When I look through some of the glossy "high end" hifi magazines, I get the impression the hifi world isn't any different
I think scientists and engineers, having achieved virtual perfection in Hi-Fi amps, mostly left the world of Hi-Fi some forty years ago. After that, the inhabitants took over running the asylum - for the most part, the technically illiterate enthusiasts and audiophiles took over, with a few exceptions here and there. So now Hi-Fi is mostly a world of superstition and nonsense and false advertising, not science or engineering....impression the hifi world isn't any different
There have been technical advances since then - amps got smaller, much cheaper, and when class D arrived, much more efficient. But not better-sounding, because good amps were already better than the human ear's limitations five decades ago, so they were already perfect then.
You can't get better than perfect, but, it turns out, you can argue about the "sound" of red wire insulations v.s. blue plastic insulation for forty years.
The funny part is that loudspeakers and listening rooms are still very much imperfect. That's where we might potentially be able to actually make a difference, if we tried. But speaker engineering is difficult, and requires expensive tools.
-Gnobuddy
I 'm with you there Gnobuddy but I'd go seven decades.
Since then we've had to put up with snake oil salesmen trying to convince us that their cheaper, nastier sounding versions are somehow better. Worse, we had "engineers" (and I am one) designing "perfect" stuff every time a cheaper technology turned up, when in reality it takes a decade or two to work out the "gotchas" in each new technology. And we still have expensive products being released based on the first "typical circuit" in the datasheet. Oh, and remember, speakers are an eight ohm resistor.
"Perfect sound forever" Well, until the marketing blokes decided it was time to sell something else.
My marker for when "the Hi Fi industry" and actual pursuit of sound quality officially separated paths is in April 1961 when we were foisted with an inferior option for FM stereo. It's been a story of ever-worse compromises ever since. And the brains and money drifted to other markets, first integrated chips, then digital and now software. So the %ge of EEs who actually have a clue is steadily diminishing. Particularly as very few will have heard actual live music (as in, with no electronics, per GnoBuddy's other thesis)
But in the market cheap and convenient have triumphed every time. Throw in some shiny and whooo and you're really going to win. I'm sure the white van speaker scam is still going strong.
It's thus hardly surprising when someone has a "Emperor's clothes" moment
(for example, hearing some well maintained BBC studio grade equipment from the '60s) that they tend to overreact to The Official Truth and are at risk of drinking someone's get-rich-quick Kool-Aid.
(Particularly when they don't know what a thousand pound would buy the BBC in 1960.
)... and still there's a huuuge section of this forum dedicated to building new hifi amplifiers ......
back to the ruby: I put the prototype circuit in a wooden box and rapped it with tin foil. it's now absolutely quiet - no noise pickup at all. before i did that i found out that the circuit is especially noisy (weird screeming sounds) when there's nothing connected to the input. these noises go away as soon as a guitar is plugged in or even if the input is shorted to ground with a finger.
i now use it in the bassman configuration with 22ohm headphones and a 680R resistor to ground for attenuation. as gnobody said, the LM386 needs to be driven hard to distort - so a 10R or 20R resistor to ground would certainly NOT be enough to attenuate sufficiently.
back to the ruby: I put the prototype circuit in a wooden box and rapped it with tin foil. it's now absolutely quiet - no noise pickup at all. before i did that i found out that the circuit is especially noisy (weird screeming sounds) when there's nothing connected to the input. these noises go away as soon as a guitar is plugged in or even if the input is shorted to ground with a finger.
i now use it in the bassman configuration with 22ohm headphones and a 680R resistor to ground for attenuation. as gnobody said, the LM386 needs to be driven hard to distort - so a 10R or 20R resistor to ground would certainly NOT be enough to attenuate sufficiently.
All our houses are glass: I'm busy deciding which 50 year old guitar amp I'm going to almost-clone next... and still there's a huuuge section of this forum dedicated to building new hifi amplifiers ......
After all, the few tweaks I will make A Very Important Change(tm) to the sound 
(If I'm lucky I'll both not kill myself and learn something )( I did actually spend more time today playing guitar than messing around with circuits. But perhaps not as much time as I spent on the various guitar forums )
Last edited:
This is normal. Guitar amps have sensitive, high-impedance inputs, and those pick up all sorts of noise from the air.
That noise can be very annoying, and fortunately, there is a pretty simple fix: use a switching jack at the input, and wire it so that the input of the amp is shorted to ground (via the jack's switch contact) when nothing is plugged in. Now your amplifier will stay quiet when the guitar is not plugged in.
-Gnobuddy
By the way - if you find the Ruby sounds too harsh in your headphones when the amp is distorting, you could try adding a capacitor across the headphone jack. For 32-ohm headphones, with left and right wired in parallel, somewhere between 2.2uF and 4.7uF will probably be about right.
The capacitor will roll off very high frequencies, reducing harshness. (A real guitar speaker does the same job even more effectively.)
Most amplifiers don't like having a lot of capacitance connected between the output and ground, but in your case, the 680R resistor you're using will keep the amplifier from "seeing" the additional capacitor. (The cap is wired across the headphone jack, so the 680R resistor is in series with it.)
Ideally, the capacitor you use will be a non-polarized type. Film capacitors will work (but are bulky and expensive). My suggestion is to use a multi-layer ceramic chip (MLCC) capacitor, which is small and inexpensive. Like this one: FG18X5R1E225KRT00 TDK Corporation | Capacitors | DigiKey
Here are some more similar caps ranging from 2.2uF to 4.7 uF: Capacitors | Ceramic Capacitors | DigiKey
-Gnobuddy
The capacitor will roll off very high frequencies, reducing harshness. (A real guitar speaker does the same job even more effectively.)
Most amplifiers don't like having a lot of capacitance connected between the output and ground, but in your case, the 680R resistor you're using will keep the amplifier from "seeing" the additional capacitor. (The cap is wired across the headphone jack, so the 680R resistor is in series with it.)
Ideally, the capacitor you use will be a non-polarized type. Film capacitors will work (but are bulky and expensive). My suggestion is to use a multi-layer ceramic chip (MLCC) capacitor, which is small and inexpensive. Like this one: FG18X5R1E225KRT00 TDK Corporation | Capacitors | DigiKey
Here are some more similar caps ranging from 2.2uF to 4.7 uF: Capacitors | Ceramic Capacitors | DigiKey
-Gnobuddy
Thanks for the suggestion regarding the capacitor across the headphone jack. I'm experimenting with that as we speak.
This amp sounds great but if i want to distort it fully it gets wayyyy to loud - also using a 8'' speaker cab. So i'll use a pot at the output for sure to control overall volume. Now which is the 'better' way to wire it? Attached are two drawings. V1 is the "series resistor" version and V2 would be a voltage divide "in front" of the speaker. What are the pros and cons of those options?
This amp sounds great but if i want to distort it fully it gets wayyyy to loud - also using a 8'' speaker cab. So i'll use a pot at the output for sure to control overall volume. Now which is the 'better' way to wire it? Attached are two drawings. V1 is the "series resistor" version and V2 would be a voltage divide "in front" of the speaker. What are the pros and cons of those options?
A year or two ago I built a tube guitar amp with about 0.1 - 0.2 watts of output power. Connected to the 10" guitar speaker in one of my Fender amps, in my apartment, it was too loud!
So yeah, 0.5 - 1 watt or more of distorted LM386 output into a sensitive speaker might be much too loud for a home or apartment.
The simplest thing to try is to run the LM386 from a lower voltage. It's rated for operation down to 4 or 5 volts, so you might try 4.5 volts from three alkaline cells in series, or a generic 5V USB charger / power supply you probably already have sitting around. This will reduce the output power a lot, and may make the amp quiet enough for your purposes (with the loudspeaker). With the headphones, you'll still need the series resistor.
(The datasheet says 125 milliwatts into a 4 ohm speaker at 4 volts power supply; depending on your speaker sensitivity and the SPL acceptable to you, this may still be too loud.)
If you like the USB power supply idea, you might need one of these tiny breakout boards: SparkFun microB USB Breakout - BOB-12035 - SparkFun Electronics
This is too big a question to answer without writing a small book in the process, but I'll give you a quick personal opinion, and then point you at some resources for the facts.
The opinion part: for this little guitar amp, it's not worth trying to design the right voltage divider. Stick with V1, and for the loudspeaker, my best guess (I don't know how loud you want it) is to use a 50 ohm or 100 ohm linear pot, if you can find one. Ideally the pot would be rated at 1 watt or more. For headphones, as you've already found out, you need much more resistance, maybe as much as 1 kilo ohm.
More opinion: if you can find an L-pad in your country, this is a specialized type of voltage divider (actually a pair of them in one housing) already designed to do exactly what you want: Speaker L-Pad Attenuator 15W Mono 1" Shaft 8 Ohm
Enough opinion, on to the facts!
1) The first key concept is the Thevenin resistance of a voltage divider, and here's a video that will answer that. It tells you what the voltage divider (your diagram V2) does by itself: YouTube
2) The second key concept - and a much more complex one - is finding out what happens to a loudspeaker when it's driven by the Thevenin impedance of a voltage divider. This requires understanding the Thiele-Small loudspeaker model, and then understanding how that intereacts with the output impedance of the amplifier (or, in your case, voltage divider V2).
Getting a grasp on the Thiele-Small model is a fairly big task, which I'll leave to you, but in a nutshell, the frequency response of a loudspeaker changes when you increase the source resistance feeding it. With a single full-range speaker, like the one you're using, you can expect a bump in the bass, and an increase in the treble.
Here's a somewhat biased and subjective article on the subject, from a man who apparently likes boosted bass and boosted treble, and thinks that is a big improvement over a flat frequency response: Amplifiers: Solid State amps verses Valve amps
For Hi-Fi, unnaturally boosted bass and treble, along with boomy one-note bass, are not desirable at all. But an electric guitar into a little 8" speaker is unlikely to suffer too much from this - this is not Hi-Fi, this is simply about the sound you like. If it sounds good to you, then it's good!
-Gnobuddy
i managed to finish this project last week. so here's my ruby headphone / speaker amp
I ended up using the original schematic with the following additions:
1.) output can be switched between headphones (680R to ground) or speaker (no additional resistor)
2.) there's a 1k pot to adjust output volume (in series with the 680R for headphones)
3.) 2k5 gain pot
4.) ceramic cap in parallel with the 220uF electrolytic between V+ and GND
5.) 10uF electro between pins 1 and 8
it wasn't easy to fit it all in one box but i managed and i think it looks fine. most of all i just loooove the sound of it. the cleans are fantastic and the overdrive as well .... with these modifications and boxed in the enclosure the circuit is as quite as a mouse. no noise pickup or anything.
things i like:
-) the sound
-) the additional output volume. i think the circuit is a bit useless without an output volume as the distorted tones will we way to lound for anything but a stage ...
things i'll add for the next version
-) i used a 2k5 pot for gain but i really kind of use it as an on/off switch. maybe it'd be better to use a switch for high / middle / low / gain
-) the 1k log pot at the output doesn't give me a useable range - it's too sensitive toward the end. I think someone mentioned using a reverse log and that would probably be the better choice.
again: thanks for all the help! this certainly has been a fun project and i'll be back with a version 2 soon
I ended up using the original schematic with the following additions:
1.) output can be switched between headphones (680R to ground) or speaker (no additional resistor)
2.) there's a 1k pot to adjust output volume (in series with the 680R for headphones)
3.) 2k5 gain pot
4.) ceramic cap in parallel with the 220uF electrolytic between V+ and GND
5.) 10uF electro between pins 1 and 8
it wasn't easy to fit it all in one box but i managed and i think it looks fine. most of all i just loooove the sound of it. the cleans are fantastic and the overdrive as well .... with these modifications and boxed in the enclosure the circuit is as quite as a mouse. no noise pickup or anything.
things i like:
-) the sound
-) the additional output volume. i think the circuit is a bit useless without an output volume as the distorted tones will we way to lound for anything but a stage ...
things i'll add for the next version
-) i used a 2k5 pot for gain but i really kind of use it as an on/off switch. maybe it'd be better to use a switch for high / middle / low / gain
-) the 1k log pot at the output doesn't give me a useable range - it's too sensitive toward the end. I think someone mentioned using a reverse log and that would probably be the better choice.
again: thanks for all the help! this certainly has been a fun project and i'll be back with a version 2 soon
Well done! I love my Ruby, it's an insanely simple circuit which is a whole pile of tone fun.
(Now if only I can fit in Gnobuddy's flat tone circuit. And a reverb chip. And a Tremolo. And....
)I wanted reverb for a small solid-state battery operated guitar amp project I've been slowly building. The reverb solution had to be compact and affordable, and the solution I chose was to buy one of these: Amazon.com: Donner Digital Reverb Guitar Effect Pedal Verb Square 7 Modes: Musical Instruments
IMO it sounds good, and seems well built, all for under $50 USD.
Here's a You Tube video review from Shane ("InTheBlues"): YouTube
-Gnobuddy
Thanks. GWills built an slap echo/reverb around a PT2399 e-bay module so I was "gunna" use that when I get ä "round tuit. But life has been "other than planned" recently so a cheap, good sounding pedal may just be the ticket.
Amazing! Another piece of electronics drops to 10% of the price we had formerly been trained to pay for it. And without having to DIY and hand-solder a QFP chip, either.
Ain't that the truth. And the second wave of the pandemic will probably hit later this year, now that we're starting to relax restrictions and increase the coronavirus transmission ratio again. Oh happy days.
-Gnobuddy
Again, I'm searching for Jfets. This time for a tillman amp or fetzer valve. Both of them recommend the use of a J201. Can the J111, J112, and J113 JFETs also be used in those circuits.
Probably not. I base that on having the J111 data sheet in my browsing history and not having bought them - but check the curves for yourself (IIRC the V vs I curve is way out)
But never fear! SMD package J201s are all over the place and you can get conversion PCBs - some people are flogging pre-made boards (like GuitarPCB)
If you really, really must have TO-92 then you're limited to NOS via reputable vendors (and be prepared to pay accordingly).
Found another little intro article on how to solder SMDs. I like to glue them first then add solder but this will work too... read on Coda Effects on Supro clones and SMD soldering
Firstly: All JFETs available to us DIY types have extremely sloppy tolerances. The J201 datasheet, for example, says that VGS(off), a key JFET parameter, can vary from -0.3V to -1.5V. That's a five-to-one range!
What this means is that if you plug fifty different J201s into the same Tillman amp circuit, chances are many of them will barely work, simply because of the random manufacturing variations between different J201s.
Same story with the Fetzer Valve. I've read through the math in the original white paper, and a proper Fetzer Valve uses resistances fine-tuned to that particular FET - there is no "one size fits all" resistance that works properly with all J201s.
The Runoff Groove guys put together lots of interesting FET circuits, but their "Fetzer valve" is a mess; they clearly didn't understand what they were doing, and to compensate for the sloppy JFET tolerances, they put a trimpot in the drain circuit of each JFET. This is the wrong place - the trimpot should have been between source and ground, not between drain and Vcc.
In short: there is no guarantee that the (poorly designed) Tillman circuit, or the (poorly designed) ROG Fetzer Valve will work with a J201! The only way to make these circuits work properly is to individually tweak the resistance values to match the specific J201 you happen to use.
And now we come to the question you asked. The answer is that the one of the key JFET parameters (VGS(off)) has a lot of overlap between the J201 and the J113. The other key parameter (Idss) is a bit different, in general the J113 will "want" to draw a bit more current than a J201 in the same circuit.
BUT - the same caveat applies as before. As we've just discussed, these primitive JFET circuits are not guaranteed to work with the original J201 in them, and you have to tweak resistance values anyway. As long as you're tweaking resistance values, you might as well use a J113 rather than a J201, as the J113 is still available in through-hole packaging (no faffing about with SMD devices and adaptor boards necessary.)
There are two ways to go about doing the tweaking:
1) Measure Igss and VGS(off) for the JFET you have, and calculate the proper resistance values from that. You have to learn a little about JFET theory to do this. The measurements are very easy with a solderless breadboard, a power supply or 9V battery, and a cheap DMM.
2) Stick a trimpot between ground and JFET source, ground the gate through a 1 meg resistor, and twiddle the trimpot until your circuit behaves the way you want. For audio circuits, this often comes down to adjusting the drain voltage so that it is roughly halfway between the source voltage and Vcc.
Good luck with your projects, whichever approach you choose to take!
-Gnobuddy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.