Why even have a step down transformer?

[sourcecharge]

So here a question...

Why even have a step down transformer when the output is regulated?

So I posted a design for a regulated power supply here using mosfets:

https://www.eevblog.com/forum/projects/1800w-50v-power-supply/

and basically, if a 110ac output was full wave rectified into a filter cap, and you have a 8KW generator, why even have a step down transformer?

Of course the input to the bridge would be fused for 40 amps, but here's the question...

Is the power drawn from the generator going to have the current calculated in I(rms) or I|avg| ?

If it's I|avg|, would the 110ac mains also be calculated in I|avg|?

Is this where I finding there is some weird inefficiency?

[Benta]

Why even have a step down transformer when the output is regulated?

To avoid killing people.

[Zero999]

Isolation from the mains of course!

[sourcecharge]

Why even have a step down transformer when the output is regulated?

To avoid killing people.

 

Isolation from the mains of course!

I'll use a 8KW generator...it's isolated...

[LeoTech]

Why even have a step down transformer when the output is regulated?

To avoid killing people.

 

Isolation from the mains of course!

I'll use a 8KW generator...it's isolated...

Well, true that, but why only an 8 KW generator, let's use a nuclear power reactor!

[Raj]

maybe,to prevent house fuse from blowing up and blow up local fuse at min, and transformer itself at max.Plus ,yes,mains isolation and ground isolation.

[sourcecharge]

Why even have a step down transformer when the output is regulated?

To avoid killing people.

 

Isolation from the mains of course!

I'll use a 8KW generator...it's isolated...

Well, true that, but why only an 8 KW generator, let's use a nuclear power reactor!

ok, but we'll still use the 40 Amp fuse before the full wave bridge, is that alright?

BTW, would that 40 amp fuse blow at 40 amp (rms) or 40 amps |avg| ?

or to put it another way:

Is the power drawn from the generator (sorry, nuclear power reactor, well on second thought, let's use the generator for now) going to have the current calculated in I(rms) or I|avg| ?

maybe,to prevent house fuse from blowing up and blow up local fuse at min, and transformer itself at max.Plus ,yes,mains isolation and ground isolation.

it would be isolated from the house.

the generator creates the isolation.

the current would be limited with a fuse before the full wave bridge.

[sourcecharge]

before anyone says it, the regulation mosfets, and components would of course need to be changed somewhat to allow the higher filtered voltage.

[Benta]

BTW, would that 40 amp fuse blow at 40 amp (rms) or 40 amps |avg| ?

Always RMS, which represent the heat value of the current. Noone uses average current for anything.

[Nusa]

Also, fuses are not precision devices; they're safety devices. A so-called fast-blow 40A fuse might still tolerate 100A for a full second before it blows. Or 42A for half an hour. Their rating also changes based on ambient temperature -- the stated rating is usually at 25C. Datasheets will usually give you those details in chart form.

[sourcecharge]

BTW, would that 40 amp fuse blow at 40 amp (rms) or 40 amps |avg| ?

Always RMS, which represent the heat value of the current. Noone uses average current for anything.

You know that make perfect sense...

The only weird part is, when I simulate the circuit in B2 spice, the AC rms current is like almost 3x higher than the DC avg current, but the AC absolute average current, or |avg|, is almost exactly the DC avg current...

So that's why I not sure about this...

Also, fuses are not precision devices; they're safety devices. A so-called fast-blow 40A fuse might still tolerate 100A for a full second before it blows. Or 42A for half an hour. Their rating also changes based on ambient temperature -- the stated rating is usually at 25C. Datasheets will usually give you those details in chart form.

Ya, that's a problem...safety of this thing would have to be strict.   

Maybe start off with testing the fuses first to see what their tolerances are the day and time the converter is used?

[dmills]

The RMS current will tend to rise as the conduction angle of the bridge falls, this is why all modern power supplies above very modest output have power factor correction to make the input current sinusoidal.

The bigger you make the input caps the worse the effect will become, try it in spice.

But seriously, telecom rectifier!.... I like the Eltek stuff, but HP also have 48V server power supplies that are cheap as chips surplus.

Regards, Dan.

[Brumby]

Also, fuses are not precision devices; they're safety devices. A so-called fast-blow 40A fuse might still tolerate 100A for a full second before it blows. Or 42A for half an hour. Their rating also changes based on ambient temperature -- the stated rating is usually at 25C. Datasheets will usually give you those details in chart form.

Ya, that's a problem...safety of this thing would have to be strict.   

Maybe start off with testing the fuses first to see what their tolerances are the day and time the converter is used?

If that's the case, why are you still hanging on to a fuse solution?  Current monitoring circuit breakers have a much more predictable performance characteristic.

[sourcecharge]

The RMS current will tend to rise as the conduction angle of the bridge falls, this is why all modern power supplies above very modest output have power factor correction to make the input current sinusoidal.

The bigger you make the input caps the worse the effect will become, try it in spice.

But seriously, telecom rectifier!.... I like the Eltek stuff, but HP also have 48V server power supplies that are cheap as chips surplus.

Regards, Dan.

Man, where are you guys getting these telecom rectifiers?  48V and 40A for 15 bucks? There's just no way... I would buy like 4 of them if I found that deal..

Ya, I think it's the big caps that causing the inefficiency...... that and the low inductance.

Also, fuses are not precision devices; they're safety devices. A so-called fast-blow 40A fuse might still tolerate 100A for a full second before it blows. Or 42A for half an hour. Their rating also changes based on ambient temperature -- the stated rating is usually at 25C. Datasheets will usually give you those details in chart form.

Ya, that's a problem...safety of this thing would have to be strict.   

Maybe start off with testing the fuses first to see what their tolerances are the day and time the converter is used?

If that's the case, why are you still hanging on to a fuse solution?  Current monitoring circuit breakers have a much more predictable performance characteristic.

That sounds like a great solution.

Thanks, I'll look into that!

[dmills]

"hp 48V server power supply" on Ebay seems to give good options (Nothing quite that good, but still very, very cheap, the bitcoin mining thing seems to have somewhat queered the field), or I can have a brand spanking new Eltek SMPS 4000 4kW 48-53V jobbie for well under 200 quid (Under 100 if I don't mind second hand and having to go get it).

Either way it completely stomps having to roll your own, especially as with all due respect you clearly do not really know what you are doing in this space.

Do be careful, even 48V when these current levels are in play has major damage potential.

regards, Dan.

[Red Squirrel]

I'm new to this stuff myself and I actually want to build some high power SMPS eventually (I should probably start with small ones first) but my understanding is that it's for at least these 2 reasons:

1: Isolation.    By having a transformer you now end up with isolated power on the other end.  Now you can choose to tie one leg to ground if you need the rest of the circuit to be ground referenced, or you can leave it floating, or you can even have a centre tap and tie that to ground, or declare that as ground, if your device requires negative voltage.  Basically it gives you freedom to redefine the ground or simply let it float. For example for an audio amp you would probably have a centre tap secondary, and the centre would be the system ground.  In lot of situations you just want to leave it floating, for safety. Ex: phone chargers and such.

2: To step down to a more suitable voltage for the regulation components.  If you are making a 5 volt power supply, it makes sense to step it down to like 20 volts or something like that.   Then the components on that side can be rated lower.  The capacitors and mosfets for example. You can use physically smaller capacitors to get the same capacitance.    It also makes that side of the circuit much safer.

Someone do correct me if I'm wrong on these points though, I'm still learning myself.

[tpowell1830]

So here a question...

Why even have a step down transformer when the output is regulated?

So I posted a design for a regulated power supply here using mosfets:

https://www.eevblog.com/forum/projects/1800w-50v-power-supply/

and basically, if a 110ac output was full wave rectified into a filter cap, and you have a 8KW generator, why even have a step down transformer?

Of course the input to the bridge would be fused for 40 amps, but here's the question...

Is the power drawn from the generator going to have the current calculated in I(rms) or I|avg| ?

If it's I|avg|, would the 110ac mains also be calculated in I|avg|?

Is this where I finding there is some weird inefficiency?

Okay, for some reason, you want 110 X 1.414 VDC. Maybe you want to drive a DC motor or one helluva bench power supply at ~40 amps. You want to drive this directly with a 8KW genset. No problems doing this so far.

I am concerned that you are not taking the 156 VDC @40 amps serious enough from a safety perspective. That yields 6224 Watts. The other problems, and many they are, such as a diode failing closed would put 110VAC onto your power supply.

You may know exactly what you want to accomplish, but your line of questions says that you do not have the knowledge or experience to manage this safely.

Please re-consider this for awhile until you have a clear idea of all of the ramifications involved. I am a bit afraid for you. 

Please take this with a grain of salt, as it is just my 2 cents...

[Gregg]

If you do get your induction heated 3D metal printer running, the power source will probably need to run continuously for many hours to make anything sizeable.  Telco rectifiers are built to run 24/7 for years and already have all of the cooling, EMF mitigation, harmonics filtering etc. engineered for you.  I seriously doubt that a proper power supply is the biggest hurdle in the path of your project. 
Telcos use lots of small 48V nominal that float batteries 53V to 54V and they are slowly phasing out the heavy transformer units for smaller and more efficient SMPS making the older technology units relatively cheap and relatively unwanted by the telco industry. 
Where are you located?  I’ll have to look in my storage and see if I have something suitable for your project.

[Gregg]

I just looked in my storage and found I have 6 SMPS modules that output 19 amps each at 48VDC as well as a shelf to hold 3 of these in parallel.  I'll let them go cheap; PM me if you are interested and I'll send more pictures.

[sourcecharge]

I'm not familiar with Power supplies that are called rectifiers.

Why are they called rectifiers instead of just power supplies?

If they are PS but with a rectifier on the output, why don't they just say power supply with recified output?

If these things don't need batteries or capacitors between the load and the PS, then I might want to get a couple of these things, but I am still trying to get the "roll my own" PS done...


I would like to find one that does 48V and 50A powered from a 220AC line if I have to give up on my DIY PS...

I've just put too much work into it to give up on it without at least trying to get rid of the small bugs..

[Brumby]

The term "Power supply" is a generic term.  It covers any and all possible solutions for providing power to a device.  Anything from a battery to a bench supply to industrial scale stuff.

"Rectifier" is, technically, something that creates DC - but sometimes it can get used as a less-than-precise substitute for "power supply", implying DC output.  My impression is that this has come from historical origins - something which other members may be able to shed more light.

[sourcecharge]

So here a question...

Why even have a step down transformer when the output is regulated?

So I posted a design for a regulated power supply here using mosfets:

https://www.eevblog.com/forum/projects/1800w-50v-power-supply/

and basically, if a 110ac output was full wave rectified into a filter cap, and you have a 8KW generator, why even have a step down transformer?

Of course the input to the bridge would be fused for 40 amps, but here's the question...

Is the power drawn from the generator going to have the current calculated in I(rms) or I|avg| ?

If it's I|avg|, would the 110ac mains also be calculated in I|avg|?

Is this where I finding there is some weird inefficiency?

Okay, for some reason, you want 110 X 1.414 VDC. Maybe you want to drive a DC motor or one helluva bench power supply at ~40 amps. You want to drive this directly with a 8KW genset. No problems doing this so far.

I am concerned that you are not taking the 156 VDC @40 amps serious enough from a safety perspective. That yields 6224 Watts. The other problems, and many they are, such as a diode failing closed would put 110VAC onto your power supply.

You may know exactly what you want to accomplish, but your line of questions says that you do not have the knowledge or experience to manage this safely.

Please re-consider this for awhile until you have a clear idea of all of the ramifications involved. I am a bit afraid for you. 

Please take this with a grain of salt, as it is just my 2 cents...

I appreciate your concern, and that you are open to the idea of doing such a feat with the proper safety precautions.

Let's keep this as hypothetical for now to keep your mind at ease.

One of such precautions is to calculate the current carrying capacity of the wire being used.

So, I have posted 2 simulations that show the same circuit with the AC current meter simply measured differently.

One shows the Absolute Average AC current, while the other shows the RMS AC current.

Which one is correct?

Keep in mind that this is an IDEAL B2 spice simulation with a 60hz 170Vp sinewave source input, and it is not realistic where heat is calculated.

This source would be the generator, so don't worry, it's isolated everyone. 

All energy is conserved, and if the correct current calculation is the RMS current, where does wasted energy go? 

Does anyone here know?

[james_s]

"Rectifier" in this context is a rather archaic term for a DC power supply. I've heard it used in fields that have been around for a long time, telco, railroad, electroplating, stuff like that.

[sourcecharge]

Thanks but I'm really only interested now in what the correct current measurement is from a AC Voltage source would be.

Please see above...

Is it the absolute average current

Or

Is it the RMS current

If it is the RMS current, where does the wasted energy go?

The only path is through the 1 load, so how does 500A(rms) magically transform in to 160A(avg), which represents a difference in the calculated power.

So somehow, all that energy on the load magically vanishes?

Does anyone know the answer to this?

Is this AC voltage source actually putting out only the absolute average current?

This current is about the same as the DC average current, which conserves energy.

If that is true, then why is power at our houses being measured as RMS current?

[dmills]

They are BOTH correct, but measuring different things!

The RMS measurement is measuring the heating effect, and is very much larger then the average because the conduction angle of the diodes is so small, and the heating is proportional to current squared.

The average is the integral of charge transferred with respect to time measured over 1 second.

If you look at the current waveform thru one of your diodes you will see very short, very large pulses just as the waveform reaches peak, this is the time at which the diodes are forward biased, and it is this massive current pulse that gives the very high RMS value. Smaller capacitors, series inductance, supply impedance (not negligible on most small generators) all help.

This is why sane modern power supplies of this power class have power factor correction and are typically designed to have far larger rectifier conduction angles.

Regards, Dan.