How to size up a mains transformer for a certain regulated output voltage?

Hello.

I remember a thread I started a while back in which people pointed out that the transformer I had picked out to use in a project wasn't sufficient for the voltage I wanted, but I can't find that thread again weirdly enough.

In my current project I am going to need a ±5,5V power supply, ±1A.
I have a choice between 2 transformers, both are 16VA but the output voltages are 6VAC or 7,5VAC.
I can't find any straight answer to this but initially I chose the 6VAC transformer but I would really hate buying that only to find out that I should have gone with the 7,5VAC.

A 6VAC(I think that is equivalent to 6Vrms) rectified should output something like (6V*√2)-2V=6,485V
Given that I am planning to use Low Drop-out regulators that sounds as a suitable output voltage, but I might as well flip a coin since I don't even know if this is correct.

Making the same calculation with 7,5VAC = 8,6V and that sounds good for using standard voltage regulators and maybe I should go with the saves option and use 7,5VAC and accept the extra power being dissipated.

Regards

David_ said:

(6V*√2)-2V=6,485V

Click to expand...

The -2V is in the formula as a typical drop through a voltage regulator. Therefore you should be okay to use the 6V transformer.

What about your house voltage? Is it stable at the normal figure? What is the normal figure where you live? Suddenly it's not easy to predict your final output.

If your house voltage is noticeably below normal, then that is a reason to consider going with the 7.5V unit.

You want +/- 5.5V at 1A, i.e., 11W. Add 10% for efficient loss (90% efficiency) and 11W will be 12.1W.

You also be a good designer and keep 10% margin on the head; that means 12.1W*1.1=13.2W.

Your transformers are rated 16VA (I assume pf to be 0.8) or 12.8W- does not make it. But perhaps you are not running on the limit all the time and it may be ok (if cost and space are critical).

Now come to the output voltage. 6V*sqrt(2)=8.49V; subtract one or two diode drops and it becomes 7V (two diode drops) or 7.7(one diode drop); good enough for LDO regulators.

Now we can talk about the current: just remember that the input filter capacitors get the peak voltage of 7.7V only a couple of ms. The ripple frequency is 100 Hz (full wave rectifier).

If you do not have a jumbo capacitor after the rectifier, you will go out of regulation if you draw 1A current for more than a few ms.

Well, you are being economical.

The 6 Volts should be OK, But If you need 5.5 Volts at 1 Amp, Your Transformer should have a Minimum 1.5 Amp Rating.

chemelec said:

If you need 5.5 Volts at 1 Amp, Your Transformer should have a Minimum 1.5 Amp Rating.

Click to expand...

Because the 1A of DC current is not at 5.5V x 2, instead it is at 7.78V x 2 or more.
Also I think the transformer voltage must supply the extra voltage used by the rectifiers (the high peak current causes a forward voltage of 1V instead of 0.7V) and caused by ripple from the filter capacitors.

Hi,

I only know that the secondary VAC is multiplied by 1.414, and the IAC is multiplied by 0.62 to get the respective unloaded DC voltage and loaded DC current the transformer can provide.

I have a tiny 1VA per secondary, 12V transformer, it's a novelty indeed... the actual DC current each secondary can provide is about 48mA, not the equally "low" 76mA inexperience led me to believe a year and a half ago!

The Hammond "Rectifier Relationships" pdf is good for sizing up transformers.

Sorry, I didn't get the notification that any answer had been posted which is really weird.

The -2V in my calculation was a generous estimation for the voltage lost over the diodes, as to the voltage in my home, just now I measured 234V with my DMM. I have had the intentions to design a device that would be put in the wall socket to be left there for a year or two in order to plot and analyze the actual voltage fluctuations, but I have forgotten about that. I feel it to be a surprisingly exciting project to do but having done a few measurements(something like 5 times over 24 hours) manually seems to indicate that my mains voltage is pretty stable, apart from that day of those 5 measurements I might have measured the mains about 10 times over the last 2 years and every time the result have been the same. I haven't actually measured any variation at all, it's always 234VAC.
In Sweden we have a 230VAC nominal voltage amplitude that is allowed to vary by +5%/-10%, actually the specification says that the voltage has to be within 207VAC and 244VAC which is +6%/-10%.
But I live quite near one of our largest power plants and there are a few industrial district in my town, which I theories means that I live in one of more stable areas. I read something saying something akin to stability is related amongst other things to how many users there are related to the area of land, so if you live far out in the bush you probably won't have as stable voltage supply as I have who live in one of our largest city's.

So it would seem as 234VAC is a pretty constant value, and the transformers are mad for 230VAC, take a look.

Hahn EI541 1128:
Power = 16VA
Secondary(s) = 6VAC(2x)
Output current = 2x1333A
Primary = 1x230VAC, 50-60Hz ±10%


Hahn EI541 1129:
Power = 16VA
Secondary(s) = 7,5VAC(2x)
Output current = 2x1070A
Primary = 1x230VAC, 50-60Hz ±10%

So there isn't really any choice here, the 7,5VAC with its 1,07A output current isn't sufficient.
This will work because this is for a power supply for my LCRZ-meter project that have been discussed in another thread and I have wished to enable a ±5V@±1A excitation signal(that is it's absolute maximum values), but I would not consider the design a failure if it didn't reach quite that current. But since I have set that goal that is what I'll design for.

The other circuits in the LCRZ-meter design doesn't need many mA at all, 0,333A will definitely suffice(meaning there are a margin), I would have liked a larger margin(thank you for telling me that, I know that I always should add margins but it does slips my mind more often than I'd like) but I'll consider the excitation signal current as a spare reserve that will have to act as the margin if needed be, by assuming that the full ±1A will never be reached completely.

The transformer is on sale you see so I'll go with that, I will when I have accomplished the objectives of the LCRZ-meter design design a version with a revised power supply anyway, I really want to end up with a battery driven device with a LCD to be optionally used stand alone as opposed to in conjunction with Matlab.

As for the circuit, here it is:


I've added the first pair of inductors as an just in case measure, those inductors footprints will be easily shorted. though I haven't picked out the values for them.

As for the output inductor of the negative regulator, it looks like a pi filter but may I use the regulators output cap as the first cap in a pi filter?
The negative regulator is much noisier than the positive regulator.

There will be also some margin on the line voltage, e.g., 230V +/- 10% or so. The voltage can be low when the demand is high and it does up when the demand is low. In that case, you should compute twice with 230V+10% and 230V-10%.

In most part of the world, the short term line frequency variation is less than 1% and the long term line frequency variation is very very small (it means I do not know).

Hence 50-60Hz+/-10% is absurd.

A mains voltage variation of +/- 10% is absurd. My electrical network was designed properly to have pretty good voltage regulation. It is 121V when everybody is using it and is 123V when they are not.
But many immigrants and refugees are coming. Lots of babies are born and us old people are living longer so the use of electricity keeps rising.

My electrical utility company gave away for free compact fluorescent light bulbs, solar garden lights and LED Christmas tree lights to save power. All traffic lights and street lights are LEDs. They also gave away for free smart meters so they can charge more for electricity used at peak times and also gave away smart thermostats where they can shut off my air conditioner instead of causing a total brown out. They say I will not notice if they turn off my air conditioner for an hour.

A mains voltage variation of +/- 10% is absurd. My electrical network was designed properly to have pretty good voltage regulation. It is 121V when everybody is using it and is 123V when they are not.

Click to expand...

It only depends on the setting of the two relays: under and over voltage cut-off relays in your national standards. Usually this is +/- 5% or +/- 10%.

Setting this to a tighter limit will trip the power too often. No one likes the power to trip often.

But many immigrants and refugees are coming. Lots of babies are born and us old people are living longer so the use of electricity keeps rising.

Click to expand...

Most immigrant and refugees (well, practically all) live at the bottom of the pyramid and contribute more to the society that they receive. Per capita, they contribute the highest. If they disappear, by some magic wand somehow, your life will never be the same again! This is nothing but globalization at your own door step.

Another class of immigrants bring talent and knowledge and money (at least one of them) and they contribute to the economy far more (else they will not be welcome). Babies are the future workers and not a burden, they are a potential asset. Old people living longer also contribute something to the society (else what is the point in the quality of life in the old age?- or is that only a potential for the old age home industry)

My electrical utility company gave away for free compact fluorescent light bulbs, solar garden lights and LED Christmas tree lights to save power. All traffic lights and street lights are LEDs.

Click to expand...

An industry is trying to reduce the consumption of a product they are producing. Something is clearly not right.

They also gave away for free smart meters so they can charge more for electricity used at peak times and also gave away smart thermostats where they can shut off my air conditioner instead of causing a total brown out. They say I will not notice if they turn off my air conditioner for an hour.

Click to expand...

Why this technology is not used by the manufacture of the air conditioners? What is that *great* thermostat?

My electrical utility company is (was?) non-profit and is (was?) owned by the government. It costs more to make an additional nuclear generating station than to give away cost savers.
Wikepedia discusses smart meters and their associated smart thermostats. It says the utility has the ability to connect/disconnect service which is what they do to reduce electricity consumption when demand is high from air conditioners.