Dealing with the no-load voltage of standby transformers

H9KPXG uses a single Triac to connect or disconnect the equipment to the AC mains. It switches the "Line" wire but the "Neutral" wire remains connected at all times. Then, one second later, it uses a (single pole) relay to bypass the Inrush Current Limiter disc, after the inrush has mostly completed.

If you wish to connect and disconnect the AC mains on both Line and Neutral, you'll need two Triacs. Or, you'll need to replace the Triac(s) with a second relay; the new one needs to be a double pole relay with high current contacts and a coil which is compatible with your standby power source.

If you choose the zero-Triacs, two-relays approach, the total coil power is probably going to exceed 2 watts, so you'll need a somewhat bigger transformer for standby.
Yes, the circuit requires 2 relays. However, as the relays are not switched simultaneously and the power consumption is low in steady state, I decided that about 2 watts would be sufficient. I will review my idea. Good point!
 
If you need 12V then you should use a transformer with 12V nominal output voltage. Not 9V.
If you need 2W, then your transformer should be about 2.8 VA minimum. But not 1.5 VA
Don't forget these little transformers have a high no-load voltage because the voltage drops significantly under load. So if you load over the limit, the voltage is below specification.
The no-load voltage is considerably higher. Use a capacitor suited for that. That is, 12 V x 1.5 = 18V peak. +100% for no-load and 35V is good, 50V is better.
Regulate the voltage down to 12V using a 7812.
The standby power is determined by the transformer construction. Not by the voltage on the capacitor. The capacitor is virtually without leak and only the stand-by circuit itself draws a little bit of power. The main consumption is by the transformer primary. That is the price you pay for a linear stand-by supply. You are right in preferring a linear supply instead of SMPS when part of an amplifier.
 
All the small transformers lower than 5VA I have have way higher voltages at low loads. For 12V DC relays I use 9V 2.3VA transformers and still have over 13V at light load. Unregulated that is.

It all depends on the load and the requirements of allowed/desired maximum ripple voltage. For a small circuit and 1 relay you will likely always have too much voltage with a 12V AC transformer. The simplicity of just a transformer, Graetz bridge and 1 oversized filter cap is nice. Relay circuits usually are not that picky but being close to the right voltage is better.
 
Last edited:
You should check the relay specs carefully. Many have a "must operate" volate of around 70% of rated. So for a 12V relay that could be as low as 8.4V Conversely most will tolerate 150% of rated or 18V for a 12V. The top end is affected by ambient temp though. Again check the datasheets on your relays. Relays as Jean-Paul says are not usually picky. Oh, one final note, they are mechanical and gravity can have an impact, so some will spec the position of the relay for the must operate voltage.
 
And a double-pole relay has a noticeably greater armature mass than a single-pole relay. So the electromagnet must be made stronger, and usually that means the required coil power is greater for double-pole relays than for single-pole.

Perhaps it might be theoretically possible to wind the electromagnet coil upon a different core material having greater magnetic-pizazz per turn, so the core power stays the same while the magnetic pulling force doubles; but that seems not to be offered as commercial products, in practice.
 
I don't think they'd change the magnet structure if the relay series offered DP and SP models. There is a great deal of variation in relay power for a given purpose as well as footprints. I'm amazed at what they can do with some of the recent models I've purchased. Super small and yet still 240VAC with an amp or 2 of contact capacity. And so as I said, check the specs carefully. I've got one relay for a pool pump(hi power DPDT) that is pretty power hungry normally and it only allows 110% over voltage.
 
Indeed. Many relays are double pole (probably most produced and sold too) and the single pole version being a derivative of the double pole version. So in fact they usually have the same coils. Really nothing to worry about. Even if it were one better switches off both L and N. Question of priorities then. I regularly see the standard double pole versions being used in single pole applications.

Anyway I would again recommend to only use fully sealed versions with bifurcated contacts for a long service life and to always use way higher current rated contacts when switching inductive loads.
 
Last edited:
These are typical standby transformers:

Screenshot 2024-11-29 062002.png


You can see by just how much the off load voltage rises. For really small ones it is 1.8X so 12 volts AC output becomes 21 volts. Design with that in mind and its not an issue and even a small loading will quickly pull that back down.

Screenshot 2024-11-29 062159.png
 
Yes please let us not make too much of an issue of the simple and circumventable issues that come with real switches, transformers, relays and other simple, reliable and matured passive items. About everything active that takes their place has besides key parameter lower cost more technical drawbacks amongst which shorter service life is one. Buy cheap, buy twice.

Besides that it maybe is a good idea to not choose the smallest of transformers. Recently had a bunch of those 0.35VA in automatic shutter controls that all had turned brown as they are too small to have adequate heat disposal. Strange thing is that the types till about 2.3VA cost practically the same?!
 
Last edited:
Efficiency is usually not the key parameter to worry about when it is about standby power.
Erm, its the main parameter. Standby is by definition on 24/7, 1 watt on standby is 31.5 megajoules a year (9 kWh). If you run a 100W input amplifier 2 hours a day and it has 3W of standby, the standby circuit consumes about 27% of the power - reducing the standby power has a noticable effect on the bills usually.

Also if the standby power has a poor power factor you get billed several times over for it too, so details matter.
 
I think it was someone at Intel years ago that coined the phrase, a dollar a watt, for 24x7x365 usage. And that was awhile ago. Probably around a buck fifty to 2 bucks now depending on location. I think some areas of CA/HI are around 40c/kwh so that is closer to 4 bucks/year/W for the idle juice.
 
Erm, its the main parameter.
It is indeed 🙂 I worked out how much energy our doorbell transformer has used since the house was built in 1953. The transformer is quite a big thing, these days probably around the size of a 30va block. Its also heavy, 500 gram {ish} I guess.

Lets say 70 years at 1 watt, That's 613,000 hours, 613kWh which at todays prices would be £150

(is 1 watt reasonable or no?)

A battery operated bell (say 4C or D cells) would probably last 10yrs a set... decent alkalines bought from Poundland... makes you think.

A bit like this:

Screenshot 2024-12-01 180031.png
 
  • Like
Reactions: Deenoo
For a 2W max. PSU 70 (transformer) or 90% efficiency (SMPS) won't break the bank as the usual power draw is way lower.

Mooly that is a LARGE doorbell transformer. Here they usually are 8V 0.5 or 1A.

Edit: yours is only 5VA. Batteries will not last 10 years and they will be worse to the environment. Also they are always depleted when YOUR package is about to arrive. A while ago I replace dry fridge for an A+++ one and then calculated the difference in energy over 10 years.....Anyway the new one is pure garbage despite a former quality brand name so its total lifetime power consumption will be even greener 😉.
 
Last edited:
  • Like
Reactions: Deenoo and Mooly
Also if the standby power has a poor power factor you get billed several times over for it too, so details matter.
Not true. Most countries in EU and AFAIK in USA as well bill by the use of U x I x cos phi. Bad power factor means low cos phi so this product becomes lower the worse power factor is.

A transformer gets warm during standby. This temperature increase is caused by real (as opposed to virtual) power usage. Indeed, the usually less quality the transformer is the higher the ohmic and iron losses. Quality both as in manufacturing quality as Q quality: ratio between L and R. A perfect transformer has no real current flow during no-load. All current is virtual (cos phi = 0), the transformer stays cold. And you don't pay for virtual current.

The higher the quality of a transformer, the worse the power factor is, the less you pay.
 
  • Like
Reactions: Mooly and jean-paul
Things may be different when a smart kWh meter is installed.....
It doesn't have to do with a meter being smart or not. It is dictated by regulations and tariff structure what you pay.

Indeed a classic electromechanical induction meter would not be able to register virtual or imaginary power at all. A smart meter is able to measure and register it. But it doesn't mean say that you are billed by U x I. I have a smart meter and I am still billed by the kWh, not by the KVAh.

Elaborating on that power company would like to bill you for the use of imaginary power because imaginary power or imaginary current (with cos phi = 0 or sin phi = 1) does cause losses in the distribution system. Because most of the transmission lines are ohmic, and in ohmic lines voltage and current are in phase and cause real loss.

In addition, the installed capacity of power generation and distribution is counted in kVA and not in kWh. Which means the installations are larger than the amount of kW being generated and distributed.

The first, additional losses, is much less than the cost of the generated power. I mean if you use 1 kVAh of fully virtual power (cos phi = 0) the losses in the distribution system are by far not equal to the energy cost 1 kWh or real power (at cos phi = 1).

The second, installed power is a capital expense for the power company. These capex is already included in the kWh price you pay. Power systems are laid out for a given cos phi, and the kWh price is calculated based on the capex. Whether or not you use it.

AFAIK only large consumers like companies have to pay additionally for a bad cos phi. It would be unfair if kVAh price would be the same as kWh prices. As for virtual power, only losses have to be paid, not generation costs.
 
Last edited:
As jlinkels says, today residential pays for kwh not kvah. I would not be surprised if it changed. As an example, I have panels, and when I first installed them the utility used net metering. Around 10 years ago they switched to they pay me X for all power generated by my panels and I pay them Y for all power I consume (even from my own panels). X < Y, Net result is I pay for my own generated power. And I reviewed the docs that were first done when I installed the panels. Unsurprisingly there was no guarantee they would do net metering.
 
The whole misconception of solar panels is that one produces "clean" energy and puts that in the grid only to use that energy in the evening. It does not work that way. You don't pay for the energy that your panels created and use yourself, you compensate your energy use by generating a part of it yourself (at your own cost to buy/install the stuff). Here grid connections of new large solar installations are now even refused as there is too much energy fed back at times that no one needs it.

Whole point is that solar energy (just like wind energy) was pushed too fast as a solution to problems with help from governments and a bunch of non technical persons. It is just business and not in the interest of energy producers. The drawbacks were noticed late despite technical people objecting. Particularly safety and metering were just put aside when the hype started. I recall having to create parking lots for E-cars and when I mentioned to install kWh metering per charging station everyone looked at me if I was crazy. Of course having the parking lots painted green with nice lighting was more important 🙂
 
Last edited:
  • Like
Reactions: gerrittube
I experimented with some philips/nexperia flyback controller chips for usb powering. The standby power could be as low as 50mW, if a large output cap was used. The electrolytics were the only life limitng factors. If one day we could use ceramics, then the e-waste could be reduced, probably at the cost of a larger standby power. film caps also have a limited life, sometimes called self healing, or the better word should be self destructing.