Making an alternator that charges ‘properly’

[Robert Atkinson 2Participant @robertatkinson2]

Posted by Mark Rand on 11/07/2023 20:21:46:

Posted by Robert Atkinson 2 on 11/07/2023 12:28:08:

There are two problems with using a VFD to power aircraft avionics other than the safety of no mains isolation.

1/ The output is "delta"
2/ Harmonics from the PWM.

1/ means it is OK for balanced loads like fans but most avionics need a neutral return. It can be solved by a delta star transformer. This also fixes the isolation issue.
2/ needs a filter, maybe just series inductors, but may be specific to the VFD model.

Been there, done that.

Robert.

Point of order:-

If the inverter has balanced positive and negative DC busses, then it can produce star output as easily as delta. That's just a matter of a centre tap on the HF (DC-DC or AC-DC) converter transformer that generates the DC bus voltage in the first place.

Regards

Mark

That is not correct for a VFD. There are number of issues. Firstly the VFD does not have a HF converter transformer and does not use balanced positive and negative buses. They just rectify the incoming AC to a single DC bus The topology relies on an isolated load.
Additionally the arrangement you describe will only work for a balanced load. Any inbalance between phases will affect the voltages so the "neutral" will not be at 0V with respect to all phases.
To provide a star output without a transformer requires active control of the star point (neutral) voltage. This requires it's own power driver effectively adding a fourth phase driver and additional control complexity.

Robert.

[John MCParticipant @johnmc39344]

No idea what size or output alternator is being discussed, I'm guessing the article refers to making one? To be driven by steam or IC engine of what size?

Rather than make one, how about using one of these.

Known generically as a "Kubota" alternator, I think it has an output of around 150watts. Would probably need 1/3hp to drive it. Will give useful output at around 1500rpm. Price would be around the £90 mark with a suitable rectifier/regulator. The largest diameter is about 85mm. The mounting bracket can be easily modified.

Clearly no good for a Stuart10V/H but good for the larger engines in their range. Could easily be dressed up to look "period". Or sat on the front of a showmans engine?

If anyone needs dimensions of one of these let me know and I will measure one up.

[Mark RandParticipant @markrand96270]

Posted by Robert Atkinson 2 on 12/07/2023 14:51:25:

That is not correct for a VFD. There are number of issues. Firstly the VFD does not have a HF converter transformer and does not use balanced positive and negative buses. They just rectify the incoming AC to a single DC bus The topology relies on an isolated load.
Additionally the arrangement you describe will only work for a balanced load. Any inbalance between phases will affect the voltages so the "neutral" will not be at 0V with respect to all phases.
To provide a star output without a transformer requires active control of the star point (neutral) voltage. This requires it's own power driver effectively adding a fourth phase driver and additional control complexity.

Robert.

I assure you that it can and does work. Cheap VFDs won'd have the capability, but ones from the likes of Danfoss manage it. For that matter, if you've got an inverter TIG welder, it will be arranged that way as well. Unbalanced loads aren't a problem for any inverter, since the three generated phases are independent of each other, apart from timing. Some will fault if the unbalanced load limits are set too tight, but their purpose is to detect loss of phase in the load, not to protect the VFD.

I don't, personally, have those problems in the shed, since I use a seperate delta-star transformer downstream of the VFD/inverter in order to allow for a grounded neutral.

Regards
Mark

PS:- sorry for the thread drift everyone.

[Robert Atkinson 2Participant @robertatkinson2]

Mark,
I'm happy to be proved wrong but I've been working professionally with this stuff for years.

One thing is you refer to inverters, these are not the same as VFDs even though they may have a lot in common. Can you provide a model number for a VFD which can be used for unbalanced star connected loads? The point is that they are not intended to drive star or single phase loads so the electronic design and control software are not designed to support them. The phases are generally not independent of each other. All the current suppled by one phase must return through the other two. Most VFD inbalance protection is to protect the output devices not to protect the load from loss of phase. Inbalnce requires at least one phase to produce more power and excessive voltages. A device may survive inbalance when running at less than it's rated load (<33% for example) but that is not efficent.
Why do you use a Delta-Star transformer in your workshop?

Robert.

[Author]

Posts