Quick query ref VA ratings

[bikerhifinut]

If I want to get a transformer to power up a set of 6.3V valve heaters and I know the total current draw is 4 amps, which I think is what 4 x EL34 will take, then a quick sum suggests that 6.3 x 4 = roughly 25VA needed?

So on the basis of under running is a good idea and that the next popular VA rating is 50VA I should be fine buying a 50VA toroid?

Can anybody see a flaw in my logic (or lack of?)

Andy.

['LIVEWIRE?']

Seems O.K. to me, Andy.

[G8UWM-MildMartin]

EL34 heater is 1.5A.

[kalee20]

At a superficial level, it's fine.

At a slightly deeper level, transformers aren't perfect. The output voltage sags slightly with increasing load. Output voltage is usually specified at the nominal load - with less load it'll be higher.

So a 6.3V 25VA transformer should put out 6.3V when loaded with 4A. But a 6.3V 50VA transformer may put out 6.6V when loaded with just 4A (it would drop to 6.3V at 8A).

Whether this matters, is up to you! Personally I tend to aim for running valve heaters at, or 1% below, their nominal voltage, at nominal mains.

[bikerhifinut]

Originally Posted by G8UWM-MildMartin

EL34 heater is 1.5A.

Of course it is! Sorry a bit of brain fade there as I dug out my original fag packet scribble and there it is........... 6 amps needed for 4 x EL34.

ooops.

A

[bikerhifinut]

Originally Posted by kalee20

At a superficial level, it's fine.

At a slightly deeper level, transformers aren't perfect. The output voltage sags slightly with increasing load. Output voltage is usually specified at the nominal load - with less load it'll be higher.

So a 6.3V 25VA transformer should put out 6.3V when loaded with 4A. But a 6.3V 50A transformer may put out 6.6V when loaded with just 4A (it would drop to 6.3V at 8A).

Whether this matters, is up to you! Personally I tend to aim for running valve heaters at, or 1% below, their nominal voltage, at nominal mains.

I get you kalee.

I was actually thinking in terms of using a 6V toroid as they are cheap and as our mains tends to run a bit on the high side of nominal 230V I will likely get 6.3V anyway. or close enough not to matter within the valves tolerances. I figured 8A for the choice if I wanted to use KT88/6550 at a later stage as they draw 1.9A heater current.

Thanks for the reassurance chaps.

[G8HQP Dave]

The bigger the transformer, the bigger the 'cold heater' switch-on surge. Therefore no advantage in making the transformer bigger than necessary. You need about 38VA for EL34 or 48VA for KT88 so go for 50VA.

[broadgage]

IMHO, it is preferable to oversize transformers in the interests of safety and reliability, particularly in the case of cheaply produced imports that may be rated in Chinese watts.

Don't overdo this though for the reasons already given of voltage rise on part load.

This voltage rise may be mitigated by selecting a 6 volt transformer rather than 6.3 volt, a 5% reduction, and by utilising the 250 volt tapping on the primary if available.

If in doubt, build the circuit and measure the heater voltage under typical working conditions.
If found to be too high then considering reducing this by inserting a small dropper resistance in either the heater circuit or in the mains supply to the transformer.
If a 6.3 volt 6 amp heater circuit is found to be say 7 volts, then adding a 0.1 ohm resistance in the heater circuit will bring it close to the nominal value.
The mains current into the heater transformer will be about 200ma and a dropper resistance to drop say 20 volts here could be considered.

[bikerhifinut]

Thanks.

The main thrust of my enquiry was to check my calculations on VA ratings and on that score I am content. I don't wish to have too large a transformer on this job as the all up weight will be a big factor and thus the maximum size of 50VA will be adhered to. That gives me a bit of future proofing should I redesign the output stages to use KT88/6550 types.
If the heater voltage does drop to 6V again I ain't too worried, but given the average 252V mains voltage measured here I figure I'm going to get 6.3V or thereabouts so no worries and seeing as the heaters are designed to work in the range of 5.7V to 6.9V, the odd 0.1 to 0.3V difference isnt going to be an issue.
Thanks for the info folks, it's going to make my life easier on the final design, I know Joe Bog is going to be smacking his head in dismay but my feeling is I am introducing an unnecessary complication into the circuit by rectifying 9V and then regulating it down with 5Amp regulators, which would need 1 reg for each channel/pair of valves plus the associated heat sinks. And it should be quite possible to get no heater induced hum using AC heaters on power output pairs. Heck I have a single ended EL84 amp here that uses AC heaters and zero hum, just by careful routing and twisting of the heater wiring. I'll use the 9V winding to supply 6.3V DC to the preamp and phase splitters as I have a pair of nice little PCB LM317 types (think I am usung LT1086 jobs) that will be repurposed for that. Proven technology in the previous mono versions.
Just need to find some help in cutting chassis top plate holes now but again I should manage that on my own with care and arthritis permitting!
A.
read about it in homebrew thread when i get going!

[G8HQP Dave]

I was thinking of the safety and reliability of the valve heater, rather than the transformer. Provided that the transformer can cope with the steady load ('Chinese watts' permitting) there is some advantage in heater lifetime if the transformer is not too overspecified.

Consider the difference between 5% regulation and 10% regulation i.e. the effective series resistance from the transformer is 5% or 10% of the hot heater resistance. Cold heater resistance might be only 20% of the hot resistance, so the total initial circuit resistance could be 25% or 30% of hot. This means 20% more initial current surge for the bigger transformer so 44% more power in any weak spot in the heater.

Voltage regulated DC heater supplies can be much worse in this respect. In some cases people find that the regulator goes into current foldback and sulks every time you switch on, unless provision is made for coping with cold heaters.

[Herald1360]

Just how susceptible are 6.3V heaters to switch on surges anyway? Worse than GLS lightbulbs? Weren't they originally intended to be slapped straight onto a fully charged 6V lead acid battery anyway?- not much surge limiting there!

I have never had a heater go on switch on, my FM only EKCO lights the UCC85 (I think that's what it is) very brightly upon turn on, it is still OK.

[G8HQP Dave]

For some reason the bright light on switch-on seems more common with some dual triodes. What happens is that the part of the heater which is not in the cathode tube heats up more quickly than the rest (because it doesn't have any cathode to take the heat away) so its resistance increases so it heats up more. Eventually the rest of the heater catches up and so reduces the current.

[kalee20]

Dave's got another good point - a transformer with a soggy output characteristic is likely to be kinder to the heaters than a heavily-rated, but under-run, big transformer.

At the end of it, what defines a transformer rating anyway? Your 25VA transformer may be perfectly happy supplying even 30VA, continuously, if it is bolted to a big piece of aluminium mounted in a position of good airflow. It's the temperature rise which sets the limit. If you are at all worried, measure the primary resistance cold, then switch it on with the intended load for 4 hours, switch off and immediately measure the resistance again. You can calculate the internal temperature rise from:

Rise (°C) = (Rhot/Rcold - 1)/0.00393

and if this is less than 50°C you should have no need to worry, even with a cheap Chinese transformer!

[G6Tanuki]

Originally Posted by Herald1360

Just how susceptible are 6.3V heaters to switch on surges anyway? Worse than GLS lightbulbs? Weren't they originally intended to be slapped straight onto a fully charged 6V lead acid battery anyway?- not much surge limiting there!

My experience is that 6.3V heaters are extremely robust!

In times-past we had TVs with series-connected valves and I can remember seeing a little 6AL5/EB91 double-diode light up as bright as an incandescent bulb within about a second of power being applied, then dim down to normal over the next ten seconds as the other, higher thermal-inertia heaters got up to temperature and the series-chain voltage distribution settled down.

I don't recall there being any significant mortality of EB91s a a consequence of this.

[ms660]

They did suffer from heater/cathode leakage.

Lawrence.

[turretslug]

I've wondered about the typical/classic combined HT/LT transformer's ability to supply surge LT current in the time before the HT winding is significantly loaded and whether this "spare" capability represents a significant hazard to valve heater durability compared to a closely-rated LT-only transformer. I've noticed flaring with some valves such as the EB91 and double-triodes but it's generally brief and not to the 3000K+ whiteness familiar with incandescent bulbs. One anecdotal case that sometimes crops up is open-circuit heater in the separately-supplied EB91 noise-limiter in numerous Eddystone receivers with the common 3937P transformer- this appears to supply nearer 7V than the nominal 6.3V anyway, aggravated by mains generally being higher than the 230V tapping provided. Wherever possible, I fit a 220 ohm ("ambient", 25 degree Centigrade) inrush limiter thermistor to the primary side of my thermionic stuff, this at least providing a few seconds of "slugging" as witnessed by dial lamps. This typically adds 5-7V running burden, which may help with "generous" mains.

Originally Posted by kalee20

At the end of it, what defines a transformer rating anyway? Your 25VA transformer may be perfectly happy supplying even 30VA, continuously, if it is bolted to a big piece of aluminium mounted in a position of good airflow. It's the temperature rise which sets the limit.

And it'll be a benign sinusoidal loading, too- I wonder if transformer makers allow a little headroom anyway for users not always taking jaggedy-waveform rectifier/large-electrolytic-capacitor type loading into consideration?

[bikerhifinut]

This is all very interesting stuff and enlightening.

However.

Are we perhaps overthinking this just a little?

I'm happy now that my "back of an envelope" calculation for a suitable transformer to heat up a set of valves was in the ball park and I can look for a suitable transformer either toroidal or E I type to suit the job planned.

But the discussion has been interesting.

Andy.

[G8UWM-MildMartin]

One obvious advantage of unpotted toroidal transformers is that it is very easy to overwind a couple of turns to adjust the output voltage.

[Al (astral highway)]

Hello, there's another current thread on V-A ratings,

Mods, I wonder would it be helpful to merge these?

The thread has some practical guidelines on the extra overhead needed to take into account the difference between watts and V-A, mixed in with some details and debate on the theoretical side. So something for everyone, whether we need a back-of-the envelope calculation or whether we need to consider what possible physical characteristics a transformer could have.