Great job.
I need help with ripple eater. Guess what? I've connected grounded radiator to transistor... I have replaced transistors on one side: BD139 MJE15032. Before connecting AC, I've removed and tested diodes in1407, all were ok on multimeter.
When I connect it to 25v x2 300va toroid, multimeter connected to vneg-vpos displays voltage slowly increasing to 39V dc. I disconnected when reached 39.5V.
What to do now?
Before that i had it tested on both channels one by one. It was working, but at 25% higher villages than in pdf
I need help with ripple eater. Guess what? I've connected grounded radiator to transistor... I have replaced transistors on one side: BD139 MJE15032. Before connecting AC, I've removed and tested diodes in1407, all were ok on multimeter.
When I connect it to 25v x2 300va toroid, multimeter connected to vneg-vpos displays voltage slowly increasing to 39V dc. I disconnected when reached 39.5V.
What to do now?
Before that i had it tested on both channels one by one. It was working, but at 25% higher villages than in pdf
Some quick points on the PSU’s etc
Nx-Amp PSU includes DC offset and switch ON/OFF muting protection so this is why it is more complex than the other PSU’s. I did not include any ripple eater function so the ripple on the nx-Amp power rails will be between 1 and 2 Volts at full power (100 watts into 8 Ohms).
The original six Amp is just a straight supply with no protection features. Since the supply rail is only +-22V and there are two pairs of 250 Watt transistors per channel, they see very little stress so I kept it really simple. I have shorted the output of my sx and kx amps a few times during development and the rail fuses (5A) just pop but the amps work fine after replacing the fuses.
The ripple eater PSU was designed specifically for use with the kx and sx Amps. No protection is included, but both amps have very overrated output stages as mentioned. In a class A amp, the ripple is high even at low outputs because of the high continuous current load. The ripple eater reduces power supply rail ripple to a few tens of mV typically.
Nx-Amp PSU includes DC offset and switch ON/OFF muting protection so this is why it is more complex than the other PSU’s. I did not include any ripple eater function so the ripple on the nx-Amp power rails will be between 1 and 2 Volts at full power (100 watts into 8 Ohms).
The original six Amp is just a straight supply with no protection features. Since the supply rail is only +-22V and there are two pairs of 250 Watt transistors per channel, they see very little stress so I kept it really simple. I have shorted the output of my sx and kx amps a few times during development and the rail fuses (5A) just pop but the amps work fine after replacing the fuses.
The ripple eater PSU was designed specifically for use with the kx and sx Amps. No protection is included, but both amps have very overrated output stages as mentioned. In a class A amp, the ripple is high even at low outputs because of the high continuous current load. The ripple eater reduces power supply rail ripple to a few tens of mV typically.
Last edited:
Yyy? I dont get it?
How long should it take to stabilise voltage and what value that voltage should be? My RE without load needs 30-40seconds to teach 39V on vneg-vpos, is it ok? Maybe I'm wrong and it should be 60V from both halves?
How long should it take to stabilise voltage and what value that voltage should be? My RE without load needs 30-40seconds to teach 39V on vneg-vpos, is it ok? Maybe I'm wrong and it should be 60V from both halves?
If your transformer is 25V AC you will get about 38 or 39 V DC out with no load.
That voltage can only be used with the kx-Amp in class AAB mode and not in class A mode. You should have c. 28 to 29 V DC on each side so 56 to 58V no load across both PSU’s.
My PSU takes about 6-7 seconds to come up to full supply - ie 90-95% of the final reading. My supply rails are +-28VDV before the ripple eater and +-26.6 after the ripple eater.
Put a load on the output (say 100-200 mA) and then redo the readings.
That voltage can only be used with the kx-Amp in class AAB mode and not in class A mode. You should have c. 28 to 29 V DC on each side so 56 to 58V no load across both PSU’s.
My PSU takes about 6-7 seconds to come up to full supply - ie 90-95% of the final reading. My supply rails are +-28VDV before the ripple eater and +-26.6 after the ripple eater.
Put a load on the output (say 100-200 mA) and then redo the readings.
Surely I will do.
I forgot to say that touching anywhere on mje15032 and bd139 legs lights up diode on that side of pcb.
I forgot to say that touching anywhere on mje15032 and bd139 legs lights up diode on that side of pcb.
Tested.
My dummy load were 2x 330ohm paralell in glass of water.
Vplus-0 4.5v
Vneg-0 5.5v
Vplus-vneg 6.8
When i moved cable connector at mje15032 between vneg vplus and 0 red diode stops emiting as long as cable is connected.
My dummy load were 2x 330ohm paralell in glass of water.
Vplus-0 4.5v
Vneg-0 5.5v
Vplus-vneg 6.8
When i moved cable connector at mje15032 between vneg vplus and 0 red diode stops emiting as long as cable is connected.
Princetemi
You can fix it ,Ike this. Print the circuit off and then measure the voltages around the circuit and write them on the diagram and then post it up here so we can help.
Remember to load the output of the PSU.
You can fix it ,Ike this. Print the circuit off and then measure the voltages around the circuit and write them on the diagram and then post it up here so we can help.
Remember to load the output of the PSU.
Hi all,
Excuse my ignorance but is 'Class AAB' basically running Class AB with higher bias currents or something else? (Can't find much info on a 'net search' but also came across 'Adaptive Auto Bias' as regards amplifiers).
Thanks
P.S. I have a 25/0/25 500VA transformer that's looking to meet up with an amplifier circuit - perhaps the kx 🙂
Excuse my ignorance but is 'Class AAB' basically running Class AB with higher bias currents or something else? (Can't find much info on a 'net search' but also came across 'Adaptive Auto Bias' as regards amplifiers).
Thanks
P.S. I have a 25/0/25 500VA transformer that's looking to meet up with an amplifier circuit - perhaps the kx 🙂
Yes - class AAB means higher than usual class AB biasing so you get a larger class A region than you would with conventional class AB beading.
In the kx-Amplier, class AAB sets each output pair at 300 mA, so 600 mA per amplifier, and in class A is c. 600 mA per pair, 1.2 mA total per amplifier.
Your transformer would work well with a kx-Amp in class AAB mode.
In the kx-Amplier, class AAB sets each output pair at 300 mA, so 600 mA per amplifier, and in class A is c. 600 mA per pair, 1.2 mA total per amplifier.
Your transformer would work well with a kx-Amp in class AAB mode.
Last edited:
Thanks for the quick answer Bonsai.
Do you know/have estimated the 'transition point' at which the amp goes from 'A' to 'AB' in the kx?
Anyway, I'll order the boards from Jims (have already ordered 2 by 'ripple eaters' a week back - so I'll use one per channel and have one spare).
Do you know/have estimated the 'transition point' at which the amp goes from 'A' to 'AB' in the kx?
Anyway, I'll order the boards from Jims (have already ordered 2 by 'ripple eaters' a week back - so I'll use one per channel and have one spare).
Hi, still running my splendid SX amp. Considering moving to the KX next year.
Is it advisable to implement speaker protection on these amps or are they that "bulletproof" 🙂
I have not in my SX.
Sincerely
Is it advisable to implement speaker protection on these amps or are they that "bulletproof" 🙂
I have not in my SX.
Sincerely
@ polsol are the psu boards supplied with the kx amp boards not different from the ripple eater boards sold seperately. I know that Jim's audio do sell the kx psu pcb on its own as I have just ordered one as I intend to make monoblocks at some point. It is an excellent sounding amp with just the one psu board.
Looking at the pics on Jim's both boards (sold separately and with the kx) are named as 'Multiplier Cap PSU' so it seems they are the same. A quick look seems to confirm that but as the boards are inverted between the two items so it's a little difficult to compare visually.
There might be component differences though.
Bonsai will probably confirm your question.
