Direct Drive VS Serpentine Belt Drive in Small Power Applications

Here’s a question I was recently asked:  “George if a engine is directly coupled to a generator head is it easier on the motor power wise?
Darryl”

I do scratch my head when I see the use of the term ‘motor’ VS engine, but when I see the use of engine and motor in the same sentence, I almost want to ask for clarity.  There is reason to frame the question further, but let’s march on.

Direct drive forces you to operate the >engine< at synchronous speed. (generator speed). This means you are forced to do the work with a fixed number of power strokes per minute regardless of your average loads IF you are interested in running near target frequency, which might be 60hz for an example.

With direct drive you have no opportunity to tune your machine for the best Fuel/Kwh conversion. We also know that some of these prime movers will live twice as long if we move the RPMs from 3600 to 3000 for instance.  I remember a study of one popular British Leyland Auto engine years ago.  Lowering the shift points to 300 RPM below the factory red-line doubled the miles driven between rebuilds VS using red-line as the shift point on the race course.

People look at any drive system outside of direct drive as a loss, but seldom do you see them investigate how little loss there is in some drive systems.  You will  see a lot of interest in 900 rpm generators that might be direct driven off a slow speed prime mover, and you wonder how much effort has been put into evaluating the advantage?  At the end of the day… the Engineer realizes he is now either buying or building a very specialized generator head that will be very costly to produce an/or repair because of the limited market in both supply and demand.

Using a low loss drive and making RPM adjustments to the prime mover can make all the difference in fuel economy, and it allows you to operate with more common and less expensive equipment, this is at the heart of KISS design.

At the end of the day; the longer term cost of making a KWH of electrical energy is the measure and the comparison you should make.  A lightly loaded diesel generator does not produce the better fuel/KWH figures, if you change the ratio and load each power stroke a little more, there is an oppurtunity to produce a better Fuel/KWH figure while running your average daily loads.  In addition, a low loss drive gives you options in the field, you can quickly remove the drive and look for the source of noise, replace the engine or generator head with a spare and be back working in short order. You can also perform much more of the work operation with gloves on, which can be nice if you are in cold country.

It’s not so easy with quill shaft (single bearing generators) which are the majority in small generator power.  As for other direct drive units, we need be careful to note how alignment was attained and note any shims unless we will perform a new alignment.  These things get overlooked in the bush when getting back on line quickly is the focus.  An example is a small hospital in Africa, or a family living off gird anywhere with temperatures well below zero F.  Another thing to note, when we drive generator heads like the ST 4 pole at our North American speeds, we are at the rotational limits for this design. Since the rotor has a lot of mass and diameter, we need be careful how we couple it.

I have noted a few backyard engineers post that they designed a direct drive coupling that wouldn’t break. With such a coupler, if you were to parallel grid power and your generator power (in error) close to 180 degrees out of phase, it would look like a train wreck.  IT >could< literally strip the bobbins off your generator’s rotor or break the crankshaft in the engine.  A lot of the better direct drive couplers have a rubber  spider or similiar to absorb some of this hammer like blow from the power stroke, and others have a roll pin or link that will fail when that ‘train wreck’ happens.  A properly designed coupler should act as the fusible link, it is far easier to replace the coupler than a rotor or crank shaft.

There may be another thing to consider, diesels are more efficient than most other prime movers, BUT they produce a lot of torque during their power stroke and it looks less friendly to some loads than gasoline engines produce. The Serp drive can mitigate the train wreck, and may help somewhat with the peak power stroke.  Let me give you an example.. there are some really nice diesel powered aircraft, picture a V8 four valve turbo charged diesel stuffed in a Cessna. Ron Pointer (a pilot serving mission work) has flown such an aircraft and was really impressed with performance, especially the ability to take a good load off a short jungle run way in the Congo for instance.  Of course the higher your climb rate, the less likely you are to be shot at.  Ron tells me of all the problems they have keeping quality petrol fuel in stock. The diesel power in Africa would be wonderful asset BUT it requires the use of a composite prop.  The metal props are best suited for Africa and it’s many primitive dirt and sand run ways eat composites quickly!  Since metal props are not certified to be compatible with these violent diesel power strokes,  there were no diesel powered air craft in African mission use at the time of this writing.  The fear is that the metal prop will work harden prematurely and cause a violent failure in the air, (According to Ron’s research).

Some people are able to quantify the pros and cons of direct drive VS a low loss belt drive while out for a walk.  I might take a stab at it after a long walk… maybe all the way around Australia? At this time, I see advantages to the low loss belt drives for the third world and serious “off grid >SMALL< power plants”.  One of the first lessons to learn is the serious error we make in buying or building a power plant too big,  you lock yourself into a poor fuel/KWH figure IF you can’t load the generator to 80% or better. A drive system that can be tuned to match your average load and find that sweet spot in RPM and number of power pulses required could be key in gettign far superior fuel economy over the direct drive configuration.

Other studies worth your time is the continued use of motorcycle chains over shaft driven bikes in performance motorcycles. The chain is a far more efficient drive, an no one wants the extra losses of  shaft and ring a pinion in a all out performance bike.  It’s more expensive and heavier.  The Serp drive is right in there with the chain as per efficiency without all the problems.  Of course those big ole power strokes might help stretch a chain, and there are plenty of maintenance issues with chains.

Opinions vary, but at the end of my walk, I think the Serpentine drive has a lot to offer us off griders and serious back up power folks. Direct drive does not offer the flexibility or ease of replacement or repair.  Opinions will vary…  some will want to deploy a more expensive, hard to find,  less flexible drive, we who believe in KISS would ask why?

Here’s an email I received 11/18/2010

George-

Question:

Do you know if anybody ever used a cog drive (ie, Gilmer belt) to couple a Lister to a ST head?  Just trying to think outside the box.  Seem to recall that cog drive is more efficient than plain old belt like serpentine? Thanks, Al

Answer:

“Plain old belt like serpentine?”

Wait a minute! “Plain and old?” Is that a bad thing?

Al, Part of KISS is designing around most available, inexpensive and proven  components that do the Job well.

I remember a hard drinking union laborer who bought an Industrial Serpentine Drive Pulley and belt at twice the price of our stuff BECAUSE he thought it would be better 🙂 All he did in our opinion is make it far harder to source parts when he needed them and pay double.

But in your case, seems you want to buy two cogs and two bushings instead of one pulley and one bushing?

There’s also the issue of carrying a higher radial load further out on the Lister shaft. In the case of the incredibly inexpensive and available  automotive serpentine belt which makes use of the outer diameter of the flywheel to drive off of, we place such a small radial load on the crankshaft bearing that it is likely difficult to calculate a lower MTBF (of the bearing) than the no load running. Part of the advantage is placing the load right next to the bearing on the inside of the flywheel rim. With the cog, we need fit out on the shaft past the Gib key head, and we create a need to transfer torque VS higher feet per second of belt travel WHICH the Serpentine design does well.

To try and answer your question more completely, yes…the cogged belt had been used, it will work, but I doubt anyone will beat the performance VS cost of the serpentine drive in the application you describe. One customer finally changed his $28 Serp belt after 16,000 hours plus of documented running. He didn’t change it for any reason other than scheduled maintenance and wanting to be ready of a predicted cold winter at his off grid residence.

As for efficiency…. I doubt you’ll measure that.

George B.

All the best,

George B.