I've got a Fender Champion II 25 which at 12 pounds is a good match to my 67 years. It does an excellent job simulating a Fender Deluxe Reverb clean and slightly overdriven. It probably excels at other things, but that is what I use it for. In some situations, I would like to connect it to a PA without losing the speaker output. The headphone output sounds very good into my JBL IRXBT108 but mutes the speaker. Unlike the previous generation Fender Champions that have a differential class D output, the Champion II has an old-fashioned class AB output with one side grounded (looks like an unmarked clone of the discontinued TI LM1875). I think I can get away without a transformer. I've found this on the Internet and might try it:
Seems it is a -27 dB pad with a -3dB@1.4kHz single pole low pass filter, if I did the arithmetic right. When I read free Fender Deluxe Reverb cabinet IRs into Audacity and check the spectrum with a rectangular window function it does not look much like a single pole low pass, though 1.4 kHz might not be wrong. I'm not certain how much of that junk is the short snippet and that window, but it seems any other window function would clobber the IR's main tap. When I try a Blackman window a simple low pass filter does not look that far out of place.
The passive cabinet simulating Friedman "Amp No Mo" DI has a potted can which no doubt obscures their secret sauce. Buying a $99 DI for a $129 amp is just too crazy for me. Any ideas for something more sophisticated than the above, but still passive? There seems to be a rich selection of Chinese branded inexpensive active DI's with "cabinet simulation" in their bullet list of claims, but sticking with something passive appeals to me and I'm not a shredding connoisseur of high-end high-gain tube amps.
Thanks!
Jon
Seems it is a -27 dB pad with a -3dB@1.4kHz single pole low pass filter, if I did the arithmetic right. When I read free Fender Deluxe Reverb cabinet IRs into Audacity and check the spectrum with a rectangular window function it does not look much like a single pole low pass, though 1.4 kHz might not be wrong. I'm not certain how much of that junk is the short snippet and that window, but it seems any other window function would clobber the IR's main tap. When I try a Blackman window a simple low pass filter does not look that far out of place.
The passive cabinet simulating Friedman "Amp No Mo" DI has a potted can which no doubt obscures their secret sauce. Buying a $99 DI for a $129 amp is just too crazy for me. Any ideas for something more sophisticated than the above, but still passive? There seems to be a rich selection of Chinese branded inexpensive active DI's with "cabinet simulation" in their bullet list of claims, but sticking with something passive appeals to me and I'm not a shredding connoisseur of high-end high-gain tube amps.
Thanks!
Jon
Basically yes as far as cabinet emulation it is just some sort of EQ.
Usually just rolling off the top end is all that is needed.
For a very simplified passive pad, for speaker level to line.
It is a simple passive filter. With distortion the roll off is likely much more desired.
Just clean tone not as much as a issue. Or a higher filter.
Another classic DI / Preamp was the Sansamp which had opamp cabinet emulation.
Same thing basic EQ or high pass filter with a little Q to emulative the bass peak or common guitar speakers.
The again a higher order filter to roll off highs for distortion. Then a notch filter to emulate the typical
3k dip that many guitar speakers have. They have used all pass filters as well to emulate a microphone being
on or off axis. Wouldn't call it a gimmick, but often not completely necessary. Just a bunch of filters.
Most the important is the high end roll of so distortion doesnt sound awful on the mains,
Of course now it is full on FFT emulation, which they sample many real speakers and whatever response curve
they have is emulated at every frequency. Or much like the older direct effects pedals had many more presets
using different filters. And a cool name for the preset like bassman 4x10 or tweed 1x12 blah blah just filters.
The passive 1 pole filter in your design is set at around whatever it is. Being 1 pole to have enough cutoff
at around 5k to make distortion sizzle sound ok. So being 1 pole the filter would be set at around 800 Hz
to 1.2 kHz to make 5k actuated enough. Being a simple 1 pole filter.
With clean tone could be opened up a little. as with the opamp emulators they can use 2 pole or 3 pole
filters .More easily so the cutoff is sharper and the filter more open, then brick walls that sizzle at 5k
I would assume it is possible the Duncan magic can is just a 2 pole or 3 pole passive filter.
It is just harder to make higher pole filters with passive components.
Instead of using a 800 Hz to 1.5 kHz 1 pole filter. Is likely a 3 to 5k 2 pole or 3 pole filter.
Again same thing getting a sharp cutoff for high frequency for the sizzle and noise of typical guitar distortion.
Since speakers have natural roll off. He might have added a notch filter to emulate a mid dip or the classic
2 to 3k dip that many guitar speakers have. Some like it some dont. changes the sound of the distortion.
Usually just rolling off the top end is all that is needed.
For a very simplified passive pad, for speaker level to line.
It is a simple passive filter. With distortion the roll off is likely much more desired.
Just clean tone not as much as a issue. Or a higher filter.
Another classic DI / Preamp was the Sansamp which had opamp cabinet emulation.
Same thing basic EQ or high pass filter with a little Q to emulative the bass peak or common guitar speakers.
The again a higher order filter to roll off highs for distortion. Then a notch filter to emulate the typical
3k dip that many guitar speakers have. They have used all pass filters as well to emulate a microphone being
on or off axis. Wouldn't call it a gimmick, but often not completely necessary. Just a bunch of filters.
Most the important is the high end roll of so distortion doesnt sound awful on the mains,
Of course now it is full on FFT emulation, which they sample many real speakers and whatever response curve
they have is emulated at every frequency. Or much like the older direct effects pedals had many more presets
using different filters. And a cool name for the preset like bassman 4x10 or tweed 1x12 blah blah just filters.
The passive 1 pole filter in your design is set at around whatever it is. Being 1 pole to have enough cutoff
at around 5k to make distortion sizzle sound ok. So being 1 pole the filter would be set at around 800 Hz
to 1.2 kHz to make 5k actuated enough. Being a simple 1 pole filter.
With clean tone could be opened up a little. as with the opamp emulators they can use 2 pole or 3 pole
filters .More easily so the cutoff is sharper and the filter more open, then brick walls that sizzle at 5k
I would assume it is possible the Duncan magic can is just a 2 pole or 3 pole passive filter.
It is just harder to make higher pole filters with passive components.
Instead of using a 800 Hz to 1.5 kHz 1 pole filter. Is likely a 3 to 5k 2 pole or 3 pole filter.
Again same thing getting a sharp cutoff for high frequency for the sizzle and noise of typical guitar distortion.
Since speakers have natural roll off. He might have added a notch filter to emulate a mid dip or the classic
2 to 3k dip that many guitar speakers have. Some like it some dont. changes the sound of the distortion.
Yes as noted -27 dB pad with a simple 1 pole low pass filter.
If we mark the -3 dB and -6 dB points or technically -30 and -33 dB
-30 is 7.57k and -33 is 13k
So a very open non intrusive filter for clean tone.
Sounds about right to me I usually make clean tone filters if wanted to be about 12k at -6 dB
And for high distortion usually about 5 to 6k
Of course with distortion as noted a sharper cutoff really cleans things up and is pleasant with
a sharper filter 2 pole or 3 pole set at the same cutoff 5 to 6k
If we mark the -3 dB and -6 dB points or technically -30 and -33 dB
-30 is 7.57k and -33 is 13k
So a very open non intrusive filter for clean tone.
Sounds about right to me I usually make clean tone filters if wanted to be about 12k at -6 dB
And for high distortion usually about 5 to 6k
Of course with distortion as noted a sharper cutoff really cleans things up and is pleasant with
a sharper filter 2 pole or 3 pole set at the same cutoff 5 to 6k
Last edited:
This padding is also already done in the amplifier for the headphone jack.
As noted the headphone jack has a switch or switching jack as well which shuts off the speaker.
You can bypass that switch in the amp. So the speaker stays on.
Likewise add you own switch for the option of speaker on or off.
The switch is integrated with the headphone jack, so find the solder points for the speaker on the board.
As noted the headphone jack has a switch or switching jack as well which shuts off the speaker.
You can bypass that switch in the amp. So the speaker stays on.
Likewise add you own switch for the option of speaker on or off.
The switch is integrated with the headphone jack, so find the solder points for the speaker on the board.
I have the schematic for the previous generation Champion, which is a completely different digital design.
Champion II: analog integrated 480 MHz floating point ARM uC with mystery 28 pin ASIC (maybe effects, it is connected to the same JTAG port as the uC). Same chipset is in Mustang LT series. I suspect they went back to class A-B for the Champ II for cost and simplified FCC certification. It plugs into the wall, so power savings is not an issue. The Mustang LT25, which shares the same digital front end, does use a modern low part count class-D IC with no heatsink beyond the package's exposed paddle.
Champion: separate Codec, 56x DSP, low end ARM uC, TDA8950J-1 class-D output stage differentially connected to the speaker with a large number of passives plus a heatsink.
In the older version the headphone detect goes back to the uC and a mute line goes to the TDA8950 via a cable wire between boards. If the headphone detect does not trigger a speaker IR in the DSP, it would be fairly simple to intercept the power amp mute. There's no publicly available schematic for the Champ II's and with their 10 year newer design, the trivial compute power to do a 1024 point IR, the large number of analog IO pins on the uC, I suspect the mute is implemented in the uC with separate outputs going to the power-amp/power-supply board and the headphone jack. Also the muted signal might include a speaker IR, as the headphone output to a FRFR sounds excellent. In any case a hack of that multilayer surface mount PCB is not something I'm going to try.
In your above sim you show a 1nF cap, and the schematic from the internet I posted has 5nF. If the 5nf 1.4 kHz -3dB frequency is correct, at -6dB per octave, it would be roughly an addition 12 dB down at 5.6 kHz.
Does this seem reasonable for the problem (mild overdrive and clean usage) of connecting the speaker terminals to a PA line input?
Champion II: analog integrated 480 MHz floating point ARM uC with mystery 28 pin ASIC (maybe effects, it is connected to the same JTAG port as the uC). Same chipset is in Mustang LT series. I suspect they went back to class A-B for the Champ II for cost and simplified FCC certification. It plugs into the wall, so power savings is not an issue. The Mustang LT25, which shares the same digital front end, does use a modern low part count class-D IC with no heatsink beyond the package's exposed paddle.
Champion: separate Codec, 56x DSP, low end ARM uC, TDA8950J-1 class-D output stage differentially connected to the speaker with a large number of passives plus a heatsink.
In the older version the headphone detect goes back to the uC and a mute line goes to the TDA8950 via a cable wire between boards. If the headphone detect does not trigger a speaker IR in the DSP, it would be fairly simple to intercept the power amp mute. There's no publicly available schematic for the Champ II's and with their 10 year newer design, the trivial compute power to do a 1024 point IR, the large number of analog IO pins on the uC, I suspect the mute is implemented in the uC with separate outputs going to the power-amp/power-supply board and the headphone jack. Also the muted signal might include a speaker IR, as the headphone output to a FRFR sounds excellent. In any case a hack of that multilayer surface mount PCB is not something I'm going to try.
In your above sim you show a 1nF cap, and the schematic from the internet I posted has 5nF. If the 5nf 1.4 kHz -3dB frequency is correct, at -6dB per octave, it would be roughly an addition 12 dB down at 5.6 kHz.
Does this seem reasonable for the problem (mild overdrive and clean usage) of connecting the speaker terminals to a PA line input?
Whoops, yes indeed 5n
Mild overdrive, yes -12 dB at 5 to 6 kHz is about right to cut down the very harsh harmonics of distortion.
Basically crackle noise at that point.
Far as simple 1 pole filter to achieve that you lose a little more at 2 or 3 kHz
Which is where the last amount of detail or presence players typically like with distortion.
So some will like it others will say it lacks presence or say it sounds a little dark.
So you can boost the treble control on the amp , to bring it back. But run into the classic issues with.
Now the 5 to 6k harsh tones sneak back in with a simple treble booster or shelf filter.
Or others are just fine with it.
So often a narrow boost is added at around 2 to 3k, like a peaking filter.
Or as mentioned earlier, smooth distortion with good presence. Is often a 2 pole or 3 pole filter set higher.
Around the same 3 to 4k to get presence, then 5k to 6k is just brick walled -12dB of even lower.
So short answer, yes about 1.2 to 1.5 kHz for a 1 pole is about right tradeoff for less harsh, but not too dark.
Famous " Tube screamer" distortion is set too dark about 730 Hz to 800 Hz 1 pole.
Then the " Tone" control is actually a treble booster so users can choose how dark or brighter they want that.
Some mods done to that circuit with countless copies. Is exactly what you getting here.
They open that filter up from 730 Hz to 1.2 kHz.
You could essentially play with what you like by experimenting with 2.2n, 4.7n ,5.6n and 6.8n
to see or hear what you like with your distortion amounts. Or even add a 3 way switch to make it dark to bright.
And still have variance with the amp Treble Control as well.
Mild overdrive, yes -12 dB at 5 to 6 kHz is about right to cut down the very harsh harmonics of distortion.
Basically crackle noise at that point.
Far as simple 1 pole filter to achieve that you lose a little more at 2 or 3 kHz
Which is where the last amount of detail or presence players typically like with distortion.
So some will like it others will say it lacks presence or say it sounds a little dark.
So you can boost the treble control on the amp , to bring it back. But run into the classic issues with.
Now the 5 to 6k harsh tones sneak back in with a simple treble booster or shelf filter.
Or others are just fine with it.
So often a narrow boost is added at around 2 to 3k, like a peaking filter.
Or as mentioned earlier, smooth distortion with good presence. Is often a 2 pole or 3 pole filter set higher.
Around the same 3 to 4k to get presence, then 5k to 6k is just brick walled -12dB of even lower.
So short answer, yes about 1.2 to 1.5 kHz for a 1 pole is about right tradeoff for less harsh, but not too dark.
Famous " Tube screamer" distortion is set too dark about 730 Hz to 800 Hz 1 pole.
Then the " Tone" control is actually a treble booster so users can choose how dark or brighter they want that.
Some mods done to that circuit with countless copies. Is exactly what you getting here.
They open that filter up from 730 Hz to 1.2 kHz.
You could essentially play with what you like by experimenting with 2.2n, 4.7n ,5.6n and 6.8n
to see or hear what you like with your distortion amounts. Or even add a 3 way switch to make it dark to bright.
And still have variance with the amp Treble Control as well.
Last edited:
Thanks! I'm going to try out the original circuit to see how it sounds and proceed with your suggestions if it doesn't make the grade.
Well sure enough I did make a math error and the -3 dB frequency is 1.5 kHz. Since my powered speaker has a 10K ohm input impedance I need to scale down the resistors to lower that network's output impedance. Going with all standard components: dropping resistor R1=15K, load resistor R2=1K, and C=.11 uF (.1 and .01 in parallel), I get a 24 dB pad and the same 1.5 kHz -3 dB frequency. I think the following is right. I'm should have looked it up in a book rather than do the algebra.
Vo/Vi = R2/(R1 + R2 + j2PiFCR1R2)
-3 dB F = (R1 + R2)/(2PiCR1R2)
Vo/Vi = R2/(R1 + R2 + j2PiFCR1R2)
-3 dB F = (R1 + R2)/(2PiCR1R2)