Structo wrote:I just don't "hear" that kind of attenuation in my amp.
Remember that the 24th fret (if there is one) of the high-E string on a guitar is only 1320 Hz.
With the 0.001u and 250K trimmer, and that frequency, the full range of the trimmer only goes from -3.6 to -4.1 dB, probably not audible. That seems consistent with your observations, doesn't it?
Sorry about the scale on the plot. Note that the hearing of most humans goes south somewhere between the 10kHz line and the fist small division following it (20kHz).
Structo wrote:I installed the 250K with a .001uf cap.
Doesn't seem to do a whole lot, a very subtle difference as I roll through the rotation of the trimmer.
I still have some very high frequency hash that I thought could have been my speakers but now I have two EVM 12L's so I think I can rule that out.
I guess what we can conclude from this is that the high frequency stuff you want to get rid of is entering the signal path downstream of the OD output, because this thing should kill it.
If it is present on the clean channel too, that would confirm that theory wouldn't it?
pine wrote:Am I right in saying that the value of the resistor changes how much of the highs you remove and the value of the cap changes which specific frequencies you remove? Would lowering the value of the cap to 470pF target higher frequencies than the 1000pF or lower frequencys? "
After a little more thought and a little more math, I think I have this figured out. The thing has a shelving response (trebble cut) with the min and max attenuation above the cut frequency set by the 150K series resistor and the trimmer setting. The -3dB turnover frequency defies a nice tidy formula, but you can get pretty close using 1/(2*Pi*R*C), where R is the series resistor. The attached plot shows the response for the proposed 0.001u and 250K trimmer.
This was an interesting exercise, but maybe I'm easily amused.
MPM
Things are not that simple, in my view. Standing at the junction of the 150K output resistor and the treble bleed network, you're assuming the only resistance is 150K, however, that is not the case. If you look back from the 150K resistor, your possible path to ground would look like: 150K in series with the output resistor of the stage, and all of that would be in parallel with the 100K pot (full value). As for the output resistance of the stage, it is the parallel of the plate load resistor and the internal resistance of the 12AX7 stage, which is more or less on the order of 63K.
So, assuming a 100K plate load, you have: 100K//63K = 39K. Now, that is in series with the 150K resistor and you get 189K. Then, that is in parallel parallel with the 100K pot, and so the equivalent resistor at the treble bleed junction is 100K//189K = 65K. At zero setting of the trimmer and with a .001uF cap, the frequency breakpoint for the equivalent network would be: 1/(2*Pi*65*.001) = 2.4KHz. So that would be the lowest 3dB point achievable.
My method was pretty easy, at least for my own situation. With the amp on the bench, one could adjust it 'til blue in the face and still not know where to put it. But setting it and then taking the amp to a gig with a full mix on stage and playing it -really focused whether there was too much or not enough top on the lead sound with the bleed engaged. I generally don't use it, but when the room is bright it comes in handy--
My method was pretty easy, at least for my own situation. With the amp on the bench, one could adjust it 'til blue in the face and still not know where to put it. But setting it and then taking the amp to a gig with a full mix on stage and playing it -really focused whether there was too much or not enough top on the lead sound with the bleed engaged. I generally don't use it, but when the room is bright it comes in handy--
For me personally.. if you voice the amp to where it sounds good at home or on the bench by the time you get it to a gig with a band behind you it can usually stand to be a little brighter...The presence helps a bit but you reach that point where it gets a little rough around the edges..I love the idea of the concentric pots on the front with the trigger (OD entrance) and taper(treble bleeder) like Todd's 150w...complete control...Anyone have a source for these pots..
butwhatif wrote:My method was pretty easy, at least for my own situation. With the amp on the bench, one could adjust it 'til blue in the face and still not know where to put it. But setting it and then taking the amp to a gig with a full mix on stage and playing it -really focused whether there was too much or not enough top on the lead sound with the bleed engaged. I generally don't use it, but when the room is bright it comes in handy--
Yeah, I've noticed this as well. I haven't gigged with my D'Lite.
So it's bedroom tone for now.
But I do turn it up quite high sometimes.
It could even be my ears, since I'm getting old and haven't protected them as good as I should have over the years.
Plus a lifetime in construction doesn't help either.
But I love discussing this stuff because that's how you learn.
Even old dogs like me learn something new here just about everyday.
ayan wrote:Things are not that simple, in my view...
So, assuming a 100K plate load, you have: 100K//63K = 39K. Now, that is in series with the 150K resistor and you get 189K. Then, that is in parallel parallel with the 100K pot, and so the equivalent resistor at the treble bleed junction is 100K//189K = 65K. At zero setting of the trimmer and with a .001uF cap, the frequency breakpoint for the equivalent network would be: 1/(2*Pi*65*.001) = 2.4KHz. So that would be the lowest 3dB point achievable.
Hope this clarifies things a bit.
Gil
Okay, cool, I'm with you; thanks! Aero-Thermo-Mechanical guy learning new chops ...
I put the equivalent resistance into my long-hand solution in place of the 150K series resistor and left the trimmer resistance in the shunt position (in series with the cap to ground). It now shows -3dB at about 2kHz. What do you think?
(Edit: Not sure if this transform is legit, but it looks right... the -3 dB frequency agrees pretty closely with the estimate, much closer if I drop the trimmer resistance to 1%, and the attenuation with frequency vs. trimmer setting looks right based on shorting the cap.)
Tom,
I wouldn't worry about having a 500K trimmer in there; it's only a variable resistor so you can dial it down to whatever you want.
MPM
You do not have the required permissions to view the files attached to this post.
Last edited by martin manning on Fri May 15, 2009 11:16 pm, edited 6 times in total.
So according to your graph, you would get -11db attenuation at 12 Khz
with 25K ohms.
Not only that, but you would get over -9db attenuation at 8 khz.
But, it sure doesn't sound like it.
I suppose it doesn't shunt everything above a certain frequency but attenuates it by the db's recorded on the graph.
Electric guitar reportedly has a range of 82-1050 hz on the root notes.
So what would 10 khz be, 10th order harmonic?
I don't know enough about harmonics to figure it out but it sure seems as if a high pass filter such as the treble bleed circuit would really clamp down the high frequency hash that I am hearing.
But as mentioned I don't know what it sounds like in the real gigging world.
Funny thing is I was hearing the same high freq crap on my little 6v6 amp.
That is why at first I suspected the CL 80 speaker I had paired with the EVM 12L. But now I have two EV's and although it isn't as bad as it was it's still slightly there.
I've changed guitars, cords, tubes, etc in order to try and pin it down.
I don't have golden ears by any stretch but I am hearing this.
Structo wrote:So according to your graph, you would get -11db attenuation at 12 Khz
with 25K ohms.
Not only that, but you would get over -9db attenuation at 8 khz.
But, it sure doesn't sound like it.
I suppose it doesn't shunt everything above a certain frequency but attenuates it by the db's recorded on the graph.
Electric guitar reportedly has a range of 82-1050 hz on the root notes.
So what would 10 khz be, 10th order harmonic?
I don't know enough about harmonics to figure it out but it sure seems as if a high pass filter such as the treble bleed circuit would really clamp down the high frequency hash that I am hearing.
But as mentioned I don't know what it sounds like in the real gigging world.
Funny thing is I was hearing the same high freq crap on my little 6v6 amp.
That is why at first I suspected the CL 80 speaker I had paired with the EVM 12L. But now I have two EV's and although it isn't as bad as it was it's still slightly there.
I've changed guitars, cords, tubes, etc in order to try and pin it down.
I don't have golden ears by any stretch but I am hearing this.
Each of the small divisions is equal to the value of the first major division to it's left, so each small division after 10 kHz is 10 kHz. The last symbol on each curve is ~33 kHz, so it's about -11 dB at 33KHz.
The second-to-the-last symbol is 16.4 kHz, and only people with pretty good (i.e. young, un-abused) ears can hear that. The stuff that matters for guitar is probably 8 kHz and below, the third-from-the-last symbol. The plot shows about -9 dB at that point. Gil said he sets his trimmer at about 112K (45%), so he is taking about 4 dB off the highs.
The symbols on the curves are one octave apart, a factor of two in frequency. 1 kHz to 10 kHz, a factor of ten or a "decade," is 3 and 1/3 octaves.
MPM
Last edited by martin manning on Wed May 13, 2009 1:01 pm, edited 2 times in total.
Funny thing is I was hearing the same high freq crap on my little 6v6 amp.
That is why at first I suspected the CL 80 speaker I had paired with the EVM 12L. But now I have two EV's and although it isn't as bad as it was it's still slightly there.
We've built quit a few 6V6 amps early on (our Derringer Model uses 4x 6v's in HRM configuration)
We had to use the JJ's that stand up to those higher plate voltages..
To my ears when you approach those higher voltages on the 6V's 450-470 I notice an unpleasing anomaly that kind of chases the note...It rides on the top end and has a sort of un even breakup to it..It's not as bad with the 4 6v's as it is with 2 but is still there..With the HRM stack in there you can dial some of it out but once you hear it your ears still focus in on it.. I don't know if it's the JJ's or what since some amps did it worse than others..When we modified those amps and set them up with 380/400 on the plates the artifact was gone..A 4X6V HRM cranked up sang like a bird nothing like it....
