"Knal" bij schakelen standby > on

[HaroldA]

Mijn beide Sovteks knallen als ze voor het eerst van "standby" naar "on" gaan. De indicator light dimt ook helemaal, dus ik vermoed dat de voeding te weinig vermogen kan leveren om de 7 filtercaps te vullen (2x 220nF in serie, 2x (2x 47uF in serie) en 1 losse 47uF).

Zou ik dit op kunnen lossen door over de standby switch een weerstand te zetten zodat de caps altijd "half gevuld" zijn als de amp in standby staat?

Improved standby switching:
Most standby switches simply turn the HT on and off, allowing the amp to be muted while the valves are still idling hot. When the standby is switched to "on" all the power supply capacitor will instantly draw inrush current;- this 'hard switch on' and inevitably shortens the working life of the rectifier and capacitors. However, if we are willing to forgo the complete muting facility we can instead redesign the standby switch for a 'soft start' to truly help extend component life (especially the rectifier, whether valve or solid state). If we arrange the switch to allow the anode current to increase very gradually at the same time as the heaters warm up, then both inrush surge and cathode poisoning can be completely avoided. This can be done simply by placing a resistor in parallel with the switch. Obviously this works best if the switch is placed before the reservoir capacitor.

Finding a suitable resistor is relatively simple because the amount of current we allow to flow on standby isn't that important, and it is in fact easier to decide instead how much power we want it to dissipate. Because the amp might be left on standby for some time we don't want the resistor to get very hot, and equally, we would like to avoid using a large and expensive metal clad resistor.
A 2W resistor would be ideal. To avoid excessive heat we will allow it to dissipate no more than 1W. At the moment the mains switch is turned on, the resistor will have to drop the full HT voltage. If the standby switch is to be placed after the rectifier we can use the value of HT voltage. If the switch is placed before the rectifier then use the AC (rms) voltage of the transformer.
In this example the switch is to be placed after the rectifier and the HT is 300Vdc. For 1W dissipation:

P = V^2 / R
(Note, ^2 means "squared".)
R = (V^2)/P
R = (300^2)/1
= 90 000 ohms
The nearest standard is 100k, but since it only has to suffer the maximum power dissipation at the moment of switch on, half this value or 47k would be fine too.

The initial current will mainly be swallowed by the reservoir capacitor so most of the HT will be dropped across our standby resistor, and the voltage on the valves' anodes will be negligible for the first few seconds while the heaters begin to warm. As the reservoir and smoothing capacitors charge the anode voltages will rise gradually and a small amount of anode current will be allowed to flow as the heaters reach full temperature. Once the capacitors are fully charged and the valves warmed up the amp will remain on standby, allowing a trickle current to flow at all times (so the amp will not be totally muted; there may be the faintest of strangled sounds if we tried to play). The current flowing will be small (and very difficult to calculate!) but sufficed to say that it will probably be about half the initial surge, and the anode voltages will all idle at somewhere in the region of half their normal value.



This is an enormously simple modification to any amp, and any value over about 47k (2W) or 150k (1W) should do. It is worth pointing out that this mod was hastily added to later production runs of the Vox AC30 reissue, to fix the problem with rectifier flashover.

Ceriatone has a slight variation. They put a .047uF cap and 100K resistor across the standby switch.

Is die 47nF cap nog een interessante toevoeging?