SSS #002 REDRAW DRAFT

[martin manning]

I have seen one or two 12AX7 data sheets that mention a preferred section (1, the 678) for lowest noise, but I suspect that may have been peculiar to those manufacturers. AFAIK most data sheets use 1 and 2 for the sections, not a and b. The GE 7025 data sheet description says "for use in low-level stages of high-gain audio amplifiers," and it does include noise figures, but it does not differentiate between the sections. GE 12AX7A does not differentiate for noise either.

Having the pins match the layout would be most important for anyone checking voltages, the a and b not so much. The layout has no section designators, so just swapping the symbols in your schematic would make the pins match. I believe it would be easy enough to follow Fender's convention and edit the symbols to make 123 the "a" side... An opportunity to up your KiCad game, perhaps?

[ViperDoc]

I have seen one or two 12AX7 data sheets that mention a preferred section (1, the 678) for lowest noise, but I suspect that may have been peculiar to those manufacturers. AFAIK most data sheets use 1 and 2 for the sections, not a and b. The GE 7025 data sheet description says "for use in low-level stages of high-gain audio amplifiers," and it does include noise figures, but it does not differentiate between the sections. GE 12AX7A does not differentiate for noise either.

Having the pins match the layout would be most important for anyone checking voltages, the a and b not so much. The layout has no section designators, so just swapping the symbols in your schematic would make the pins match. I believe it would be easy enough to follow Fender's convention and edit the symbols to make 123 the "a" side... An opportunity to up your KiCad game, perhaps?

No doubt. I'll make a reversed tube and put the "layout matched" triodes and the voltages on the next version. Thanks, man!

[martin manning]

Is the schematic's bias circuit correct?

Yes, looks good to me.

Voltages with 5751 reverb driver tube. Ryan's layout has voltages for a 12AT7, which changes the numbers for that tube and its supply node, but not much else.

[ViperDoc]

Here's a version with layout-aligned noval socket pins and voltages. What is "J1"(DRAIN) in the voltage list? The JFET drain lead? Once it's tidied up, I'll post it at the beginning of the thread.Thanks.

[DELETED] [POSTED ABOVE]

[martin manning]

Looks great! J1 is the JFET, Drain is the drain pin.

Oops, I think you have the reverb input and recover triodes swapped a/b.

[ViperDoc]

Looks great! J1 is the JFET, Drain is the drain pin.

Oops, I think you have the reverb input and recover triodes swapped a/b.

Good eye! Thanks. SSS #002 V7.1 fixed and POSTED!

BTW, I see both SSS002 layouts I've seen disagree on which side sends and recovers reverb. Which layout should be the benchmark?

[martin manning]

I see both SSS002 layouts I've seen disagree on which side sends and recovers reverb. Which layout should be the benchmark?

I'm assuming Ryan's layout is the reference. What other layout is there? One other "layout" I can think of would be Erwin's PCB's.

[ViperDoc]

I see both SSS002 layouts I've seen disagree on which side sends and recovers reverb. Which layout should be the benchmark?

I'm assuming Ryan's layout is the reference. What other layout is there? One other "layout" I can think of would be Erwin's PCB's.

The "Layout by Aaron" in the Dumble files section is primarily the one I used. Yours looks like a legit DIYLC doc that I've never seen. That must be Ryan's? I would *love* to get my hands on that. The other is a SSS-inspired layout, looks like, so not necessarily 002, so my mistake.

[martin manning]

Ryan's layout is on Github. There is a link in his SSS 002 thread.

[erwin_ve]

I see both SSS002 layouts I've seen disagree on which side sends and recovers reverb. Which layout should be the benchmark?

I'm assuming Ryan's layout is the reference. What other layout is there? One other "layout" I can think of would be Erwin's PCB's.

The "Layout by Aaron" in the Dumble files section is primarily the one I used. Yours looks like a legit DIYLC doc that I've never seen. That must be Ryan's? I would *love* to get my hands on that. The other is a SSS-inspired layout, looks like, so not necessarily 002, so my mistake.

Aaron's layout is a merged sss002/sss004 amp.
Here is the github for Ryan: https://github.com/colganr/steel-string-singer-sn-002

[martin manning]

Here's the layout I have locally:

[ViperDoc]

Thanks, Martin and Erwin. Looks great.

NOTE: I have a growing list of questions:

1) The Reverb Driver cathode resistor and Reverb Send/Recovery plate load resistors are 2W parts. Why is that?

2) The power supply in the layout uses a dual section 33/33uF cap in parallel for 66 uF supply for the screens. Why is that? I used the power supply off the original schematic posted in the files section. Is this a problem?

3) Mixer CF plate load and cathode resistors are 5W rated. This must be on the 002 as well?

[martin manning]

1) The Reverb Driver cathode resistor and Reverb Send/Recovery plate load resistors are 2W parts. Why is that?

No good reason, AFAIK. Those could be 0.5W

2) The power supply in the layout uses a dual section 33/33uF cap in parallel for 66 uF supply for the screens. Why is that? I used the power supply off the original schematic posted in the files section. Is this a problem?

The hand drawn schematic had a larger cap down the line from the screen, and there was some debate over whether that one might really be the screen cap. On my schematic I have a 47u on the screen node and 22u for all the rest.

3) Mixer CF plate load and cathode resistors are 5W rated. This must be on the 002 as well?

The 100k cathode R's should be 1W, but the others could be 0.5W.

[ViperDoc]

I noticed the Colgan layout has an 8K2 off the FET source lead, whereas the other schematic had a 2K7. This is the FET bias resistor, correct? I imagine this is where you could put a 2K7 in series with a trim pot to adjust that, right? Or find the desired value and swap, but my point is: is the 8K2 appropriate for the 2N3823, and the 2K7 for the original NTE452? Or is this simply to taste? Thanks.

[martin manning]

I noticed the Colgan layout has an 8K2 off the FET source lead, whereas the other schematic had a 2K7. This is the FET bias resistor, correct? I imagine this is where you could put a 2K7 in series with a trim pot to adjust that, right? Or find the desired value and swap, but my point is: is the 8K2 appropriate for the 2N3823, and the 2K7 for the original NTE452? Or is this simply to taste?

No and yes. JFET's are not known for consistency, so the source resistor is selected to get the desired operating point. The 2N3823 model in the sim required a 5k9 to more or less center the operating point, and it is likely some kind of average representation of that part. Some might prefer a hotter bias for more distortion.