A higher wattage SMPS might work slightly better, but after reading some posts by the designer of the SMPS series by Connexelectronic, it seems we can indeed use any size capacitors after the SMPS we want, we only have to take care to limit the in-rush current. The over-current protection is likely to kick in with any wattage SMPS if the cap bank is large enough.
I'm building mine with a SMPS300RE and if i find the bass lacking i will definitely look into bigger caps & in-rush current limiter. Just need to find a solution which ideally doesn't involve a mechanical relay and doesn't cost too much.
On a side note i decided to try using a BF862 for T1. Anyone else besides me and aspiringv trying that?
I would be tempted to go with Power NTC Thermistor.
It is precisely this duty they have been designed for.
If your supply ramps to 50Vdc quickly, then initially when the capacitor bank is discharged the full voltage is available to charge the caps.
If you know what output current the PSU will tolerate as start up current, then simply find the resistance from: Vdc / Imax at start up.
Then calculate, or estimate, the time required to substantially charge the cap bank from RC and tapering off of the charge current.
Finally bypass the resistance to bring the PSU fully on line.
This should all be inside the SMPS desinged for audio power amplifier duty.
If it's not then the SMPS is not suitable for audio power amplifier duty.
I recall a discussion on modifications required to allow one of these SMPS to start up when output bias was set highish. That again points to a poorly designed SMPS for audio power amplifier duty.
I get the feeling that the two major proponents of the audio SMPS on this Forum do not understand the transient requirements of Audio Power Amplifiers.
It is precisely this duty they have been designed for.
If your supply ramps to 50Vdc quickly, then initially when the capacitor bank is discharged the full voltage is available to charge the caps.
If you know what output current the PSU will tolerate as start up current, then simply find the resistance from: Vdc / Imax at start up.
Then calculate, or estimate, the time required to substantially charge the cap bank from RC and tapering off of the charge current.
Finally bypass the resistance to bring the PSU fully on line.
This should all be inside the SMPS desinged for audio power amplifier duty.
If it's not then the SMPS is not suitable for audio power amplifier duty.
I recall a discussion on modifications required to allow one of these SMPS to start up when output bias was set highish. That again points to a poorly designed SMPS for audio power amplifier duty.
I get the feeling that the two major proponents of the audio SMPS on this Forum do not understand the transient requirements of Audio Power Amplifiers.
Last edited:
from this thread.Cristi said:
Cristi is the designer of the SMPS300RE, elsewhere in that thread he states that the soft-start time constant is chosen to be just enough to charge the on-board capacitors. He says it's 100-200ms now. So with a total capacitance (SMPS+Fetzilla) roughly equal to the original design capacitance of the Fetzilla, assuming a linear ramp-up of voltage, what time constant should we shoot for?
(I'm looking at you, AndrewT)
All the best,
Kris
I realise now that AndrewT's comment likely refers to the discussion about adding a Power NTC or a regular resistor to limit in-rush current, i'm still not convinced it has come to that, there may still be a way to tweak the SMPS300RE in a way which will allow it's built in ramp-up to last a bit longer, but for aspiringv's SMPS it may prove necessary. I did find this from the SMPS300RE page:
It's not entirely clear if he means that the OC protection kicks in at 520 Watts, i.e 6.8 Amps per rail(260W/38V), or 390 Watts(i.e 5.1 Amps per rail) but i guess it's something that can be figured out if need be.
Aspiringv, i found this page:
Charging a Capacitor
It could help with figuring out how big a resistor is needed. I put in values of 9701uF and 38V, and using a 20 ohm resistor appears to limit the current to 1.9 Amps and still gives us a sub-200 ms charge time(ideally, my guess is it will take slightly longer to charge).
Now all we need is a way to calculate the charge times with a rising voltage, but ideally we need more details, but for argument's sake let's assume a linear ramp-on over 200ms. With an SMPS300RE "Imax" of 6.8/5.1 Amps(per rail) we should be able to calculate a maximum cap size that is safe. This feels like an integral equation, let the math wizardry begin!
It's not entirely clear if he means that the OC protection kicks in at 520 Watts, i.e 6.8 Amps per rail(260W/38V), or 390 Watts(i.e 5.1 Amps per rail) but i guess it's something that can be figured out if need be.
Aspiringv, i found this page:
Charging a Capacitor
It could help with figuring out how big a resistor is needed. I put in values of 9701uF and 38V, and using a 20 ohm resistor appears to limit the current to 1.9 Amps and still gives us a sub-200 ms charge time(ideally, my guess is it will take slightly longer to charge).
Now all we need is a way to calculate the charge times with a rising voltage, but ideally we need more details, but for argument's sake let's assume a linear ramp-on over 200ms. With an SMPS300RE "Imax" of 6.8/5.1 Amps(per rail) we should be able to calculate a maximum cap size that is safe. This feels like an integral equation, let the math wizardry begin!
Last edited:
Another SMPS shutdown when trying to start with a 600mA output bias setting (600mA Ib).
They discovered that by reducing the bias setting to around 300mA that the SMPS would start up and fully power the amplifier ready for use.
I don't know whether they have a method that allows the original 600mA Ib to start up.
They discovered that by reducing the bias setting to around 300mA that the SMPS would start up and fully power the amplifier ready for use.
I don't know whether they have a method that allows the original 600mA Ib to start up.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.