Ground Isolation for Daisy-Chains

[PRR]

voltage on an unloaded transformer / rectifier wall wart will be SQRT(2) times the loaded output

Not universally so.

Yes, there is always a sag. and 41% regulation is not too unusual for very small transformers, which favor low cost over efficiency. Also a lot of vacuum tube rectifiers seem to be "designed" to this neighborhood. But they don't always sag to the RMS of the AC, which is what SQRT(2) implies.

[peterc]

I saw an internal shot of Gig Rigs Virtual battery
Thegigrig.com/power and it uses the same chip as in Jack Ormans schem on p2
Could that be a solution?

http://www.muzique.com/lab/9v_iso.htm

[Rob Strand]

Not universally so.

Yes, there is always a sag. and 41% regulation is not too unusual for very small transformers, which favor low cost over efficiency. Also a lot of vacuum tube rectifiers seem to be "designed" to this neighborhood. But they don't always sag to the RMS of the AC, which is what SQRT(2) implies.

Yes we often see something like 40% drop under load for smaller wall-warts but the cause is regulation not the sqrt(2) increase output from AC rms to AC peak due to the rectifier + cap filter (the DC roughly follows the AC peak).

For a wall-wart the *average* DC output is roughly equally to the label when the full *DC load* is applied.   However the output voltage can often be 10% lower than the rated DC voltage at full DC load.   There is a standard for DC power supplies which gives an allowed voltage tolerance of +/- 10% on the rated/label voltage at full load.   The standard also caps the maximum under no load which prevents DC wall-warts producing any old voltage with no load.

If you make your own DC supply with a small transformer with very poor regulation then you can see substantially higher no-load to fully load voltage ratios.

A weird effect you can see is the difference between no-load and very light loads.   With no load the voltage can rise up because the diode drops decrease with low current.    It can be anything from 0.2V to 2V.

I have some data on 17 wall-warts and the ratio of no-load voltage to nominal is 1.47 and the ratio of light load to nominal is 1.35.    I don't have the loaded voltage on those but another set of data showed at full load the voltage was more like 90% of nominal:
https://www.diystompboxes.com/smfforum/index.php?topic=121357.msg1140395#msg1140395
(Clearly some wall-wart do not comply with the standard.   They could be old units before the standard evolved.  I know some of the wall-warts in that link were very old.)

[R.G.]

There's not much of a standard here where I am. The maker tells you their transformer provides X voltage at Y current. What that doesn't tell you is what the transformer provides at lighter currents. Some few makers tell you the "regulation" of the transformer; that being the amount that the voltage on the secondary drops from no load to full load, and they only tell you the full-load voltage.
Ideally, the maker would tell you the open-circuit voltage  of the transformer and the full load voltage and current. Those numbers tell you how to calculate the losses.
In general, for AC mains transformers with no active circuitry, the "regulation", the loss from no load to full load, increases the smaller the trannie gets. I've seen them as bad as 50% - losing half the no-load voltage at full load.
We also have new-ish laws effectively outlawing mains-frequency wall warts in the name of saving the planet. There are energy efficiency laws which effectively mean it's illegal to sell new trannies that are not switching regulators, and not all of those. So mains-frequency step-down wall warts are a dying breed here.

[Rob Strand]

There's not much of a standard here where I am. The maker tells you their transformer provides X voltage at Y current. What that doesn't tell you is what the transformer provides at lighter currents. Some few makers tell you the "regulation" of the transformer; that being the amount that the voltage on the secondary drops from no load to full load, and they only tell you the full-load voltage.
Ideally, the maker would tell you the open-circuit voltage  of the transformer and the full load voltage and current. Those numbers tell you how to calculate the losses.

We might have a standard (essentially following the EU standard) but the reality is the situation isn't much different to yours.   The standard applies to transformers and DC supplies (wall-wart or some sort of enclosure).

Our labeling is the same, only nominal voltage and max current.   For wall-warts: the actual voltage at rate current (at nominal mains voltage and frequency) is +/- 10% on that.  For transformers +/-5%.

No other info. Transformers, may or may not give regulation.

All the info I have is from my own measurements.   I've even gathered up average winding temperatures at idle for wall-warts and some transformers.

A real pain here is the *nominal* mains voltage was 240V but they changed the voltage to 230V with skewed tolerances.   On top of that the actual voltage either remained at 240V or in many cases dropped to 235V or so.  I believe the same thing happened in the UK.  The equipment on the other hand was rated at *nominal* 240V and is now rated at nominal 230V.  So if you plug in old stuff and the mains is low the voltages are lower and if you plug in new stuff and the mains is high the voltages are high.

In general, for AC mains transformers with no active circuitry, the "regulation", the loss from no load to full load, increases the smaller the trannie gets. I've seen them as bad as 50% - losing half the no-load voltage at full load.

As I recall the no load voltages in the standard are very loose.   Possibly even as loose as your measurements but it has a number of VA categories which tighten the upper limit as the the VA increases.  (Also different specs for standard and short circuit proof transformers - which is why I can't remember the numbers.)

We also have new-ish laws effectively outlawing mains-frequency wall warts in the name of saving the planet. There are energy efficiency laws which effectively mean it's illegal to sell new trannies that are not switching regulators, and not all of those. So mains-frequency step-down wall warts are a dying breed here.

Same here.   Wall-warts with transformers getting harder and harder to come by.

The good thing about the switchers is they are regulated. Compared to the old transformers you can pretty much ignore the whole issue of regulation.

[PRR]

In general, for AC mains transformers with no active circuitry, the "regulation", the loss from no load to full load, increases the smaller the trannie gets.

This is inherent in transformer geometry, thermals, and marketing.

The winding parameters are more favorable as the VA gets bigger. Small transformers "must" be lossy.

Big transformers "must" be efficient because surface/volume ratio works against them. Above 100KVA we find radiators or enormous tanks.

Organizations who buy and feed many-many-KVA transformers watch their huge electric bills and shop/negotiate for low losses. You and me buying dozen-VA lumps have no leverage and must take what they offer.

[R.G.]

Yep. Small amounts of iron need many turns to get primary inductance up enough to prevent large magnetizing currents. Many turns in a small core means very thin wire, which means high resistance. High resistance primary means high voltage drops from both magnetizing current and transformed secondary current. Many turns of primary means correspondingly many turns of secondary, also in a small space, and that leads to high secondary wire resistance. The wire resistances are effectively in series with the transformer.
Getting X volts out at Y current means that the winding ratio has to be set up to be higher at no load so that Y current times the wire resistance sags to X volts at full load, so the designer makes the turns ratio X volts plus Y current times the wire resistance. Putting ever more turns on the secondary to keep the secondary voltage up can lead to rapidly diminishing results as more turns requires smaller and smaller secondary wire. Eventually, it becomes impractical.
I ran into this in spades trying to use two small trannies hooked secondary to secondary to run a pedal tube amp. I tried two 120V:12V transformers hooked 12V to 12V. The 12V fed heaters, the isolated 120Vac backwards primary made B+. Sagged like crazy, as the "regulation" losses were twice as bad to the B+ output.