DCAudioDIY.com

A number of posts have mentioned the problems with high inrush currents affecting the line voltage and causing stress on components when initiating a cold start. Here is a circuit I designed about 20 years ago that addresses that problem. I use it in my main system that has 4.6KW of power amplifiers plus all my other equipment. It all runs on a shared 15 amp circuit in my home and I have never tripped a breaker nor had a component fail from inrush currents.

The output is divided into two sections. One is soft start and the other is an instant on for devices that are sensitive to the rate of rise of the power supply voltages. I had some components where the LCD display would not come on with a soft start. In addition, it looks at the line voltage and if it is above or below set limits the supply turns off. It is a latching circuit so if it detects a line anomaly it will stay off until the start button is pushed.

The soft start sections run through a phase control circuit much like an automatic wall dimmer. When the phase gets to full, the triac that drives the phase control is bypassed by a large power relay and the phase control circuit is disabled to minimize any switching noise. There are LEDs that indicate the various states including power on, soft start, and run.

All the voltage sensing and timing for the phase control are done using LM339 quad comparitors. The triac is a 40 amp device mounted on a relatively small heat sink because it is used for only a few seconds. To provide the highest di/dt protection and to assure no dropout of the triac with inductive loads, the triac is driven by a synchronized DC pulse with a high initial current charge to the gate by a capacitor that bypasses the gate current limiting resistor.

All of these devices or more modern equivalents should be readily available. Total cost should be relatively low. The chassis and outlets should be some of the highest costs. To minimize chassis work in mounting a lot of outlets, I simply mounted 4 multi outlet power strips on the chassis and wired them to the internal circuitry. I'm including images of the schematic, the power outlets on the chassis and the control panel.

_________________
Walt

I've had issues turning on power amps on some circuits in the past. I ended up using X10 modules. One dimmer to get the capacitors charged up, then an appliance module to supply full power. I use Girder, a PC based home automation program, to sequence the dimmer on, appliance module on, dimmer off, to bring the amps up. Worked great. The appliance module is a 20 Amp mechanically held relay so there's nothing in the circuit that wouldn't normally be there.

I'm about done converting all my amps to switching supplies with soft start built in so it's no longer an issue.

_________________
I have too much stuff - https://www.pleasebuymystuff.com

Any mains supplied electronic equipment should have sufficient internal storage to handle the peaks in the music regardless of the cord or line that is feeding it. The power supply has a time constant where it can continue to supply power even after the mains connection is removed. If you are going to use the equipment for continuous sine wave testing, yes, then the demands on the mains supply are great and all the connections are important.

With music, the crest factors greatly mitigate the average power required from the mains. You have high current demands when the kick drum slams but after that the demand decreases allowing the mains to recharge the supply. In that case, the internal supply provides the necessary peak power and the capacitors recharge between the peaks at a lower average rate. In fact, if you have a very large storage capacity in the power supply, any impedance in the mains helps to decrease the magnitude of the peak current surges in the mains as the supply caps are recharged and the mains current rate gets closer to the average rather than the peaks.

The downside of a well designed high storage capacity supply is that you have a high initial charge time constant. I used to work on a high power commercial electronic flash unit that had an energy storage capacity of 40,000 joules. It used a phase control circuit to prevent damaging charge currents when switched on. Once charged the current dropped to a very low value and only the leakage currents of the oil filled capacitors plus any control circuitry were reflected to the mains. When the flashtubes were fired, they would essentially short out the capacitors for period of a few milliseconds and then the phase control circuit would again mitigate charge current until the bank was fully charged.

An amplifier will hopefully never short out the power supply. If it does, you have other problems. With an amp, if the supply is large enough, the time constant of the load during the peaks is sufficiently long enough that the supply will not sag very much. Think of it as an RC (or RL if a choke supply) time constant where the R is the speaker and the C is the total capacitance of the supply. To prove a point, lets assume the supply was really mongo and had 100 farads of capacitance. The time constant for an 8 ohm speaker load would be 800 seconds. It could supply the current to the kick drum with virtually no drop. The time to totally recharge the caps would be longer than if you had a very small storage capacity but the good thing would be that the supply voltage would not change much or very rapidly.

So it really boils down to having a good supply in the equipment. The higher the capacity in the supply the more the equipment acts like it is battery operated and only relies on the line for recharging. If you had a battery operated preamp for example that took 24 hours to recharge you could probably recharge it from a small wall wart power supply. The capability of the mains as far as the ultimate operation of the preamp would be totally immaterial as long as it could supply sufficient power to recharge the batteries during the charge cycle.

So I have always concentrated on the power supplies in the equipment I have either built of bought. If I purchased a unit with a weak supply I fixed the supply rather than trying to fix the mains. Regardless of what you do with the mains you can only go as far as the utility supply to your meter. All the many things that happen to mains power are in the domain of the utility. A well designed power supply will thumb its nose at any mains problem because the time constants are so long in relationship to the waveforms you will encounter in music. Whew, more than enough said. How do you think I really feel about this issue?

Both really good ideas for two different situations Dave.

_________________
Walt

Walt. This is not word salad and it really does work that way. First the analogy to your oven although I understand it was a joke has nothing to do with music and more with continuous sine wave testing which I already mentioned. As far as audio power supplies, there is no such thing as too much capacitance. If putting large values of capacitance on the output of a 3 terminal regulator seems to slow the rise in voltage to the audio stage it is because of the characteristics of the capacitors not their capacitance value. Large electrolytics usually have a large value of series inductance which will limit the rate at which current can be delivered to the load. Properly bypassing the big cans, properly meaning a series of stepped values and types, not just one small 1uf film cap, will eliminate the HF shortcomings while allowing for the long time constant that will give isolation from the mains and good results. Also using a lot of smaller electrolytics which have lower inductance which further decreases when use in parallel works better than a single large cap and minimizes the amount of bypassing which has to be done. It is a trade off as usual. Proper bypassing is also important in battery supplies as the internal impedance of the battery at HF can also be problematic.

As far as a higher VA rating on a transformer used in a preamp power supply where total power and efficiency are not an issue, it is will help somewhat but is not the best solution. Remember that if the output voltage of the transformer is the same a higher VA rating means more current. All of your charging of the filter caps still has to take place at the same point of the AC waveform where the secondary voltage exceeds the voltage on the caps plus the rectifier drops. With the lower impedance of the secondary the caps can charge faster but that also results in higher current spikes.

If you are not worried about a little more power dissipation across the regulator, you are better off going with a transformer with a higher secondary voltage rating with enough current that under maximum current load and lowest expected line voltage, the voltage of the filter caps will not decrease to the point where the regulator will drop out of regulation. The extra voltage of that design will provide more headroom for the inevitable sags you encounter in the line. Also, since the secondary of the higher voltage lower VA transformer will overcome the voltage on the caps and the rectifiers earlier in the cycle, there will be more time for the caps to charge and because of the higher secondary resistance lower current spikes. Overall since the impedances are not as low there will be more power loss, but when you are talking about a preamp it is inconsequential. No one solution works for all scenarios and a high power amp cannot be done this way because the dissipation will be excessive.

_________________
Walt

Walt:

At 74 and having been in this since a kid I have a lot of both theoretical and practical experience. In fact if you don't have both you are at a disadvantage. My published works on both electronics and speakers have also been scrutinized for over 20 years by a lot of people and have passed muster in a much larger public domain than this forum. So please don't try to minimize what I have done in both areas. You may disagree with me as do others which is healthy but my experience is not minimal.