Bob said "My paycheck however is still the same."
As the French say plus ca change! What has happened is that there's been a raft of cheap (almost junk) "tube" audio products on the market, such as pre-amps and headphone amps. These can be raided as a source of inexpensive tubes, sockets and other bits. Vintage radios, particularly those with damage, remain a useful source of OPTs and PS transformers too
There's also been a range of really cheap valve guitar amps out there, some of which are close to the target $ on deep discount and screaming out for some rebuilding/modification. This was not the case when this thread starte
As the French say plus ca change! What has happened is that there's been a raft of cheap (almost junk) "tube" audio products on the market, such as pre-amps and headphone amps. These can be raided as a source of inexpensive tubes, sockets and other bits. Vintage radios, particularly those with damage, remain a useful source of OPTs and PS transformers too
There's also been a range of really cheap valve guitar amps out there, some of which are close to the target $ on deep discount and screaming out for some rebuilding/modification. This was not the case when this thread starte
I've a Jaycar KC5370 valve preamp kit which uses a DC/DC booster for B+ on a single 12AX7 (good lord, that kit was from 2003!).
The Seymour Duncan twin-tube series used tiny weeny power transformers back-to-front, generating B+ from about 16VAC, which came from a wallwart
On the more traditional front, don't forget the various "lamington" designs, such as the AVA101 minilamington
The Seymour Duncan twin-tube series used tiny weeny power transformers back-to-front, generating B+ from about 16VAC, which came from a wallwart
On the more traditional front, don't forget the various "lamington" designs, such as the AVA101 minilamington
I like that idea!
A JFET is the lowest-noise device you can find for an e-guitar input stage, you can get some decent sounds out of one even with no valves, and using a solid-state power amp takes away a lot of weight, bulk, and cost.
There were once lots of small or smallish 9-pin triode/pentode valves which might go well between the JFET and the SS power amp, giving you one more valve stage for more "valvey" tone. Availability of those long-obsolete valves might be an issue these days.
I still have a few 6JW8 triode/pentodes in the junk box, err, treasure chest. I used one in the last valve amp I breadboarded - you get both a triode preamp stage, and a pentode output stage, all from one tiny bottle.
I used a Hammond replacement Fender reverb transformer, but a line transformer as Thoglette suggested would do, too. I think the reverb transformer was far and away the most expensive single item in my build.
If one is open to solid-state solutions, a Flamma FS06 Preamp and a class-D power amp board make a surprisingly versatile and good-sounding combination. The Flamma is around $75 USD, currently close to $100 Canadian.
-Gnobuddy
LM1875 chip amp might be a good way.
Supply need only be +/-12V.
There is thread of LM1875 in the chips forum.
Last edited:
I totally agree with you. As a front end amplification stage it's really a great option. I've built a few Jfet boosters in pedal format, and run through a tube amp they can give a good volume boost on the signal, helping drive the tubes but remain very quiet. Also not loading down the e-guitar signal which will preserve it's frequency response. Small signal mosfets can do big gain, but I do find them noisier so I'd tend to use them for subsequent gain stages.
That, sure. And also the input noise current is very low for a JFET.
The input noise current runs through the output impedance of the guitar, and turns into a noise voltage (Ohm's law). For a source like an e-guitar (which has high source impedance), very low input noise current is a good thing. Both triodes (valves/tubes) and JFETs are good at this.
Total noise at the input is a combination of the noise from input noise current, and input noise voltage.
JFET input current noise is always low, so if you pick a JFET that also has a low input voltage noise, then you'll get a good noise figure even with a high source impedance signal.
Funny thing about BJTs and e-guitar loading: though early BJTs really couldn't generate a high enough input impedance to keep, say, a Les Paul guitar happy, that changed several decades ago, when high-gain silicon small signal transistors came along.
As it happens I grew up with lovely little silicon planar epitaxial transistors like the BC149C and it's more modern cousins (BC550C, for instance). These have a current gain of over 500, so it's not hard to get a high enough input impedance to work well with an e-guitar. All you need is an unbypassed 2.2k emitter resistor!
In a practical circuit there's also the loading of the base bias resistor(s), which will lower the input impedance. The easy fix is to use a 4.7k emitter resistor, and a suitable high-value base bias resistor to the positive supply rail. A base bias resistor value in the vicinity of 2.2 - 2.7 meg ohms will be roughly right for an emitter follower using a BC550C and a 4.7k emitter resistor.
But - even though this will give you an input impedance over 1 megohm, a BJT has more input noise current than a JFET. That noise current flows through the high source impedance of a guitar, turning into noise voltage as a result. And so you end up with more input noise than a good JFET input stage.
As for valves, they're handicapped by the high operating temperature. It's an unbreakable law of nature that the higher the temperature, the more thermal noise the device generates. The vacuum tubes we use for guitar have an operating cathode temperature in the range of 1300 degrees Kelvin. A JFET at room temperature is sitting at around 300 degrees Kelvin - roughly one fourth the valve's cathode temperature. Electrons boiling off the triode cathode have about four times as much thermal energy as electrons emitted by the JFET source!
The actual thermal noise voltage is proportional to the square root of the temperature, which means, all else being the same, a hypothetical perfect JFET will only have about half the thermal input noise voltage of a hypothetical perfect triode valve - that's 6 dB less noise.
In real life neither the JFET nor the triode will be perfect, and both will have other sources of excess noise. But the JFET starts out with this rather significant 6 dB thermal noise advantage in its corner.
I agree with Shanx that MOSFETs tend to be noisy little beasts. I'm not sure why, but it has something to do with the way they're manufactured, and we'd probably need a semiconductor engineer from a major manufacturing plant to explain the problem to us.
I've tried both BJT and JFET input stages with e-guitars, and I don't hear nasty gritty clipping from either one. From my point of view, both are better than using an op-amp.
There is a catch, though. The days when semiconductor manufacturers made discrete JFETs intended for low-noise amplification seem to be behind us. Most of the discrete JFETs that are currently available seem to be designed for switching applications, not for low noise.
On the other hand, there are still low noise JFET-input op-amps being manufactured, and these may soon be our only source for low-noise JFET input stages.
Unfortunately op-amps tend to play poorly with e-guitar signals, as we've discussed on another thread, so we're faced with a bit of a Hobson's choice here!
-Gnobuddy
For e-guitar use, op-amps sound bad (harsh and gritty) if allowed to clip, and they also sound bad (too clean) if not allowed to clip.
To my ears, a simple single JFET sounds VASTLY better on both counts.
If you want to use an op-amp and generate good e-guitar sounds, you have to try and force additional circuitry onto the feedback or input networks to try to keep them from clipping, and also to try and make them generate small amounts of subtle nonlinearities.
There is an expression that probably originated in the American South: "That's like putting lipstick on a pig."
IMO, trying to hog-tie and force an op-amp into sounding good with an e-guitar plugged into it, is very much like putting lipstick on a pig.
From previous experience, I know you are unlikely to agree. No worries.
-Gnobuddy
To my ears, a simple single JFET sounds VASTLY better on both counts.
If you want to use an op-amp and generate good e-guitar sounds, you have to try and force additional circuitry onto the feedback or input networks to try to keep them from clipping, and also to try and make them generate small amounts of subtle nonlinearities.
There is an expression that probably originated in the American South: "That's like putting lipstick on a pig."
IMO, trying to hog-tie and force an op-amp into sounding good with an e-guitar plugged into it, is very much like putting lipstick on a pig.
From previous experience, I know you are unlikely to agree. No worries.
-Gnobuddy
Agreed.
From previous experience, I know you are unlikely to agree. No worries.
-Gnobuddy