Voltage Variation Mystery

1969 (MEP-017A) Military 10HP 5KW, 60 Hertz (11 hours total runtime). When I check the AC output one leg reads higher than the other (loaded or otherwise). For example: if I adjust the voltage so that the higher of the two is 120 V, the other leg will be around 112 V. I ran all of the specified voltage regulator checks and it seems to be within specs. Is this unusual? Will it cause problems when I run appliances etc. off of it? Any help will be appreciated.

Hi Steve, It sounds like the voltage regulator is sensing one winding (120 volt)instead of both (240volt). When they are setup like that, the one winding that is being sensed by the regulator is bang on the voltage while the other sort of floats around within the specs deemed acceptable by manufacturer. They are okay when you are running a balanced 240 volt load but tend to drift apart when you run unbalanced loads on them. If you move the sensing to the other winding you will have the reverse happen and have a voltage increase on the other winding while loading the winding with the sensing on it. Hope that explains it. John

Well, yes the V/R does sense only one of the 120V output legs so that could be part of the trouble. But a couple of questions: You're not relying on the set's internal voltmeter for these measurement, right? - that meter's not very accurate and is subject to a lot of vibration, making its readings suspect. So where exactly are you making the voltage measurements - on the L0, L1, L2, L3 terminals? Need the details. Also, how do you have the phase selector switch set? I know the MEP-017A doesn't have a 120/240 position - not sure about the MEP-017, though.

Offhand, your readings are surprising - I wouldn't think there would be that much variation, especially as you say the variation shows up even with no load connected.

As for damaging appliances - I don't think there'd be a problem - set the average of the two to 120V and you'd be OK.

Output selector is set for 240 Volt single phase. Taking readings with a digital multimeter on the L2 and L3 posts. (The onboard voltage meter readings generally agree with my VOM readings). Output at the 120 V receptacle is consistantly the higher reading. Is it possible that the output voltage adjustment rheostat is defective? Thanks for your input.

I understand taking the reading on the 120V receptacle. Where, exactly, are you taking the other reading?

Does this unit also make 3 phase? In order to get 120/240 single phase some pretty weird connection of the windings must be done.

My 5ABG mil spec Onan is a 6 wire unit that can make 120, 120/240, 120 3 ph , 120/208 3 ph.

In 120/240 mode it looks like 2 windings fight each other. Perhaps this is why your voltage is different across the two legs.

Anyone have any info on a 5 ABG or a 3 ABG? These are 1953 era units that use a all aluminum flat head 4 cylinder Continental N56 engine. There is not any manual # on the unit and nothing on the web for these units.

Thanks Harold

Yes, The MEP-017A is similar, it uses a 6-wire alternator. One surprising limitation, though, is that it does not have a switch setting for 120/240V single phase. It can be obtained, of course, but requires a modification.

Hi Steve,

Any update on this? Do you need any help setting up for 120/240?

Jim

Sure, I'd appreciate any help you can offer. I'm not sure I understand how 120/240 single phase differs from 240 single phase for the purposes of wiring into a transfer switch. Any light you can shed on this would be a good start. -Steve

The difference is that 240 single phase, like that coming from the MEP017A, does not have a neutral. In other words, it can't provide 120V - just 240. It is not 120/240 - just 240.

When you set the MEP017A to 240, The L2 and L3 are the two output lines and there's 240V across them. There is no neutral provided at the main output lugs (surprisingly). The L0 output is not used in this mode. L0 is a neutral only when using the 120/208 position. (as you found out, there is a neutral at the 120V outlet on the panel, though) Why they didn't bring that neutral out to the L0 terminal is a mystery to me.

So there's two options for you. One is to use the 120/208 position. You'll have a neutral at L0, and use either 2 of the other 3 outputs as your 'hot' lines.

The other option is to modify the set to provide a neutral at L0 all the time. Not that difficult to do. Do you have a TM or can you read the schematic on the set?

(note that my experiece is with the MEP017A, I'm assuming the MEP017 is the same)

Jim

Jim, Thanks for your patience with my ignorance on this subject. I guess because I don't really understand how 3 phase technically (or practically) differs from other phases, I'll venture this question: Can I in fact run my house off of the 120/208 3-phase mode by just choosing the two 120V legs and the neutral and ignoring the 208V leg? If so, why not choose that route? If I've misunderstood your last post or that route is not feasible (or advisable) for some reason, then I am prepared to modify the genset to have true 120/240 single phase. I do have the TM and the schematics. Thanks again for your help and patience. - Steve

You certainly can run your house from the 120/208 output from that set. I have a MEP004A (15kW) and, like your set, it does not have a 120/240V mode. When you do this, the loads in your house that need 240V will get 208 while the 120V loads will get exactly the 120 they need. We have 240V central A/C and a 240V well pump at our house and they both run fine. Generally, I adjust the voltage a bit high, so that the set is actually putting out about 125V, which raise the 208 to 216.

If you want 120/240, set the selector switch to the 240 position and mod the set by installing a heavy (6 gauge wire or so) from the L0 output lug to S1-6. If you look at the set schematic, you'll see that this bypasses the S1 "F" contact, so that L0 is hooked to the neutral point for all 4 switch settings (though you must not use the L0 output in either the 120 1 phase or 120 3 phase settings)

[bitsnpieces1]

Just ran across this thread and thought I would chime in with this. Most three phase commercial power comes with one high leg, that is slightly higher voltage. This is done so that control systems could be run off that leg (120v, 1ph) and that would drop the overall volts to the same as the other two legs.

[Jim Rankin]

3 phase power voltages are balanced as perfectly as possible given the (sometimes) changing load on the individual phases. This is because on a 3 phase motor, a small variation in supply voltage between the 3 phases will create a large imbalance in current drawn on each line so that motors have to be derated if voltage is very far off. Fine tuning an installation sometimes involves "rolling the phases" (A to B, B to C, C to A )to match the load phase which is slightly higher to the supply phase with the highest voltage for the best possible balance of amperage drawn. This takes things like the control transformer load and any other single phase loads into account

The only 3 phase source that normally has a lot of difference (sometimes called a "wild" leg) is the 120/240 3 phase delta, 4 wire service which has one leg that will measure something like 208V to the neutral. The other phases both measure 120 to the neutral and all 3 phases are 240 (any combination of 2 lines A-B, B-C, A-C). That is, until the un-balanced single phase loads on the service drag the voltages down on one or two of the lines. These services are commonly supplied by an open Y/open delta transformer bank using 2 cheaper single phase transformers, so the power co can supply 3 phase cheaply to rural customers, and are noted for voltage imbalance because of the open delta transformer bank and the large imbalanced single phase load.

[bitsnpieces1]

I worked for a municipal wastewater system with lift stations running 1.5hp to about 150hp motors powering pumps. The city also owns/runs the power plant. We had a HIGH leg, not low, it ran just a few volts higher than the other two, maybe 2-3. I'd have to check to be sure of the numbers. This also is from stepdown transformers, ~17kv to 480 or 240 volt. Then step down to 120ac 1ph & 24/12v ac 1ph. It is quite possible that the high leg was a specialty hookup, albeit it widely used.

[Jim Rankin]

I expect that the higher voltage phase was the one across the whole city that was usually the lightest loaded. There could also have been automatic transformer tap changers or power factor correction capacitors that were affecting the voltage on that phase.

In practice, a few volts would have little real effect especially on a 480 volt system (less than 1%) (480 + 480 + 483)/3= 481 483-481=2, 2/481 X 100%=0.5% imbalance. Imbalance in voltage creates between 5 and 10 times more current imbalance which is what leads to the derating.

I guess it was always easy to ID the phases with that one being consistently high.

[Jim Rankin]

I think this has strayed a long way from the original poster's question about measuring imbalanced 120 volt line to N circuits from a generator set.

As far as I know, A perfectly balanced output would be the most desirable and normal thing for a generator set.

Some possible reasons for the voltages not to be balanced have been suggested, but I expect the original problem was that there was really no neutral to measure to on that generator set when set for 240 volt single phase (except maybe in the convenience outlet).

Voltage is always measured between two points. Line to line, line to N (or Line to ground (IF! the N is bonded to ground). When measuring voltage on a set without the N being available, you will always be able to measure voltage from line to ground because of capacitance, but it will not always be correct and will not supply a load because there is no circuit back to the windings if the Neutral is not bonded to ground or connected through a load somehow.

The most common problem with this situation is when someone connects a 240 volt ONLY set to a single phase 120/240 service or fails to connect the neutral properly to a 120/240 set. Without a neutral, the 240 will divide between the two "120" volt sides of the panel according to the balance of the connected loads. The heavily loaded side will drag down and the lightly loaded side will rise. Connect up the neutral and the voltage will divide evenly unless there are problems with oveloading.