SSS CF Driver Compression Questions

[Guy77]

A part of the compression is coming from the LNFB loops. Especially when you drive the first volume pot a little more.
Not all SSS are the same as are the LNFB loops.
Which one do you have?

Erwin

My build is adapted slightly from MHarman's Layout which I believe is based on #004.

So I see that I did rip out the LNFB completely from V1 because I did have the switch on the front panel (instead of an FET input) and I felt after a while that it wasn't doing enough to warrant a weird looking switch above the input. I may put that part of the circuit back in since I'm now chasing subtleties that I may not have noticed before...

Glad to hear the drop in Phase Inverter plates has improved things a bit for you!

Cheers
G

[bcmatt]

It's all very subtle. I feel like I need a bit more something. Maybe I need to record a demo and see if I'm chasing something that doesn't actually exist. It's not like it sounds bad, but I feel like some people claim these things want to sustain almost into feedback with a clean tone, and I'm not getting near that. It's possible that I just am getting overall lower gain from my amp than some people get from their SSS models. I basically have to turn on my PAB and Mid Boost switch as well as turn up the volume above 9 to get any sort of breakup when digging in.

My current plate voltages:
6L6s: 438V
CF Driver: 188V
PI: 208V
V1, V2, V3: 178V- 182V

This is with just changing the 2.2K Dropping Resister to about 20K (15K/10W + 5K/5W in series).

I also put a couple separate 1 ohm cathode resisters on for the 6L6s so that I can try the PI balance method mentioned above. (as well as see how well my output tubes are actually matched while biasing).

I guess I will likely next re-implement the LNFB for V1 again... But wouldn't that just lower the gain more? I feel like my problem is perhaps an overall lack of gain. I feel like I need it pushed closer to the clipping point on each of these triodes...
Perhaps I need to add those extra resisters below the Cathode Resister/Cap pairs on the preamp tubes that some SSS models have? Mine is like the MHartman layout that doesn't utilize those.

[Stephen1966]

It's all very subtle. I feel like I need a bit more something. Maybe I need to record a demo and see if I'm chasing something that doesn't actually exist. It's not like it sounds bad, but I feel like some people claim these things want to sustain almost into feedback with a clean tone, and I'm not getting near that. It's possible that I just am getting overall lower gain from my amp than some people get from their SSS models. I basically have to turn on my PAB and Mid Boost switch as well as turn up the volume above 9 to get any sort of breakup when digging in.

My current plate voltages:
6L6s: 438V
CF Driver: 188V
PI: 208V
V1, V2, V3: 178V- 182V

This is with just changing the 2.2K Dropping Resister to about 20K (15K/10W + 5K/5W in series).

I also put a couple separate 1 ohm cathode resisters on for the 6L6s so that I can try the PI balance method mentioned above. (as well as see how well my output tubes are actually matched while biasing).

I guess I will likely next re-implement the LNFB for V1 again... But wouldn't that just lower the gain more? I feel like my problem is perhaps an overall lack of gain. I feel like I need it pushed closer to the clipping point on each of these triodes...
Perhaps I need to add those extra resisters below the Cathode Resister/Cap pairs on the preamp tubes that some SSS models have? Mine is like the MHartman layout that doesn't utilize those.

If you have tried V1 without the LNFB and you still can't get it to bite I don't think restoring is going to help and yes, closed loop gain (with LNFB engaged) is going to be lower. But not from perspective of the signal entering the stage which instead sees more headroom and more swing on the grid resulting in a cleaner sound. The LNFB also filters noise and intermodulation distortion and improves the linearity of the stage though, so reconnecting it is not a bad idea.

If you want to drive a stage to break up earlier, your lower (than MH's) B+ voltages are helping but you might want to try biasing slightly warmer (1k bias resistor instead of 1.5k) or colder, with a higher value resistor. With a bias resistor at 1.5k and a 4.7uF bypass cap and 100k plate resistors it is probably close to centre-biased.

Experimenting with the value of the cathode bypass cap you can also trim the tone (like a high-pass shelving filter) and increase the gain of the low frequencies if you increase its value. On the other hand, a smaller value cap will increase the gain of the high frequencies - giving you an apparent treble boost. But it will be at the expense of the lowest frequencies which actually deliver more apparent power for the same strength input signal.

If its the howl of feedback you are looking for, then you are probably better off looking to augment the higher frequency part of the spectrum later in the tonestack and those filters which - honestly - I don't understand.

[bcmatt]

If you have tried V1 without the LNFB and you still can't get it to bite I don't think restoring is going to help and yes, closed loop gain (with LNFB engaged) is going to be lower. But not from perspective of the signal entering the stage which instead sees more headroom and more swing on the grid resulting in a cleaner sound. The LNFB also filters noise and intermodulation distortion and improves the linearity of the stage though, so reconnecting it is not a bad idea.

If you want to drive a stage to break up earlier, your lower (than MH's) B+ voltages are helping but you might want to try biasing slightly warmer (1k bias resistor instead of 1.5k) or colder, with a higher value resistor. With a bias resistor at 1.5k and a 4.7uF bypass cap and 100k plate resistors it is probably close to centre-biased.

Experimenting with the value of the cathode bypass cap you can also trim the tone (like a high-pass shelving filter) and increase the gain of the low frequencies if you increase its value. On the other hand, a smaller value cap will increase the gain of the high frequencies - giving you an apparent treble boost. But it will be at the expense of the lowest frequencies which actually deliver more apparent power for the same strength input signal.

If its the howl of feedback you are looking for, then you are probably better off looking to augment the higher frequency part of the spectrum later in the tonestack and those filters which - honestly - I don't understand.

I did notice that Viperdoc's 002 schematic has some cathode voltages and his V1 cathode's are at 2.15VDC while my v1 cathodes are down at about 1.25VDC.

I'll try a few more different 12ax7s in V1 and see if things change but perhaps I need to get my cathode voltage up a bit. That would be biasing warmer? using lower value resisters? I'll experiment with 1K and see what happens.

[martin manning]

Do you have a known good set of tubes in this? Those voltages came from my LTspice simulation, so they are "perfect" specimens.

[Stephen1966]

I did notice that Viperdoc's 002 schematic has some cathode voltages and his V1 cathode's are at 2.15VDC while my v1 cathodes are down at about 1.25VDC.

I'll try a few more different 12ax7s in V1 and see if things change but perhaps I need to get my cathode voltage up a bit. That would be biasing warmer? using lower value resisters? I'll experiment with 1K and see what happens.

Good idea to start with a bit of tube rolling first. 1k will warm the bias but for more noticeable assymetric bias you may need to go over to the cold side. I wouldn't take the bias resistor any smaller than say 820 Ohm for stability reasons, but you can go much higher for a cold clipper effect.

[Guy77]

I believe the Mhartman version of the SSS has the mixer tube that mixes the clean and reverb signals like the JM Signature clone.
I wonder if changing this can get you what you want.
The #005 SSS does not use that method of mixing the clean with the reverb and it has lots of gain on tap. You can listen to various videos of it on YouTube.

Its a bit of work to change it to a #005 but it may get you what you want.

Cheers
G

[bcmatt]

Do you have a known good set of tubes in this? Those voltages came from my LTspice simulation, so they are "perfect" specimens.

OK, I rolled a number (5 or 6) of tubes through V1 (5 or 6) and they were all about the same. I left in the one that seemed to sound the best to me and that also gave me the highest Cathode Voltage of about 1.5V (so they were al pretty similar).

I reinstalled the LNFB on V1 with the 3 position switch and went back to the single input underneath the switch (since a low input is ridiculous with how clean this amp is already). I do think the LNFB switch does help take the edge off and round off the attack of the note a little bit while maintaining the same clean sustain, so I guess that is a step in the right direction. I didn't have the 10M resisters, so I used my last couple 8.2M resisters, and then 2x 4.7M (9.4M) resisters for the one that connects to the second stage grid.

20240408_165301.jpg

Speaking of the cleanliness of these SSS amps, how clean do most peoples' SSS amps stay throughout their volume pot (without FET input)?

I removed the 20K dropping resister and brought it back down to a little more reasonable 13.6K (2x 6.8K in series). This helps keep the V1 and V2 plates at about 188VDCish while the PI plates are at about 235V.

It just occurred to me that I should roll some tubes through V2 as well, and see if that one is affecting things. I'll try that next. I'm a little reluctant to start changing cathode resister values since it seems that all iterations of SSS models stay at 1.5K throughout with mostly 4.7uF bypass caps. I may have chosen the wrong version of SSS to mostly follow but I've definitely heard some models of SSS amps of uncertain lineage, that have that special bloom and sustain that I am looking for. (I'll try to find the youtube samples of the ones that get what I'm after)

I see a schematic of #002 with added 100R resisters. Can anyone explain what those are going for?

OK, I'm going to roll some tubes through V2 and then I ought to tackle the PI balance method next.

[bcmatt]

Well, I did the PI balance trim adjustment. It made me feel like a bit of a mad scientist, HAHA. I Used a USB-C to headphone jack dongle from my phone using a frequency generator at just over 400Hz. I had to go into my Full-Drive 3 pedal with just the boost turned on and I turned it up until I measured 1.5VAC. I couldn't get even half that without the boost pedal.. but it should be clean. Then I plugged the output of the pedal into the power-amp in jack on the back of the amp. I had the speaker jack out to my Badcat Unleash as a reactive load (The actual re-amp part of of the Unleash isn't currently working. I had it repaired once on warranty, but that's long over. I think it somehow keeps burning out some part of it's solid state amp because of some finicky part of the design).
Anyways, the first attempt only got me down to about 5mV with the trimpot, so I swapped the two Power Tubes. Then I had to turn the other direction. I found the point where it would pass from -0mV to +0mV and left it there. I did notice it would never start climbing above +0mV. even though there was a bit more turning to go. Is that to be expected?
I measured the PI plates after and noticed that they are now reading 236.8VDC (Input side) and 234.6VDC (Poweramp side). Anyways, that was a fascinating procedure. \

Anyways, the amp still seems about the same. It can sound very even and will let clean notes and chords ring out well and clear. I'm going to go search for examples now to see if there's something more I ought to get. It still is incredibly clean... similar to my Deluxe reverb when it had that bad old tube in V1. I have nothing against clean amps ( I often use a cleaner tone than pretty much all the other guitarists I know), but I do wonder if this amp should be able to run closer to it's clipping point so that it's not too hard to keep the volume from jumping like crazy when you click on a mild overdrive pedal. (you really have to have the pedal volume dialed extremely precisely). But, I also understand if that is the point of this amp... maximum headroom... give it a boost and it will just give you more clean signal.

If I do decide to experiment with a little colder clipping, is there an ideal triode to try it on that lines up with any version of an SSS? The Trainwrecks save it for the 3rd triode...

[Stephen1966]

I see a schematic of #002 with added 100R resisters. Can anyone explain what those are going for?

Do you have a schematic you can share?

If I do decide to experiment with a little colder clipping, is there an ideal triode to try it on that lines up with any version of an SSS? The Trainwrecks save it for the 3rd triode...

As you say, what you're contemplating doesn't seem to have been tried before (that's never stopped me personally speaking!) but if you consider it a rule of thumb that "overdrive should come late in the signal chain," you might try that. It doesn't have to be a hugely invasive process though as you only need to lift one leg of the bias resistor and tack in another until you hear the effect you are seeking. Easier still if you have a resistor decade box.

Edit: check out Rob Robinette's section on the cold clipper. https://robrobinette.com/Tube_Guitar_Am ... ld_Clipper I think there's a difference between this and a mildly off-centre, asymmetric bias but the principle of keeping the distorted lobe in phase is sound.

[erwin_ve]

I see a schematic of #002 with added 100R resisters. Can anyone explain what those are going for?

Those cathodes have a connection to the LNFB loops. Otherwise the LNFB loops would be shunted to ground.

[bcmatt]

I see a schematic of #002 with added 100R resisters. Can anyone explain what those are going for?

Those cathodes have a connection to the LNFB loops. Otherwise the LNFB loops would be shunted to ground.

Thanks! That makes sense. Mine Doesn't have that because its LNFB plays between plates and grids and doesn't connect to the cathodes at all. 100R seemed like it would be way too subtle of a difference anyways, but I wasn't sure what going in series below the bypass cap would do.

My biggest difficulty is that I have adapted a variation of the MHartman layout that is supposed to be based on #004 (not that I knew it when I built it), but I have never seen a schematic of that version...(if anyone has one I can see, that would be hugely appreciated). Therefore, I'm only ever looking at schematics of #1, #2 and more recently #5.
I even went as far as drawing my needed power adaptations on the #1 Schematic:
https://ampgarage.com/forum/download/fi ... &mode=view

This was my build thread:
https://ampgarage.com/forum/viewtopic.p ... S&start=15

[bcmatt]

Do you have a schematic you can share?

Ya, that's been my struggle. Every time I want to look at a schematic, I start searching around the forum until I find one. (never found a #4 yet that actually should go with my amp).

Here is the #002 I more recently found this morning. Sorry, I don't remember where on the forum I found it now. I had to search again landed on this one. So I appreciate when people put info on it, (so sorry if I am reposting it against your will to whoever it belongs to!)

SSS#002Schematic.pdf

As you say, what you're contemplating doesn't seem to have been tried before (that's never stopped me personally speaking!) but if you consider it a rule of thumb that "overdrive should come late in the signal chain," you might try that. It doesn't have to be a hugely invasive process though as you only need to lift one leg of the bias resistor and tack in another until you hear the effect you are seeking. Easier still if you have a resistor decade box.

Edit: check out Rob Robinette's section on the cold clipper. https://robrobinette.com/Tube_Guitar_Am ... ld_Clipper I think there's a difference between this and a mildly off-centre, asymmetric bias but the principle of keeping the distorted lobe in phase is sound.

Thanks! Ya, Robinette is sort of my guide for any thoughts I have about trying to add cold clipping... but thanks for the link, it would take awhile for me to wander around trying to find it again.
I realize he talks about the advantages of different stages to consider putting it. I didn't remember that.
I also see that it should not accompany a bypass cap. Do you think adding resistance in series between the regular cap/resister and ground would avoid the buzzy pitfalls of increasing the regular bypassed resister value?

[bcmatt]

I believe the Mhartman version of the SSS has the mixer tube that mixes the clean and reverb signals like the JM Signature clone.
I wonder if changing this can get you what you want.
The #005 SSS does not use that method of mixing the clean with the reverb and it has lots of gain on tap. You can listen to various videos of it on YouTube.

Its a bit of work to change it to a #005 but it may get you what you want.

Cheers
G

Thanks! I am digging a bit more into #5 to try and understand the differences. I may mess with the Reverb circuit down the road. Right now, I think I'll experiment with the global Feedback value (mine seems to have the most feedback compared to other SSS circuits, seeing as my resister is 2.7K. I pulled two tubes, so I now have an 8ohm tap only (instead of 4) but I should have half the power. So, the value would go up a bit because of the 8 ohms... but down a bit because of the 50Watts. I'll try something....

[Stephen1966]

I also see that it should not accompany a bypass cap.

It might be true when the idle point is shifted so far towards cutoff but you can try a mildly asymmetric bias and see how that works, even small changes in value have an effect on the headroom. In my case I used a 2k7 resistor with an unremarkable 4.7uF cap. It's a different amp I'm working on but the idea is the same. Mine isn't buzzy at at all, a fuller sound I would say.

https://ampgarage.com/forum/viewtopic.p ... 50#p465050