Reasonable derating for a power transformer ?

[00940]

I'm about to order a toroidal transformer for a small project and I wondered how much of a derating is reasonable to keep heat in check.

I usually overspec the xformers I order by aiming for a continuous draw at around 60% of their rating.

In this case, the amp will consume 2.8A @6.3V for the heaters (on AC, so a resistive load) and about 80mA DC (from a 140VAC secondary) at iddle for the B+. For a bridge rectifier with cap input, on average, the ac current is 1.8* the DC current drawn. So, at the very least, I'd need a 40VA transformer (2.8*6.3 + 0.08*1.8*140=37.8).

As a reflex, I'd go for a 63VA xformer, at something like 6.3V/4A, 140V/270mA. But would it be really unreasonable to go for a 50VA (6.3V/3.5A, 140V/200mA)?

 

[rayma]

I wondered how much of a derating is reasonable to keep heat in check. I usually overspec the xformers I order
by aiming for a continuous draw at around 60% of their rating.

For heaters, I'd worry more about the actual delivered voltage.

 

[00940]

That's what resistors in series are for. 😉 Even though these days, I usually ask for a highish voltage secondary and use a small switching reg.

 

[Tom Bavis]

At 60% of VA rating, total heat from I^2*R will be 36% of that at rated power. If resistive loss were the only factor, 40VA in a 40VA transformer might reach 90C, 50 VA transformer about 65C, 63VA transformer about 50C. Though temperature rise won't improve QUITE that much by derating, since core loss will go up with a larger core.

 

[marcelop]

Be careful not to excessively overspec the transformer, or otherwise the output voltages will increase (assuming the rated voltage of the windings is under full load) and heater damage could result. If you do so be sure to include series dropping resistors on the secondaries to ensure appropiate operating voltage.

Otherwise as tom said, the only major concern is operating temperature, but bear in mind that those who designed the transformer did so it would perform well under full load and even took into account a safety margin just in case.

 

[AndrewT]

AC power into a resistive load is used to determine the rating of a transformer.

The continuous DC rating after feeding through a capacitor input filter is VERY different.

Expect the continuous maximum DC current rating to be roughly half the AC current rating.
This will make the transformer run hot, probably right up at the maximum temperature that the designer used to select the enamel temperature rating. Most transformers are rated by maximum operating temperature.

If you want the transformer to have a reliable long life then I usually suggest that the continuous DC current drawn from the transformer be ~ 25% of the AC current rating.
Using this and working back to the primary, you would use 4times the continuous DC current to arrive at the required AC current rating.
Then that determines the primary AC rating.

Using the post1 values of 2.8Aac @ 6.3Vac and 80mAdc @ 140Vdc one would end up with 2.8*6.3 + 4*0.08*140 = 62VA
and the current rating of the two winding would be 2.8Aac and 320mAac
This would run the heater winding @ 100% of rating, but that seems to be the norm for the valve/tube Builders.
I might choose a slightly higher rating for the heater winding and up the transformer to around 80VA

 

[00940]

If you want the transformer to have a reliable long life then I usually suggest that the continuous DC current drawn from the transformer be ~ 25% of the AC current rating.
Using this and working back to the primary, you would use 4times the continuous DC current to arrive at the required AC current rating.

You're even more conservative than I... A 3times ratio usually gets me barely warm toroidals. The 3 times ratio comes from the 1.8 ratio in between IAC and IDC for a cap input bridge rectifier (given by manufacturers such as plitron), combined with the 60% load aim. Nelson Pass suggest using twice the VA for a given continuous load on the DC side, which is about a 2.8 ratio IAC/IDC.

Be careful not to excessively overspec the transformer, or otherwise the output voltages will increase

Yep, fully agreed. Voltage regulation is about 10% at those sizes.

Though temperature rise won't improve QUITE that much by derating, since core loss will go up with a larger core.

That's the kind of info I was hoping for, thanks, gonna check that.


Just for info, I attach the specs of the xformer I'm looking at. They're class B rated btw (max 130°C).