Improved Soft Start + DC Blocker aka iSSDCB
Joined 2003
Paid Member
I've updated the original SSDCB to offer the following benefits on the new iSSDCB
  1. Greatly simplified wiring required only on the mains supply side of the transformer
  2. Double the full AC mains load capability of the original SSDCB – 4.6A vs. 2.3A
  3. Allows use of attractive anti-vandal power control switches – max required switch current rating is 30mA DC
  4. Very low standby power draw of c. 0.2W
  5. In-rush current limiting of 10A for 1 second
  6. Small compact, easy to assemble PCB

Here is the link to the project download page on

If you have any questions, I'll be happy to answer.

Joined 2011
Paid Member
Sorry if this is a dumb question, from someone who doesn't live in Europe:

Does this meet regulatory requirements that on/off switches MUST disconnect both legs of the AC mains from the rest of the equipment?

I ask because I think AC neutral appears not to be switched off and completely disconnected. There's a current path from U2 (AC neutral IN), through D4, through the Inrush Current Limiter thermistor, to U8 (AC neutral out) . . . ? Is this okay in the EU?

Thanks for any clarification!
Joined 2003
Paid Member
There is no regulatory requirement to isolate both hot and neutral on consumer audio gear - few pieces of commercial gear have this in them. A good example is a Schurter (or similar) switched IEC receptacle found on the back of a lot of equipment- just the hot is switched.

On the iSSDCB the in-rush limiting is placed in the neutral line because the cans of the main bypass caps for DC isolation will then be at around ground potential so it’s safer. You must tape the cans over in any event. The hot is switched through the power ON/OFF control relay, so that’s pretty much conventional.

I would always in any event recommend a switched/fused receptacle like a Schurter for mains entry on the rear panel - I’ll add a note to that effect in the presentation. As always, never use equipment that is not earthed (safety grounded).

One approach to answering this question is to calculate the total energy stored in the PSU in Joules.

You stated 50V secondaries. With 50V secondaries, when full wave rectified (full wave bridge), we are talking at least 71V DC, maybe a little more or less to allow for DC mains variations. So to be conservative make sure you use 80VDC capacitors in your PS for your Wolverine. That’s a juicy amplifier!

If your cap bank is 60mF total per channel (assuming 30mF per rail), let’s calculate the energy storage per rail, then per channel, and then double it for the final.

The energy stored is calculated by: 0.5*C*V*V; where C is in Farads and V designates the RECTIFIED DC voltage. That’s why I stated 71V DC in the previous lines. There is a (+)71V DC rail and a (-)71V DC rail.

0.5*3.0E-2*(+71)(+71) = ~75J. That’s the total for just one rail. Two rails is now 150J, and that’s just one channel. Other channel will add another 150J, so now the total is 300J approximately.

The Ametherm ICL that is listed in the BOM for this softstart design at a 230VAC has a max rating of 250J. So your 300J exceeds that. Moreover, you also have a 1200VA transformer which probably adds its own load.

So unless modified, this may not be soft start you are looking for. Perhaps an ICL with a higher joule rating would be useful, or individual softstarts for each channel may be useful, i.e. turn your big stereo amp into a monoblock amp. 600VA with 71V rails and 60mF total per channel. Or decrease the capacitance in the power supplies. Or use lower voltage secondaries (for lower rectified voltage rails). There are probably other possibilities.

I’m sure @Bonsai and others with the technical acumen will correct me if I am wrong as I am still learning. And there may be other limitations within this softstart design that I am unaware of to accommodate the HUGE inrush currents we are talking about here!

Joined 2011
Paid Member
Both Cordell and Self advise connecting "N" inrush current limiter discs in series (NOT in parallel !!). If each one is rated for X Joules, the series arrangement can handle (N * X) Joules. Of course you and your circuit designer would need to recalculate the desired cold-resistance and the desired steady state current for the individual ICL discs in this new connection topology. Next, either find a way to bodge it on the existing PCB, or else lay out a new PCB with N footprints for inrush current limiter discs.
  • Thank You
Reactions: 1 user
Joined 2003
Paid Member
Been on a road trip for the last 7 days :)

For the in-rush, you could reduce the in-rush time since the joule rating is always predicated on the assumption that it is measured over 1 second. In almost all cases, most of the in-rush energy is in the first 5-10 mains cycles. You can reduce the in-rush surge bypass time to 0.5 seconds by replacing C7 with 470 uF and R6 and R7 at 68 Ohms each. This will give around 0.6 seconds.

I would agree that placing Amertherm NTC's in series is a better solution than in parallel.

Re the energy calculation, I think the figures used by Ametherm are for continuous use - ie not bypassed as we are doing here after a few seconds. The device rupture energy is stated as c. 2x the rated energy.

Might be an idea for someone based in the US to ask Ametherm about this.

@NSP - The continuous full power load of the iSSDCB is 1100W at 240 VAC and half that at 120 VAC so you should be good to go. I can imagine listening to a class AB amp drawing that much power and coming away with your ears still intact!
  • Like
Reactions: 1 users