Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

6kVA autotransformer upstream of bidirectional grid tied inverter?

Status
Not open for further replies.

cupoftea

Advanced Member level 5
Joined
Jun 13, 2021
Messages
2,545
Helped
54
Reputation
108
Reaction score
111
Trophy points
63
Activity points
13,355
Hi,

I have just seen a grid tied inverter for connecting a battery to the 240VAC UK mains. Its bidirectional. Its 5.5kWh rated. The mains comes to it via an autotransformer, which knocks the mains down by just 15-20V or so. The reason for the autotransformer said to be because the GTI needs protection should the mains go too high.

This autotransformer was like a "ring donut" type of diameter approx. 20cm and depth 7.5cm

Would you agree this doesn’t seem big enough for 6Kva? (The one below is much bigger)

Also, it seems strange to have such a thing?...dropping the mains by so little?

(i wondered if the real reason for it might be too stop mains transients getting to the bidi GTI?)

6Kva Autotransformer
 

Also, on the subject of GTI's, how do they sense when the mains has failed? (as you know by law they have to)...is it by slightly changing their frequency, and then seeing if the "incoming" mains frequency changes?
 

There are several differing methods for loss of mains detect, freq pull is common.

Auto transformers are typically half the size of isolating transformers for obvious reasons.
 
0.33 kVA/kg sounds about right for high end CRGOS laminated autotrans. That has to be derated for low pf. But the critical issue is thermal runaway has low margins for excitation overvoltage in some designs. This is when inductance drops like a rock and it turns into a small incendary device if that is used on the BiDi GTI.

But this is not ferrite judging by the weight specs.

- iron losses of the toroidal core are typically very small, typically 1.1W/kg at 1.7 Tesla and 50Hz

But I have no idea what your specs are and cannot state it is compatible to your requirements.
 
Last edited:
for a ferrite autotrans
not sure you'll find too many 50/60Hz auto-transformers made out of ferrite ....

Certainly high mu, silicon steel toroids draw significant peak currents near the end of each half sine, if you run them over design voltage
 
autotransformer, which knocks the mains down by just 15-20V or so.

This usage of an autotransformer acts by means of a 'small' winding (few turns) bucking mains AC. It's in series with the load and carries full current. The voltage drop is by reactance... thus there is little heating effect.

The 'large' winding (many turns) steals a couple Amperes from mains, to step it down to the 'small' winding (and provide bucking effect). Thus the large winding need not be thick wire.

Some net power is wasted but it's slight in view of overall power going through the system.

Power factor error is reduced by designing the transformer with adequate Henry value in primary and secondary, as well as step-down ratio.
That is the benefit of transformer behavior. If instead windings were re-arranged as a coupled inductor, then it would cause large power factor error.
 
Last edited:
Sorry Brad, you analysis is a little bit wide of the mark - take a 230 : 115 Vac auto Tx, supplying a 115V 20A load

the auto Tx carries only 10A in each wdg - but there is 20A in the load .... neat trick - but that is they way they work ....

no power is really wasted - it is all performing a function ( p.s. many people think the lower winding supplies 20A but this is not the case as a simple drawing of the real ckt indicates )
 
Thanks, and i woudlnt minfd betting any mains transient would find it more difficult to get through with the autotxfmr there?
 

Hi,

the 240VAC UK mains. Its bidirectional. Its 5.5kWh rated. The mains comes to it via an autotransformer, which knocks the mains down by just 15-20V or so.
let´s say an autotransformer for stepping down 20V from 240V...
just needs to be rated for this "20V x load_current"

Thus it´s rather the size of a 6kVA x 20V / 240V = 0.5kVA

Klaus
 
My source is Rod Elliott's website, where he goes into depth about construction and operation of the autotransformer. He tells how it can easily transition to a bucking transformer (or else boost transformer), with the adjustment of a few specs. He states:

A bucking transformer is really a modified version of an autotransformer. The difference is that only a small part of the winding has to carry the full load current.


(His site specifically asks that his home page be linked):
sound-au.com

Several schematics are in the article. These gave me a clearer picture how to simulate an autoformer (starting with the ordinary transformer model). It brought out the fact that the unit can get by with a power rating at a fraction of the overall Watt throughput (per Klaus post #11).
 
Last edited:
bucking transformers have isolated secondaries, separate to auto-transformers which do not ....

the iso sec can be placed in phase with the mains - boosting - or antiphase - bucking ....
 
My source is Rod Elliott's website, where he goes into depth about construction and operation of the autotransformer. He tells how it can easily transition to a bucking transformer (or else boost transformer), with the adjustment of a few specs. He states:




(His site specifically asks that his home page be linked):
sound-au.com

Several schematics are in the article. These gave me a clearer picture how to simulate an autoformer (starting with the ordinary transformer model). It brought out the fact that the unit can get by with a power rating at a fraction of the overall Watt throughput (per Klaus post #11).
"This is probably as close as you can get to the much hoped for (but disallowed by the laws of physics and the taxman) 'something for nothing'."

I could use more anti-phase mutual coupling on my tax-return to recoup my losses.
--- Updated ---

bucking transformers have isolated secondaries, separate to auto-transformers which do not ....

the iso sec can be placed in phase with the mains - boosting - or antiphase - bucking ....
If you start with an isolated transformer with isolated windings of different guage wire and connect them in series , to make buck/boost type auto-transformers, they are no longer isolated and the high load current only passes thru the 2ndary non-isolated winding.

While only the excitation current goes thru the primary winding to create the shared flux cancellation (buck) or addition (boost) for fractional voltage ratios. I believe the tradeoff is added series impedance and voltage regulation error for a great savings in the cost of windings and core size from the lower primary current. This and other reasons limited the % range for this small fractional voltage change method.

I may have this backwards, so please correct
 
Last edited:
a buck or boosting Tx has its isolated secondary connected in series with the mains, in phase or anti phase, the primary, is of course driven from the mains.
--- Updated ---

Fig.4 in Brads link with the explanation given as the "proper" way is a little misleading, as it removes the solid output of the LV wdg, and allows the voltage to be pulled down by the load, which is less the case for a true bucking connection.
 
Last edited:
a buck or boosting Tx has its isolated secondary connected in series with the mains, in phase or anti phase, the primary, is of course driven from the mains.
--- Updated ---

Fig.4 in Brads link with the explanation given as the "proper" way is a little misleading, as it removes the solid output of the LV wdg, and allows the voltage to be pulled down by the load, which is less the case for a true bucking connection.
Perhaps it's semantics or my lack of experience on this application, but once the isolated secondary is connected, it's no longer isolated so primary share currents from mutual coupling of secondary current and flux. I tried a simulation but could not show a significant primary current reduction at 2:1 , perhaps too much.
--- Updated ---

Ideal DC transformers variable Lp sliders 240V:120V simulation , does not match expectation for reduced primary coil current.

What is the actual equivalent circuit of a Buck Autotransformer? A fixed Variac ?
I'm trying to imagine it but incorrectly.
 
Last edited:
The LV sec has to be isolated ( as in its construction ) to allow either a bucking or boosting function - if it was permanently connected at some point during manufacture ( i.e. non isolated ) then such connections would not be possible - and usefulness would be limited.
--- Updated ---

Any common step down mains Tx can be used as a bucking or boosting Tx, e.g. 230:24 Vac, the higher the 2ndary current rating the higher the total power to the load one can have.
--- Updated ---

Auto-transformers are an interesting case, as said above for a 50% tapping, Iout = 2 x I on any wdg, which is useful.

As you move away from this split more maths is involved, for example, if you wanted 230:23 Vac auto Tx at 40A out, then the LV part of the wdg ( near neutral ) will have to supply 36 amps, with the remaining 4A coming from the 230V wdg ( there will be some Imag drawn from the source too )

and so on ...
 
Last edited:
Your answer reveals a lack of simple analysis, any AC volt source - and this includes the LV wdg on a Tx can have load current in any direction, and in fact any shape - dictated usually by the load.
Usually - and in this case - the Imag is supplied by the higher driving voltage ( relative to turns ) i.e. the 230VAC
--- Updated ---

Isolation is achieved in the usual way.
 
You did not specify the parameters needed to simulate.


I finally figured out the ratios for simulation for 10% drop https://tinyurl.com/yzro8ww5

Since these are ideal XFMR's (DC) use reset to null remanence as both grids start at 90 deg.

and observe excitation + W and difference in currents and -W generated.

Ultimately flux is mutual so there is no isolation when conduction is bridged. To reduce voltage 10% becomes trivial with 10:1 turns ratio not accounting for k and conduction losses.
 
Last edited:
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top