Digital Safe Soft start unit

[BMW850]

Hallo,

Now a have the LC Audio soft start (Digital Safe Soft start unit max. 10 Amp.) but these soft start is now broken for the second time.

I will replace this soft start to a heavily one, and I need 2 soft starts for my Amplifier Monoblock, Toroid Transformer 42 volt 2000Va.
I can’t find a website that sell these heavily soft starts. (I need a fully assembled kit because I am not a electrician)
Please can somebody help me where to find a good soft starts that can do the job easily.

Rudy
P.s Sorry for my bad English

[Ian Millar]

I have made 5 or so ESP soft starts and they are super reliable after many years of use. I use one on a 2kVA toroidal and used a bank of six by 5W 120 Ohm resistors for the ballast. It works VERY well. I have an LC Audio soft start and it failed after 1 hour due to a very cheap SMD diode after the string of voltage-reducing SMD resistors. The LC audio soft start is a useless device. If it worked, it would only limit inrush current to around 0.75 Amps, so is really nothing more than a delay switch. It uses a puny 330 Ohm resistor. Too high resistance and very low current. I am overhauling my LC audio with better ballast and a ner larger diode this week.

[star882]

Just use a NTC. And a bypass relay if high efficiency is important.

[Ian Millar]

Please excuse my typo: "ner" => new

Originally Posted by star882

Just use a NTC. And a bypass relay if high efficiency is important.

I understand "NTC" to mean "Negative Temperature Coefficient" for thermistors (temp up => resistance down), but this has nothing to do with the LC Audio diode failure problem. The diode sits in the power supply path before the timer-control circuitry and when it fails the relay opens so current is directed back through the ballast. My LC Audio soft start had an NTC thermistor (for reasons which remain a mystery) in series with the ballast, but it is not in thermal contact with the ballast and just sits there like an ornament alongside the relay. I replaced it with a 167 degree C thermal fuse bonded to the ballast. The fuse blew due to failure of the diode.

[Ian Millar]

Oh! I guess that post (2-up) was just spam going by the hyperlink in the sign-off? (excuse me if I am mistaken)

Anyway substitution of a regular 1N4004 diode, a bank of proper ballast resitors, a thermal fuse bonded to one of them and beefed up (higher-current) PCB tracks seems to correct many of the problems with that LC Audio "device". A fixed-up version is featured toward the bottom of my blog page for an autotransformer in which one was used.

Compact Soft-starter/Autotransformer Audio Projects

[Redshift187]

Using NTC thermistors in place of the balast resistors is not a bad idea. At the currents they pass, they self heat quickly, giving more of a curve to the current flow, allowing it to "start soft" and ramp up close to maximum, so that when the relay bypasses them, it is switching a smaller voltage difference. In addition, the thermistors should handle the heat better too.

[Ian Millar]

Originally Posted by Redshift187

Using NTC thermistors in place of the balast resistors is not a bad idea. At the currents they pass, they self heat quickly, giving more of a curve to the current flow, allowing it to "start soft" and ramp up close to maximum, so that when the relay bypasses them, it is switching a smaller voltage difference. In addition, the thermistors should handle the heat better too.

I thought that the voltage is set so the current ramps up as the resistance falls with rising temperature.

The guy wants to soft-start 2kVA transformers. It's a pretty tall order for a consumer soft start. A conservative approach is needed IMO.

According to the ESP site NTCs can explode if a fault condition arises. I imagine a relay going open (which happened with my LC Audio device after just 1 hour) concurrent with an amplifier fault causing an NTC to reach an explosive equilibrium condition wherupon the unsuspecting takes a close look to have it explode in his face all without the main fuse blowing.

A bank of resistors in contact with and in series with a thermal fuse with a temperature spec. of 167 degrees C (which is lower than the solder melting point and resistor high temp rating) will not do the nice ramping act, but seems like a more robust compromise. Better still would be to use a high-current aluminium-encased resistor of say 22R and bolt it to his chassis. Connect the thermal fuse to one of its terminals and epoxy it to the aluminium case. Run teflon-insulated (ie heat resistant) wires back to say an ESP soft start main board.

That "digital safe" (whatever that's supposed to mean) device has an NTC in series with a single 330R ballast for some odd reason. It seems dumb because the resistance will never fall below 330R which is too high to start with - only allowing around 700 mA through. Useful perhaps for a small toroidal which doesn't need a soft start anyway = useless.

[Redshift187]

Yes, your approach is safe enough, I just don't like the idea of subjecting the resistors to such thermal cycling. When I was going to design my own soft start, I was going to use something like a pair of 40R 50W resistors (in parallel) which had a peak power dissipation allowance of something like 250W.

I also definitely agree with the thermal switch for such a high VA transformer. That would obviously not be useful with thermistors.

That said, my (commercial) soft start uses a separate transformer of its own and a 555 timer IC to ensure the relay *always* activates.

[Ian Millar]

With 22R on a 230V supply the resistor(s) must dissipate around 2400W so the timing has to be fast to avoid overheating the ballast - say 500 ms max. I found that the ESP 100ms timing was too fast for a 2kVA toroidal because the 20A circuit breaker at my meter box still tripped every so often. I suspect that the inrush surge peaks later than 100ms on such a big transformer. I changed a cap to increase timing to around 500ms. The mains voltage sections of the PCB tracks of the soft start must be beefed up with extra wire for a 2kVA toroidal too. Mine has been working well for over 5 years now.

Another intersting thing is that the transformer will growl momentarily with different levels of ferocity depending on where in the AC cycle power is first applied (and then again at the relay activation point). Some days mine growls twice, other days just once and other days not at all. Some sites suggest (perhaps wrongly) that switching at the zero crossing point is desirable. According to ESP this is wrong and switching should ideally occur at a peak. I haven't gotten my head around that yet.

While I don't see how having a 555 timer will guarrantee relay action (the ESP soft start uses a simple RC timer and an auxilliary transformer too but a fault could still cause the relay action to fail). I am interested to know about your commercial soft start. My CJ preamp uses a 555 in its warm-up muting circuit

[Redshift187]

I don't remember exactly what resistance I had planned on using, but A) I'm not on 230V, I'm on 110V and B) I only use a 330VA transformer. I had calculated it and selected resistors that would be within their limits.

Basically, the 555 timer is part of a digital circuit to enable the relay, instead of an analog one. Anything can fail. But by having a separate transformer and circuit to enable the relay, minimizes the risk of a failure outside the soft-start causing an additional failure in the soft-start.

The documentation states that it uses 4x 10R thermistors in a series-parallel configuration. This starts at 10R but drops to 3.6R at 80C. After about 500ms (a rough guess), a relay shorts the mains around the thermistors, and the relay connecting the thermistors opens.