Quote from: vices like vipers on January 20, 2011, 06:05:59 PMHey everyone, I have been lurking around here for a little while.
I would like to build my own amp, but I have no idea where to start.
I was going to build a Tube amp, but then I saw the price of tubes, and that just turned me off of that idea. So now I'm looking into solid state.

Nothing wrong with that. The higher voltages of tubes turned me away, as I'm clumsy enough to really hurt myself. It's been a while, so I'd feel confident now but when I started I didn't have the safety knowledge or equipment to be effective.

Quote from: vices like vipers on January 20, 2011, 06:05:59 PM1. I have a bunch of computer parts (and computer power supplies), I was wondering if there is any way I could use them?

Sort of, but it winds up being a lot of work to salvage things. The best computer parts to salvage from are really old ISA and PCI soundcards and occasionally old CD-ROM drives, as they sometimes have an amp chip you can use (google the part number for a datasheet). Most of the stuff will be surface mount and very difficult to work with. It makes for great soldering and desoldering practice, but tends not to be worth it in the end unless you're very, very low budget (as I am).
You may find folks out there using old computer power supplies to power their projects, but I'd recommend caution for two big reasons and a little one:

First, they're capable of extremely high power compared to the needs of most of the circuitry you'll be working with. This means that if you have a short somewhere the computer power supply will happily pump out several to tens of amps and cause parts and wires to smoke and glow before the current limiting kicks in. Wall wart power supplies and batteries are much safer to work with until you're ready to start working with mains powered power amps, and a fault will likely only cause your parts to get hot or glow for a moment as opposed to the fireworks a high current supply is capable of providing.

Second, they're switching based supplies with the wrong sort of voltages (mostly). An older AT supply (one with a switch attached) will usually have high current supplies for +12V, +5V, and +3V, and low current (or absent) supplies for -5V and -12V. Newer ATX supplies (big 20 plus pin motherboard connector) will often omit the low current supplies and crank up the amperage well beyond the older AT versions. These are really inconvenient voltages when working with audio, which often requires +/- 15V or higher, or 30V plus in a single ended supply.

Third, they're switchmode power supplies that aren't really designed for low noise. When you're working on a preamp and trying to minimize noise it's very irritating to work with a noisy supply. This is less of a concern, and can be solved somewhat with filtering and post regulation.

Quote from: vices like vipers on January 20, 2011, 06:05:59 PM2. How would I go about building my very own solid state amp?
(I have dial up, so it is kind of hard to download teemuk's book. But if I need it, I will try and download it tonight)

I can not recommend Teemuk's book highly enough. I'm on my third read through of it, interspersed with other books. It's worth taking the time to download. It's over 400 pages of everything you'll need after you've mastered the basics.
I'm with Joecool on the Little Gem. It's an excellent, simple project to get you started and it'll be useful forever. It works great as a sanity check, something you're sure works just fine and you can use as a go-to when you want to be sure the signal you're using is good. That, and if you hook it up to a proper guitar speaker it'll amaze you how good it sounds for being so simple.

Quote from: vices like vipers on January 20, 2011, 06:05:59 PMBefore we get started on the amp, we are going to make a Breath conttroler like the one Meshuggah uses.
Tat way I will get some more experience with wiring and stuff.

I had to look that one up. I couldn't find a schematic, but it sounds like a pretty advanced project to start with! It's always good to have a big idea pushing you to learn all the little bits, so if it starts to look like too much don't despair, just set it aside for a bit and build something simple to get more practice.

Hope that helps some.

I think I found the schematic.

Poking through it now.

And the parts list.
And the manual.

Looks like it uses:
Reverse audio (anti-log) - gain, treble, and bass
Audio (log) - volume
Linear - master, mid, reverb

I think I should take Teemuk's advice and poke through more schematics. I was looking at the output stage and thought, "Oh wow, one of those Sik. . Sizl . . not Darlington pairs!" (sorry Mr. Sziklai)

[lots of edits as I added stuff]

I'm not sure, but I'll take a wild guess.
Do the replacement pots have the same taper as the old ones? It sounds like what would happen if you swapped a linear pot for a log taper pot.

I'm not sure what your level of expertise is, so I'll toss in a link on pot tapers. My apologies if it's something you've mastered.

Hope that helps.

Quote from: KMG on January 18, 2011, 07:05:11 AM

QuoteLooks like at such high voltage these projects could cost close to real tube versions...

But the sound is also close to tube prototypes, especially while live playing. The main advantage of fet versions is their reliability & long time sound stability (no effect of emission loss).
During the "blind" tests guitarists usually say "I can`t tell what is the specific amplifier, but certainly, tube one".

Yep, this is how I read these two projects. Lot's of care and testing resulting in a design that's a very close (or better) match for tube characteristics, but with all the durability benefits that come with solid state. It should sound the same from day one until the electrolytic caps leak (figure twenty years with good caps), barring damage or bad luck with one of the semiconductors. The FETs won't age, degrade (enough to matter), and require replacement like tubes would. Definitely an advanced project, but a real masterpiece.

Quote from: joecool85 on January 17, 2011, 01:37:18 PMYou don't play?

Not even a little. I'm a fix-it guy and science nerd through and through, and the world's a better place when I don't try to make music.
I've found guitar audio fascinating in it's difference to most of the rest of electronics. Not only is distortion a good thing, it has to *sound* right, so it's a real challenge. I really enjoy squashing misconceptions and voodoo among the musicians I know, and it's saved them a lot of money. They were in dire need of someone who could do the simple things like swap out some pickups without getting dinged for $150US every time for experimenting, and they're starting to pick up the basics.
As good as the symbiosis is, I sometimes wish I could close the feedback loop on a project by myself. If I wasn't so strapped for cash I'd have picked up an old beater guitar by now just as an 'analog arbitrary wave generator'. 

Quote from: J M Fahey on January 18, 2011, 04:39:16 AMDJPhil, you're not alone !!

It is a great comfort to hear. I feel a lot less handicapped in such good company. :tu:

Quote from: J M Fahey on January 18, 2011, 04:39:16 AMAnyway, if your friend does want a pedal or anything, he should chain himself to your workbench until finished.
Personally I do not believe very much in the "just make it and tell me when it's ready" school of thought

I'm learning that lesson the hard way. The folks I know are a bit wild compared to me so it's hard to understand sometimes. I always get an itch when something's half done and stay up nights finishing it, and it's hard to imagine not having that sense of urgency. My compromise with them is that no progress is made beyond a day or two until I see them again, that way I can set it aside and go back to tinkering on my own stuff.

The Professor is one of the very few I've actually built so far. I just wanted to throw in my two cents in case it's useful.

It's the type of FET emulation circuit that takes a bit of tweaking to get right. If you decide to build it I'd recommend not skimping on the bias pots like I did. I used 500Ω single turn trim pots I had lying around, and it was wicked difficult to bias the FETs properly. With proper multiturn pots it'd be much easier. You don't necessarily have to hit the voltages they list, but if you tie a voltmeter across the pot you'll be able to tweak the pot til you're running at 1/2 the supply across the pot. This is one of those times you'll really need a multimeter! I used 5457s for the FETs, though I don't think it'd be a whole lot different with 102s. FETs are squirrely parts with huge manufacturing variations, that's what makes it so important that you manually bias them.

In the end it's one of those projects that got away from me, as I never had the fellow I made it for long enough to sit down with him and work out the bugs. It's very hard to make effects for people when you aren't capable of plugging in a guitar and making your own noise!  xP

Excellent work!
That's a very well thought out and documented project. I bet you'd be able to sell boards or kits for it. I don't think I've seen such a well made DIY high voltage FET tube sim circuit before.
I'll bet that you'll have folks who are trying to make their own asking about the power supply board soon. It might get tricky for those of us working from 115VAC. I bet I could sort something out from older tube circuitry, after all, it makes sense.

You've got a really good article on your site about FET simulation as well! (Google translate link to english)

Thanks for all this, I've got some reading to do.  :tu:

Quote from: joecool85 on January 15, 2011, 04:56:49 PMOk, what about the resistor and cap that teemu had on the end of the circuit that I didn't add in?

I went back and found a schematic in this post, let me know if I should be looking at another.
R12 and C11?
I believe these act as a lowpass filter and help set the output impedance. It starts rolling the impedance down at about 14kHz, and this helps keep ultrasonics out of the following stage (presumably a power amp) which aids in stability. It certainly wouldn't hurt to put it in, and if need be you can just solder the parts to the output jack.

Quote from: joecool85 on January 15, 2011, 04:56:49 PMThe schematic had a resistor inline as well as one to ground and a cap to ground originally right before going into the poweramp circuit - should I put these in?

I'm not sure about this one. Do you mean preamp as opposed to power amp? Could you say which components in the above schematic?

Hope that helps.

Yes, stay safe over there guys!
If you need help or need to check on someone you know I'd recommend contacting the Australian Red Cross. I worked with the Red Cross here locally when we had some flooding in 1993 and there's no better bunch of folks when you need a hand.

Wow you guys work fast! Well, that and I'm long winded too.  :duh
Here's a link to a page on cap codes, and a quick online calculator. The letter codes can be a number of things, but the letters used for voltage ratings are from an old military scale that's probably not used here. If so, it would mean 4KV rated cap!
The letter codes are often tolerance ratings. An 'M' would be +/- 20%, which is the standard tolerance for cheap caps. Sometimes you'll see a 'J' or 'K' rated cap (5% or 10%) to signify that the cap needs to be better than cheap. Capacitors will definitely teach you to forgive sloppy tolerances, especially when you imagine how many things out there are stuffed to the gills with +/- 20% parts and 1% resistors! It's a rare circuit that will need better than 5% on a cap, though it's a bit more common to need a well matched pair (twin T filters for example).

Included is a picture from when I thought I'd be the first to respond. The codes discussed above are commonly used for through hole small caps of all types.
You never know what you'll run into on an old schematic.

Quote from: joecool85 on January 05, 2011, 03:57:55 PMCould you elaborate on this "audio ground"?  Also, why would having a small .1 or so ceramic cap help?  I've seen it a lot in pedal board PSU designs but never knew why.

Audio ground means the same as 'virtual ground'. In a single supply system it sets a point near or at the middle of the supply for the AC audio signal to ride on. Because it will have to source and sink (a little) current you want it to have a very low impedance. If it were high impedance then the ground would follow the signal as it moved up and down, anywhere from slightly (causing frequency dependent attenuation) to horribly (almost no signal and badly distorted). The lower the impedance the better the virtual ground's compliance as a (sort of) regulator.
The choice of resistors is a balance between getting that low impedance and wasting lots of power on smoking hot ten ohm resistors. 10k is a nice middle ground (sorry for the pun) and is often used. The 22uF capacitor and it's 100nF friend help keep noise off of the virtual ground by providing it a low impedance path to ground. They serve a similar purpose to the bypass capacitors in a voltage regulator, and help insure that the virtual ground is as flat and stable a DC value as possible. The extra smaller cap helps with higher frequency noise and RF interference in a manner I can't easily describe. This is where joecool85's simulations come in very handy, otherwise you find yourself doing the hard math.
This all becomes academic if you choose to use a 2x9V battery setup, as you can leave the circuit as it was originally. The virtual ground link above has some additional details on various methods that go beyond a simple voltage divider if you like.

Regarding simulation, I'm a fan of LTSpice (also free), though it took me a few weeks of tinkering to get used to. I'd recommend playing around with a few of the free ones out there and seeing what works for you. Simulators shouldn't be trusted for everything, but they are indispensable for quickly working out simple things like the above. Their ability to graph output gives you access to the sort of information you might not have the equipment to see in practice (oscope, spectrum analyzer, etc.).

I hope that helps some. As usual, always double check my work against the pros to be sure.

Sure, there are dozens of places to start learning electronics. Everyone has their favorites.

There's a series of interactive tutorials hosted by the University of Texas at Austin that really surprised me.

I stumbled on this one while looking for something that would appeal to a friend of mine who's very visual and hands on. It seems like a good intro just to get the terms and ideas rolling around in the mind. I think it'd be excellent for those without a strong math or science background, or perhaps those who work better through analogy than abstraction.

Just thought I'd mention it in case anyone found it useful.

Quote from: kvandekrol on January 02, 2011, 06:52:48 PMWould it be possible, using a charge pump like a MAX1044 along with regulators, to get +/- 15V from an 18v wall wart?

Possible yes, but charge pump regulators can't handle much current draw. Usually the output is limited to 10mA or so. I'm not extremely familiar with the current demands of the average preamp, but this seems low. The positive supply should be fine, however.

I think the linear supply already described would be simplest. It seems involved, but that's really the nature of split supplies. They're not as common as they once were. I think the only thing you could do to make things more simple would be to buy a premade 'brick' external power supply, but that may be more expensive in the long run. They tend to be switchmode as well, which may require a bit of filtering for sensitive preamp stages depending on the circuit and how noisy it is.

Just my two cents, hope that helps some.

These are roughly what you're looking for, I believe. I do have some other considerations to add though.

Smaller transformers like this have very poor regulation, meaning that they'll likely only hit their rated voltage very near their max load. For lighter loads the voltage will stray higher. You can calculate the specifics, but often it's worth simply building a mock up so you can test everything at once (like operating temp, etc.).

A 20VCT transformer will give you a 14.14VDC bipolar supply. Given the voltage variability from using a small transformer and the fact that you'll likely need solid rails for a preamp I'd recommend post regulation. This means you'll need to account for the dropout voltage of whatever regulator you use. The 7815 and 7915 require about 2.5V to be comfortable, so you'd need roughly a 25VCT transformer (25 / 2 * 1.414 = 17.675VDC). Complicating this is the poor regulation of small transformers mentioned above, which may give you some breathing room with lighter loads.

Finally, be very careful with mains over pcb traces. I'd recommend not allowing anything near mains traces closer than the distance across the transformer. In such a small case you'll probably need to isolate the area under the mains traces with plastic card or the like to ensure it can't arc to the chassis.

If all else fails you can try to find a brick style power adapter that's already taken all this into account. You can sometimes salvage them from inkjet printers, though they're often marked as +16V/+32V.

Hope that helps some.

Quote from: J M Fahey on December 11, 2010, 11:05:44 AM
They are protection diodes as correctly stated above, but they are not grounded, please check that the "free" end goes to either +15V or -15V; so the actual threshold is a whopping 15.7 V .
They will not clip a guitar signal, they are there to protect the input op amp from those guys who enjoy plugging a speaker out into another amp's input.
You say those don't exist?
I know a few !!!! :duh
They do not even be too  :loco , most people think that headphone out can be used as line out.
Definitely not, most bean counting designers save a 0.001 cent resistor and headphone outs slam the full speaker voltage (think 16 to 40 V RMS) into high impedance loads (pre inputs).

Ah yes. To get the effect I was describing regarding clipping you'd have to connect them to ground, not the power rails. I misread the schematic, I apologize.