High frequency current source power supply for LED lights hanging off the power bus?

[treez]

Hello,

Suppose i have a load of LED luminaires that i want to light from a power supply.

I can do it with a voltage source power supply, and have current regulators connected up to the voltage bus.....................OR

I can have a high frequency (50KHz) sinusoidal AC power CURRENT source, and simply couple the led luminaires to the power bus via current transformers.

The current source way means having completely waterproof connection.........as opposed to the voltage source way, where lots of non waterproof connectors would be used.

So, do you think this is a viable proposition?.......and if so, what steps would you take to assure a good power factor in the current source bus?

 

[chuckey]

if you use a AC current transformer then the LEDs have to be used two at a time, wired back to back to stop the DC component saturating the core.
Frank

 

[mtwieg]

Getting a good power factor with such a scheme is likely not possible with a nonlinear load at 50KHz. Why do you think a good PF is important? What is the actual AC current source?

If isolation is important, than it would probably be easier to just use a DC bus and have isolated DC-DC in the LED drivers.

 

[treez]

chuckey

if you use a AC current transformer then the LEDs have to be used two at a time, wired back to back to stop the DC component saturating the core.
Frank

We dont put them two at a time, its just a transformer, driven by a current source, and so the secondary current is by NPIP=NSIS, as normal.

mtwieg

Getting a good power factor with such a scheme is likely not possible with a nonlinear load at 50KHz. Why do you think a good PF is important? What is the actual AC current source?

The AC current source is a switch mode current source.........just like a kind of LLC converter but instead its current regulated, so that it is a current source (AC 50KHz)

If you think about it, the connection is via magnetic field, and there are no connector gaps for water to seep into.

Power facotr is always necessary when you have AC.......and we have AC.....but how would we best assure a good power factor......?

The lamps work fine, the system works fine, but i am looking into the power factor...do you know how to correct the power factor with this set up?

 

[mtwieg]

If you think about it, the connection is via magnetic field, and there are no connector gaps for water to seep into.

Power facotr is always necessary when you have AC.......and we have AC.....

This isn't necessarily true. PF is typically desired because it causes line distortion which can interfere with other loads, and it increases transmission losses over very long distances. But in your case you have many identical devices, so interference shouldn't be an issue, and if you're sourcing current through a high Q resonant tank, then the nonlinearity in the load probably won't be very noticeable to the source.

It's the same situation with wireless power transmission, which uses high Q resonators loaded with rectifiers. The rectifiers always have very poor power factor (they appear either as a very low impedance capacitance or high impedance inductance), but that has basically no relevance to the overall transfer of power.

 

[treez]

mtwieg, thanks .......regarding this, i can assure that the capacitor C1 in the following is there to correct the power factor...........

https://i46.tinypic.com/28jxsgg.jpg

......please ignore the actual inductance values, they wouldnt be this high in our product.
.......please also ignore that the load is a battery......the load can also be a led lamp.

Our senior engineer said that C1 is there to correct the power factor.

Our power bus is about 8 metres long, and so we still need to reduce conduction losses, and so getting a good power factor is worth while, especially since we transmit at 50KHz, so there is skin effect to worry about too.......so we do need to correct the power factor.
Anyway mtwieg, you agree that the waterproof nature of this inductive coupling method is very good?...the connection is by magnetic field which can go through the waterproof plastic conductor coating.

 

[mtwieg]

mtwieg, thanks .......regarding this, i can assure that the capacitor C1 in the following is there to correct the power factor...........

The function of C1 in that circuit is effectively to resonate with the inductance of the CT, causing the voltage to increase. But once those rectifier diodes begin to conduct to a low impedance load (like a battery, a large capacitor, or a LED), that capacitor will be effectively shorted out and becomes irrelevant. As long as those diodes have a large conduction angle, the circuit will inherently have poor PF, regardless of C1.

You said yourself before that your circuit was similar to a LLC converter. In a LLC converter, the rectifier circuit effectively presents a very bad PF to the primary, but that's irrelevant to the operation of the LLC converter. An LLC converter doesn't need a secondary resonant capacitor, because the secondary rectifier diodes are forced to conduct almost continuously.

Our senior engineer said that C1 is there to correct the power factor.

Our power bus is about 8 metres long, and so we still need to reduce conduction losses, and so getting a good power factor is worth while, especially since we transmit at 50KHz, so there is skin effect to worry about too.......so we do need to correct the power factor.

Have you actually simulated the whole system, including the resonant tank on the primary? Can you actually demonstrate the impact of poor power factor on the system, in a simulation?

Anyway mtwieg, you agree that the waterproof nature of this inductive coupling method is very good?...the connection is by magnetic field which can go through the waterproof plastic conductor coating.

Sure, waterproof is good, but there are lots of ways to achieve it. I can't really say what the optimal method would be without a better idea of what the whole project looks like.

 

[treez]

Thanks mtwieg,

Its just a couple of LED luminaires (10W to 40W) "attached " to the HFAC bus.......they are attached by inductive couplers. The HFAC bus is powered by a SMPS, which produces the constant current which is used by the LED luminaires..the bus is a loop of wire, but obviously its twisted together so the loop area is as small as possible.........the couplers literally just hook into it (you manually "pull" a bit of the bus out and hook it into the coupler....that gives the primary connection, and the secondary is inside the coupler, as you would expect )...thats all there is. (obviuosly theres the battery and the emergency led driver to power the lumiares when the mains fails...the mains provides power to the smps that produces the 50KHz HFAC bus.

Its not possible to get anywhere near as good waterproofing as this "Inductive coupler" method with any other practical method......as i mentioned, magnetic field can couple through the waterproof insulative coating of the wire......thers no connector gaps for water to run into.

I cannot demonstrate the impact of poor power factor yet.

But i assure you that the most senior engineer on this project tells that C1 is needed to give good power factor...and this guy knows his stuff very well indeed.

In my simulation i get primary voltage and current being in phase with no capacitor......and when i add the cap i get them going out of phase...so i am confused.........I note (especially with the higher voltage led strings) that the primary voltage waveform is significantly non linear......(has harmonics of the 50khz sine in it)

 

[mtwieg]

Thanks mtwieg,

Its just a couple of LED luminaires (10W to 40W) "attached " to the HFAC bus.......they are attached by inductive couplers. The HFAC bus is powered by a SMPS, which produces the constant current which is used by the LED luminaires..the bus is a loop of wire, but obviously its twisted together so the loop area is as small as possible.........the couplers literally just hook into it (you manually "pull" a bit of the bus out and hook it into the coupler....that gives the primary connection, and the secondary is inside the coupler, as you would expect )...thats all there is. (obviuosly theres the battery and the emergency led driver to power the lumiares when the mains fails...the mains provides power to the smps that produces the 50KHz HFAC bus.

Its not possible to get anywhere near as good waterproofing as this "Inductive coupler" method with any other practical method......as i mentioned, magnetic field can couple through the waterproof insulative coating of the wire......thers no connector gaps for water to run into.

Right, I understand the method, and yes if the enclosure is designed properly then it will be very robust against moisture. It's hard to imagine a more sealed approach without resorting to very expensive sealed connectors.

Are the CTs fully closed, meaning you have to thread the HVAC through it, or do they clamp open and shut?

But i assure you that the most senior engineer on this project tells that C1 is needed to give good power factor...and this guy knows his stuff very well indeed.

I respectfully disagree. Anecdotal qualifications aren't really relevant to me.

In my simulation i get primary voltage and current being in phase with no capacitor......and when i add the cap i get them going out of phase...so i am confused.........I note (especially with the higher voltage led strings) that the primary voltage waveform is significantly non linear......(has harmonics of the 50khz sine in it)

Could you share the simulation files, assuming they're in LTspice? I'm betting the phase shift is just due to a change in resonant frequency (which is expected, and probably unavoidable to an extent), but that just means you should adjust your resonant frequency.

What is actually producing the AC current? Is it just a half/full bridge driving the resonant load, with a modulated output voltage or duty cycle?

 

[treez]

Yes i will share them, but to be honest with you, i only did the emergency light, and i have never seen the schem for the HFAC current source power supply, or the detail of the couplers......just that i know they are the size you can grip with a hand.

This is what ive been using to simulate the coupler.....but the L values are almost certainly too high here....at least i believe so.

https://i47.tinypic.com/ws30y0.jpg


....one reason i look into it is they've asked me for a method of producing a phase cutting of the 100KHz haversines so we can have the same coupler "set" to produce different output currents, and ~300Hz PWM dimming can be also applied to make the range even more wide.....and ideas for the 100KHz phase cutter?

phase cutting
https://www.edaboard.com/threads/276625/

.....i speak of battery charging here, but really this phase cutting method is for the luminaires.




I must admit, i'm amazed there are no objections to my post here........i have just put forward a 50KHz, yes, 50 KiloHertz power bus, and nobody appears to be batting an eyelid.................this power bus is 1000 times higher than the normal mains power bus..........and yet nobody is batting an eyelid...........surely you should be coming up with "hey...you can't do that" ?

BTW , this system works and is operational.