Utterpower Test Bed

Hands on experience with a product is so important, there is just no substitute, when you get the product; it doesn't help much just to take it out of the box and stare at it, but there are testimonials and product reviews given on the WEB based on just that! A guy takes it out of the box, looks it over and reports that it's a quality product and nicely made! Now how useful is that report?

If it's a gun, I say shoot the darned thing, and you might even try hitting a target. If it's a fishing reel, go fishing, if it's a generator, I say you need to have a plan to test it, and to compare it to other offers. I'm expecting that this investment will allow me to measure performance and compare.

With that said, I've been busy building a test bed for generators, and after learning more about the "VFDs" I knew that had to be part of my Test Bed.

The heart of the Bench is a large three phase motor driven by a Variable Frequency Drive that can be controlled by a lap top. The bench allows me to quickly mount a new generator, and with a torque arm and sensor I can make comparisons and measurements that should help me identify the efficiency of a particular generator and compare it to other offerings. March 21, 2008 I started to compare a good make of ST3 against the latest version of the Utterpower PMG, the PMG is only available in limited quantities right now, it will take a while for production to catch up with demand.

I first mounted the ST3, and hooked it's output into my test loads and measurements. I sent an order to the VFD to bring the unit up to speed slowly. When it reached running speed I noticed it made the bench hum, that large diameter rotor has wound bobbins, laminations, and I would imagine it's a little prone to imbalance at speed. China does balance the better made ones, and the sign this was done is the metal washers attached to one of the metal fan blades. If yours has a plastic fan, it's likely this step was omitted. We know the STs are crude and heavy, and we know they do the job for a lot of people. Still,.. the efforts to balance this particular unit did not leave it turbine smooth. I plugged in my 2000 watt load, noted the amps, frequency and total watts, lots of audible noise from the vibrating bench.

I then mounted the PMG and ran it up, I put the same load on it, and noted it is turbine smooth, and very quiet. BUT, at that moment I realized there was plenty of data I did not record when I was running the ST3! It's all about planning and having a protocol to stick to. I have the torque arm sensor, a DB meter, a way to measure the voltage the current, and even total watts and power factor. I'm going to take my time an be methodical about it, but I can see that there will be many things I can measure and compare IF I am methodical in my approach.

The VFD gives me total control to change RPM even by a fraction of a cycle, but this TEST BED is all new to me, and I will be looking for friends and fellow DIYers I know to validate my findings however basic they are.

As per the PMG, I don't expect to find far greater efficiency, it seems there is an opportunity to save a few watts in excitation, and we know a tight air gap between rotor and Laminations is pro and con. The PMG I want will be at least as efficient as the ST, and will have none of the failure points of typical generators. So far my long term and intense study of failures identifies the following.

Voltage regulators are a leading failure point in small generator sets, it's often the case that these VRs are not well engineered enough to handle the environmental and some of the situations the power plant and loads present. An example is running out of fuel, there are many repair shops that will warn you about this, and it has been attributed to the death of many VRs in less expensive generator sets. As the set is running out of fuel, the VR attempts to boost field current above reasonable, and this leads to the destruction of the VR, in some cases, it literally creates enough heat in the field winding to break down the insulation varnish in the windings, this can sometimes create a 'see saw' situation, as the VR will continue it's attempt to compensate, and the field will continue to degrade as the current is increased. Expensive generators often have smarter VRs that limit current, or even kill the excitation source to the winding when it knows things are less than reasonable. You can't put a $300+ VR on a gen set designed to compete in the hardware store generator market yet along with Chinese made EBAY and internet offerings. (so I think).

Other common failure points are brushes, slip rings, rectifiers, and basically everything associated with the rotating field. Brushless fields? Well, there's failure components there too, diodes and more, some are easy to replace, and others are not.  We have heat in this rotor, and conductance thru the air is the typical method of getting rid of this heat before the electrical insulation breaks down. We have another force working against us too, and that is centrifugal force, RPM is not the only factor, it is also the diameter of the rotor that determines just how much force these windings and the insulation has to mitigate. In the salient rotor design, we are at the outer limits of the design at 1800 RPMs, (according to many engineering manuals including US NAVY manuals) and considering we have wire wound bobbins, and large diameter rotors, we might understand why we don't see 3000 and 3600 RPM ST designs. Some PMGs are designed to run a jet turbine shaft speeds, and they do it for a long time.

There are many things I think we can explore with this test bench, and the many auto alternators modified, and used for DC charging as compared to using an AC generator and efficient and smart battery charger may be an interesting study. since we can measure the POWER necessary to drive the generator, we will be able to compare the work done, and the power required to get that work done. 

I plan to learn a lot with this test bed, but I am already worried about what I don't know, and what I need to know in order to share useful and accurate information.. I'm certainly in no hurry to publish facts.. I'll likely just give you my observations and beg for assistance in validating the data I collect.. check back when you have time.

All the best,

george B.

home