Axe-Fx III Firmware Version 5.02

AC-30's have a poorly designed cathode follower. The bias point is totally wrong and crushes one side of the signal considerably. IMO either the cathode resistor should be 100K or the plate resistor of the preceding stage should be higher, 200K. The clipping is extremely asymmetric. Too much IMO. A little asymmetry is good because it warms things up. Too much causes excessive even-order harmonics which makes things fuzzy and indistinct. The clipping is so asymmetric on an AC-30 that it's almost a half-wave rectifier.

You can tweak this by adjusting the Preamp Bias point and/or lowering the Cathode Follower Compression. Or you can lower the Harmonics value which reduces the asymmetric distortion. The downside of that is that it then overdrives the phase inverter causing blocking distortion from excessive bias excursion. Another thing to try is to increase the Grid Clipping value which will add a little headroom. Start with the Bias point. AC-30's are very sensitive to the tube type and part tolerances. A tiny change in the bias point can make a big difference. The default bias point is based on Mullard ECC83 tubes.

I prefer the AC-20 because it doesn't have a cathode follower so doesn't suffer from these problems. The cathode follower in an AC-30 doesn't even do all that much. Normally you use a cathode follower to preset a low-impedance source to the tone stack but the tone stack in an AC-30 doesn't present that great of a load anyways.

Hi Cliff, you mentioned before that by adjusting the CF parameters in a non CF amp, we are essentially adding a CF to that amp. How would you characterize the effect of this on a high gain amp, like the Energy Ball or any other non CF high gain amp?
 
AC-30's have a poorly designed cathode follower. The bias point is totally wrong and crushes one side of the signal considerably. IMO either the cathode resistor should be 100K or the plate resistor of the preceding stage should be higher, 200K. The clipping is extremely asymmetric. Too much IMO. A little asymmetry is good because it warms things up. Too much causes excessive even-order harmonics which makes things fuzzy and indistinct. The clipping is so asymmetric on an AC-30 that it's almost a half-wave rectifier.

You can tweak this by adjusting the Preamp Bias point and/or lowering the Cathode Follower Compression. Or you can lower the Harmonics value which reduces the asymmetric distortion. The downside of that is that it then overdrives the phase inverter causing blocking distortion from excessive bias excursion. Another thing to try is to increase the Grid Clipping value which will add a little headroom. Start with the Bias point. AC-30's are very sensitive to the tube type and part tolerances. A tiny change in the bias point can make a big difference. The default bias point is based on Mullard ECC83 tubes.

I prefer the AC-20 because it doesn't have a cathode follower so doesn't suffer from these problems. The cathode follower in an AC-30 doesn't even do all that much. Normally you use a cathode follower to preset a low-impedance source to the tone stack but the tone stack in an AC-30 doesn't present that great of a load anyways.
Thanks, @FractalAudio . I love it when you share deeper insights.
 
Anybody else noticing strange behavior using S/PDIF output?

Had audio cut out for a second a couple of different times.
The Silver Jubilee patch I've been playing lately seems significantly thinner.
Went into the Axe's I/O setting and switched the AES/SPDIF option to AES and back to S/PDIF.
After doing so, the output level of the S/PDIF output seemed hotter.
I'm not overloading via the Axe-FX meters, but the S/PDIF output is clipping the S/PDIF input of an Antelope Zen Tour.
This wasn't happening until I switched the AES/SPDIF option from AES and back to S/PDIF.

I don't recall any of this with the beta versions.

Loading v4.03 to compare...
Does it relate to this at all?
https://forum.fractalaudio.com/thre...-spdif-after-fw-5-00-beta-3-read-this.148908/
 
yeah i haven't been checking too closely, but my main preset that doesn't change much is in the 60% range, which caught my eye. it was definitely upper 70s a while ago.
 
I was just jumping around to various presets. Love FW 5.02 by the way but I did find on Preset 047 Royale With Cheese (Tip of the hat to Tarantino) that the High Cut in authentic doesn't seem to have any affect on the tone. Can someone confirm? I did reset the amp block and still same behavior.
 
Last edited:
Completely happy with this release, as others have said the Bandmaster and Voxes (two amps I've never enjoyed before) are now just sublime.

Interesting on CPU usage, as part of the FW tweak process I tried to reduce CPU usage - amazing the difference removing shunts which are there to make it look neat in Axe-Edit and reducing a few IR mixes for a pre-mixed one can make.

I've gone from 85-6% down to 69% on the same sound with this release and the above
 
Back
Top Bottom