Is this the only way to get bidirectional, isolated SMPS with step-up and down?

[treez]

Changeover switch to switch between circuits that are mains isolated from each other?

Hello,

We need to do a (small as possible) 2kW SMPS car battery charger which is bi-directional so that it can also put the battery energy back into the mains.

Do you think the attached is a way to do it?
If so, can you state if the needed “changeover switch" exists anywhere?

Preferably the changeover switch could be electronically controlled, and could switch between circuits whilst preserving their mains isolation from each other.

Please find schematic attached.

 

[BradtheRad]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

At 360V the battery voltage will just barely overcome mains peak voltage. The question arises, what waveform do you plan to send onto the grid? Will it be sine PWM?

(I'm not saying it has to be a sinewave although that would be the expected waveform.)

 

[treez]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

thanks, as in the schematic shown, the waveform going back from the battery would go via a grid tie inverter back into the mains. Do you reckon the full bridge SMPS could be made bidirectional by four changeover relays in the schematic? -Literally swapping over the input and output connections.

 

[BradtheRad]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

Regulations in your area may require a transfer switch. The purpose is to island your system from street power if the grid blacks out. This may need to be installed between grid and inverter, or between inverter and battery bank.

If you can incorporate this function in your changeover switch, then it will simplify things.

Of course you only need to island your system when it is in the mode of sending power onto the grid. So in the event of a blackout your changeover/transfer switch should automatically either:
(a) go into charge mode, or
(b) disconnect battery from inverter, or
(c) disconnect your entire system from the grid.

Another question is what will operate your changeover/transfer switch? Will it be manual? Automatic?

It could be a relay, controlled by the grid, or else the battery bank.

Or it could be manual, perhaps with 2 positions, or else 3 positions where the middle position disconnects everything.

 

[treez]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

the purpose of this changeover switch will just be to make the full bridge smps bidirectional. I am seeking a mains isolated changeover relay, 4 of them. That saves having to design a bidirectional full bridge SMPS.

For example, the following changeover relay datasheet doesn't say whether there is mains isolation between the points that it switches between..in fact none of the relay datasheets seem to say this.

CHANGEOVER RELAY
https://media.digikey.com/pdf/Data Sheets/Tyco Electronics Aerospace/K40P_Series.pdf

 

[FvM]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

The usual way to achieve what you want is to have a bidirectional PFC (also called AFE - active front end) and a H-bridge output stage instead of a passive rectifier. So the converter is able to transfer energy in both directions. No switchover is required in this case, just set a negative current setpoint to the controller.

 

[chuckey]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

Treez, your relay is good for 6 KV at 50 hertz, between its contacts, but only 500V between the coil and earth. In the output of diesel generator sets that take over supplying mains to transmission sites, special contactors have to be used. there are two of them which close depending if mains or generator is selected BUT they are mechanically interlocked so if the contacts weld on one of them, the other will not pull in.
I think you will need a lot of mains filtering to stop spikes seeing off your electronics.
Frank

 

[mtwieg]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

The usual way to achieve what you want is to have a bidirectional PFC (also called AFE - active front end) and a H-bridge output stage instead of a passive rectifier. So the converter is able to transfer energy in both directions. No switchover is required in this case, just set a negative current setpoint to the controller.

Agreed, if want maximum density then using two unidirectional converters with a switch is not optimal. Rather you should use only bidirectional converters. So an AFE followed by a bidirectional full bridge converter would be best.

 

[treez]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

Thanks, we tried a bidirectional full bridge SMPS , but found that when we tried to drive "backwards" through it (ie, as a current fed full bridge), then we got large high voltage spikes on the "primary" side fets, due to the leakage inductance. We have a simulation of a Full bridge LLC converter which happens to run bidirectionally, without those problems. The only problem is that the frequency range of the LLC in "backward" mode is going higher than we would like...its a matter of making sure that all of our loads can be handled by the LLC, without too much switching loss. The converter has now been spec'd up to be a 7kw LLC converter (& bidirectional).
The attached is the ltspice simulation of the bidirectional 3.5kw LLC converter.


thanks for the info on the relay. By the way, its not two converters and a switch, its one converter, with the changeover switch module to make it bidirectional. -However, another problem is that getting all the high current tracks to the 4 changeover relays is going to be a problem...specially with all that current, and the fact that each "end" of the converter is isolated from the other.....however, we will still look into the changeover switch method of making one full bridge converter bidirectional.

 

[FvM]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

I'm not sure if bidirectional LLC is a suitable way. The bidirectional power supplies that I know, e.g. battery testers up to 180 kW are following a more conventional design concept..

But anyhow, I just wanted to mentioned that "grid friendly" drives and power supplies with recuperation capability are widely used in industry and green energy applications. Don't reinvent the square wheel.

 

[treez]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

I'm not sure if bidirectional LLC is a suitable way. The bidirectional power supplies that I know, e.g. battery testers up to 180 kW are following a more conventional design concept..

LLC, is, we have found, the only topology that facilitates bidirectional power flow. The basic full bridge smps offers up too many problems in terms of overvoltage spikes due to transformer leakage when going "backwards".

But anyhow, I just wanted to mentioned that "grid friendly" drives and power supplies with recuperation capability are widely used in industry and green energy applications. Don't reinvent the square wheel.

Thanks, we definitely don't want to reinvent anything, just want to get the job done as cheaply and simply as possible. Are you saying that these are off-the-shelf items, that we can just buy? We haven't seen any.

What we want is a 7kw automotive battery charger, which can also put battery energy back into the grid. It must be PFC'd in battery charging. Battery voltage is 310v-410vdc. We would like the SMPS to be a single bidirectional unit and not two in parallel. Also, we want the PFC/grid tie inverter to be a single bidirectional unit. May i ask Do you think all this is possible? Is it available as an off-the-shelf item?

 

[FvM]

Re: Changeover switch to switch between circuits that are mains isolated from each ot

A google search for "bidirectional battery charger" or similar suggests that everybody is doing it now.

I understand that LLC is a clever solution to combine isolation and voltage variation over a large range. But there's no thing like a free lunch. You pay for it in some way, surely with a complex control scheme, and above a certain power level, the advantage is vanishing. That's no fixed limit, it may still make sense for 7 or 10 kW. I don't know, presently LLC is not my business.

But if you have problems to implement it, you should take a more straightforward way, separate AC/DC (AFE), bidirectional isolator (some kind of synchronous converter, but not necessarily LLC) and a half bridge to connect the battery to the secondary DC bus.

 

[treez]

Hello,
The attached is a bi-directional, mains isolated LLC converter, 7kW. (LTspice simulation and schematic)
Its for a electric vehicle application whereby the car battery can also be discharged back into the mains.

As far as we know, (do you know otherwise?) this configuration of LLC converter, with the LLC components doubled up and populating both sides of the transformer, is the only way to do it to be sure of getting sufficient voltage gain in both directions. There is no other topology which can be mains isolated and bi-directional, with step-up and step-down in both directions.

It’s obvious what the Upper resonant frequency is, but what inductance value would you give to the magnetising inductance for the equation’s of AN-4151 by fairchildsemi? (L10 // L12 or just L10?)

AN-4151 (page 4 has the “Lm” value for magnetising inductance)
https://www.fairchildsemi.com/application-notes/AN/AN-4151.pdf

 

[D.A.(Tony)Stewart]

Full bridges on either side of a transformer is perhaps the only way to make an efficient bi-directional isolated converter. But Modulation and magnetic designs vary widely.

Here is a 1MW bi-directional converter.

https://www.google.ca/url?sa=t&rct=...71hIJaoq0pY6bcQ&bvm=bv.99804247,d.aWw&cad=rja

 

[treez]

Thanks, as you know, that one only steps up one way and down the other.
The electric vehicle application requires either step up or down in both directions, and as far as we believe, the LLC with double populated LC's is the only way to do this. (as shown in post #1)

 

[mtwieg]

Cuk converter comes to mind.

 

[treez]

Thanks, but I don't think its isolated? I know there is an isolated version of a cuk converter but I don't think that version is bidirectional..woops...alternatively..

https://www.power-mag.com/pdf/feature_pdf/1310569074_Teslaco_Feature_Layout_1.pdf

...fig 14 of the above though, as the article says, is a bidirectional , isolated cuk if implemented with fet switches over the diodes. -but the cuk has problems with fet overvoltages and needs some special snubber, active type I think.

 

[mtwieg]

The isolation doesn't prevent bidirectional operation. And as far as I know the Cuk doesn't require any extraordinary amount of snubbing, if built properly.

 

[Easy peasy]

At 2kW a bi-directional Cuk is just do-able, with Si or SiC mosfets and SiC diodes, some snubbing will be required, the ease of drive is often a deciding factor in this approach... of course the series caps required do bulk up the total volume a little.

 

[treez]

thanks, I still think the changeover switch approach of the top post is the way forward for bidirectionality, its so much simpler. All that is needed is switches (relays etc) that disconnect from one side of the isolation barrier and connect to the terminal on the other side, so that if any switch is wrongly actuated, we don't get the non-isolated side contacting with the isolated side. I am sure this must be possible with some type of multiple pole switch (contactor etc)?