240v or 480v power supplies.

[Yzordderrex]

Greetings,

I need to spec some multi-kilowatt 48v power supplies to run large LED arrays.

My choices are 240v or 480v - both 3 phase. My gut tells me that the 240v power supplies may be a little more reliable. I would think that 20 years ago this would be much more so, but with advances in mosfets and so forth the gap may have narrowed.

Would you go with the 240v or 480v supply?

thanks,
Bob
N9NEO

 

[Cydrome Leader]

Greetings,

I need to spec some multi-kilowatt 48v power supplies to run large LED arrays.

My choices are 240v or 480v - both 3 phase. My gut tells me that the 240v power supplies may be a little more reliable. I would think that 20 years ago this would be much more so, but with advances in mosfets and so forth the gap may have narrowed.

Would you go with the 240v or 480v supply?

240 (or 208) stuff is off the shelf, which would make things easy to
source.

 

[tm]

Greetings,

I need to spec some multi-kilowatt 48v power supplies to run large LED
arrays.

My choices are 240v or 480v - both 3 phase. My gut tells me that the 240v
power supplies may be a little more reliable. I would think that 20 years
ago this would be much more so, but with advances in mosfets and so forth
the gap may have narrowed.

Would you go with the 240v or 480v supply?

thanks,
Bob
N9NEO

Please define "multi-kilowatt". Is it 2 kW or 2000 kW?

2 kW is off-the-shelf easy. Even with N+1 for some backup.

Will you have a float battery backup?

Don't forget the distribution and protection system.


tm

 

[Yzordderrex]

12kW total

There are various manufacturers that can supply 480v input supplies up to at least 6kW. There are 8 Lamps that need about 1.5kW each. No battery backup. I wont forget anything.

Just need to know if the 240v MAY be more reliable than 480v.

Bob

 

[whit3rd]

I need to spec some multi-kilowatt 48v power supplies to run large LED arrays.

My choices are 240v or 480v - both 3 phase.

Unless there's very long wires going to the site, 240V is fine (actually, in
three-phase, isn't that 208V?). I'd worry about multi-kilowatts at 48V,
but unless the multiple is a hundred or so, not at 208 or 240.

 

[Phil Allison]

"Yzordderrex"

I need to spec some multi-kilowatt 48v power supplies to run large LED
arrays.

My choices are 240v or 480v - both 3 phase.

** In the USA, 240V power is two phase: ie 120V - 0 - 120V.

There is also 3 phase 208V power, with three 120V lines and a common
neutral.

480V power is 3 phases, with three 277V lines and a common neutral.

One advantage of having 3 phases is that when rectified, the output is DC
with only modest ripple at six times the line frequency.

A 240V PSU design can be used in the USA on two phase power or in the rest
of the world on 230/240V single phase power.

Makes it the conservative and likely much cheaper approach.



.... Phil

 

[[email protected]]

Greetings,

I need to spec some multi-kilowatt 48v power supplies to run large LED arrays.

My choices are 240v or 480v - both 3 phase. My gut tells me that the 240v power supplies may be a little more reliable. I would think that 20 years ago this would be much more so, but with advances in mosfets and so forth the gap may have narrowed.

Would you go with the 240v or 480v supply?

Do you need Power Factor Correction (PFC) ?

The traditional three phase 6 pulse rectifier will produce quite nasty
looking current waveforms.

If iron core 50/60 Hz transformers are used, various delta/wye
connections can be used for 12 or 18 pulse rectifiers, with quite
clean current waveforms even without PFC.

Using a switcher, but making a three phase input side PFC makes things
quite complicated.

One thing at least worth studying is to use three separate single
phase PFC switchers, connecting the inputs in delta (wye would be
acceptable, if the PFC is good, in order to avoid excessive neutral
loading) and simply put the +48 Vdc outputs in parallel. Of course,
there is some risk for oscillations, when a single inverter "sees" a
varying "load" due to the current fed by the other power supplies.

Three off the shelf inverters might be cheaper than a purpose built
three phase system.

 

[Neon John]

12kW total

There are various manufacturers that can supply 480v input supplies up to at least 6kW. There are 8 Lamps that need about 1.5kW each. No battery backup. I wont forget anything.

Just need to know if the 240v MAY be more reliable than 480v.

I'd go with the 240 volt version. Even if the unit is PFC'd which
requires the DC bus voltage to be higher than the line's peak voltage,
600 volt transistors can be used. That's where most of the research
and improvements have been made over the past few years.

A 480 volt PFC supply would have to use 900 or 1200 volt transistors.
IGBT voltage drop is much higher than 600 volt ones. The only
resistance for 1200 volt FETs approaches half an ohm.

So a unit using 600 volt transistors will be more efficient and
probably more reliable.

One other thing to consider. On a 240 volt system, a fault may blow
up the faulted component and maybe some circuit traces around it but
it's generally repairable. A fault on a 480 volt system generally
destroys the whole unit.

John
John DeArmond
http://www.neon-john.com
http://www.fluxeon.com
Tellico Plains, Occupied TN
See website for email address

 

[David Lesher]

12kW total

There are various manufacturers that can supply 480v input
supplies up to at least 6kW. There are 8 Lamps that need about
1.5kW each. No battery backup. I wont forget anything. Just
need to know if the 240v MAY be more reliable than 480v.

480 is an area where many extra rules/precautions are required. I'd skip
it if possible.

 

[[email protected]]

480 is an area where many extra rules/precautions are required. I'd skip
it if possible.

While I understand that the US NEC article 490.2 defines the border
between LV/HV at 600 V, However, I was not able to determine, if this
is phase to neutral, phase to phase, RMS or peak or even peak-to-peak
and does it apply to DC also ?

The 277/480 Vrms (Y/Delta) feed is well below 600 V. In a simple 6
pulse rectifier the rectified DC voltage is only 580 Vdc.

So what so special about 480 V (Delta) compared to say 120-0-120 V ?

In Europe, the IEC definitions are used, dividing between LV/HV is
1000 Vac RMS and 1500 Vdc. Thus the ordinary 230/400 V feed as the
quite common 690 V (Delta) for big motors fits well within the LVD (LV
Directive) and VFDs for these voltages are widely available.

 

[Tim Williams]

480 is an area where many extra rules/precautions are required. I'd skip

While I understand that the US NEC article 490.2 defines the border
between LV/HV at 600 V, However, I was not able to determine, if this
is phase to neutral, phase to phase, RMS or peak or even peak-to-peak
and does it apply to DC also ?

Traditionally, three phase voltages are specified line-to-line RMS.

Dunno about DC, or about the use inside a machine (as AC or DC), but the
main concern is arc flash.

I don't think 240V can really flash over. Haven't seen an example before.
Don't know if it's something about the physics, if a certain amount of
voltage is required to support a plasma dense enough and wide enough that it
grows without bound, or if it's just the extra short-circuit current
available.

I do know that domestic 240V circuit breakers and fuses are rated for 10kA
clearing; 480V are required to clear 100kA and up.

The practical issue is: does the machine conform to NEC and NFPA
regulations? (Or the equivalent CE, IEC, etc. in other countries.)

And what most of those rules come down to is, as long as you have the fuses
rated for 480V duty (they're more expensive, even the smallest are $10
each), and/or breakers to clear it, and sufficient clearance and protection
on the wiring, you're fine.

Ultimately, what you do inside a box, as long as it's properly fused, isn't
too important -- having that fuse at the input is what makes it safe.

The 277/480 Vrms (Y/Delta) feed is well below 600 V. In a simple 6
pulse rectifier the rectified DC voltage is only 580 Vdc.

My experience has been 700VDC open circuit, maybe 620-650 under load (light
filtering, PF > 0.93). Big caps will draw spikier current (PF < 0.9) and a
slightly higher voltage, which might be important in VFDs.

In Europe, the IEC definitions are used, dividing between LV/HV is
1000 Vac RMS and 1500 Vdc. Thus the ordinary 230/400 V feed as the
quite common 690 V (Delta) for big motors fits well within the LVD (LV
Directive) and VFDs for these voltages are widely available.

Yeah, 400/480 is in that uncomfortable range where it's still technically
"LV", but it likes to flash over and burn or kill people, so you have to be
more careful around it.

480V 3ph is approximately 1 HP per amp (1 HP ~= 3/4 kW), and the human body
can only put out maybe 300W peak mechanical power. It doesn't take many
amperes at 480 to overwhelm what the human body can resist (mechanically or
otherwise).

Tim

 

[Cydrome Leader]

Do you need Power Factor Correction (PFC) ?

The traditional three phase 6 pulse rectifier will produce quite nasty
looking current waveforms.

If iron core 50/60 Hz transformers are used, various delta/wye
connections can be used for 12 or 18 pulse rectifiers, with quite
clean current waveforms even without PFC.

Using a switcher, but making a three phase input side PFC makes things
quite complicated.

One thing at least worth studying is to use three separate single
phase PFC switchers, connecting the inputs in delta (wye would be
acceptable, if the PFC is good, in order to avoid excessive neutral
loading) and simply put the +48 Vdc outputs in parallel. Of course,
there is some risk for oscillations, when a single inverter "sees" a
varying "load" due to the current fed by the other power supplies.

Three off the shelf inverters might be cheaper than a purpose built
three phase system.

there are many telecom grade rectifiers that can do all this and more, off
the shelf.

 

[Lasse Langwadt Christensen]

Den mandag den 7. oktober 2013 21.53.20 UTC+2 skrev Jon Elson:

Phil Allison wrote:






There are actually 120/208 Wye systems and 240 V (usually delta) systems

used in the US. Our building at work incredibly has 120/208 Wye,

240 V delta, and 480 delta, all in ONE building. Stuff like this just

grows like topsy.



The 120/208 Wye is used a lot as you get 120 V line-neutral for

office loads like lights, computers, etc., and can use 208 for many

heavier loads like big window A/C or central A/C units. 240 V delta

might be used for larger computers, central A/C, etc., and used one

phase at a time for big heating elements and such.



Then there is the historical use of 240 V open-delta systems where

two 240 V single-phase transformers are used to provide three-phase

service to shop buildings. You can get corner-grounded open-delta,

which gives you the ability to run 240 V 3-phase motors and also

supply 240 V single-phase devices. This scheme allows you to

get 3-phase with two-pole breakers, as one of the 3-phase wires is

neutral. Or, there is also center-grounded

open-delta, where one of those transformers is center-tapped like a

residential 120/240 transformer. This gives you a typical residential

type 120/240 single phase service, PLUS 240 V 3-phase, all with one

pair of transformers. You use two separate power panels for the

two uses (single- and 3-phase.)



Jon

how did the US end up with such a mix(mess) of difference supplies?

-Lasse

 

[Cydrome Leader]

Den mandag den 7. oktober 2013 21.53.20 UTC+2 skrev Jon Elson:

how did the US end up with such a mix(mess) of difference supplies?

-Lasse

I've wondered the same with 50/60Hz outside the US.

 

[Jasen Betts]

Den mandag den 7. oktober 2013 21.53.20 UTC+2 skrev Jon Elson:

how did the US end up with such a mix(mess) of difference supplies?

They were first, and there were competing private suppliers.