Utterpower Test Bench: 3KW PMG

If you look at the PMG testing, you can see that you can prove it good or bad in a few minutes even if you can’t spell ‘lectricity. Most are not aware that there is a high end generator set available (called an APU) that uses a PMG, it sells for about $7000 and is used in a mission critical application where conventional smaller generators are considered too failure prone. Building your own PMG is a possibility if you have the basic skills.. read on.

Once you’ve reached some level of maturity in your thinking, you’ll wonder how the term ‘whole house generator’ was ever phrased?

One quickly realizes that you manage loads, you don’t buy a huge generator just in case you want to run everything at once. There is seldom a reason to run your two largest loads at the same time and you can set up a simple relay to assure that you strip one load automatically before the other starts. It all boils down to efficiency, running a 15KW generator at 10% load will quickly break the fuel bank, and of course the initial cash layout of the generator will be significant. Lots of people learn that having too big of generator is a big problem.

People with experience know they must have a generator that can start their largest load. They also understand that it takes a lot more power to start some loads than it does to run them, so they factor this.

People with well pumps often make their generator purchase based on what it takes to start and run their well pump. In North America, most people have a one horse power submersible pump or something easier to start like a 1/2 HP surface mounted jet pump. If you have too small of generator, you can damage the pump. Going by the name plate rating of the generator is not always a good idea, as some generators rated at 5kw or more fail to start one of these pumps, and others rated at 3kw continuous can start them with ease.

Most of us understand we need the power source (prime mover) to develop the power to drive the generator through that inductive motor start up which can draw current 3 to 5 times as much as what it takes to run the pump motor. If all is right, this start up is short, maybe four tenths of a second. When you understand this, you can see the stored energy in a heavy flywheel can help deliver the power to cover the large demand for current. If you have a generator with a very light flywheel, the engine will need to develop this energy on its own, and the governor will need to respond quickly. In the case of the heavier flywheel, it is helping cover this short event, and the governor has more time to react and open the fuel rack.

Let’s look at a two wire 1 hp submersible pump.  In my opinion people buy these because they are a little cheaper.   The 3 wire down the well pumps often have a little better power factor and take less current to start, so I think this is a good test for the little PMG.  If it can start this pump, it can start most of them.

In this case, we are going to use the Yanmar Clone.  Our drive ratio is 5:4, so we are running the engine at less RPM and lower power than its horsepower rating of 10 HP at 3600 rpms. We do this to increase the longevity and the efficiency of the generator set. This 5:4 ratio is about as low as you want to go with the PMG and still start a big load like this with other loads on line.

UtterPower Test Bench Graph

Click to Enlarge

Above: Here’s a graph produced by a power analyzer.  It’s hard to see, but there are two inputs recorded here.  These are both of the 120 volt legs.  You can see that the pump looks a little like a crow bar has been dropped across the power leads of the generator, but about two tenths of a second into the start, we see there’s some CEMF shutting off the current as the pump rotor starts to build RPM.  At about four tenths of a second the pump is running near maximum speed and current has dropped off with voltage stabilized.  Setting up your governor on the CF186 or other prime mover to produce a voltage of 127 volts at the distribution point gives you an ANSI standard voltage and allows the PMG to produce power all the way to 4.2 KW continuous and still be ANSI legal. Please note, the Power Factor on this Pump is not optimum, and adding some run capacitance across the pump leads between the pump relay and the pump could lower the start current and use even less generator capacity. Bill Rogers has written an excellent book, Home Power Producers Guide to Electrical Reality, about what Off-Gridders and those who run back up power frequently should know, and you can learn how to optimize your loads and why you do it by reading his book.

The above example has about 675 feet of wire between the pump and the generator.

Update 7/3/2010: I recently added more load bank to the Utterpower test bench.  This allowed me to explore further the real load starting capacity of the Utterpower 3KW  PMG. With the PMG soaked to a temperature of 110F, the power analyzer saw power at >6KW for 10 seconds delivered at >220 volts. I don’t suggest that all PMGs are certified to deliver this kind of starting current, but I picked this unit at random for this test.

As we know, we typically need far less than a second to start the typical ‘hard to start’ inductive loads, and then the rotation of the shaft quickly cuts off the high current. Circuit breakers take a little time to react to high current, and the properly sized breaker will handle start current and still protect the PMG. Of course there are other considerations, we know that our prime mover MUST be up to the task of delivering the torque to maintain RPM through this short duration when we start a large induction motor. In stationary applications, this is where high mass flywheels shine, it is why so many Mechanical Engineers are fascinated with the old prime movers, the high mass stores energy, many loads take three to five times as much power to start as it does run, this info is usually provided for pump motors and other loads we must power during outages or off grid. What the ME and seasoned DIYer understands is the high mass flywheel allows us to provide high starting torque, and still run a power plant that is generally loaded at eighty percent or more where we see the best efficiency. If we deploy a prime mover with a low mass flywheel with larger cubic inches OR at far higher RPM to assure enough power to start these big loads, we typically lose efficiency and the amount of fuel we burn to produce a measured number of KWHs over time goes down. I remember being a far younger man and having all too little appreciation for the marvels created by men of past generations who had a deep understanding of the science long ago. Of course there are other considerations, and one being that some of these old engines were designed to be rebuilt in place with simple tools, no need to unbolt them from their mounts, or haul them to a repair facility, many have access doors big enough for the largest hands to gain access. the best of field mechanics might replace a cylinder liner, rings and new big end rod bearings in less than an hour. Keep water out of the fuel, keep the sun from shining directly on your gen set, and the injection pump might outlive you. Still we need consider the application, we all understand that these engines are not best suited for use in a motor home, or a number of other applications where light weight and portability and other factors are a consideration.

Imagine a motorcycle with a 24 inch 200 pound flywheel.   The first corner you try and make should be on video…

Here’s an example of dependable, efficient, easy to maintain off grid power located in our Cascade Central Mountains.

Note: As a comparison, I started a one horsepower 3 wire down the well pump (more typical) 130 feet down with 200 feet of wire between generator and pump. In this case, the CF186 was fitted with a 5″ : 4.2″ Utterpower drive. The no load voltage was 254volts (within ANSI standard), and the pump brought the voltage down to 240 volts for running. In this case, we have more horsepower available, and in this situation I can run a lot of stuff in my shop and still start the pump when it wants to run.

If there’s a point to the story, it’s that you can tune your drive system to start your largest load safely and often you can do this with a little less rpm and greatly increase the longevity of your machine and save fuel as well.  The danger is going too far and not developing the power necessary for starting inductive motors. What’s obvious to you is the direct drive doesn’t allow you to do this, there are advantages to low loss belt drive systems 🙂 dropping the belt also isolates the engine and generator which can be valuable in location a noise like a bearing going bad, vibration, etc. Having two totally separate units greatly simplifies repair and replacement when necessary. The belt driven PMG is also a very compact package and should fit nicely into the old engine compartment should you decide to make your own replacement.

One of the worst case stories I have told in past years happened about 12 years ago. A friend had a blower motor that fed a glass blowing furnace powered by natural gas. The motor was rated at only two horse power but it was part of a custom blower designed by a glass blower. The blower fan was made of heavy cast iron, it was directly mounted to the motor shaft and took a long time to spin up to speed. The owner needed a backup power for this blower and ended up buying a natural gas powered 8KW rated generator with auto transfer.

Over the next year, there were several restarts of the blower on commercial power and each failed, the breaker tripped out, the motor found bad and expensive replacement ordered and down time for the furnace. The cause was a very poorly designed blower that was nearly cooking the start windings with great commercial power. The 8KW generator would start the motor, but unknown to the customer, it was burning out the start winding in the motor when doing so! Of course the motor might run for six months or a year before shut off, so no one suspected the motor was actually bad all that time. This is more evidence that you should know your loads. This was rare, but the generator salesmen was consulted to size the generator required and naturally he assumed the 2hp motor blower would start with the 8KW generator just fine, as did I!  🙂 Moral of the story, glass blowers are not always good engineers?

Had someone put a power analyzer on that blower and consulted the motor manufacturer, they would have realized they had a product that was on the edge of destroying itself during the first 10 starts on good commercial power. In all likelihood, a few starts back to back would have taken it out at the factory had they done so. Motor specs give you all the data you need to assure you have a proper start. I am left wondering how much quicker the start would have been had they used aluminum instead of cast iron for the blower?

You can apply this story to air compressors, and even a fridge! If these items are started before the pressure is bled off the output of the compressor, they can draw a lot more current than normal to start.  Unplug some refrigerators in the middle of a run, and then plug them back in, and you’ll see the lights dim.

This entry was posted in Generator Realities, Test Bench. Bookmark the permalink.

Leave a Reply