I disconnected one of the zeners because I also didn't trust them but I couldn't see a difference. I'll try to experiment a bit more with Spice.
But at this moment I'm enjoying the amp with the New headphones and a Laphroaig. So me happy.
But at this moment I'm enjoying the amp with the New headphones and a Laphroaig. So me happy.
BZX84C6V2 is 200p, so 100p for two in series.
That together with 22k is 72kHz.
There is also no reason why you need to use 220k/22k for feedback, and not a lot less.
Patrick
That together with 22k is 72kHz.
There is also no reason why you need to use 220k/22k for feedback, and not a lot less.
Patrick
Mmm. I used a different zener from Nxp, but this one also has 200pf.
The high resistorvalues I used to have a symetric differential pair as advised by Zen Mod.
The high resistorvalues I used to have a symetric differential pair as advised by Zen Mod.
You can use 1k / 10k on both legs as far as I am concerned.
Any modern day source will have no problems driving a 11k input impedance.
But whatever you are happy with.
Patrick
Any modern day source will have no problems driving a 11k input impedance.
But whatever you are happy with.
Patrick
Actually balanced impedance on diff pair legs are far less critical for JFETs then BJTs.
But you know already that I am sure.
Patrick
But you know already that I am sure.
Patrick
Using the new headphones I have a problem that can be fixed with these resistorvalues but I'm a afraid to do so.
These headphones have a far higher sensitivity and don't need a lot of gain. The Alps is only a few degrees open. So I need a way to lower the gain big time. But of course not fixed. I won't be selling the Sennheiser.
Of course I can change the feedback, but I'm afraid the soundstage will get smaller. The tubeamp still beats the ss amp even with the 30 ohm Philips. Anyone experience with changing the gain and/or feedback?
These headphones have a far higher sensitivity and don't need a lot of gain. The Alps is only a few degrees open. So I need a way to lower the gain big time. But of course not fixed. I won't be selling the Sennheiser.
Of course I can change the feedback, but I'm afraid the soundstage will get smaller. The tubeamp still beats the ss amp even with the 30 ohm Philips. Anyone experience with changing the gain and/or feedback?
Last edited:
you can try with 10k/10k instead of 22k/220k , then decide
there are always few more tricks to decrease gain , while keeping nfb in same bracket ..... if needed
there are always few more tricks to decrease gain , while keeping nfb in same bracket ..... if needed
Mmm, there are a lot of headamps being designed on the Pass forum. I see a trend 😛
But back to my amp. I've been enjoying mine for a couple of weeks now with the new Philips headphone. This headphone is a lot better then the old Sennheiser, but the sound is not perfect: it is a bit "sharp". To find out if it is the headphone or the amp, I dug up an old amp I build some 20 years ago. An OP27 for the voltage gain and a BD139/140 Class A push pull pair. Elektor design.
This Elektor amp has an even fatter bass, but nowhere near the control of the Pass. Bass has no depth or details with these Opamps. Also the soundstage of the Pass is way, way better. But the highs are a bit softer on the Elektor design. Sailing to Philadelphia (Knopfler) is Sailing to Philadelphia. On the Pass it is SSSSailing to Philadelphia. If you know what I mean. You can hear the strong S from Knopfler on the Elektor design, but it is not too much. It can be on the Pass. Also the Pass sometimes sounds like it is walking on its toes (close to clipping). It doesn't matter if I'm listening loud or not. My CD player has a DAC with integrated attenuator so I'm not overloading the input.
Is this sound signature part of the higher resolution or shouldn't this be the case? I know the headphones are a bit on the edgy side but not this much. On the scope everything measures fine and as simulated in Spice. The maximum output is as expected. So the Aleph gain is about 0.5.
There is one thing I'm not certain about. The rails are at 13.5 Volts. The Idss of the Toshiba's is around 7.5 ma at this voltage. The CCS is at 8.5. Devided by 2 gives 4.25 per Jfet. Maybe a bit high.
Also the Fairchilds have a very high Vgs. For a 180 ma current I need a Vgs of almost 6 volts. This means the Jfet resistor is a lot higher than 1k (probably close to 1.5k). Can this give an imbalance in the Jfet pair?
But back to my amp. I've been enjoying mine for a couple of weeks now with the new Philips headphone. This headphone is a lot better then the old Sennheiser, but the sound is not perfect: it is a bit "sharp". To find out if it is the headphone or the amp, I dug up an old amp I build some 20 years ago. An OP27 for the voltage gain and a BD139/140 Class A push pull pair. Elektor design.
This Elektor amp has an even fatter bass, but nowhere near the control of the Pass. Bass has no depth or details with these Opamps. Also the soundstage of the Pass is way, way better. But the highs are a bit softer on the Elektor design. Sailing to Philadelphia (Knopfler) is Sailing to Philadelphia. On the Pass it is SSSSailing to Philadelphia. If you know what I mean. You can hear the strong S from Knopfler on the Elektor design, but it is not too much. It can be on the Pass. Also the Pass sometimes sounds like it is walking on its toes (close to clipping). It doesn't matter if I'm listening loud or not. My CD player has a DAC with integrated attenuator so I'm not overloading the input.
Is this sound signature part of the higher resolution or shouldn't this be the case? I know the headphones are a bit on the edgy side but not this much. On the scope everything measures fine and as simulated in Spice. The maximum output is as expected. So the Aleph gain is about 0.5.
There is one thing I'm not certain about. The rails are at 13.5 Volts. The Idss of the Toshiba's is around 7.5 ma at this voltage. The CCS is at 8.5. Devided by 2 gives 4.25 per Jfet. Maybe a bit high.
Also the Fairchilds have a very high Vgs. For a 180 ma current I need a Vgs of almost 6 volts. This means the Jfet resistor is a lot higher than 1k (probably close to 1.5k). Can this give an imbalance in the Jfet pair?
Last edited:
you can try even higher current through input JFets , up to 90% of Idss
that will demand lower drain resistors
also - try to implement some source resistors for JFets , thus introducing some degeneration , with goal to decrease OLG , so amount of NFB , which you already have plenty
that will demand lower drain resistors
also - try to implement some source resistors for JFets , thus introducing some degeneration , with goal to decrease OLG , so amount of NFB , which you already have plenty
Hi Zen Mod, thanks for the reply.
I think I have seen this degeneration sollution on the babelfish tread. I'm not sure if I have any space left to fit in a couple of resistors (the whole PCB is a bit tiny).
Doesn't increasing the current to 90% of the Idss make the amp more sensitive for clipping? You only have 10% headroom left.
I think I have seen this degeneration sollution on the babelfish tread. I'm not sure if I have any space left to fit in a couple of resistors (the whole PCB is a bit tiny).
Doesn't increasing the current to 90% of the Idss make the amp more sensitive for clipping? You only have 10% headroom left.
if you observe it with scope , there is no exact clipping
you likely have problems associated with sort of excessive feedback , needing some time to try sweet spot for OLG
though - ref to sch in you first post - how you chose 0R12 for output resistor ?
if you want PapAleph operation , you need to maintain similar ratio of output and source resistances
this way , you're close to pure SE operation
also - some fiddling with gate resistors of outputs can tame things a little
you likely have problems associated with sort of excessive feedback , needing some time to try sweet spot for OLG
though - ref to sch in you first post - how you chose 0R12 for output resistor ?
if you want PapAleph operation , you need to maintain similar ratio of output and source resistances
this way , you're close to pure SE operation
also - some fiddling with gate resistors of outputs can tame things a little
The version I'm using right now is in post 26. The output resistor is 0.47 ohm and the source resistors are 2.2 ohm.
This way the damping is high. The ratio between this resistor and the Aleph CS (R8) is 10. This gives a nice working Aleph CS at 30 ohms.
BTW: by OLG you mean: Open Loop Gain?
This way the damping is high. The ratio between this resistor and the Aleph CS (R8) is 10. This gives a nice working Aleph CS at 30 ohms.
BTW: by OLG you mean: Open Loop Gain?
Last edited:
Just a brain wave: I didn't use a NFB cap. The gate of the second Jfet is connected to ground with just a 22k resistor. I read somewhere (from Nelson himself) this cap isn't needed. It only helps to have less temperature drift.
The amps does drift from cold to warm (from 180 mV DC to 10 mv DC). This is not an issue for the headphones, because I wait a few minutes before I plug them in. Most of the DC is gone in a minute or so.
Can this skipping of the NFB cap also cause strange behaviour of the Jfets? And cause an edgy sound? 😕
The amps does drift from cold to warm (from 180 mV DC to 10 mv DC). This is not an issue for the headphones, because I wait a few minutes before I plug them in. Most of the DC is gone in a minute or so.
Can this skipping of the NFB cap also cause strange behaviour of the Jfets? And cause an edgy sound? 😕
Hi,
Yes, I've done some things with the amp. But I have to think. It was a long time ago, and I've been on vacation (always functions like a reset 😀 ).
Some 15 years ago I build a simple opamp preamp. I decided to slaughter it and use the metal case and some other stuff. The attenuator was replaced by an old Noble balance potentiometer. A simple resistor made this lineair pot a bit more log.
The original preamp input selector is still in place (so I've a headphone amp with 4 inputs).
The rec-out selector was completely rewired. It is a 4 pole, 3 position switch.
Position 1 = no extra input or output resistor. So 20 dB gain and very low output resistance.
Position 2 = 20 dB input divider and no output resistor. So no voltage gain.
Position 3 = no input divider and 120 ohm (or something) output resistor. This works fine for the old Sennheiser.
The nfb is not changed. Also the power supply is the same. For some reason the hars SSS is gone. Couple of possible reasons:
* the headphone has burned in
* my ears got used to it
* the Noble pot is a better than the Alps
* wiring is a lot better
* I've added a couple of MKP's to the power supply
Most likely it is a combination of all. I do know I like this amp a lot and it is my daily driver. I don't use the previous Tube amp anymore. This amp has a nice bass, very good soundstage and a lot of slam.
Yes, I've done some things with the amp. But I have to think. It was a long time ago, and I've been on vacation (always functions like a reset 😀 ).
Some 15 years ago I build a simple opamp preamp. I decided to slaughter it and use the metal case and some other stuff. The attenuator was replaced by an old Noble balance potentiometer. A simple resistor made this lineair pot a bit more log.
The original preamp input selector is still in place (so I've a headphone amp with 4 inputs).
The rec-out selector was completely rewired. It is a 4 pole, 3 position switch.
Position 1 = no extra input or output resistor. So 20 dB gain and very low output resistance.
Position 2 = 20 dB input divider and no output resistor. So no voltage gain.
Position 3 = no input divider and 120 ohm (or something) output resistor. This works fine for the old Sennheiser.
The nfb is not changed. Also the power supply is the same. For some reason the hars SSS is gone. Couple of possible reasons:
* the headphone has burned in
* my ears got used to it
* the Noble pot is a better than the Alps
* wiring is a lot better
* I've added a couple of MKP's to the power supply
Most likely it is a combination of all. I do know I like this amp a lot and it is my daily driver. I don't use the previous Tube amp anymore. This amp has a nice bass, very good soundstage and a lot of slam.
Last edited:
Maybe I should make a new Thread. The title of this one doesn't cover the current setup. The Jfet is different and the powerfets too.
But I don't want to "pollute" this forum by making to much Threads.
If someone thinks it is an addition to this forum, I will start a clean Thread with the latest status. I do believe this amp is quite OK. It's also stable (haven't changed anything in the last couple of months).
But I don't want to "pollute" this forum by making to much Threads.
If someone thinks it is an addition to this forum, I will start a clean Thread with the latest status. I do believe this amp is quite OK. It's also stable (haven't changed anything in the last couple of months).
- Status
- Not open for further replies.