is it the same thing), per previous explanations. Cram more air into the intake. If not exhaust driven, the speed is slower and the hp declines....from 3,000 to 2,200 when derating a 40-2 to a 38-2S. And the sound would decline as well.
The video corroberates what I understand your theoretical conclusion is, that the whine is present but less prominent than if it were spinning with exhasut drive.
All of the turbo EMDs sound similar to me, with some nuances based upon how prominant the turbo whine is. The normal 567 and 645 sound a lot different than the turbos, and different from each other.
Thanks for your efforts and explanations. The goal has been to determine what sound file on the Tsunami2 decoder best represents the GP38-2S.
Since I have the video, and may try to look for more, that stops the need for further speculation. Thanks again.
BTW, my testing of the various PM settings on the T2 says that setting CV123 to 7 (a 710, not sure which one yet until I look it up) produces the sound most closely resembling the video. Maybe sampling different videos will produce different results.
So for the purist, I guess setting the PM CV to a Tsunami2 710 is a bit of a "kludge" in solving this problem, but the sound seems close enough.