In general you cannot do this - you need to have grounds of all the devices connected one way or another. The reason is that the voltage by definition is not an electric potential, but it's a difference of potentials between two points (that's what you mean when you say "voltage across a resistor" - it's a difference of potentials between two ends of a resistor). We always assume a point of "zero potential" to be the ground point of our circuit, but in reality this is rather artificial - you could say that the ground is at 100v potential, and it still wouldn't change anything because what we're measuring is a difference of potentials. It's just convenient to say that the ground has zero potential.

You want the reference points in all the circuits (by that we usually mean grounds of the circuits) to be at the same potential, so you need to connect them by a piece of wire - then you cannot say that the ground of one circuit is at 100v and the ground of another circuit is at 200v, because they are electrically connected, so they must have the same potential.

The problem with your circuit wiring was that the preamp and power amp grounds were connected not only by one path (that is good), but by TWO paths: one going straight from preamp to power amp, and another one going like this: preamp->power supply caps->power amp - which introduces a ground loop which may cause trouble (look at the attached image).

When you remove one of those two connections there is still ONE left, so all is good! Ideally you want all of ground points of all the circuits in your amp (power amp ground, preamp ground, power supply ground and what not) connected to one, chosen by you, ground point of your amp (usually the point between the filter caps) by just one path.


Back to the pedal-preamp case: of course it may happen that the pedal power supply is not "floating", which means that it's ground is in some way connected to the mains ground and also your amp common is connected to the mains ground - then you don't need the ground part of the guitar cable going from the pedal to the amp, but it's a rather rare situation so don't cut that cable  :trouble

Man, I fell like I'm really awful at explaining anything so I'll just wait for Phil or someone to clear all the confusion I surely introduced  :loco

you have to remove either a ground path from the preamp to the power amp (to achieve more of a "star grounding"), or you can remove a ground path from the preamp to the PS filter caps (to get something in the mold of "Ground bus" - you will have a return path starting at the lowest current part (input), then gradually increasing the current - preamp, then power amp, then power supply).

Like it was mentioned, you cannot keep both of those connections.

If the ground points of the preamp and power amp are connected to each other directly and also there are two separate connections from the preamp and power amp to the common point, then you're asking yourself for trouble - like you said, you're creating a ground loop and also there's a danger that a high current return from the power amp would travel to the common point through the pre-amp stage, which of course is not desired. You can read more about this in this excellent post: http://www.ssguitar.com/index.php?topic=288.0

no, I left the gain settings more or less unchanged.
I haven't yet tried it with master pot cranked up at maximum, I did however max the "volume" (or "gain") pot and noticed no squeal. Did you test your amp with different guitars? Sometimes a noisy/high output pickup + a lot of gain = squeal, so grab another guitar and see if the problem persists.

Welcome back!
I finally bought myself a little dremel tool for cutting metal which allowed me to (almost) finish my project!
The head section of the amp is more or less ready. I didn't have a shielded wire at hand so I used a regular wire for all the wiring, but once I get myself some I'll surely fix it - however, the noise doesn't seem to be an issue so far. I may make some shielding for the preamp in the future if I feel like it. I don't have a dedicated guitar speaker, so I just connected it to the speaker from my stereo system - sounds quite nice, and it's going to sound only better with a real guitar speaker!

I live in a block of flats so I couldn't really test it at anything close to full volume, but after an hour of low-to-mid volume playing the heatsink didn't really get warm at all, so I think it's a good sign for the things to come. I think that heatsink should be enough for a single LM3886, right?

I haven't yet decided whether it's going to be just a standalone guitar head or a combo amp, so the speaker wiring is just temporary for test purposes, don't rip me for that!  :trouble But here are some pictures (click to make it larger):




I'm sorry exztinct01, but I cannot record any sound clips, because I don't have a microphone    It's all good though, because I'm a terrible guitar player so you don't have to listen to my terrible attempts to catch a clean chord  :lmao: however, when I'll figure out a way to record a sample with a good audio quality I'll make sure to invite a friend who can play well and post something here.

I tried it with a guitar plugged straight into the amp and with a modded tubescreamer I made and it sounds fine either way  :dbtu:

Considering the fact, that the signal goes through the FET, three capacitors, a resistor and LM386 that's not expected. Are you sure you are measuring the continuity with the circuit NOT connected to the battery/DC?

It would also help if you posted some photos of your build!

If you do have a digital multimeter you can of course check if battery measures around 9v (but if it's not 100 years old then it shouldn't really be a problem, as this circuit should work with supply voltage as low as 4-5v I suppose). If it does (and if you have any other audio amp) then I'd advise setting up a simple audioprobe (you can read about it here) and probing through the circuit starting from the input. If you do get the signal up to pin 2 but not on pint 5 of lm386 and pins 6/4 are connected to +9v/GND, then there is a chance that your chip is dead.

But the easiest way to troubleshoot would be to provide us with some pictures of your build  :dbtu:

I didn't have in mind putting a booster in the same box, just add a booster pedal to the end of your signal chain if you want to get your distortion from overdriving the amp rather than from a distortion pedal. I prefer to get distortion from my pedals rather than amp, but I liked the overdrive from this preamp so it's just an idea.

Well, I like a lot of bass from the guitar amp so I actually increased the amount of bass in this preamp  :duh but if you want less, try a bigger value for C6 capacitor for starters.

Think twice before building him a BMP if you don't like bass as this thing has TONS of it and not much of treble response.  :grr Of all the overdrives/distortions I've build one of my favorites is BSIAB II, check it out if you haven't already!

Quote from: exztinct01 on May 22, 2016, 09:50:17 AM
- Our guitarist even asks me why there is no gain pot and I don't know what to answer him. All I know is that the preamp already has two volume controls and the Master Volume at max produces a little bit distortion. But he asks for more, I don't know how to add a circuit that will integrate a gain control there.

I've build this preamp and I think of the "volume" control on the schematic as a gain control. This control sets the level of the signal before the second op-amp stage with a set gain and clipping diodes, so using this pot not only changes the volume of the pedal, but also amount of clipping.

If you want more distortion out of this preamp, I think the easiest way is to put some booster pedal before the amp in the signal chain - with hotter signal at the input the diodes would clip more easily. If you want to be able to control the gain of specific op-amp stages, as those work as non-inverting amplifiers, you can lower the values of R11 or R6 to 2,2k (that's the lowest value you should use according to Rod) and wire a 5k resistor in series. Although in my opinion that would be redundant to the volume control.

But keep in mind that I'm as new as you to the area of building amps so take this comment with a grain of salt! Better wait for someone more qualified to answer before you make any changes (and the best idea is to put the circuit on a breadboard and then experiment)

you can use two batteries: connect + end of first battery to the Vs+ point on the schematic, - of the first battery to the ground, + side of the second battery to the ground and - of the second battery to V-. That way you get a +-12V power supply and everything should work just fine.

Like this:

Quote from: J M Fahey on May 02, 2016, 04:51:01 PM
You'll lose audience (I think)  if you fill page after page of obscure (for us)  code and what you expect it to do, including endless "version mods" :trouble, but a nice finished project for a voltmeter or any other useful add on , including a PCB design and easy to download and burn code , a fully finished and tested package, will in fact be welcome. <3)
Just open a new thread for it so those interested can have all together in one place

Sure thing - I won't bother you with any posts about this until I'm done with the whole thing (but hey, the code will be awfully simple! You can measure 0-5V using AD converter, so you have to bring the voltage down using a voltage divider and then view it on LED display  ).

Quote from: exztinct01 on May 02, 2016, 05:41:54 PM
don't power supply wires, pots wiring, need insulated wires too?

Damn it, of course they do, what I meant was shielded wires - I always confuse those words for whatever reason.

Both of you are probably right and I think this will be an overkill, but when I think about it, I think I prefer to have the meter built in the enclosure - it takes less space, you don't need to connect it every time you want to use the supply - it'll just be more convenient for me. And it's just a fun thing to do! Most reasonable people (that doesn't make amps for a living) would probably buy a pre-made guitar amplifier instead of wasting time and money on a DIY thing, but that's not why we're here on this forum 

About the price - as we all probably have a lot of resistors and capacitors (and this project will call only for a few) the only components you really need to buy are micro controller atmega8 (I think it costs around 1$) and a LED display (even less than 1$). When you factor in the PCB, etchant, wires I think it's reasonable to assume the whole thing won't cost more than 5$ so it's not the end of the world.

But I'll finally shut up about this stupid idea I had and get back to amp-related stuff  :dbtu: sorry for littering this thread

I have a few AVR microcontrollers lying around, so I think I'll add a digital voltmeter circuit and a LED screen to the bench supply so I don't have to measure the voltage myself everytime I want to use it. If you're interested I can post the code and schematics when I'm finished with this.


Whooops, I just realised this is getting terribly offtopic and not really solid state amp related - I better get back to the amp wiring (I still don't have insulated wire for the inputs and outputs)  :duh

sure thing! Sorry for not replying for so long, here are the files.

Regarding the bench supply - you probably need to drill a few more holes in the "GND" section from what I remember, also make sure to use a 10uF capacitor instead of 2200uF AFTER the regulator, as suggested by photo! Keep in mind also that the PCB is designed for a specific radiators I had for the 317 and 337. If you want to make some changes to the layout I can post a whole EAGLE project.

As for the preamp - sorry man, I just don't remember the specific values i put in there because I soldered most of the elements right after I took them from my breadboard  but I don't think I've changed much, maybe a few slight changes in values of caps or resistors here and there to achieve more bass that I like, I don't think I removed any components. But here's the layout I used - I don't think it's any good, I just tried to keep it small, but maybe it'll be of any use to you. Sounded alright when I tested it so maybe it's not that bad after all.

EDIT: Oh, and one thing - there's a misleading caption on the bench supply layout - the TRANS-GND pad isn't actually the 0V from the secondary winding of the transformer - it's the MAINS ground that is connected to the chassis. The 0V from the transformer should be connected somewhere to the big ground area in the middle, probably somewhere between the four 2200uF filtering caps.

Thank you! But there's nothing to be envious about - it looks nice because as for now it's just a PS, once I wire everything up I'm afraid it's going to look like a hot mess, "neat wiring" is something I've always struggled with 

Yeah, I know how hard it is to finish - before I started I thought the whole thing wouldn't take me more than month to finish - after all, it's just three simple circuits (PS, preamp and power amp) - I mean, how hard could that be to put it all together? But when I think I'm finally ready to proceed I find out that I'm still missing something - be it a 0,1R 5W resistor, a zener, connectors or something else - it's already been more than 1,5 month and the end still isn't near 

Oh, and speaking about Ruby - I couldn't wait for the whole project to finish, so in the meanwhile I made myself a modified version of the Ruby called Noisy Cricket:

I guess it could be less noisy if I put it in a metal enclosure, but whatever, I only use it for playing on my headphones and it works just fine  :dbtu:

Good luck on your project!

hello again!
Just stopping by to say where I'm at right now with my amp:
once again I didn't really have much time lately so the project haven't moved that much, but at least I finished my bench supply! Here's some photos:
that's how it looks on the inside:

I left the fan from the PC supply in place, and even got it working - not sure if it helps dissipate any heat at all, but hey, at least it looks cool!

here's a look at the fan:


after I finished this thing it was terribly ugly, so I decided to make it look less sad. The quickest way is of course using letter-stickers!

it's not that it looks beautiful now, but hey, at least I tried!

and a look from the front:



On another note, I've also finished the power supply for the amp:


It works as it should - with no load it measured +-37VDC, phew!

Since the pre-amp and power-amp are already done, I just have to wire it all together now and the "head" part will be ready - but it's also the part that I dislike the most, so it'll probably take me lots of time! Ah, and I have to get myself a heatsink for the LM3886.