its a poor PSU, it is self astable oscillating that only suitable for low power. and has some burned resistor at its astable loop module. better replace it another, linear psu is ok. they said ir only irf740, and it won't help. the irf740 will suffer from heat and over current overtime that will damage it again.
Not a lame question at all 
All these checks are with the power OFF. Its possible with some faults for the reservoir caps to remain charged to 340 volts (on UK mains) and so you need to be sure that they are discharged before working on the board. Always check that there is no residual voltage on those big caps first.
All these checks are with the power OFF. Its possible with some faults for the reservoir caps to remain charged to 340 volts (on UK mains) and so you need to be sure that they are discharged before working on the board. Always check that there is no residual voltage on those big caps first.
As it has been pointed out previously, a self oscillating push-pull like that one, governed by transformer saturation (and FET desaturation), will stop if overloaded, and if designed properly the process won't be destructive. The diac causes periodic startup attempts. This is pretty similar in function to push-pull SMPS found in other low cost equipment, or early SMPS equipment, like Behringer INUKES and plate amps, stuff made by Loud Technologies, early QSC, etc. These circuits go one step ahead modulating the whole push-pull on and off in case of overload.
Power rail goes through a current limited buck regulator just after the push-pull, so sophisticated overload protection in the push-pull is not essential (it would be essential for more powerful amplifiers where components would be used closer to their limits, in this case there is some margin). Overload protection in audio output stage is not required either, as the buck regulator that powers amplifier rail can do a folded current limiting. Clever simplification, where the class G contributes more than one advantge.
I have seen far worse sub amplifier modules in that plate format, same price range, with all thru-hole parts. These BASH modules are similar in concept to Labgruppen, in low cost flavor.
Fear to SMPS is fear to understanding the steady-state behavior of circuits as AC, waves and frequencies, something that is AC in steady state, not just DC, not just single freq AC steady state either, oscillators, cross-talk, how oscillators interact one with another. Of course these same mental calculations are used to understand the AC steady state nature of life, for the ones not fearing to do some tuning of the loops they are immersed in.
Last edited:
After revising the schematics in detail I think there are a couple of circuit sections missing, one dealing with ON/OFF and another with idle class G rail voltage adjustment and current limiting. I can't tell if there is foldback current limiting.
If I was the designer I would do the following additions/changes to these modules:
- Quality capacitors.
- Foldback current limiting of output stage implemented in the buck, half current limit at 0V at speaker terminals.
- DC protection, latching push-pull SMPS off. One opto from secondary to primary required.
- Low power standby mode, shutting down the buck, making most of the +/-15V stuff switched, and operating push-pull SMPS in hiccup mode. Another opto from secondary to primary required. I don't know if it could get as good as 500mW standby, probably not, due to resistive dividers and stuff in primary side.
If I was the designer I would do the following additions/changes to these modules:
- Quality capacitors.
- Foldback current limiting of output stage implemented in the buck, half current limit at 0V at speaker terminals.
- DC protection, latching push-pull SMPS off. One opto from secondary to primary required.
- Low power standby mode, shutting down the buck, making most of the +/-15V stuff switched, and operating push-pull SMPS in hiccup mode. Another opto from secondary to primary required. I don't know if it could get as good as 500mW standby, probably not, due to resistive dividers and stuff in primary side.
Last edited:
A small flyback for the low voltage solves the stdanby power dilemma, using a relay to switch power to main SMPS section (and its dissipative resistors), and another relay for voltage selection.
Next evolutions could be merging the buck into the push-pull, maybe as a resonant flyback or LLC, valid for LF only, or just going class D, but all that would require newer parts and production techniques, so the threshold for a good low cost product that could be manufactured with insertion machines and single wave soldering (old facilities), is probably between the last improvement suggested and evolving to more sophisticated SMPS and class D.
Next evolutions could be merging the buck into the push-pull, maybe as a resonant flyback or LLC, valid for LF only, or just going class D, but all that would require newer parts and production techniques, so the threshold for a good low cost product that could be manufactured with insertion machines and single wave soldering (old facilities), is probably between the last improvement suggested and evolving to more sophisticated SMPS and class D.
Right I have done some testing with the results as follows :-
Q401 Pinout G - Pin 1, D pin 2, S pin 3 (with DMM on diode setting Red lead on Gate)
G to D 1139
G to S 825
S to D 797 (Red lead on S)
Q402
G to D 1139/1140
G to S 825
S to D 795
R421 - 10R
C418 56.67 uf ESR 58 Ohms
C419 46.7 uf ESR 58 Ohms
C118 104.2 uf ESR 59 Ohms
C119 (not accessible)
Q401 Pinout G - Pin 1, D pin 2, S pin 3 (with DMM on diode setting Red lead on Gate)
G to D 1139
G to S 825
S to D 797 (Red lead on S)
Q402
G to D 1139/1140
G to S 825
S to D 795
R421 - 10R
C418 56.67 uf ESR 58 Ohms
C419 46.7 uf ESR 58 Ohms
C118 104.2 uf ESR 59 Ohms
C119 (not accessible)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.