I own an XA30.8 and noticed that the gate resistors for the output devices seem quite high in value (don't recall the exact value - would need to open the amp up to check, but I remember they seemed high, maybe several 100 ohm each).
Since I have previous experience with lowering gate resistors which seemed high in value with favourable results (significantly more transparent sound), I was wondering if I should do this here too. Is there anybody here who has tried this. on any recent Pass amps?
Please share your experiences!
Thanks,
Alexander
Since I have previous experience with lowering gate resistors which seemed high in value with favourable results (significantly more transparent sound), I was wondering if I should do this here too. Is there anybody here who has tried this. on any recent Pass amps?
Please share your experiences!
Thanks,
Alexander
if they look original/not messed with, I wouldn't touch that
I've simulated circuits that needed several hundred ohms or more. I'd be surprised that sound quality would be affected by reducing values, unless stability is reduced--probably not a good thing.
in circuits with low OLG, without or with small amount of NFB, change of value of gate resistors can be perceived in some amount
with circuits having hefty (enough) amount of NFB, hardly
with circuits having hefty (enough) amount of NFB, hardly
Interesting responses. Indeed, the amp that responded well had super low nfb. Does this insight follow from simulations?
Also, I didn't realise that several 100s ohms might be required for stability. I had assumed that 50 -100 ohm would be enough to surpress any parasitic oscillations.
Also, I didn't realise that several 100s ohms might be required for stability. I had assumed that 50 -100 ohm would be enough to surpress any parasitic oscillations.
I don't see the mechanism where lowering gate impedance affects signal quality unless it starts to degrade stability, but your results seem to track mighty ZM's sage advice, which I would be a fool to challenge. 
I suppose there's a HF pole there, but would that affect the audio band? 🤨
I would have assumed the same about minimum values, and I'm not sure I've encountered that in real test circuits, but I have in LTspice.
I suppose there's a HF pole there, but would that affect the audio band? 🤨I would have assumed the same about minimum values, and I'm not sure I've encountered that in real test circuits, but I have in LTspice.
MOSFET gates need to be charged and discharged with dynamic signal.
So there are current going through them.
The gate capacitances form a low pass filter with the gate stopper resistor.
But depending on the device, those capacitances have varying degrees of non-linearities.
Increasing gate resistors have 2 consequences :
a) it reduces open loop bandwidth and therefore improves stability, especially with reactive load;
b) it increases distortion due to the above mentioned non-linear capacitances, especially at high frequencies.
Gate stopper resistors up to 1k ohm is not out of this world.
They may well be needed for stability.
But that can come at the cost of distortion.
Unless you have devices with linear capacitances, like lateral FETs.
My 2 cents,
Patrick
So there are current going through them.
The gate capacitances form a low pass filter with the gate stopper resistor.
But depending on the device, those capacitances have varying degrees of non-linearities.
Increasing gate resistors have 2 consequences :
a) it reduces open loop bandwidth and therefore improves stability, especially with reactive load;
b) it increases distortion due to the above mentioned non-linear capacitances, especially at high frequencies.
Gate stopper resistors up to 1k ohm is not out of this world.
They may well be needed for stability.
But that can come at the cost of distortion.
Unless you have devices with linear capacitances, like lateral FETs.
My 2 cents,
Patrick
What they did not say is that you can achieve the same corner frequency with a lower value gate resistor by adding an extra capacitor either between gate / source or gate / drain.
The advantage is that this extra capacitor is usually much more linear than those inside the MOSFET.
Patrick
The advantage is that this extra capacitor is usually much more linear than those inside the MOSFET.
Patrick
What EUVL said! Funny is that the amp I mentioned used lateral mosfets...
I hope that noone is using gate stoppers to set the dominant pole though. I am a bit rusty here, but wouldn't increasing the corner frequency of a secondary pole (effect of gate resistors) have the effect of stabilising the amp?
IIRC, the gate stoppers are solely used to dampen parasitic oscillations. I have never heard (and would not consider) the use of gate resistors to set a deliberate pole in the open loop function.
What brought up memories was the suggestion in the article to use different resistor values for the P and N devices. We used to do this some 30 years ago already. I now wonder if Pass is doing this..
I hope that noone is using gate stoppers to set the dominant pole though. I am a bit rusty here, but wouldn't increasing the corner frequency of a secondary pole (effect of gate resistors) have the effect of stabilising the amp?
IIRC, the gate stoppers are solely used to dampen parasitic oscillations. I have never heard (and would not consider) the use of gate resistors to set a deliberate pole in the open loop function.
What brought up memories was the suggestion in the article to use different resistor values for the P and N devices. We used to do this some 30 years ago already. I now wonder if Pass is doing this..
The XA30.8 is a sweet 'little' amp. It doesn't leave a lot of people longing for a more transparent sound. The highs are pretty darned detailed on it.
If your amp is new, give it time to work in. It takes a while.
If your amp is new, give it time to work in. It takes a while.
As far as I can tell from photos and from older threads on the XA30.8, the devices used are :
2SK170/2SJ74
2SK2013 / 2SJ313
2SC4793 / 2SA1837
IRFP240 / 9240
None of the above are laterals ??
Cheers,
Patrick
2SK170/2SJ74
2SK2013 / 2SJ313
2SC4793 / 2SA1837
IRFP240 / 9240
None of the above are laterals ??
Cheers,
Patrick
No laterals in the xa30.8. I meant the other amp in which I changed the gate resistors with good result.
That amp sucked btw. It had low feedback, low linearity, little psrr/cmrr and I could tell.
I now wonder why anyone ever thought it was worth ging that route. You can search for Plantefeve. Some of the info is still up.
That amp sucked btw. It had low feedback, low linearity, little psrr/cmrr and I could tell.
I now wonder why anyone ever thought it was worth ging that route. You can search for Plantefeve. Some of the info is still up.
I had already concluded this. Consider this thread a trauma response from the other amp 😄
Planeteve's site no longer exists.
And the XA30.8 has 8~18dB NFB, according to Nelson.
About -80dB THD @ 1W1kHz, also not the lowest around.
Doesn't say anything about the sound though.
Patrick
And the XA30.8 has 8~18dB NFB, according to Nelson.
About -80dB THD @ 1W1kHz, also not the lowest around.
Doesn't say anything about the sound though.
Patrick