Voltage effects on DC offset and bias

[albertNL]

The reason I started this post was that I wondered what happens when supply AC changes.
If DC offset and bias depend on DC voltage (given the type of power supplies in use) they depend on the AC supply.

As AC supply can change by quite a bit (+/- 10-15% or so?), this would mean that if you set DC offset and bias when AC supply is at (or close to) either extreme (i.e. lowest or highest you ever actually get from the AC company), you could end up with trouble when AC supply swings towards the other extreme.

I found no reference to check AC voltage when setting DC offset and bias. I thus concluded* it could not be such a big deal (even with a potential 20-30% AC voltage difference).

So, while the builds show it should not matter, theory says it might.
Perhaps a more experienced builder or engineer could comment.

Thanks,

Albert


* Given the number of DIY amps built, it is likely that some will have set their DC offset and bias at the extreme supply AC voltages.

In case you were wondering, yes, I have worked with digital systems for the last 30 years.

[Djenka018]

AFAIK, European countries had +10/-15% specs for supply and any design should account for such span.

In theory, Watt's and Ohm's Law are applicable as a part of a complex multi dimensional equasion for bias and DC offset change with supply rail change.

Multidimensional nonlinearity of semiconductors command change of parameters with change of supply volts but circuit design may muffle or amplify these effects.

Uneven bias change on each half of the PP amp (due to different supply volts change for each half) would create DC offset.

[PRR]

This gets back to whether the designer considered voltage variation.

IMHO, voltage variation MUST be considered.

Old-time designers usually left allowance. Back in days of tubes the "Design Center" voltage spec might be 300V, but the "Design Max" spec might be 330V.... the Center spec covered reasonable wall-voltage variation and was used in routine designs, the Max spec was for designers who carefully considered (or measured, or regulated) supply voltage.

And design methods were mostly approximate, so it was un-wise to get too close to any limit.

OTOH, some designers trust computers and SPICE un-reasonably. I have seen posted circuits which would quit or burn if supply varied even a few percent.

> setting DC offset and bias

If the offset trim method is well choosen, 10% variations of supply have little effect. OTOH sometimes you see a transistor offset "balanced" by a fraction of the supply voltage. These will go out of balance for very small supply change.

> AC supply can change by quite a bit (+/- 10-15% or so?)

The real limit on what comes out of the walls, in most areas, is Incandescent Lamps. The life of a white-hot filament varies as the 13th power of voltage. A 10% rise cuts life to 1/3rd. While people do not keep track of lamp life, which is always random, they do notice if a lot of lamps die very quickly. I have complained to my utility company about steady 127V. While there has been a century-long trend from 100V to 110V to 120V, the voltage is pretty steady year to year in most utility systems, and usually much closer than 10% of nominal.

In DIY, the instructions to set bias and offset without regard to line voltage assume that you are adjusting with nearly the same power you will use it at. Maybe 119V in your workroom at midnight and 118V in the music room in the evening; no big difference. When a manufacturer is setting up amps for wide distribution, they would think about worst-case wall voltages and adjust appropriately.

[AndrewT]

Maybe I am an old-timer, but I always look at survivability during worst case operational conditions.

In the UK I assume that sometime during the life of my equipment it will have a 254Vac mains voltage as the supply.
It must be designed to survive this without damage. It would be nice if the equipment could operate normally over the full range of 216Vac to 254Vac of mains supply.

[brianco]

Andrew,
Down here near Kelso ours runs at a constant max of 248V.....it sometimes goes as low as 242V. I have measured 250V a couple of times.

[albertNL]

OK, let's see if I understand this correctly:

Problem 1.

Assumptions:

• the F5 it is a well designed amp
• my PSU has 63V rated caps
• I have very, very large heatsinks (no heat issues)
• PSU uses 2 x 18V secondary transformer
• voltage on both secondaries identical to within 0,1%
• I have a spare 2 x 24V secondary transformer
• voltage on both secondaries also identical to within 0,1%
• JFETs can handle 34V

If the above assumptions are true then I should be able to drop in the 2 x 24V secondaries transformer without any problems.

DC offset should be OK because voltages in secondaries are identical to within 0.1%.
Bias should be higher that before but still OK (assuming MOSFETs can dissipate the heat).


Problem 2:

Assumptions:

• AC supply at my house is 197V
• AC supply at my friend's house is 250V
• heatsinking is adequate to allow for the extra heat generated (ca. 60%!)

If DC offset and bias are set at my house, the amp will work correctly at my friend's house.


Comments?