A trick is that if you bake the photo resist at 120°C, they will become solvent resistant.
Essentially becoming solder resist.
www.diyaudio.com/community/threads/new-passdiy-headphone-amp-now-available.271662/post-4602517
Patrick
Essentially becoming solder resist.
www.diyaudio.com/community/threads/new-passdiy-headphone-amp-now-available.271662/post-4602517
Patrick
Last edited:
But it is now Vgs + bias x Rsource.
And the optocoupler needs feedback connection from one of the Rsource, not ?
You need to draw it out, all 3 in parallel, then you will see.
Of course it will be easier to use the no-Rsource bias control and still be able to use three in parallel.
www.diyaudio.com/community/threads/the-f5pi-f5p-voltage-gain-stage-m2-output-stage-perfectly-integrated.406024/post-7686147
When paralleling, you definitely can't leave out Rsource.
Patrick
And the optocoupler needs feedback connection from one of the Rsource, not ?
You need to draw it out, all 3 in parallel, then you will see.
Of course it will be easier to use the no-Rsource bias control and still be able to use three in parallel.
www.diyaudio.com/community/threads/the-f5pi-f5p-voltage-gain-stage-m2-output-stage-perfectly-integrated.406024/post-7686147
When paralleling, you definitely can't leave out Rsource.
Patrick
Last edited:
back tp topic:
i dont get it because with 2 MOSFET each side it is working . the opto and the sensing is working.
hmmm...
have to read again...
i dont get it because with 2 MOSFET each side it is working . the opto and the sensing is working.
hmmm...
have to read again...
He shorted R13,14 on the main PCB.
Because you have multiple MOSFETs in parallel, they each have to have their own R13,14 to prevent current hogging.
But now you need to take one of those to allow the opto to sense.
And of course each new R13,14 should be 3x their previous values.
Patrick
Because you have multiple MOSFETs in parallel, they each have to have their own R13,14 to prevent current hogging.
But now you need to take one of those to allow the opto to sense.
And of course each new R13,14 should be 3x their previous values.
Patrick
Last edited:
You need to watch how you connect both ends of those source resistors to the main board.
Shorting R13,14 there is a bad idea.
But of course using the circuit linked above is a much better solution.
www.diyaudio.com/community/threads/the-f5pi-f5p-voltage-gain-stage-m2-output-stage-perfectly-integrated.406024/post-7686147
Just need to add a frontend board of your own.
Patrick
Shorting R13,14 there is a bad idea.
But of course using the circuit linked above is a much better solution.
www.diyaudio.com/community/threads/the-f5pi-f5p-voltage-gain-stage-m2-output-stage-perfectly-integrated.406024/post-7686147
Just need to add a frontend board of your own.
Patrick
I know it's a small overkill...but sounds great ....on opa 445..put it down to 18.9 V rails.....
@Dan K this opa445 is a good one...and can do more rail voltage +-45V
@Dan K this opa445 is a good one...and can do more rail voltage +-45V
Last edited:
Opamp compare
I finally close this year with experiences of mine with opamp rolling on different M2OPS with different output MOSFETS which I tried.
These findings are supjective and include no overall „truth“ ….
Pre amp is always Wayne with Toshiba´s as output BJT Transistors…….
With the combo Opa828 the system have a lot of deep low end bass. Bass sound like a tsunami and rolling over the floor to the listener…..impressive…but missing upper bass sometimes.
Also this combination (Toshiba and opa828) has a great separation in the far back and sounds dark and lay back and slitghtly mid accentuated.
A change to opa1611 has not so much resolution in the highs and sounds a bit faster. Maceo Parkers-childrens world remastered give not better sax truth in the mids.
OPA1655—one a my very good alternatives---more detection of all actors with a very good bass in the low end and more upper bass…..ultra high Resolution is less as opa28.
LT1793 – highs very fine like opa828 but not so „warm“…mid slightly forewarding as opa828 when loud.-more bass punch…funny guy!
Opa827 – very close to opa828 and he loves to play mid´s, mids not so in front as 828
Opa1641- OK but not very impressive –stage ist too close together and mix it up while it gets complex in sound mix and action –„too cool and lovelyness“
TL071CP -OK and much better as opa1641—soundstage smaller, less low end, high´s less resolution as top competitors.
OPA627BP -yes I got 4 of them, bought 2014…..€18 each—today lot more….
beyond any doubt ……the best one….super high resolution, more detection of actors and nothing is accentuated….
So the upper mentioned sound charateristics are the same as above but lets point out some differences in detail…..
Opa1655 compared with opa828
e.g. Marceo Parkers-children´s world –I like the mids more with 1655 more „orange-gold“ feeling with the saxo.—bass less attack softer highs.
Opa1611 and LT1793 close together—LT1793 mid´s are close to 1655 but soundstage is smaller..1611 always layback
Opa1641 –soundstage very small, detection difficult…..OK
TL071CP -better as 1641 more space as 1641----in general less resolution in the mids….
Opa827- top end seems louder as 1655 but with more resolution
OPA445AP (+-45V guy)---next real M2OPS class A/B project!!
Very good detection, thinner bass in very low end, but very controlled (more defined), which is shown very impressive in George Bensons´s- Rainy night in Georgia. Nothing is lost in the bass low end mudd here.
I think in generell pre amps could have more influence in sound..
wish you all the best for 2025
I finally close this year with experiences of mine with opamp rolling on different M2OPS with different output MOSFETS which I tried.
These findings are supjective and include no overall „truth“ ….
Pre amp is always Wayne with Toshiba´s as output BJT Transistors…….
M2OPS with curve traced Toshibas from EUVL(XEN audio-Patrick)
With the combo Opa828 the system have a lot of deep low end bass. Bass sound like a tsunami and rolling over the floor to the listener…..impressive…but missing upper bass sometimes.
Also this combination (Toshiba and opa828) has a great separation in the far back and sounds dark and lay back and slitghtly mid accentuated.
A change to opa1611 has not so much resolution in the highs and sounds a bit faster. Maceo Parkers-childrens world remastered give not better sax truth in the mids.
OPA1655—one a my very good alternatives---more detection of all actors with a very good bass in the low end and more upper bass…..ultra high Resolution is less as opa28.
LT1793 – highs very fine like opa828 but not so „warm“…mid slightly forewarding as opa828 when loud.-more bass punch…funny guy!
Opa827 – very close to opa828 and he loves to play mid´s, mids not so in front as 828
Opa1641- OK but not very impressive –stage ist too close together and mix it up while it gets complex in sound mix and action –„too cool and lovelyness“
TL071CP -OK and much better as opa1641—soundstage smaller, less low end, high´s less resolution as top competitors.
OPA627BP -yes I got 4 of them, bought 2014…..€18 each—today lot more….
beyond any doubt ……the best one….super high resolution, more detection of actors and nothing is accentuated….
M2OPS with Fairchilds (FQA10N80/FQA9P25)
Comparing with the output Toshiba MOSFETS Fairchild have (in my world) always more airiness in the highs and more upper bass as the Toshibas and slightly sounds faster or say with more dynamic. Not so lay back and dark in the far far back as Toshiba´s.So the upper mentioned sound charateristics are the same as above but lets point out some differences in detail…..
Opa1655 compared with opa828
e.g. Marceo Parkers-children´s world –I like the mids more with 1655 more „orange-gold“ feeling with the saxo.—bass less attack softer highs.
Opa1611 and LT1793 close together—LT1793 mid´s are close to 1655 but soundstage is smaller..1611 always layback
Opa1641 –soundstage very small, detection difficult…..OK
TL071CP -better as 1641 more space as 1641----in general less resolution in the mids….
Opa827- top end seems louder as 1655 but with more resolution
OPA445AP (+-45V guy)---next real M2OPS class A/B project!!
Very good detection, thinner bass in very low end, but very controlled (more defined), which is shown very impressive in George Bensons´s- Rainy night in Georgia. Nothing is lost in the bass low end mudd here.
I think in generell pre amps could have more influence in sound..
wish you all the best for 2025
Last edited:
The Toshiba K3497/J618 are high transconductance, but also somewhat high capacitance, compared to the Fairchild's.
So no surprise you find it darker.
They should go with fast frontend capable of driving some load.
If you want to compare apples to apples, use K1530/J201.
And you should give ADA4627 a go.
Supposed to be same as OPA627.
Can even try 637, since gain is high enough.
Happy New Year,
Patrick
So no surprise you find it darker.
They should go with fast frontend capable of driving some load.
If you want to compare apples to apples, use K1530/J201.
And you should give ADA4627 a go.
Supposed to be same as OPA627.
Can even try 637, since gain is high enough.
Happy New Year,
Patrick
Last edited:
Hello,
for a M2OPS Class AB with Trafo sec. 24VAC (32Vdc) project, what MOSFET are the most suitable for this task....and what kind of idle current make sence...
Opamp will be OPA445...
greeeeetings 😀
for a M2OPS Class AB with Trafo sec. 24VAC (32Vdc) project, what MOSFET are the most suitable for this task....and what kind of idle current make sence...
Opamp will be OPA445...
greeeeetings 😀
...make sense....
guessing that 2SK1530/J201 could be the best one, due the higher Gate voltage swing and lower Ciss on gate.....
What I tested on M2OPS at 4R there is also lower RdsON, because more power was possible at 4R measurements....
guessing that 2SK1530/J201 could be the best one, due the higher Gate voltage swing and lower Ciss on gate.....
What I tested on M2OPS at 4R there is also lower RdsON, because more power was possible at 4R measurements....
Hey everyone, I wanted to share my exciting journey with the M2OPS. You know, for the past year, I've been building, modifying, and testing the 'Holy Grail' with Fairchild's FQA28N15/FQA36P15 – many thanks to lhquam for this awesome project!
Now, it was finally time to try out the Toshiba 2SK3497/2SJ618, which were originally intended for the M2OPS and had been patiently waiting in my drawer for their turn. And guess what? Despite their high positive tempCo, they were working in this circuit. However, the measurement results were slightly worse than the Fairchilds, and the sound didn't quite convince me personally (https://www.diyaudio.com/community/...ower-output-stage.406850/page-38#post-7777396).
Also, I experimented with an OPA604, which unfortunately wasn't so successful. There was always a noticeable ringing that I just couldn't get rid of.
But here's the cool part: Two M2OPS boards have taken the place of the HGF in my monoblock chassis. The power supply remains the same at +/- 23.5 V. With a bias current of 1.48 A, the offset stabilized after 20 minutes from 22 mV at 20°C to +/- 1 mV at 45°C.
The measurement results are simply outstanding, and the OPA604 feels right at home in this environment! The test with a 100 kHz square wave is almost perfect. And the response to a complex load of 8R || 100 nF is no different from a purely resistive 8R. The small waving becomes invisible at 10 kHz test frequency 😉. The THD of 0.0027% at 1 W into 8R and the harmonic distribution match the predictions in Fran's test pattern perfectly.
Compared to the HGF, this means that the Toshiba MOSFETs perform better when source resistors are used. Why is that? Good question! But it's clear that the source resistors help to linearize the transfer characteristics of the MOSFETs.
At this point, I'd like to extend a huge thank you to Patrick (EUVL) and everyone else who made this project possible.
Yes, I really enjoyed this journey!
Now, it was finally time to try out the Toshiba 2SK3497/2SJ618, which were originally intended for the M2OPS and had been patiently waiting in my drawer for their turn. And guess what? Despite their high positive tempCo, they were working in this circuit. However, the measurement results were slightly worse than the Fairchilds, and the sound didn't quite convince me personally (https://www.diyaudio.com/community/...ower-output-stage.406850/page-38#post-7777396).
Also, I experimented with an OPA604, which unfortunately wasn't so successful. There was always a noticeable ringing that I just couldn't get rid of.
But here's the cool part: Two M2OPS boards have taken the place of the HGF in my monoblock chassis. The power supply remains the same at +/- 23.5 V. With a bias current of 1.48 A, the offset stabilized after 20 minutes from 22 mV at 20°C to +/- 1 mV at 45°C.
The measurement results are simply outstanding, and the OPA604 feels right at home in this environment! The test with a 100 kHz square wave is almost perfect. And the response to a complex load of 8R || 100 nF is no different from a purely resistive 8R. The small waving becomes invisible at 10 kHz test frequency 😉. The THD of 0.0027% at 1 W into 8R and the harmonic distribution match the predictions in Fran's test pattern perfectly.
Compared to the HGF, this means that the Toshiba MOSFETs perform better when source resistors are used. Why is that? Good question! But it's clear that the source resistors help to linearize the transfer characteristics of the MOSFETs.
At this point, I'd like to extend a huge thank you to Patrick (EUVL) and everyone else who made this project possible.
Yes, I really enjoyed this journey!
