'Rascal Mk1' Low Voltage Tube Amp build thread! (with PCB files & images)

[sebsongs]

Hi there.

I'm building Slackers Rascal Mk1 tube amp. I've actually already built it and tested it, and it worked first time!
Slacker (Ian) has been kind enough to guide me through a couple of steps I was unsure of, and I thank him greatly for that.

Some info:
I made my own layout based on Ian's schematics. My goal was to make it small, and it really is! 81x52 cm with the tube sockets attached to the board.
I've chosen to use the Fender 22921 as output transformer, and driving a 4 ohm speaker, it works really good.
I've ordered a Jensen C8R to use as loudspeaker in the amp, and I've also already built the cabinet for this thing.
I'm supplying the circuit with 36V B+ and about 11V for the heaters. With some tweaking of the LM317 circuit I could make it 12.6 Volts, but I'm not sure that there would be any significant difference in sound. Any tips here would be greatly appreciated.

So, I thought that I might as well share the files with you guys, so here they are:
Schematic PDF: http://sebsongs.com/files/rascal/RascalMk1_sch.pdf
Layout PDF: http://sebsongs.com/files/rascal/RascalMk1_brd_solder_side.pdf
EAGLE Schematic: http://sebsongs.com/files/rascal/RascalMk1.sch
EAGLE Layout: http://sebsongs.com/files/rascal/RascalMk1.brd

And of course I have some nice pictures also:

[Ice-9]

That looks really nice, and the cabinet looks beautifully made, I have one question about the schematic. As the schematic is Valve circuit why put in that emmiter follower TRANSISTOR before the output valves, surely it could have been kept all valve ?

[slacker]

That looks great Sebastian nice work.

Mick, the transistor is in there because I could only get the output section to work with about 1k for the grid resistor and the preceding preamp stage couldn't drive this. I'm not sure why the grid resistor needed to be this low, it might just of been the particular valves I was using or more likely it's just due to running them at such low voltage. It might have been possible to use a valve as a follower but that would of meant either losing a gain stage or adding an extra valve.

The original thread with some soundclips is here if anyone's interested.

http://www.diystompboxes.com/smfforum/index.php?topic=70802.0

[wavley]

I was looking to build an amp using a 12au7 driver, THANKS!

When I first looked at the schematic I was going to ask you what the 100n cap and 47k grid on V1 was all about (instead of the standard 68k resistor and no cap) but I looked and some princetons actually had a coupling cap here.  Guess I'm just used to a to jack input with the 1meg on one and not the other each with it's own 68k when it comes to a fender.  I've never tried it the way you have it, I'm looking forward to seeing what it sounds like.

[Johan]

I could only get the output section to work with about 1k for the grid resistor and the preceding preamp stage couldn't drive this. I'm not sure why the grid resistor needed to be this low, it might just of been the particular valves I was using or more likely it's just due to running them at such low voltage.

you experienced "spacecharge". when runing low voltages the grid wants to go negative, shuting the tube off. on the early 9volt tubedriver pedals, they tied the grid to a positive voltage to overcome this. I think it's even mentioned in their patent. PAIA has a great writeup on low voltage tube circuits
niced looking build.. 
J

[kurtlives]

Looks cool, I like!

You don't need C3 though. Also if you move R4 before R3 it still acts like a grid leak resistor and keeps the grid from going postive. ALSO then the two don't act like a voltage divider and kill a bit of gain. Maybe you know...

Have you experimented with making C12 smaller?

[Taylor]

Very nice-looking PCB. Did you get it manufactured or make it yourself? If DIY, how did you do the soldermask?

I bet if you made the boards available, you'd get some people interested. I kind of want one, even though I play bass so this amp is not that practical for me. 

[slacker]

You don't need C3 though.

C3 is there because I got some DC on the grid of the first stage, I think this is quite common with low voltage circuits.

you experienced "spacecharge". when runing low voltages the grid wants to go negative, shuting the tube off. on the early 9volt tubedriver pedals, they tied the grid to a positive voltage to overcome this.

Useful info thanks.

[wavley]

C3 is on the 5D2 version of the Princeton, I've never used that particular amp, but I'm a bit curious about how it sounds.  Low voltage tube is new to me, I've been a high voltage tube guy for years.  You have a point about C12, it seems to more of a Hiwatt value if you're doing a component for component swap, but I really feel that such a low power circuit would really benefit from a stiff power supply because I like some thump in my low end.  Some folks might really like a looser sounding amp so that might be a nice value to play with.

[kurtlives]

Ah forgot its LV.

I should know. I just built a SE amp with a VVR control. At the low registers you get DC on the grids, added a blocking cap too.

[sebsongs]

Thanks for all the kind words!
And thank you Ian for pitching in about the schematic. I have not designed it, as I said earlier. It's Ian's work totally.

As for the sound, I've yet to play it through my target speaker (should be delivered to me today) but I've tried it through a Hifi 70W mid/bass speaker and already there this thing sounds awesome! When I get my Jensen C8R (which is lightyears easier to drive compared to my hifi speaker) I'll probably get more of the real guitar tone and hopefully some more volume too.

I will record the amp when it's completed, and share the files here too.

Very nice-looking PCB. Did you get it manufactured or make it yourself? If DIY, how did you do the soldermask?

I bet if you made the boards available, you'd get some people interested. I kind of want one, even though I play bass so this amp is not that practical for me. 

Thanks Taylor! I did not make the PCB myself, I sent my layout to a company here in Sweden called PCB Manufacturing (logical name!) and had it done for something like 30 bucks including drilling. I have since coated it with some Green Coat to stop the copper from oxidizing.

I made the soldermask (by that you mean layout?) in Eagle PCB. It's a free software that is really darned good for small projects like this.
As I mentioned in my first post, I've shared the files for anyone to use and modify.
At the moment I am not going to sell PCBs but if enough people is interested it might be possible in the future.

NOTE: For anyone that's going to try this out, remember that I've put the tube sockets on the SOLDER side and all the rest of the components on the component side. If you want to change this you have to redo the layout. The reason that I did it was that I wanted to get the tubes facing out of the box and still be able to have the wires and components facing in in the box. Anyway, be sure that you recognize this when building, and also be sure that the tube sockets fits. I've made a custom fit to my sockets (purchased from musikding.de) and it may not fit with your sockets.

/Sebastian

[PRR]

Why no cathode resistor on V2?

This may relate to why you had trouble driving the grid. When grid is negative of cathode it draws no current. When grid is positive of cathode it draws large current, indeed like a 1K resistor.

Higher-voltage tube amps generally stand the cathode up on a resistor. Then it is easy to bias the grid "negative" by returning to common with a large resistor.

Low-volt tubes won't pull large current. The maximum zero-grid current for a pair of 12AU7 at 36V supply is about 13mA. Your optimum "mid-point" bias for class A1 power operation is thus 7mA. The grid-bias for 35Vp-k and 7mA per pair is 1.5V-1.8V (hard to squint, and subject to wide variation). A suitable cathode resistor is 220 ohms per pair.

This will be difficult on that PCB layout.

[sebsongs]

Why no cathode resistor on V2?

This may relate to why you had trouble driving the grid. When grid is negative of cathode it draws no current. When grid is positive of cathode it draws large current, indeed like a 1K resistor.

Higher-voltage tube amps generally stand the cathode up on a resistor. Then it is easy to bias the grid "negative" by returning to common with a large resistor.

Low-volt tubes won't pull large current. The maximum zero-grid current for a pair of 12AU7 at 36V supply is about 13mA. Your optimum "mid-point" bias for class A1 power operation is thus 7mA. The grid-bias for 35Vp-k and 7mA per pair is 1.5V-1.8V (hard to squint, and subject to wide variation). A suitable cathode resistor is 220 ohms per pair.

This will be difficult on that PCB layout.

Thanks for the info!

Indeed, that would be difficult on my PCB.
Would the changes result in any significant boost of power, or is it just to get rid of the transistor?
Anyway, I'm not making any changes to the amp now. It works and sounds good.
I may be interested in getting more juice out of it in the future, but maybe I'll just build another one then.

/Sebastian

[slacker]

Why no cathode resistor on V2?

I can't remember why I built it like that, I think I tried a cathode resistor but can't remember the results. To be honest I didn't really know what I was doing and just played about until I got something that worked. Since then I've built a real valve amp and have at least some understanding of how valves work so I might revisit the idea.

[sebsongs]

There has been some really great progress on my amp!

I had a lot of ripple noise coming from my power supply.
I solved it by putting in a LM350 Voltage regulator circuit. Now the amp gets about 32 Volts of good clean DC Voltage and it is now dead silent when I'm not playing.
The LM350 gets a bit hot despite being mounted on a heatsink. It can handle 125 degrees celsius in operational mode though and I'm sure it is cooler than that.

Here's some pictures of my now finished amp and my old Sigma Les Paul:





Next step for me is to try out another valve in the preamp section.
It should be possible to swap to 12AU7 or 12AY7 instead of the 12AT7 sitting there now, right?

/Sebastian

[Brymus]

Excellent looking amp,congrats !

[wavley]

Good job sir!

[slacker]

The finished amp looks great. I've used 12AT7 and 12AU7 in either positions, the 12AU7 is cleaner sounding especially in the power amp section. You can use a 12AX7 for the preamp section, it doesn't work well in the power amp on mine, but you could try it.

[sebsongs]

Thanks guys! I'm very pleased with the looks and sound of the amp. It has been a lot of work, but totally worth it.

Slacker: OK, maybe I'll try and get a hold of some 12AU7's then.

/Sebastian

[PRR]

> I had a lot of ripple noise coming from my power supply.

The power supply tries to do two things at once. Clever, but has limits.

The plates would like as much voltage as possible. But their current draw is VERY small. Maybe 10mA. Between the raw voltage and the "Tube amp VCC" point, insert 220 ohms in series and 470uFd 50V from VCC to ground. This will clean the plate power a LOT, with very little (2V) voltage drop.

You need power for heaters. There's several ways to wire your four (two per bottle) heaters. 6V at 0.6A, 12V at 0.3A, or 24V at 0.15A.

Heaters are BY FAR the largest power drain here. So you should wire heaters for best-match to your supply voltage. The plan shows "max 37V". The plan shows heaters wired for 12V 0.3A. If you really put 37V in and 12V out (25V drop) of the LM317 regulator, that's 7.5 Watts of heat which does need a large heatsink. If your supply is 27V or higher, it might be sweeter to re-wire the two 12V heaters in series. (This will be difficult on that PCB layout.) Then you need only drop 37V-27V= 10V at 0.15A, 1.5 Watts, 1/5th the heat and heatsinking. (Change regulator resistors for 25V out.)

And if your supply is halfway steady (even as steady as house-voltage), then instead of regulator you could just figure a series resistor to drop the voltage. Probably around 50 ohms at 2 Watts. Add 1,000uFd 35V to ground so heater leads are clean, don't buzz-up adjacent audio.

> The LM350 gets a bit hot

If your LM350 doesn't boil spit, then it's fine.

> swap to 12AU7 or 12AY7 instead of the 12AT7

Use 12AU7 12AT7 12AX7 as you please. (12AY7 does not seem appropriate.)

The V2 power amp seems to favor 12AU7. 12AT7 may or may not work "well". 12AX7 will make much less "power"; in some situations, less may be more fun with fewer complaints from neighbors.

V1a and V1b boost your 10mV-200mV guitar to the 3,000mV level the power stage needs. 12AU7 may not give enough gain to bring faint picking to full power. 12AX7 may be too much gain; in some situations, "too much" gain is just right (mondo distortion from gentle picking).