Mains transformer design for different mains frequencies

[saddlestone-man]

Hello All

I was looking at 1920s-30s adverts for a 6volt Meccano-branded (initially, at least, bought in from Ferranti) mains transformer where the buyer had to specify the mains voltage and frequency in his area. At the time AC mains was supplied at 100V-120V or 200V-250V, and frequencies of 25Hz (for example in Cardiff) to 100Hz (Bournemouth, Dover and other places). *

It's obvious that the mains voltage needs to be specified for the correct transformer to be supplied, but how would the design have changed over this frequency range? Interestingly, after a while Meccano specified that they could only supply transformers for 50Hz and above.

best regards ... Stef

* Taken from the 1933 edition of The Broadcaster Radio and Gramophone Trade Annual.

[barrymagrec]

You need a lot more Iron at lower frequencies. This would make a 25Hz transformer inefficiently heavy and expensive for 50Hz.

It is also why valve output transformers become large and expensive if you want a good low frequency response, particularly at higher power.

[broadgage]

A transformer designed for 25 cycles will be fine on 50 cycles, but the 25 cycle transformer will be larger and more expensive, so it is only worth producing the 25 cycle design if a significant number of customers have a 25 cycle supply.

In general, a transformer can be used on a higher frequency than it was designed for but not on a lower frequency.

The main exception is that a transformer intended for say 50 cycles may be used on 25 cycles if the input voltage is reduced in proportion.
So a transformer intended for 50 cycle 240 volt mains could be used on a 25 cycle system if the input voltage was reduced to 120 volts.

Lower frequencies were favoured in the past as large electric motors were then better run on low frequencies. The fact that the motor was larger and heavier was a price worth paying.

50 cycles is about the minimum for domestic electric lighting. Factories with a lower frequency supply tended to either use gas lights or to have a DC "lighting" supply as well as the AC "power" supply.

Lower than 50 cycle systems are now almost extinct.
In Belgium, some railways use 16.3 cycles.
AFAIK, London underground still have some 33.6 cycles equipment for signaling.
Parts of Australia used 40 cycles into the 1960s.
There might still be a few legacy 25 cycles supplies in the USA.

The pre-war Weir report in the UK declared that all new public supplies in the UK were to be 50 cycles AC at 240 or 240/415 volts. Additions and extensions to non standard supplies were allowed, but no new schemes.

[Trifocaltrev]

The Electricity Act of 1926 only specified that the frequency should be 50 cycles, there was no specification of voltage.

Regards
Trevor.


https://en.wikipedia.org/wiki/Electr...pply)_Act_1926

[kalee20]

The lower the frequency, the bigger the transformer, so the higher the price. While a transformer designed for 25Hz will be fine at 100Hz, it's extra weight and cost for no benefit.

The mains voltage won't really make any difference here, all that changes is the number of primary turns and the wire gauge. It might make a small difference in labour costs, though I'd expect the difference to be small enough for the manufacturer to ignore it.

[saddlestone-man]

Many thanks for all the answers, that helps me greatly!

best regards ... Stef

[Orakle42]

A photo is attached of an ADM.PATT transformer with a 50 to 2000 Hz range dated late 40ies. At 12lbs its quite heavy, and it still works perfectly!

[emeritus]

The transformer for my Hornby Dublo train set (LMS, so pre-1953 manufacture), seemingly made to the original pre-war design with double-pole secondary fuses, is for 200-240V, 40 - 100 cycles. I understand that, until the power station was destoyed in the 1952 flooding, Lynmouth was the last place in the UK with a public 100 Hz supply, but that was at 100V.

Incidentally, its original rubber mains flex is still in excellent condition, but it's years since it was last used.

[M0FYA Andy]

"One thing not yet mentioned is that, for all other factors remaining equal, the turns-per-volt of the windings is inversely proportional to the frequency, so for example a transformer requiring 30 tpv at 50Hz only requires 1 tpv at 1500Hz."

Am I missing something? This isn't intuitive.

We'd want the inductance to increase inversely proportional to the frequency, in order to maintain the same magnetising current for the same power transformer. Now, I thought inductance would be proportional to turns squared. So a doubling of frequency (at the same voltage) would allow a root-two reduction in turns.

David

[Ed_Dinning]

Hi Folks, note that admiralty pattern transformers are normally wound with high insulation levels between windings and core.
In the example above it is designed for specific volts / frequency requirements. Quite a cleaver piece of design so that one transformer could be used on several different systems in the same piece of kit by simply changing the input tapping.

Not as wasteful as might be imagined as most tappings have a similar volt/ freq product

Ed

[kalee20]

Originally Posted by M0FYA Andy

One thing not yet mentioned is that, for all other factors remaining equal, the turns-per-volt of the windings is inversely proportional to the frequency, so for example a transformer requiring 30 tpv at 50Hz only requires 1 tpv at 1500Hz.

Not mentioned explicitly, though it is implied by several posts... Higher frequencies not requiring such a large transformer. And this is why.

Actually, going from 50Hz to 1500Hz, it would in practice be pushing it to reduce tpv by 30... I'd want to reduce flux density by quite a bit, to avoid excessive core losses. So reducing by 12 would be a bit more like it.

Originally Posted by emeritus

I understand that, until the power station was destoyed in the 1952 flooding, Lynmouth was the last place in the UK with a public 100 Hz supply, but that was at 100V.

I don't know about the last place in the UK, but yes, 100V 100Hz was correct (see where I live!)

[Hartley118]

Originally Posted by emeritus

I understand that, until the power station was destoyed in the 1952 flooding, Lynmouth was the last place in the UK with a public 100 Hz supply, but that was at 100V.

The working day must have passed quickly with the clocks running at double speed!

Martin

[Ed_Dinning]

Strange that in the NE of England, we were on 40Hz and pioneered network interconnections (first 66KV transmission line), but had to wait about 4 years to change frequency so we could join the National Grid

Ed

[saddlestone-man]

I recall reading that in the early days of mains supplies, Electric Companies charged a different rate for electricity used for lighting (which was actually also used for powering various appliances) and for power via the wall socket(s). I think Chas Miller mentioned this in Radiophile at some point.

Does anyone know if this was true, and if so, which 'form' of electricity was more expensive?

best regards ... Stef

[retailer]

Here in Australia I recall my parents house had 2 meters on the main fuse board - lights and power, my father mentioned that lights were charged at a lower rate. If extra power points were needed in a house it was common practice to wire them from the light sockets - my father was very fond of doing this using figure 8 lamp cord to do the wiring, it's still like that today.

[kalee20]

Well, it costs the same to move electrons through lamp filaments as it does through heaters, motors etc of equivalent power. So no difference to the electricity company. But, I suppose the power could have been taxed differently, with lighting at a lower rate than luxury uses, which would explain it.

[emeritus]

Somewhere in the back of the shed I have a pre-war electrical textbook that briefly discusses several different dual charging rates that were then being used by electricity companies in the UK for lighting and power.

[Richardgr]

Maybe it depends on the supply to the house? Back home in Sweden I pay for a 20A 3 phase supply, but I could specify other current ratings, and the fee goes up because peak load is potentially impacted. So back in time there could have a higher tariff for the power sockets because of the peak load impact, whereas the lamp circuit would fit comfortably in base load?

[Radio Wrangler]

Originally Posted by saddlestone-man

I recall reading that in the early days of mains supplies, Electric Companies charged a different rate for electricity used for lighting (which was actually also used for powering various appliances) and for power via the wall socket(s). I think Chas Miller mentioned this in Radiophile at some point.

Some friends had a guest house built in 1980 - a sort of 11-bedroom hotel, doing B&B and evening meals.

Because the kitchen was classed as commercial, the electricity for it was metered separately at a lower rate to compete with gas. Amazing what a bit of competition can do eh?

The nearest gas network was miles away. This place was remote enough to be away from mains water and sewers. But the kitchen got cheap electricity until quite recently.

The private water supply pump had its own meter as did the stabling, so they got murdered with standing charges. Putting some armoured cables underground and combining three into one was a big saving.

David

[Vintage Engr]

Originally Posted by Hartley118

The working day must have passed quickly with the clocks running at double speed!

Martin

That reminds me of a prank I played when I was working for Rank, during the miner's strike..
The power was off for quite long periods, & we had electric wall clocks.
We'd built our own inverters etc to run some lights, which gave me an idea.
Using a couple of battery-powered 100v line amplifiers, I used an audio generator to provide 100Hz, & sneaked a cable to the clock.

My boss was very confused with the apparent time, but quickly caught on when he realised that the second-hand was going rather fast!

David.