PS: More About Dumbleator Settings

[Stephen1966]

Just some more thoughts about the output impedance of the driver and the hi-pass filter effect of the cap in combination with the send pot here. Let's see if I understand this correctly!

It seems obvious that if we use a 25k SEND (or what I call the DRIVE) pot then we might tighten up the low end by reducing the value of the coupling cap coming off the cathode. Tony suggested that the sustain may have been something of a product of cutting the bass, and there could be something to that. Tony also mentioned that he was using a 1uF send cap.

If we run the numbers for a hi-pass filter using 25k and 1uF we get cutoff at about 6.4 Hz which is very low and would certainly include all the low resonant frequencies of the guitar which bottoms out around 80 Hz.

Then again, if we run the same calculation of a "standard" Dumbleator with its send cap of .47uF and send pot of 250k we get a cutoff frequency around 1.4 Hz! Incredibly low. I think to myself that MrD wanted to exploit the full bandwidth of the system but a consequence of this is perhaps that the bass frequencies are allowed to dominate and this may explain why it's a common complaint (or feature - depending on how you look at it) the Dumbleators trim off a lot of the high end. I have the .47uF and 250k in there at the moment and I'm not sure, it definitely adds a quality of tone to the dry note in clean settings, but it still sounds fairly transparent. I don't consider myself a critical listener but it sounds pretty bright to me. Tony took it a step further though and I think he found something quite outstanding with the performance of the tube.

The fact is though, neither the Dumbleator nor Tony's more advanced version are (or were) likely to be set up with the send pots at max. So, to evaluate the impact on tone we have to consider the combinations of coupling cap and pot value with a range of pot settings. All the way from zero where, ideally, all we are looking at is the output impedance of the tube through 10/25/50/75 % of the pots value.

The numbers then:

The output impedance of the tube is calculated approximately as 1/gm. So when the transconductance of the tube is 1.6 mA/V, we get

1/0.0016 = 625R

We can say this corresponds with 0% on the pot and, with the Dumbleator's original .47uF cap and 250k send pot, this provides a frequency cut off of 542 Hz - whether you look at it as a low-pass or a hi-pass filter; the numbers are the same. This is something of a conflation though because the gm of the tube is really only instrumental in calculating the output impedance of the tube/system. It's a useful start point though because no pot manufactured today, is going to give 0R resistance when turned down to its minimum, so, lets consider 625R as the zero setting on the pot considered as a hi-pass filter instead, and this gives us:

0.47uF coupling cap/250k pot (% pot value)

0% = 542 Hz
10% = 13.6 Hz
25% = 5.4 Hz
50% = 2.7 Hz
75% = 1.8 Hz
100% = 1.36 Hz

1uF coupling cap/25k pot

0% = 254 Hz
10% = 63.7 Hz
25% = 25.5 Hz
50% = 12.7 Hz
75% = 8.5 Hz
100% = 6.37 Hz

Interestingly then, Tony's cut-off frequency with the 25k pot at minimum is almost an octave lower but no one hears anything at that level and the numbers quickly transpose so that overall, Tony's bass is being cut by an order of magnitude somewhat higher than the original. This is what he said! At 10% Tony's 25k pot is cutting the bass over four octaves higher. At 25% it looks close to five octaves higher and so on down it goes. [Correction: some faulty maths here! Suffice to say, it's mostly academic - and anything below 20Hz is mostly outside the range of human hearing, but the trend is clear!]

Does the filtering of the bass affect our impression of sustain in this case? All of these first frequency knees above perhaps 5% value of the pots are cutting bass in the range below 80 Hz but that doesn't mean they are irrelevant.

Let's see what happens with a .68uF cap and a 25k pot.

0% = 374.6 Hz
10% = 93.7 Hz
25% = 37.5 Hz
50% = 18.7 Hz
75% = 12.5 Hz
100% = 9.37 Hz

So, it's actually counterintuitive that reducing the pot to .68uF will improve the bass response by tightening it. It may do that, but now, it also encroaches on the audible frequency range of the guitar - a whole chunk of the 6th string is going to sound lifeless and uninspired.

What about a .82uF cap with a 25k pot:

0% = 374.6 Hz
10% = 77.7 Hz
25% = 31 Hz
50% = 15.5 Hz
75% = 10.36 Hz
100% = 7.77 Hz

This looks marginally better, we aren't hitting 80 Hz with a 10% pot value this time but this is midnight on the dial for most of these log pots and anything before that is going to come perilously close to cutting your audibly low fundamental notes.

The fact is, if you want your Dumbleator with a 25k pot to give the same frequency response as the original, you have to multiply the capacitors value by the same factor as the pot is reduced: 4.7uF. Maths is good like that!

The 1uF cap seems more than a happy compromise then. Tightening up the bottom end to within an inch of it's life but not so far the lowest notes being to suffer.

I guess I will find out soon enough now. The 1uF cap and the 25k pot are going in as soon as they arrive.

Stephen

[Stephen1966]

Here is the circuit with the modified 1uF coupling cap and the 25k DRIVE (send) pot and the 500k stereo mixer pot, marked RETURN (in the schematic) but labelled RECOVERY on the back panel.

SKYLINER FX loop 3.3.jpg

This is how it looks now.

1663875789753.jpg

1662668122065.jpg

All the shielded cable on the back panel is overkill I think and a pain in the neck to fit it all in, the next one I am going to try using only solid core between the jacks, switches and pots. But it's very quiet.

The bypass switch makes it easy to A/B test the effect of the loop in and out of circuit, and I found this handy to set it up for unity gain with no jacks plugged in. The sound is very nice. The sustain is, to paraphrase Tony, "cool". I really like it! What I hear as well are more pronounced mids. It thickens up the notes beautifully and it's so nice, I might never bypass the loop again. Articulate, singing notes, though it was just a rough play through today to make sure all the stuff works. It does, no problems here.

I was using it with a BOSS DD-200 delay pedal. With the DRIVE set at about 9 o'clock and the GAIN at about 3 o'clock there is no noise from the circuit but noise does creep in, the more you increase the GAIN. Doing so is unnecessary because in these positions it's unity gain and switching in and out is seamless. I don't think that is avoidable though by counter-rotating the DRIVE and GAIN by equal amounts you can probably find unity gain in other positions as well, and reducing the DRIVE will also reduce the output impedance. As it is, it is quite low now, compared to what the 250k pot would present. The 25k pot behaves well, it isn't patchy and I don't hear any clipping in the clean channel with a hot signal. OD + PAB is just YEAH!!!

The 500k stereo pot is also nice and responsive - everything from zero and a whisper of the effect at one end to full-on wet at the other. The linear pot is, I believe, the best option here. There is still a very slight hump in the signal when you travel through 50/50 but it is comparatively flat considering the hump I got using a log pot in my other experiments. I believe it to be an improvement over a mono pot in that in either series or parallel mode, it presents a more or less steady-state 250k to ground in the role of the grid leak for the recovery triode. In parallel mode the signal is all but flat across the dial then, but in series mode, it acts like a volume control and this might be combined with the DRIVE and GAIN settings for lower output impedance and a noiseless signal in those positions - don't know for sure yet, it's literally been only today I had it running. But I guess it's likely.

There isn't much else to say, except the pad on the front end (Line/Instrument) works as well and in instrument position, it reduces the signal down to a tenth of the Line position (0 dB to -20 dB). The pedal I used was fine with the Line setting though I believe it could easily start to distort if the signal was much hotter. This was all at an impressive level today and no problems.

The bright switch on the send side is likely unnecessary as Martin indicated in his mods, no difference in sound I could hear though I am keeping it in now. Who knows? This may be more effective with longer cable runs but when I tried it with a 25' cable I didn't hear any difference there either. The bright switch on the return side is very noticeable (almost too bright when engaged) and there is a healthy pop when it is thrown for the first time when the amp is started up from cold. A little resistor over the cap would probably fix that.

Everything's good then and I can button this up now. Great mods, a real improvement. Thank you.

Some final thoughts about this... If you want your Dumbleator to be transparent, stick with the 250k send pot. The smaller pot, has a marked difference in terms of tone and you may enjoy that. Or like LC, you may think it changes the tone too much and throw it in the bin