I have seen transformers designed for antique vacuum tube circuits that assume a centertapped, dual diode tube, and list the usable DC currents in this scenario. I have also seen modern transformers that did not assume anything, but listed the current the winding and magnetic circuit will carry.

 

...
Transformers are indeed rated in AC (RMS) voltage and AC (RMS) current ratings on their secondary windings, but the relationship to the DC load current is certainly not 1:1 for a FWB.

That was how they were traditionally rated, in the days when a AC transformer secondary was normally used to power a resistive load like a vacuum tube filament or an incandescent light globe etc.

These days transformers are used >90% of applications to form a DC supply with a full bridge rect. It is the *customer's expectation* that if they pay for a 12v 1A transformer, that it can be used safely in a 12v DC 1A supply.

Manufacturers understand that and label accordingly. If they did not, and a 12v 1A labelled transformer was only good for a 12v DC 0.58A supply they would cop hell from the customers.

If you doubt me buy a modern 12v 1A transformer, make a FWB + cap supply and run it at 1A DC continuous. It will be perfectly fine.

 

These days transformers are used >90% of applications to form a DC supply with a full bridge rect. It is the *customer's expectation* that if they pay for a 12v 1A transformer, that it can be used safely in a 12v DC 1A supply.

Maybe, but I have been doing electronics for 50 years and that is the first time I have ever heard anybody say that.

I used to design transformers as well as select the ones we used, and every manufactured transformer I have ever seen is specified for ACV (RMS) when delivering the specified AC current. No manufacturer will stamp a rating on something which is contingent on the external circuit and would not apply in others like a FWCT application.

For example:

http://www.signaltransformer.com/sites/all/pdf/241.pdf

these are AC specs

 

Here is how to spec a transformer:

http://www.emcod.com/custom/trans_calc.php

The specs are AC

 

well, the transformer i've got is 120VA, with two 18v (RMS) outputs, and rated at 3.33 Amps. i'm assuming the amps are AC as well because 3.33amps*18volts*2 = 120VA.

i will be using the full bridge with capacitor circuit mentioned earlier.

now i under stand that i am limited (at maximum) to 18*0.9=16.2 volts, and 3.33*0.62=2.06 amps.

now, is this 2.06 amps AT 16.2 volts? or is it a blanket limit for volts and current? i'm not sure that it is. surely i can get 25 volts if i only draw a few mA, and between 2.06 and 3 amps (3.33*0.9 = 3 = average current) at a lower voltage.

also, im having trouble finding information on this, but what should i do with earth ground from the power point?
-do i connect it to the chassis and nothing else?
-do i connect it to the chassis and the circuit ground? and do a need a resistor between the two grounds to limit current or something? (should be fine because the circuit will be isolated via a transformer from mains)
-do i connect it to the chassis and short it to the neutral line? (apparently this was done in war times to reduce manufacturing times, but it feels.... dangerous)

i'm pretty sure no.2 is the right option, but i just want to ask you guys to be double sure.

 

The second one, without any resistor.
For a bench supply, I would prefer to include an EMI filter at the input, especially when it is a switching regulator.
bountyhunter:
The average value of full wave rectified current is 0.64 Ipk, which means that Iaverage is approx. 0.9 Irms. So the AC current requirement is about 11% more than the DC value. Is the factor of 1.7 for an additional 50% safety margin?

 

The second one, without any resistor.
For a bench supply, I would prefer to include an EMI filter at the input, especially when it is a switching regulator.

i've got an old computer power supply, can i just de-solder the input caps and inductors/chokes and use them as my emi filter? i assume the filter only removes undesirable frequencies, and is usable for buck converters, and not just flyback converters.

The average value of full wave rectified current is 0.64 Ipk, which means that Iaverage is approx. 0.9 Irms. So the AC current requirement is about 11% more than the DC value. Is the factor of 1.7 for an additional 50% safety margin?

i'm not sure, but i think it must be because the inductor only supplies current when the capacitor is charged to a lower voltage than the v(t) of the transformers secondary coil. so because it isn't supplying current for a part of the waveform, but it's maximum current still has the same upper limit, the average current is significantly reduced.

 

That was how they were traditionally rated, in the days when a AC transformer secondary was normally used to power a resistive load like a vacuum tube filament or an incandescent light globe etc.

These days transformers are used >90% of applications to form a DC supply with a full bridge rect. It is the *customer's expectation* that if they pay for a 12v 1A transformer, that it can be used safely in a 12v DC 1A supply.

Manufacturers understand that and label accordingly. If they did not, and a 12v 1A labelled transformer was only good for a 12v DC 0.58A supply they would cop hell from the customers.

If you doubt me buy a modern 12v 1A transformer, make a FWB + cap supply and run it at 1A DC continuous. It will be perfectly fine.

Larger transformers all need cooling, as soon as you approach the rated Wattage. It depends on the transformer temperature if you can use it at rated Watts, if you can exceed it, or if you have to use it at lower Wattage.

 

well, the transformer i've got is 120VA, with two 18v (RMS) outputs, and rated at 3.33 Amps. i'm assuming the amps are AC as well because 3.33amps*18volts*2 = 120VA.

i will be using the full bridge with capacitor circuit mentioned earlier.

now i under stand that i am limited (at maximum) to 18*0.9=16.2 volts, and 3.33*0.62=2.06 amps.

now, is this 2.06 amps AT 16.2 volts? or is it a blanket limit for volts and current? i'm not sure that it is. surely i can get 25 volts if i only draw a few mA, and between 2.06 and 3 amps (3.33*0.9 = 3 = average current) at a lower voltage.

The 18V rating is at 3.33A. At less current it will be higher voltage, how much higher depends on load regulation of transformer (sometimes specified, not usually). For a 3.3A transformer it has pretty thick wire so it should be pretty small change. If it was maybe 10%, then the voltage would be 20VAC no load.

i will be using the full bridge with capacitor circuit mentioned earlier. now i under stand that i am limited (at maximum) to 18*0.9=16.2 volts, and 3.33*0.62=2.06 amps.

Correct for 2.06A DC load current. DC voltage depends on a lot of factors (size of filter cap and type of regulator), I think it could do about 18V regulated DC output.

surely i can get 25 volts if i only draw a few mA

Maybe, it would be close to that maybe 27-28V minus diode drops so it should be over 25V.

 

...
I used to design transformers as well as select the ones we used, and every manufactured transformer I have ever seen is specified for ACV (RMS) when delivering the specified AC current. No manufacturer will stamp a rating on something which is contingent on the external circuit and would not apply in others like a FWCT application.
...

I understand that is the correct way to do the transformer spec; ie AC voltage RMS into a resistive load.

As an example of what i am talking about with transformers supplied for the general electronics (hobby) market, I just pulled a transformer at random from my junkbox and hooked it up over lunch.

Label spec = 240v 50Hz in, 15v "Max 2A" out (see photo)

I attached a 6A rect bridge, and a large can 4700uF 40V cap, using the highest voltage tap (15v) to get the max spec power from the transformer.

Cap voltage no load = 20.48v DC 0A

My dummy load was set to 2A (shown an an old avometer);

Cap voltage 15.96v DC supplying 2A DC to the dummy load.
Ripple p/p was 2.2v, diode conduction 40% (4mS cap charge, 6mS cap discharge as seen on the DC ripple on the cap terminals).

Those specs (to me) seem to indicate a transformer designed to supply the stated 15v DC at 2A DC. Maybe with your transformer background you can work out a more accurate value for it's real spec? I don't know how to do that but would assume since it is supplying more than the rated voltage at the max rated current it is within spec?

I heat tested it for 45 minutes;
20 mins; core = 44'C, windings = 49'C (ambient 31'C)
45 mins; core = 50'C, windings = 60'C (ambient 32'C)

It was tested just sitting on the wooden bench, no fan cooling, and it was a real hot afternoon here. Temp seemed fairly stable after 45 mins, and I needed to do some work on the bench!

Generally I would never run a part at its max spec, so I'm not advising this transformer should be run at 15.96v 2A continuously. But I don't think it's a big stretch to assume the manufacturer has specced this transformer to be able to run in "a 2A DC power supply".

If you want me to do more testing (like AC current or wavefors etc) I'll leave it under the bench for now.

 

May I recommend an ATX12V computer power supply? Already made for you, more power than you'll ever need, quality ones are well regulated and have low ripple on the positive rails. You get +12V, +5V, +3.3V, and -12V, capacity typically in the tens of amps each except -12V.

Only caveat is that efficiency is generally poor at low loads, but who cares in a bench supply? Also you should cap the -12V if you use it, it's a legacy rail so they generally let it hang at +/- 5% to 10% regulation and ~100mV to 150mV ripple.



Something like the Corsair CX430v2 would be ideal. Common, cheap, fairly efficient, voltage regulation around +/-1.5%, ripple about 20mV on the positive rails.

 

i've got an old computer power supply, can i just de-solder the input caps and inductors/chokes and use them as my emi filter? i assume the filter only removes undesirable frequencies, and is usable for buck converters, and not just flyback converters.

That should normally be ok. After all, no measurements are being done about emissions. A PC Power Supply itself, as Exodus suggested, is anytime more economical than building, why, even buying another power supply. But the fun of making it is lost.

 

Those specs (to me) seem to indicate a transformer designed to supply the stated 15v DC at 2A DC. Maybe with your transformer background you can work out a more accurate value for it's real spec?

The voltage specs printed on the transformer have to be AC values. I would assume the 2A is the AC RMS rating for the transformer secondary winding, which is the same regardless of which voltage tap is used.

I attached a 6A rect bridge, and a large can 4700uF 40V cap, using the highest voltage tap (15v) to get the max spec power from the transformer.

Cap voltage no load = 20.48v DC 0A

My dummy load was set to 2A

You can run transformers above their rating, not recommended for reliable design.

using the highest voltage tap (15v)

Cap voltage no load = 20.48v

Cap voltage 15.96v DC supplying 2A DC

That proves what I said. It's a 15VAC winding which gives about 15 x 1.4 = 21V no load (minus diode drops).

When running 2ADC (which would be about 3.2A RMS on the transformer secondary winding) the voltage drops to 16V, about a 20% drop from loading.

That's a pretty big drop, it looks like a 2A (AC RMS) transformer to me.

ADD:

By coincidence, my bench supply uses a transformer rated for 14 VAC @ 4A (RMS) and the current limiter is set at about 2.6A. The no load voltage for the unregulated DC supply is about 20V, at full current it only drops about 2V because it's not being run above rating.

 

...That proves what I said. It's a 15VAC winding which gives about 15 x 1.4 = 21V no load (minus diode drops).
...

Agreed, that 21v is very typical no-load voltage of a 15v AC secondary.

...
When running 2ADC (which would be about 3.2A RMS on the transformer secondary winding) the voltage drops to 16V, about a 20% drop from loading.
...

Isn't a 20% drop in voltage from the no load voltage quite normal for a transformer? If the no load voltage is RMS*1.414 then the full load voltage would be 15v right?

At 2A DC output the full load voltage is 16v showing it is not quite at "full" load (which should be 15v).

You said in post #18 that the secondary current rating must be at least 1.7 times the DC current output. By that rule, my running the transformer at 2A DC means I have been overloading the output current at 1.7 times its MAX current.

How could it be running at 1.7 times its max possible current and still be producing MORE than the stated output voltage? How could it be running at 1.7 times the max current and still run a fairly normal temperature rise over ambient?

If the actual max DC current from that transformer was 1.7 times less (ie 1.18A) then at a huge overload like 2A it would surely be way below the stated 15v output, AND be really cooking the transformer?

From what I understand of the proper AC spec, a 15v 2A transformer should produce about 15v AC when running about 2A AC RMS. In an overload situation of 1.7 times it rated current, the voltage should be well below 15v.

Or are you saying the transformer can deliver the stated AC 30W RMS as DC 30W RMS, but it is being "overloaded" by 1.7 times? If I overloaded the transformer with an AC resistive load to 1.7 times the rated current, the RMS voltage would be far below 15v.

Are you saying a transformer rated at 30W won't produce 30W in a DC PSU and can only produce 17.6W? Sorry, I just don't get it.

 

Isn't a 20% drop in voltage from the no load voltage quite normal for a transformer?

Depends on the transformer, mainly the wire gauge. A FWB+CAP design conducts current in narrow pulses, so at each one the peak voltage is reduced by approximately the secondary winding's DC resistance multiplied times the current in the winding while it conducts. A good 3A (or even 2A) should do better than 20%, at least the ones I buy because the wire should be fairly thick.

Very low current transformers are notorious for very poor load regulation, but I would expect a 2A transformer to do better if not being over loaded.

 

From what I understand of the proper AC spec, a 15v 2A transformer should produce about 15v AC when running about 2A AC RMS. In an overload situation of 1.7 times it rated current, the voltage should be well below 15v.

No, it's basically linear since the voltage drop is due to winding resistance. If you got a 10% drop in voltage at 2A RMS, you would get about 20% drop at 4A.

Are you saying a transformer rated at 30W won't produce 30W in a DC PSU and can only produce 17.6W? Sorry, I just don't get it.

Transformers are NEVER rated in Watts, they can't be. They are rated in VA, which is Volt-Amps. VA are the product of AC voltage times AC current, Watts are VA multiplied times the cosine of the phase angle (called the power factor) so the Watts it produces is always less than the transformer VA rating which is what is printed on the transformer. Depending on circuit type, it can be much less than the VA value.

I designed offline power supplies that ran off 115VAC using a diode bridge and cap filter. The power factor is 0.6, so the VA drawn from the wall was always about 1.7X the load power + converter losses. Typically, a 1kW switcher would be drawing almost 20A (continuous) at 115 VAC (RMS) in order to supply 1kW of load power because of the big power factor loss.

 

what is a transformers current limit based on?

is it due to heat disipation from the winding resistance? (in which case, the current limit would reflect the maximum average current).

or is it due to changes in the coils inductance at high currents? (which would mean it is the maximum peak current).

That should normally be ok. After all, no measurements are being done about emissions. A PC Power Supply itself, as Exodus suggested, is anytime more economical than building, why, even buying another power supply. But the fun of making it is lost.

i very nearly considered basing it on a PC power supply. i even have 3 laying around that i could use at a moments notice. the reason i didn't use them, is because i wanted voltages higher than 12v, and boost converters proved to have undesirable behavior.

the problem with boost converters, was if they were started with a high current load, then they would attempt the drive the voltage up (to the target from the input voltage) by increasing the duty cycle (voltage=1/(1-D)). but what happened, was that the max duty cycle (something like 0.9), meant that the inductor had to handle 10x the average current that the load was drawing if it was to begin increasing the voltage above the input (the inductors i had couldn't take the current). i could easily start the converter with no load, then connect the load, and it worked fine. but i didn't like that it couldn't start with a load connected.

in the end, i just decided to go with a transformer, and to do the rectifying and filtering myself. this allowed me to stick with buck converters, and avoid the boost converters entirely.

 

what is a transformers current limit based on?

is it due to heat disipation from the winding resistance? (in which case, the current limit would reflect the maximum average current).

or is it due to changes in the coils inductance at high currents? (which would mean it is the maximum peak current).

Can be all of the above. Wire heating, core heating, core saturation.
Bottom line: transformers are rated in VA which is ACV (RMS) and AC average current.

 

No, it's basically linear since the voltage drop is due to winding resistance. If you got a 10% drop in voltage at 2A RMS, you would get about 20% drop at 4A.
...

So how are the ratings of the transformer determined? If labeled for 15v 2A and both are AC, what AC voltage would the transformer be expected to make when supplying 2A AC?

I checked the DC ohms of my transformer secondary by applying a fixed 5A DC and 2.5A DC and measuring the voltage drop and in both cases the transformer secondary showed 0.25 ohms. 5A was not enough to show any significant rise in resistance from temperature rise.

Secondary heat losses from I2R in my 2A DC test was;
2A av, 5A@40% duty, I2R losses = 5*5*0.25ohms*40% = 2.5w

and at 2A AC RMS;
I@R losses = 2*2*0.25ohms = 1W

So running at 2A DC output the secondry winding was getting an extra 1.5W waste heat. The temperature rise was acceptable enough considering this is an absolute maximum rating so the 2A rated transformer was quite capable of supplying 2A DC.

...Bottom line: transformers are rated in VA which is ACV (RMS) and AC average current.

But have very likely been labeled with a good sized safety margin so a transformer rated at 2A AC absolute max will probably be usable in a DC supply supplying 2A absolute max.

This is getting a bit off-topic so I'm happy to take this to a "transformer max safe ratings" thread if you like, i'm interested in your input finding out some good derating figures as the earlier quoted "DC amps = AC amps /1.7" is obviously way too conservative, especially on lower voltage secondary transformers with reduced secondary 12R losses.

 

Air conditioning is my day job so I have access to dozens of transformers that never thought about what a diode is. Thus, that page from Hammond is good for my uses. It might not be right for your uses.