This is about an amp, but I think the theory here is applicable to stompboxes too:
The Valvetronix from Vox uses a single 12AX7 in an ultra-low power amp configuration, to get the power amp saturation tone that we guitarists enjoy, with a clean power amp section to drive the speakers.
So here's what I'd like to see: a stripped down, bare-bones circuit fragment of the Valvetronix's low-power tube section (and a 2-tube, class AB version too, why not), an elegant way of simulating speaker load to the amp circuit, and a line level output so it could be used as an effect or in front of a larger power amp, like Vox does.
Low-volume power tube saturation tone has been discusses at length on www.amptone.com, but I have yet to see any practical examples for DIYers. Let's get this thing going, in the name of good tone, preserving hearing, and saving energy by reducing power consumption!
I have a 15 watt Valvetronix which sounds really good for such a small amp. The different amp simulations also sound great, although I put it on the modern Marshall sound and leave it there. For a 15 watt it's REALLY loud.
Axl
Hi
I, too, have a Valvetronix 15. It's got quite a few fun settings.
re:
Quoteand a 2-tube, class AB version too, why not
Because the 12AX7 is a twin triode, you only need one. That's why the Valvetronics can simulate Class A and Class AB amps with just one tube. Unfortunately, they put a bunch of op-amps between the input and the 12AX7 so that you can't overdrive it like a valve amp.
But.... the 12AX7 part of the circuit has LOTS of potential. I've just been looking at the schematic and realised for the first time that they only run that 12AX7 at 15V (grid to plate). (for some reason, they run the plate at ground and the grids at -15V. That is despite having +15V on tap.) For stompbox use, it would make sense to use +30V.
The input stage is easy enough - a long tailed pair feeding the two grids (with out-of-phase signals). The "loads" are twin op-amps in inverting form, each with 330k input resistance. Too simple. The bias is via a fixed 5.6k resistor shared by the two triodes.
To get the "tone" of different amps, there is the A/AB switch and a couple of filters (2 x SPDT switches) in a feedback network that goes from the output of the plates to the base of one of the long-tailed pairs (the out-of-phase one).
It's all potentially "doable" in a stompbox. A key ingredient will be a suitable power supply. Maybe a 12V DC wall wart. 150 mA for the heaters, and allow 50mA for the rest.
cheers
Where can you find a schematic for the valvetronix amplifiers?
Vox ?
I'm also very curious for the schematic... A quick search turned nothing, anyone got a link they could share?
I was thinking of picking one up to use as an FX unit for live gigs, that way I'd have an extra amp in case of problems.
I can get more sounds and functions from my current rack unit so I abondoned the idea.
I've been using a Digitech Valve FX for the last ten years or so.....it's got a tube and is exremely flexible and sounds pretty good after alot of re-programming.
I've thought about jacking up the plate voltage to screw sound with the tube distortion sound to improve it.
Quote from: brett on January 06, 2008, 07:47:42 AM
The input stage is easy enough - a long tailed pair feeding the two grids (with out-of-phase signals). The "loads" are twin op-amps in inverting form, each with 330k input resistance. Too simple. The bias is via a fixed 5.6k resistor shared by the two triodes.
To get the "tone" of different amps, there is the A/AB switch and a couple of filters (2 x SPDT switches) in a feedback network that goes from the output of the plates to the base of one of the long-tailed pairs (the out-of-phase one).
It's all potentially "doable" in a stompbox. A key ingredient will be a suitable power supply. Maybe a 12V DC wall wart. 150 mA for the heaters, and allow 50mA for the rest.
cheers
Thanks Brett! I am looking at the schem as well and with your comments it is starting to make more sense. Once the power supply is taken care of, the fun parts to tweak would be the input stage and the output "load."
I'm not too familiar with the long-tailed pair operation. Can you just increase the signal before the 'pair to overdrive the tube? ...Or would it work better to put a gain stage after the pair? I would hate to eliminate that section as it looks like a good way to simulate A/AB textures.
Then of course there is the output section. It seems like a small-ish transformer could be used to simulate a speaker load, but only if it would work (in terms of output impedance and frequency response) to drive an amp or effect input.
I bet a wall wart supply would work well, but I'm guessing you'd have to regulate the non-heater voltages to avoid noise. Something like what the Zen amplifiers have might work, or some other DIY-friendly approach. Maybe separate wall warts for the heater and main voltages? Or would that be overkill? Someone more knowledgeable on tubes than I could probably answer this.
Thanks in advance :icon_biggrin:
Oh yeah...
http://www.valvetronix.net/forum/link-to-ad15-30vt-servicemanual-vt794.html
please enjoy responsibly.
Thanks! :)
Hmmm.... I was planning on using an LM3886 for a guitar amp - I now see that it is indeed a good idea. :D
I thought the "Valve Reactor" circuit was supposed to have a transformer in it? ???
No DC on the plates? What's supposed to be going on there? Are they wired as diodes perhaps? Odd.
Like it would be too much trouble to just do it properly... sheesh!
Hi
A couple of comments.
Concerning the "no DC on the plates" thing: it's the voltage relative to the grid that counts, so a grid voltage of -15V and plate at ground (the Vox circuit) is the equivalent of having the grid at ground and the plate at +15V. There would be advantages of running the plate at higher voltages, and the ready availability and low cost of 7824 regulators meakes me think that 24V might be good. Do some inkjet printers use power supplies that are up around 24-48V? I'd be regulating the heaters as well, for extra low noise. (and star ground everything, etc...).
Concerning an output transformer: I don't think it will work (on its own), because there are low conductance issues at low plate voltages. A load of at least a couple of hundred kohms is needed. The Vox circuit uses a 330k load (op-amp). I've also seen a circuit where the load is a JFET with a 1M pull-down resistor (which effectively makes for a 1M load). The JFET idea is rather elegant and keeps the parts/pins count low.
I think the LM3886 is there because they are convenient and low cost, not because they are the best quality. I'd be plugging this circuit into an existing "clean" amplifier. Maybe a fat PA amp.
cheers
If I understand you correctly, you're saying I could use a jfet buffer as a load? That really is elegant. Finally this circuit is starting to make sense. For me, I want to use a class-D/class-T power section because they are so stingy with the power... with luck, the whole thing could run off of a single wall wart!
I'm going to forget this circuit. It disappoints me.
Hi
Class D works with some program material because there are only short periods that are loud.
Loudness at low power is mostly associated with efficient speakers (driven by Class AB or B amps). The LM386 is efficient (ie. low quiescent current), but produces almost 1W at 12V.
cheers
Here's a rough draft of a circuit that I drew up. It's so rough that it probably has bugs, but it's a starting point for discussion
It's a JFET source-follower input, followed by a long-tailed pair used as a phase splitter. Very simple.
The pair "feed" the grids of either one (Class A) or two (Class AB) triodes in a 12AX7. The plates of the triodes feed buffers (op-amps).
A DPDT switch could be used to (dis)connect the second triode and change the bias (bypass or parallel a 4.7k resistor with a 470 ohm resistor). The bias of both triodes needs to change, so pins 3 and 8 should be tied together.
Also, the key to getting such a circuit to actually sounding a bit like like the power section of a tube amp is feeding some of the output from after the op-amps back to the base of Q3. I suspect that the feedback should be filtered, to help the output reproduce some of the frequency "bumps" and limits seen when an output transformer is used.
cheers
(http://www.aronnelson.com/gallery/d/18488-1/Triodulator.jpg)
Nice work, Brett! Thanks for taking the time to put something together for everyone. I am itching to order some tubes to start trying out ideas. In the meantime, it looks like the best way to have both class A and AB is going the route of long-tailed pair/feedback from buffers.
So from what I can gather from Dano's post on the "9V Tube boost" the tube can be capacitor-coupled to the output and still get good tones... So for this circuit I'm guessing that the opamp buffers are mainly to keep the feedback filtering from loading down the triodes. Am I on the right track?
On another note, I'm still interested in simulating the tone of output transformer saturation somehow. It seems like a simple choke between the tube and buffers would *theoretically* saturate the same way as an output transformer (assuming you used a ferromagnetic choke), and the buffer after it would ensure it could still drive the feedback load properly.
Or looking at it differently, you could put a choke/inductor after the buffers, but before the feedback loop. Maybe that would work better. This is new territory for me, so I'm just throwing crazy ideas out there, hoping one sticks.
Thanks!
OMG i would really love to see how this project goes! the valvetronix sounded awesome!
please work on this gurus...
I have a Vox VR30R amp witch has the valvereactor technologie inside.
I looked around for schematics that does the trick and i found this really usefull.
http://digilander.libero.it/essentialaudio/hybrid_main_power_amplifier_schem_picture.htm (http://digilander.libero.it/essentialaudio/hybrid_main_power_amplifier_schem_picture.htm)
Hope it helps.
Ronald
Quote from: RonaldB on January 08, 2008, 02:24:11 AM
I looked around for schematics that does the trick and i found this really usefull.
http://digilander.libero.it/essentialaudio/hybrid_main_power_amplifier_schem_picture.htm (http://digilander.libero.it/essentialaudio/hybrid_main_power_amplifier_schem_picture.htm)
I like that one. It's very clever. :)
Ohh... it's class A. Very hot running. Maybe need some modification to be practical for guitar, but the basic concept is very good.
Someone on ssguitar.com apparently had the same idea:
http://www.ssguitar.com/index.php?topic=500.msg3507
(towards the end of the last post on this page)
...Still looking to see if it was further articulated on that page.
Best of luck,
Alex