I am about to finisch my amp, have already completed the 2 channels, all of it Amp,supply and servo is on 1 pcb/channel.
Wat worries me is that the hights or not as clear as i suspected, the rest off the amp sound very naturally and overall great.
( without the servo it sounds better but thats not an option )
or there any things i could try to impove it before i start building it into its enclosure ( witch arives in 4 weeks
).
What i did not like about it in the first place was the DC offset without the servo, is there any way to to something about this, becouse of this its needs about 10sec to stabilize the DC offset to les then 100mV. ( i allready matched the diff pairs and the resistors, this helped but not that mutch as i hoped )
Rudy
Wat worries me is that the hights or not as clear as i suspected, the rest off the amp sound very naturally and overall great.
( without the servo it sounds better but thats not an option )
or there any things i could try to impove it before i start building it into its enclosure ( witch arives in 4 weeks
).An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
What i did not like about it in the first place was the DC offset without the servo, is there any way to to something about this, becouse of this its needs about 10sec to stabilize the DC offset to les then 100mV. ( i allready matched the diff pairs and the resistors, this helped but not that mutch as i hoped )
Rudy
I would say, that C8 with 1n is quite big, i prefer 100p.
The same might be with C10.
You might work without the servo, if you put a big cap in series
with r25 (220-470uf ?)
But if the DC is that big, you should try to adjust the DC somewhere
else, maybe by adjusting r13 (with a pot ?)
What i don't really understand is the biasvoltage for T8,
Its not clear to me, what the voltage is. If it's 20Volt, this would give
~1 Amp through T8.
For good DC, its important that T7/T8 gets the same "biasvoltage".
This means if current through R9 is for example 2ma, you get a
bias of ~0.94 Volt, which results in a current through T7 of ~18ma.
The same settings should be for T8.
But for a current that big in the VAS you should add a buffer to T7/T8.
Mike
The same might be with C10.
You might work without the servo, if you put a big cap in series
with r25 (220-470uf ?)
But if the DC is that big, you should try to adjust the DC somewhere
else, maybe by adjusting r13 (with a pot ?)
What i don't really understand is the biasvoltage for T8,
Its not clear to me, what the voltage is. If it's 20Volt, this would give
~1 Amp through T8.
For good DC, its important that T7/T8 gets the same "biasvoltage".
This means if current through R9 is for example 2ma, you get a
bias of ~0.94 Volt, which results in a current through T7 of ~18ma.
The same settings should be for T8.
But for a current that big in the VAS you should add a buffer to T7/T8.
Mike
@Upupa Epops
Any suggestions on the vas current, any other disadvantages accept for the currentdraw from the diff pair.
@MikeB
Servo is not an option, as this is my third amp, and the other 2 or without servo, it was time i started experimenting the servo.
"If" i want to insert a pot, it defenetly would not be there, as with that resister the diff pair current is set.
The current in T8 is not set by that zenerdiode, but with the CCS of the diff pair. ( VAS current is 30mA ) ( total diff pair current is 5,5mA )
could the DC problem be coused by the huge difference from R5 and R25 ?
Any suggestions on the vas current, any other disadvantages accept for the currentdraw from the diff pair.
@MikeB
Servo is not an option, as this is my third amp, and the other 2 or without servo, it was time i started experimenting the servo.
"If" i want to insert a pot, it defenetly would not be there, as with that resister the diff pair current is set.
The current in T8 is not set by that zenerdiode, but with the CCS of the diff pair. ( VAS current is 30mA ) ( total diff pair current is 5,5mA )
could the DC problem be coused by the huge difference from R5 and R25 ?
Mike, don't be that fast > what the voltage is. If it's 20Volt
There is T6
Rudy,
R18 drop is 540mV,
approx. the same drop will be on R13,
as R18=R17, your input differential pair will work asymetrically,
change R9=R10 approx 2*R13
frequency compensation - don't like it and leave it for your simulation
servo - you have three poles servo! R40C7, R39C17, R38C16. Have you simulate group delay/phase shift?
There is T6
Rudy,
R18 drop is 540mV,
approx. the same drop will be on R13,
as R18=R17, your input differential pair will work asymetrically,
change R9=R10 approx 2*R13
frequency compensation - don't like it and leave it for your simulation
servo - you have three poles servo! R40C7, R39C17, R38C16. Have you simulate group delay/phase shift?
"approx. the same drop will be on R13"
not really, as the drop on T8be is bigger then T5be
You may also not forget the current for the VAS pair drawn out off/into the diff pair.
"frequency compensation - don't like it and leave it for your simulation
servo - you have three poles servo! R40C7, R39C17, R38C16. Have you simulate group delay/phase shift?"
I actually dont do simmulations, values or measured from the amp itself.
I could try to find a match on the diffpaircurrents.
not really, as the drop on T8be is bigger then T5be
You may also not forget the current for the VAS pair drawn out off/into the diff pair.
"frequency compensation - don't like it and leave it for your simulation
servo - you have three poles servo! R40C7, R39C17, R38C16. Have you simulate group delay/phase shift?"
I actually dont do simmulations, values or measured from the amp itself.
I could try to find a match on the diffpaircurrents.
Changed C10 to 100p
R17 and R18 changed to 56E for a VAS current of 11mA
but i have had not the time to listen to it yet ( with this current )
at this point the diff pair is perfektly balanced ( 1mV difference R9//R10 )yet i have a dc offset off -6V without the servo ( but i am leaving this like it is )
@dimitri R40C7 is not a pole, R4 prevents this for being one.
/edit .. also changed the zeners
Rudy
R17 and R18 changed to 56E for a VAS current of 11mA
but i have had not the time to listen to it yet ( with this current )
at this point the diff pair is perfektly balanced ( 1mV difference R9//R10 )yet i have a dc offset off -6V without the servo ( but i am leaving this like it is )
@dimitri R40C7 is not a pole, R4 prevents this for being one.
/edit .. also changed the zeners
Rudy
Comments
You do realize that because of the turn on voltages of the output devices that you are only going to be able to approach to within about 10V of the supply rails. That is, the output swing will be limited to about ±32V. This also means that if you play it at higher volume levels, the amp will run hotter than it could.
You either have to run the drivers at a higher supply rail voltage or supply a "bootstrap" setup to give the driver stage enough supply voltage to fully turn on the output devices.
The resistors in the DC sullpy rails between the bridge rectifier and the capacitors will also limit your power. They directly impact your damping factor as they slow down the ability for the filters to recharge on each half cycle of the AC. To get tight bass, you have to have as little impedance as possible between the power transformer and the speaker.
However, this does increase the peak current through the bridge which can cause noise. To prevent this, the usual practice is to put a .01uFd capacitor in parallel with each of the four diodes in the bridge.
R4 is fine. It provides a charge/discharge path for the input capacitor. This prevents a big thump from being passed through to the speaker if you plug or unplug the amplifier from the pre-amp with it switched on. The pole is low enpugh in fequency to have no effect on the sound.
You do realize that because of the turn on voltages of the output devices that you are only going to be able to approach to within about 10V of the supply rails. That is, the output swing will be limited to about ±32V. This also means that if you play it at higher volume levels, the amp will run hotter than it could.
You either have to run the drivers at a higher supply rail voltage or supply a "bootstrap" setup to give the driver stage enough supply voltage to fully turn on the output devices.
The resistors in the DC sullpy rails between the bridge rectifier and the capacitors will also limit your power. They directly impact your damping factor as they slow down the ability for the filters to recharge on each half cycle of the AC. To get tight bass, you have to have as little impedance as possible between the power transformer and the speaker.
However, this does increase the peak current through the bridge which can cause noise. To prevent this, the usual practice is to put a .01uFd capacitor in parallel with each of the four diodes in the bridge.
R4 is fine. It provides a charge/discharge path for the input capacitor. This prevents a big thump from being passed through to the speaker if you plug or unplug the amplifier from the pre-amp with it switched on. The pole is low enpugh in fequency to have no effect on the sound.
the voltage swing is about 35V@8E , witch i find fine
at 4E that will probebly will be a lot less, i know.
about the "bootstrap", that aint so easy with this kind off circuit, other supply was not an option.
about the resisitor ( accept for the fakt that they or gone ), they only limit the continuous power, not very important when just playing music, and they where indeed meant to limit inrush current and bridge noise.
Rudy
at 4E that will probebly will be a lot less, i know.
about the "bootstrap", that aint so easy with this kind off circuit, other supply was not an option.
about the resisitor ( accept for the fakt that they or gone ), they only limit the continuous power, not very important when just playing music, and they where indeed meant to limit inrush current and bridge noise.
Rudy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.