It is all here Lynn.... 🙂
Pass Laboratories XA25
go to manual download.
Nobody knows why this site is the only one to show some details...and not the Pass Labs Site.
Must have a good connection to Nelson.... :--))
Pass Laboratories XA25
go to manual download.
Nobody knows why this site is the only one to show some details...and not the Pass Labs Site.
Must have a good connection to Nelson.... :--))
It is a good custom at Pass Labs to play things down.
The 0.1% figure is nearly standard in the manuals. I think it expresses the will not to take part in the zero distortion numbers race, many other producers show.
The 0.1% figure is nearly standard in the manuals. I think it expresses the will not to take part in the zero distortion numbers race, many other producers show.
Is anyone actually seeing such low distortion measurements in their builds? My simulations show numbers closer to the .1% values. The last of my parts arrive tomorrow and I can build channels and start real measurements.
I did not measure in real the OLG and CLG.
I made a quick check in Spice and killing the 20dB feedback in the frontend and the overall feedback the unit goes to 60dB so the feedback is 40dB and might be to high for the sound.
The low values Nelson shows let me always think if the makes the output stage as common source variant.
And the first two stages only as buffers.
Would this make a very high OLG too?
I made a quick check in Spice and killing the 20dB feedback in the frontend and the overall feedback the unit goes to 60dB so the feedback is 40dB and might be to high for the sound.
The low values Nelson shows let me always think if the makes the output stage as common source variant.
And the first two stages only as buffers.
Would this make a very high OLG too?
All information says the output stage is source-follower, and the front-end has all of the voltage gain. The hocky-puck fets have very high Ciss, but very low Crss, making source follower the best choice. Variants of some of the circuits of the Sony VFET-2 appear to be relevant.
I'm getting basically same specs with around 15dB global feedback (not including feedback in the front end)
Absolutely no chance.
Results indicate source follower as I mentioned either here or other thread a long time ago.
This seems to be a reoccurring theme. Hahaha
Last edited:
You know the reason, the so very low distortion values of Nelson that even Lynn is irritated.
If I dare to say this.
If I dare to say this.
The difference here for me is I'm not even that driven to match the specs. I'm driven to build something that sounds great on my main set of speakers and can be adjusted to sound good on any speakers which will almost certainly mean changing harmonic levels for each speaker.
For me matching specs is just an academic exercise, certainly not the destination (just part of the journey).
In the end I might decide 1 or possibly 2 orders magnitude higher distortion sounds better to me. Hahah
For me matching specs is just an academic exercise, certainly not the destination (just part of the journey).
In the end I might decide 1 or possibly 2 orders magnitude higher distortion sounds better to me. Hahah
Last edited:
First thing.
Plot a distortion curve of just the output stage.
Then it's just a matter of working out what the front end needs to do/contribute to get it where you want.
You may need/want to tweak the output stage also if your devices aren't matched appropriately.
Then go to the toilet many times and ideas will flow. (That works for me)
One other thing, I know for a fact that the P Channel IXYS model is completely wrong at our operating conditions, based on that there is a good chance the N Channel models are also wrong. You need to fix your models to match reality at your operating conditions when you are using zero degeneration, which is not as critical normally when you throw large amounts of source degeneration at devices.
Last edited:
Once you figure it out you will be kicking yourself for not realizing it earlier.
The answer is staring at you in the face.
The answer is staring at you in the face.
One more tip for the spec/cloning zombies. Hahaha 😀
If you look at the slew rate of 100V/uS you can determine the bare minimum bias point used in the second stage to achieve that.
In reality it's probably higher than that based on Papa's conservative nature.
If you look at the slew rate of 100V/uS you can determine the bare minimum bias point used in the second stage to achieve that.
In reality it's probably higher than that based on Papa's conservative nature.
Last edited:
that's easy
err on overkill side
then , if you want to know , having toy in vivo , decrease 'till it clicks
err on overkill side
then , if you want to know , having toy in vivo , decrease 'till it clicks
I am having a hard time reconciling these two specs:
- Slew Rate: 100V/us
- Bandwidth: DC to -2dB @ 100kHz
Frequency response and Slew don't necessarily need to be related not to mention the true Slew Rate Spec is probably higher.
For slew you can determine experimentally. One way is clip the amp with a square wave and measure the leading rising edge of the square. Hope that makes sense.
For slew you can determine experimentally. One way is clip the amp with a square wave and measure the leading rising edge of the square. Hope that makes sense.