Accent Switch Discussion from Hoyt Axton's Dumble for sale in Nashville topic

[Charlie Wilson]

Tony, if you go to the picture in the thread and zoom in, yes you will see 104 in white writing on the side of the cap. Here is a photo of Bonnamassa's amp. Looks like there is a pot on the front and the 500 ohm trimmer. He probably just turned the trimmer all they way to least resistance and put a 1k or so pot instead. Easier than having to pull the PCB and rewire the amp.
CW

[talbany]

I'll take a shot here and put together a description of how to understand and set up your GNFB circuit.


The voltage that is applied negatively to the PI (at its tail) through a voltage divider.In the case of the Dumble original design we know the circuit basically feeds back 7.7% of the output at the 4-ohm jack to the "subtracting" half of the PI.(we also determined that the 270/3k3 & 390/4k7 were basically the same 7.7% of signal is fed-back)
From there what happens when you add a cap in series with a presence pot, all of that in parallel with the (390 or270) As you increase the setting on the presence control, the cap ends up bypassing the 390/270 as you vary the pot, which allows high frequencies in the loop to go to ground. If they go to ground, they will not get fed-back (negatively) into the PI. By not subtracting those highs means that they will appear in the final output. This is basically how turning up a presence control adds highs to the sound. Choosing different value bypass caps you can fine-tune the frequency response of the loop. Make the cap smaller and higher frequencies will be decoupled. Make the cap larger and lower frequencies will be decoupled. At some point, the presence control starts working on the midrange, rather than just the high end (which is what we believe Charlie is hearing).
What we don't know yet (or we think you are hearing) is? given the tail resistor is in parallel with the presence cap (which forms an R/C network). We seem to think that the 270/1uF values get you lower in the "midrange" frequency, causing that smoothing effect your hearing but it's hard to know for sure until we run the numbers on all the different values your throwing out. (You can always compensate for this by just changing the value of the presence cap).

At the end of the day and to make things simpler for those wanting to tweak their GNFB
1..Determine the amount of "voltage being fed-back through the NFB resistor.
2..Determine the frequency roll-off you want by choosing the value cap you prefer (.1/1/2uF)

As far as the tail resistor goes the larger the value the better balanced the PI (cathodes further from gnd). The smoother the amp will sound.This should be a set value for the PI
If your amp is 50 Watt use the voltage divider ratio (= Rt/(Rt+Rf))needs to be 1.4 times (reciprocal of 0.7) greater than in a 100W amp to have the same NFB loop. So a 3.3K would work the same in a 50W amp as a 4.7K in a 100W one, all other things being equal.

IMO How you set the GNFB is a kind of personal thing each amp is different each player is different each output transformer/tubes are different so as amp tweakers we should not be afraid to experiment here and fine-tune the overall sound (it's just 2 parts)

So no real way of telling why Dumble used a particular GNFB circuit in a given design I am sure he has his reasons. It would be a good question to ask him if you or anyone talks to him.
Tony

[Charlie Wilson]

So a progress report, with the 2k presence turned all the way down which is basically the presence cap out of the circuit, I can hear little if any change in the resistor values. With the presence turned to say 1 o'clock, I can. With the 4k7/390 values and a 2uf cap the brashness in the upper mids is pretty smoothed out. So both you guys are right. I do kind of like the idea of adding just a little sparkle to the top end, so I am going to try some different presence cap values including the .1. I think this also means that I may prefer the sound of the presence cap more with either the 4k7/390 or 3k3/270.
CW

[bcook]

Charlie, I've yet to see a picture of what treble cap is supposed to be used with the classic stack, but you might want to try adjusting that also. I've settled on a 270pf there.

[martin manning]

So a progress report, with the 2k presence turned all the way down which is basically the presence cap out of the circuit, I can hear little if any change in the resistor values. With the presence turned to say 1 o'clock, I can. With the 4k7/390 values and a 2uf cap the brashness in the upper mids is pretty smoothed out. So both you guys are right. I do kind of like the idea of adding just a little sparkle to the top end, so I am going to try some different presence cap values including the .1. I think this also means that I may prefer the sound of the presence cap more with either the 4k7/390 or 3k3/270.

Your observations seem consistent with what was said above. If the presence cap is active, changing the tail resistor will affect the frequency where the boost starts... 390 vs. 270 is about half an octave, and with a 1uF cap you are well down into the mid range. That change of 100 ohms plus or minus won’t matter in the PI balance since that’s determined by the entire resistance from the cathodes to ground, and this is a very small part of that.

Tony, I had the same thought as you re trying to explain the function of the presence circuit. There are a number of subtleties in it for such a simple network. I made up the attached chart; see if it helps.

[talbany]

Martin
Great! Looking forward to checking it out. Should have some time later tonight to look it over
Thanks for making the chart
Tony

[Charlie Wilson]

OK, sorry I am going to have to flip flop a little. Attached is a recording of what I am hearing. The 2k presence is turned all the way down. I am only going from the 4k7/390 to 3k3/270. Sorry for the Iphone recording but I still think you get the idea.
CW

[martin manning]

My phone won’t play it, but you are comparing 4k7/390 against 3k3/270, with 2k presence pot set at 2k, and a 1uF presence cap? Going out on a limb here... Did the 3k3/270 seem to have a bit more low end?

[Charlie Wilson]

Yes, and a bit flatter on the top end.
CW

[martin manning]

HF voltage ratio went up 3.6% (more HF FB and lower HF gain), and LF voltage ratio went down 1.3% (less LF FB and more LF gain). The LF/HF balance has shifted. See the .pdf I posted. Looking forward to hearing it!

Ok, yes the second go-around has less top end (and hiss), more low mid... at least that's what I hear on my notebook's speakers. A calibration of the sound to the numbers is very useful, thanks.

[Charlie Wilson]

OK, a little more experimenting with this. I lifted the presence cap and realize that even with the 2k control turned all the way down, the sound of the cap is always there. I actually like the flatter sound of the amp with no presence cap. I also realize that I don't think I like the sound of a 1uf cap in this circuit even with the presence control turned all the way down. So I tried a .1 and clipped in my resistors to get the 3k3/270 and it sounds pretty good but... you only hear a change in presence with the pot 3/4 to full up. Hmmm, about 500 ohms. So, the accent switch makes sense, there is a contrast when going from the flatter no presence cap sound to the cap with a preset on a 500 ohm trim but not so much when the cap is always in the circuit and turning the knob on a 2k pot. It makes sense that when Dumble did modify the accent to a knob presence he probably used the 1uf cap(probably for players that insisted on an adjustable presence). With the 3k3/270(.1 cap) the bass strings get a little thuddy, hmmm...a .01 V1B coupler might fix that(I have a .02 right now). So at this point I don't know what the heck I'm doing but I think I'm learning something. Martin, I am curious, how large would a presence pot need to be to completely defeat the effect of the presence cap? Hearing the Low Plate Classic without the presence cap is a bit of a revelation for me as the amp was intended by Dumble's design at the time to be heard flat and if you wanted some high end kick you flip the switch. The same with the 2nd and 3rd Generation amps.
CW

[Max]

OK, then would it be safe to say ... that the 3k3(or slightly lower)/270 with a .1 presence(accent) cap appears to be Dumbles choice for a Low Plate Classic? Given that there are at least four photo examples of amps with this configuration and none with the 4k7/390.
CW

AFAIR I remember there seem to exist some other configurations, too. Example: a low plate classic ODS 100W with this 3k3/270 configuration, but with a 1,5k presence pot and a 0.68 uF presence cap.

Cheers,

Max

[talbany]

Martin
Finally got the chance to check out the chart. Question?
Why did you run that circuit?
Here are the GNFB circuits Dumble used and I thought we were discussing
3k3/390/4K7 1uF with a 2K pot This would be the "Skyliner" FBC
2k7/270/4K7 1uf 2k pot ..This would be (according to CW) the "Classic" FBC
3k3/270/4k7? .1 (104) w 500-ohm pot. This is called the "accent" control

There is another one that just has the " switch" (no Pot) I am so confused now I don't know what that one has maybe C.W can tell you.

Tony, I had the same thought as you re trying to explain the function of the presence circuit. There are a number of subtleties in it for such a simple network. I made up the attached chart; see if it helps.

BTW.. I know that my description of the presence control was pretty basic (which was my intent) Having tried most of these different FB circuits over the years I am aware of the "subtleties"of each.

Thanks for "trying" to do the chart.

Tony

[talbany]

AFAIR I remember there seem to exist some other configurations, too. Example: a low plate classic ODS 100W with this 3k3/270 configuration, but with a 1,5k presence pot and a 0.68 uF presence cap.

Cheers,

Max

Max
Do you have an opinion on this circuit?

Thank You!

Tony

[Charlie Wilson]

Whoa, not trying to start any trouble here. Just to clarify, I have never seen a Low Plate Classic in person. My only knowledge of them is from photos mostly on this forum. So Hoyt Axton, Joe Bonnamassa, Carlos Rios, and Borderland are the only ones I have seen. My hope was that maybe there is some more variations out there that people know about and they would chime in. Like Max did. Max do you recall if the .68 was a tantalum? I find, to my ears, the shift in frequency the 3k3/270 gives the presence cap a step in the right direction to where I want my amp to sound. I just need to find a presence cap value I really like. I said it before but I find it interesting that these early transitional accent switch amps were intended to be played with the presence out of the circuit unless you want the extra high end bump. Tony, check out the GNFB/accent switch on the crazy looking high plate amp you posted a photo of.
CW