Hi,
Like many others here, I'm wondering what would be the best way to keep the AX idle, but a little warm. Yes, I've read that Nelson switches the current sources on and off, both for the front end and the power stages, but couldn't find how it is performed practically
.
My goal is to keep front end diff pair in operation all the time, and to reduce the current delivered by the Aleph current sources...
One way I found to do so (using simulation) was to increase the resistor between the power mosfets and the base of the controlling BJT, as show on the attached schematic excerpt (R16a and R16b). The additionnal resistor is shorted by a relay when the amp is turned on. Simulations show that the current is divided by 3 with the values shown, and increasing R16a lowers the current very easily.
But the question I want to ask the AX gurus is : Is it worth trying, or am I totally wrong ?
Any hints greatly appreciated
Like many others here, I'm wondering what would be the best way to keep the AX idle, but a little warm. Yes, I've read that Nelson switches the current sources on and off, both for the front end and the power stages, but couldn't find how it is performed practically
. My goal is to keep front end diff pair in operation all the time, and to reduce the current delivered by the Aleph current sources...
One way I found to do so (using simulation) was to increase the resistor between the power mosfets and the base of the controlling BJT, as show on the attached schematic excerpt (R16a and R16b). The additionnal resistor is shorted by a relay when the amp is turned on. Simulations show that the current is divided by 3 with the values shown, and increasing R16a lowers the current very easily.
But the question I want to ask the AX gurus is : Is it worth trying, or am I totally wrong ?
Any hints greatly appreciated
Attachments
No . You can use relay to parallel a fixed resistor on the summ VR2/ R17 to get you a smaller value , on each side of the amp . This will lower your bias in the 2 current sources and will leave the diff pair bias intact
It is simple: lowering the value lowers bias . Increasing the value cranks up the bias .
try values to get very low bias on the output sections , and don't forget the relay must parallel the additional resistor SAME TIME on both sides , otherwise problems . And if you use other relay to disconnect speakers little time before the bias change , it can do no harm to your speakers
Regards
Anael
Ps if you want to do like Nelson , you must have extra capacitors on the circuit CCS diff so to ensure the 3 CCS cut sama time independant of how much the draw . Calculating the exact value is difficult but doable. This way you get no turn off thumps
It is simple: lowering the value lowers bias . Increasing the value cranks up the bias .
try values to get very low bias on the output sections , and don't forget the relay must parallel the additional resistor SAME TIME on both sides , otherwise problems . And if you use other relay to disconnect speakers little time before the bias change , it can do no harm to your speakers
Regards
Anael
Ps if you want to do like Nelson , you must have extra capacitors on the circuit CCS diff so to ensure the 3 CCS cut sama time independant of how much the draw . Calculating the exact value is difficult but doable. This way you get no turn off thumps
Hi Anaël,
Thanks for answering.
Yep, it was intended to be done on both sides of the amp. And of course, the speaker relay is planned, 'cause with a reduced bias, although the diff out offset is still near 0V, I want to be sure a small delay between relays switching (one per "side" in my implementation) won't harm my speakers, since the "same time" is only theoretical with relays.
Nevertheless, your solution is much better than mine. Have you tried it already ?
Thanks again.
Thanks for answering.
Yep, it was intended to be done on both sides of the amp. And of course, the speaker relay is planned, 'cause with a reduced bias, although the diff out offset is still near 0V, I want to be sure a small delay between relays switching (one per "side" in my implementation) won't harm my speakers, since the "same time" is only theoretical with relays
.Nevertheless, your solution is much better than mine. Have you tried it already ?
Thanks again.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.