If you could show the PSU schematic, we could help you more.
Secondary (unloaded/loaded) voltage, DCR, static load current also need.
7Vpp is horrible value, something was poorly planned.
Secondary (unloaded/loaded) voltage, DCR, static load current also need.
7Vpp is horrible value, something was poorly planned.
That is amazing how it tuned out for you just with a very simple compensation technique.
I won't bet it will work for a random transformer.
If I am not wrong, the NFB is way over 20dB.
I won't bet it will work for a random transformer.
If I am not wrong, the NFB is way over 20dB.
Here is a soft start/cap-multiplier I've used - the mosfet is designed for linear use, Mouser (US) has them in stock
No free lunch....
If you load voltage doubler first capacitor with high current (4*40mA), the ripple goes to the sky.
BTW, in this case the PT secondary current also high, the PT -only HV use- dissipation about 86W.
It is no coincidence that author recommends the CLC filtering instead of voltage doubler.
p.s. if I use stacked capacitors -even in doubler- always put equalizer resistors next to capacitors.
If you load voltage doubler first capacitor with high current (4*40mA), the ripple goes to the sky.
BTW, in this case the PT secondary current also high, the PT -only HV use- dissipation about 86W.
It is no coincidence that author recommends the CLC filtering instead of voltage doubler.
p.s. if I use stacked capacitors -even in doubler- always put equalizer resistors next to capacitors.
Is not really doubler, is basicaly 2 power suply in series. Is 2 diodes for negative and 2 for positive.
Or can be cosidered a bridge with capacitors on series, but the midle is connected to half of secundar to be balanced. If I understood well. The midle could be conected also with some resistor. But the ripple yes I simulate the same on multisim and confirmed by measuring. If you add on simulator a simple secundar with 350V , a bridge and 220uf filter will see the same. 2x220Uf in series with 2x220uf is echivalent of 220uf.
Or can be cosidered a bridge with capacitors on series, but the midle is connected to half of secundar to be balanced. If I understood well. The midle could be conected also with some resistor. But the ripple yes I simulate the same on multisim and confirmed by measuring. If you add on simulator a simple secundar with 350V , a bridge and 220uf filter will see the same. 2x220Uf in series with 2x220uf is echivalent of 220uf.
Yes, I see now, it's stacked.
But the problem in both cases the same: SS rectifier charges the large first capacitor with large, wide current pulses (about 300mA/220uF capacitor, about 130mA RMS).
It's generate large sawtooth ripple voltage.
Never use this point as power supply for low/moderate PSRR device without additional filtering.
But the problem in both cases the same: SS rectifier charges the large first capacitor with large, wide current pulses (about 300mA/220uF capacitor, about 130mA RMS).
It's generate large sawtooth ripple voltage.
Never use this point as power supply for low/moderate PSRR device without additional filtering.
To soft start, I would put NTC Thermistor in serial with the primary side of power transformer.
NTC Thermistor has high resistance when it is cold. During power on, it heats up as it has high resistance initially. Once it reaches the operation temperature, the resistance drops to a point you can almost ignore.
No fancy circuitry is required.
NTC Thermistor has high resistance when it is cold. During power on, it heats up as it has high resistance initially. Once it reaches the operation temperature, the resistance drops to a point you can almost ignore.
No fancy circuitry is required.
Some improvement:
before filtering:
after filtering with mosfet (thanks longspeak ! ) :
also before and after with Arta Audio:
before filtering:
after filtering with mosfet (thanks longspeak ! ) :
also before and after with Arta Audio:
Last edited:
Interesting thingy. I'll test this
could this be posted more readable please .
Also what voltages are the zeners.
Is 9V1 Zenners.
Instead of 1M resistor I have used 390K. With 1M the voltage drop was to high, about 25V. Now is about 14V. Is enough to compensate 7Vpp ripple.
But also deppend of the mosfet I think. I have at home only one 2SK1745, is only 500V. So need some better, this was just for testing. I have 460V in and aprox 446V out from filter.
Instead of 1M resistor I have used 390K. With 1M the voltage drop was to high, about 25V. Now is about 14V. Is enough to compensate 7Vpp ripple.
But also deppend of the mosfet I think. I have at home only one 2SK1745, is only 500V. So need some better, this was just for testing. I have 460V in and aprox 446V out from filter.
Agree, the sound was perfect but now the graph also . Beeing an electric engineer I like to be perfect 🙂
This point was only for the final tubes. But beeing UL push-pull the signal was modulated in G2 of the EL34 I supouse. With now signal was not any audible hum.
Glad it helped. As istvan said, they are 9.1v zeners. Here is a clearer version
only caution is pay attention to safe operating area for the mosfet if you are going to use it for a large voltage differential in a soft start, here are the graphs for IXTH device, notice the DC amperage is <1A at 600V.
only caution is pay attention to safe operating area for the mosfet if you are going to use it for a large voltage differential in a soft start, here are the graphs for IXTH device, notice the DC amperage is <1A at 600V.