• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

A DIY Tube Preamp/Line Driver by Passinwind

#44
This is really a great work. Very nicely done, and that is a large part of the battle of tone. It does the job, and I would sure like to play it. Don't tell anybody, but I've done a bit of solid state work myself . :)

jule

Passinwind said:
This is still a definite work in progress, but here goes anyway. When I built my DIY solid state preamp, I always intended to add a line driver stage eventually. I actually found that a balancing transformer was all I really needed, but still...I can't leave well enough alone I guess.;)

I wanted to keep it very simple, but also at the same time to try out some different approaches than using just the typical 12Axx tube gain stages and basic power supply. I found a nice looking circuit board for the line driver itself here: Link Removed, and went with a kit from Welborne Labs for the power supply. The turret board is also from Welborne; it contains a biasing network that I wired up for the DC filament supply. The rack case came from Par-Metal, the output transformer from Link Removed. So far I have been going between two 6SN7s and one 6SL7 paired with one 6SN7. I don't really need the extra gain from the 6SL7, so I'll likely settle on two 6SN7s in the long run.

Initially I set the line driver up with a fairly high input impedance so I could use a passive bass directly into it, if desired. I've since decided that I prefer it with either my SS preamp or my active basses straight in, so I've optimized things for that now. I may implement a "loading" function if I find it useful in testing.

I'll post some clips and updated distortion and noise specs in a few days, after I dial in a few more things. As you can see, there's plenty of room for more features too. It already sounds pretty danged good IMHO though.:cool:

Link Removed

Link Removed

Link Removed
Click to expand...
 
#45
I finally caved in and bought a nice stash of NOS 6SN7s, settling on one Sylvania and one Motorola-branded GE pieces in the Aikido board. I also found some more Sylvania JAN 6SL7WGTs from the same year as my previous favorite ('55) -- the new ones are actually even nicer.

Link Removed

A more recent internal shot, although taken before I tidied things up a bit more and changed out a couple of components for subtle refinements:

Link Removed

Hopefully I can do some tube comparison clips soon, but in the meantime I finally found another old one using some new production E-H and Tung-Sol 6SN7s. I added a little 'verb this time in post-processing just for the heck of it : [Invalid or Expired Link Removed] for a streaming MP3 or follow the SoundCloud link in my sig for a slightly different lower res version, it's the clip called PW6B / Crescent Moon Fretless / Q-Tuners. I built this thing mainly for live usage, but it seems plenty quiet enough for recording too, eh?


Happy New Year to all my fellow TB DIY'ers...:bassist:
 
#49
MAMMOTHvolume said:
Looks very nice! just learning to use spice myself for this kind've thing. What tone stack did you use out of interest?
Click to expand...

My own, of course...:cool:

Bass and mids are active, close to Baxandall, in the feedback loop from the second 6SL7 plate back to the grid. There is substantial treble pre-emphasis involved as well. Treble control is passive, cut-only, between the second and third stages. At midpoint you get the curve shown above, since you are cutting into the baked-in boost in the previous stage. Here's the modeled frequency response curve model with the treble control nearly all the way up:

Link Removed

The idea is dead-clean highs at any setting, and the ability to have variable distortion in the lows and/or mids (or not, as desired), with lots of interactivity between all the tone controls. Works great for fretless, and cabs with tweeters, which is all I care about really...;)

My model doesn't include the input and output transformers, so the actual high shelf/bell top point will likely be a little lower than shown in the model. It'll change some depending on the control settings as well.

You might want to have a look at the Fender, Marshall, and Mesa Spice models in the Yahoo LTSpice forum if you haven't already. They were a big help to me in figuring out how to model tube circuits and how to run sound clips through my models.
 
#50
Passinwind said:
My own, of course...:cool:

Bass and mids are active, close to Baxandall, in the feedback loop from the second 6SL7 plate back to the grid. There is substantial treble pre-emphasis involved as well. Treble control is passive, cut-only, between the second and third stages. At midpoint you get the curve shown above, since you are cutting into the baked-in boost in the previous stage. Here's the modeled frequency response curve model with the treble control nearly all the way up:

Link Removed

The idea is dead-clean highs at any setting, and the ability to have variable distortion in the lows and/or mids (or not, as desired), with lots of interactivity between all the tone controls. Works great for fretless, and cabs with tweeters, which is all I care about really...;)

My model doesn't include the input and output transformers, so the actual high shelf/bell top point will likely be a little lower than shown in the model. It'll change some depending on the control settings as well.

You might want to have a look at the Fender, Marshall, and Mesa Spice models in the Yahoo LTSpice forum if you haven't already. They were a big help to me in figuring out how to model tube circuits and how to run sound clips through my models.
Click to expand...

ahh interesting stuff! Going active was my next plan of attack. Hows the wave form look with the active tone stacks? I've found them to sound pretty pleasing in the past. Interaction is something i'd like to avoid though :)

I suppose modelling the transformers isn't really worth the effort
 
#51
MAMMOTHvolume said:
ahh interesting stuff! Going active was my next plan of attack. Hows the wave form look with the active tone stacks? I've found them to sound pretty pleasing in the past. Interaction is something i'd like to avoid though :)
Click to expand...

It's probably much easier to make an active stack non-interactive, actually. This one just sounds better for my purposes at a wider variety of settings though. It's not at all like a Fender stack where one control often negates the other ones completely; mostly what happens in mine is that the mid point shifts around, which I find really useful in a 3-band EQ. I'll post some examples in a bit.

I suppose modelling the transformers isn't really worth the effort
Click to expand...

On the power amp side it definitely would be though. Jensen probably already has models for the ones I'm using, I guess I should see if they want to share 'em. I could just measure them myself, but at this point it seems kind of moot.
 
#52
Passinwind said:
It's probably much easier to make an active stack non-interactive, actually. This one just sounds better for my purposes at a wider variety of settings though. It's not at all like a Fender stack where one control often negates the other ones completely; mostly what happens in mine is that the mid point shifts around, which I find really useful in a 3-band EQ. I'll post some examples in a bit.



On the power amp side it definitely would be though. Jensen probably already has models for the ones I'm using, I guess I should see if they want to share 'em. I could just measure them myself, but at this point it seems kind of moot.
Click to expand...

Yeah your standard fender / marshall type tone stack isn't great. I hate the way they ruin the waveform aswell.

Yeah I was planning on driving each control seperatly, obviously it uses more valves but I think the results will be worth it!

Ahh didn't realise they'd already modelled them, I wind my own so I assumed it'd be a ballache putting in the data for your own
 
#53
MAMMOTHvolume said:
Yeah I was planning on driving each control seperately, obviously it uses more valves but I think the results will be worth it!
Click to expand...

Since the tone stack and main preamp part were a retrofit I didn't have that option. I actually went with a 6SL7 mainly due to filament current constraints.

My preamp is fairly low impedance at the input, since I only use active basses. My tone control scheme reflects that too, I tried to make the frequency centers especially friendly to my main fretless BG so none of the controls are too redundant. At this point I kind of have a separate dedicated rack preamp for each of my three basses, although this one sounds pretty good with all of them.
 
#55
Pocket protectors on...:ninja:

Numbering format: Bass-Mid-Treble, scale of 1-10.

6-8-5:
Link Removed

6-2-7
Link Removed

6-2-5
Link Removed

4-8-3
Link Removed

4-4-3
Link Removed

4-1-5
Link Removed

4-1-2
Link Removed

2-6-7
Link Removed

2-6-2
Link Removed

All of these are potentially pretty useful if used in conjunction with judicious boosts or cuts of the tone controls on my bass. Pretty different from most commercial pieces, true to the contrarian DIY'er ethic. :rolleyes:
 
#56
seanm said:
Wow, the spice model was very accurate. That is a very flat EQ... nice job!
Click to expand...

Thanks Sean. I designed a few dozen different tone control schemes in Spice before I settled on this one. I think I changed one resistor value slightly after building and hearing it, but that's pretty good considering how far out of my comfort zone the whole design was to begin with. Running sound clips of my own bass through the model before building anything was pretty helpful, more so than I really expected. Sounds better in real life though.
 
#59
MAMMOTHvolume said:
Hmm they look pretty great! I'll get on the spice tommorow and see if i can get some interesting things up
Click to expand...

Since I have a pre-EQ gain pot, I can make some distortion happen pretty much anywhere there's a frequency bump. More like a little subtle warmth or grind than overdrive though. As you can tell from my clips, I'm not much of a rocker really.

If I had one or two more gain stages to work with I could've done some really cool stuff, like semi-parametric EQ or nested feedback loops. Spice is really nice for brainstorming things like that without making smoke.
 
#60
Passinwind said:
Since I have a pre-EQ gain pot, I can make some distortion happen pretty much anywhere there's a frequency bump. More like a little subtle warmth or grind than overdrive though. As you can tell from my clips, I'm not much of a rocker really.

If I had one or two more gain stages to work with I could've done some really cool stuff, like semi-parametric EQ or nested feedback loops. Spice is really nice for brainstorming things like that without making smoke.
Click to expand...

Haha tell me about it! I normally map stuff out with the scope so I'm thinking spice is gonna save me alot of time trying to work things out.

I've looked at some semi parametric eq's have you found any that look decent? I kind've wanna avoid to much feedback as it normally ruins the dynamics a little doesn't hurt if you wanna get rid of some noise though :)
 

Latest posts

Top Bottom
Back