Fuses between the rectifier and capacitors are a little pointless because they will blow after sometime, and if you lose one before the other leading to only one voltage to the amp, there can be quite bad consequences on the amplifier and/or speakers.
To limit inrush the soft start needs to be before the transformers. Would recommend against using sawtooth controllers (all solid-state devices fit in this category) and instead stick to passive soft start and a big relay.
Do not use the Chinese ones, they stick thermistors in parallel instead of using larger rated thermistors - this is not good practice as thermistors are generally very poorly matched and differential heating can cause your inrush to be actually worse, and in some cases cause extreme current through one leg of the parallel pair - causing thermistor failure.
Thermistors like to blow up spectacularly because manufacturers assume the designers know what they are doing and can properly implement it. With the Ebay/Ali parts, this is almost always a fallacious assumption. You need to do it yourself.
Note that for reactive loading the transformer may see peak loads of up to 2x the actual output current, and for very large cap banks this can cause transformer humming and self-heating. 1200VA is on the edge for that sort of bank, and just about enough if you use one per channel.
To limit inrush the soft start needs to be before the transformers. Would recommend against using sawtooth controllers (all solid-state devices fit in this category) and instead stick to passive soft start and a big relay.
Do not use the Chinese ones, they stick thermistors in parallel instead of using larger rated thermistors - this is not good practice as thermistors are generally very poorly matched and differential heating can cause your inrush to be actually worse, and in some cases cause extreme current through one leg of the parallel pair - causing thermistor failure.
Thermistors like to blow up spectacularly because manufacturers assume the designers know what they are doing and can properly implement it. With the Ebay/Ali parts, this is almost always a fallacious assumption. You need to do it yourself.
Note that for reactive loading the transformer may see peak loads of up to 2x the actual output current, and for very large cap banks this can cause transformer humming and self-heating. 1200VA is on the edge for that sort of bank, and just about enough if you use one per channel.
i can use power resistor instead of chinise thermistor no?
What kind of values i can use in this case to start this psu?
What kind of values i can use in this case to start this psu?
10R will do.
A relay should shorten this resistor after the large caps have reached ca. 80% of their final voltage, depending on how much the amp behind draws when idling.
Hans
A relay should shorten this resistor after the large caps have reached ca. 80% of their final voltage, depending on how much the amp behind draws when idling.
Hans
Mongis, each amp supposedly runs class A. The schematic indicates the idle current per amp may be something like 4x 0.5 = 2A, based on 0.7Vbe and 1.2Vb and 0.5 ohm ballasts.
With +/- 51V rails, that is a load of 102Vx2Ax2 = 400W. Given a nominal 50% rectifier efficiency, that requires at least 800VA transformer, so ball-park a 1200VA PT may be ok.
PSUD2 indicates default winding current is circa 6.5Arms, so fuse needs to handle that level continuously - perhaps an 8A IEC 5x20 is next practical type. Given the 10ms rms is about 40A, which is under 10x multiplier for a Time delay 'T' rating, but over the 4x multiplier for a Fast 'F' rating, then the fuse needs to be a time delay T rating. One fuse in each secondary arm.
That is ball-park, and a better estimate would need power transformer winding resistances (which may be difficult for you to measure accurately), and better measurement of the amp idle current. Also I am talking IEC compliant miniature cartridge fuses - you may be in a country that only sells UL compliant fuses, and that would require a re-assessment as strangely not all fuses are created equal.
You may or may not need any mains feed in-rush protection, depending on whether your mains breaker trips. Imho it is not wise to add inrush limiting for 'aesthetic appeal'. I'd also recommend you install a primary side fuse (in your equipment) that is just above max primary continuous current level (eg. use a power meter to measure) and T style - not to protect mains feed wiring, but to provide as much chance for a fuse to operate if something goes wrong, given you have a 1200VA PT.
With +/- 51V rails, that is a load of 102Vx2Ax2 = 400W. Given a nominal 50% rectifier efficiency, that requires at least 800VA transformer, so ball-park a 1200VA PT may be ok.
PSUD2 indicates default winding current is circa 6.5Arms, so fuse needs to handle that level continuously - perhaps an 8A IEC 5x20 is next practical type. Given the 10ms rms is about 40A, which is under 10x multiplier for a Time delay 'T' rating, but over the 4x multiplier for a Fast 'F' rating, then the fuse needs to be a time delay T rating. One fuse in each secondary arm.
That is ball-park, and a better estimate would need power transformer winding resistances (which may be difficult for you to measure accurately), and better measurement of the amp idle current. Also I am talking IEC compliant miniature cartridge fuses - you may be in a country that only sells UL compliant fuses, and that would require a re-assessment as strangely not all fuses are created equal.
You may or may not need any mains feed in-rush protection, depending on whether your mains breaker trips. Imho it is not wise to add inrush limiting for 'aesthetic appeal'. I'd also recommend you install a primary side fuse (in your equipment) that is just above max primary continuous current level (eg. use a power meter to measure) and T style - not to protect mains feed wiring, but to provide as much chance for a fuse to operate if something goes wrong, given you have a 1200VA PT.
Last edited:
My leach am has a 2kw toroid and lots of uf, it powers up nicely with a cheap ebay softstart, I find it strange.
No that ballpark doesn't work with ss rectifier and capacitor input loads, as the mains current is not a nice sinusoid but rather an ugly peaky waveform that decapitates the sine voltage waveform for all other equipment in your house. Better to guesstimate no more that 50% of 1200VA is available as output DC power (so +/-50V at 6A).
I have made a PCB of a soft start board with remote trigger.
PowerAmpAudio/SoftStart at master * profdc9/PowerAmpAudio * GitHub
PowerAmpAudio/SoftStart at master * profdc9/PowerAmpAudio * GitHub