Few years ago, when my playing with electronics and sound just begun (and is still more or less a play), I had a bunch of old speakers which I was not really sure if worked. So I took a small 25W Yamaha amp and connected them to speaker out. I am not sure what was the reason, but I beleive I had to surpass the minimum resistence vastly, because the amp literary died on me. In few second the sound began to fade and than it went silent. I don't have the knowledge to analyze it and find the fault. At some time, when I was fiddling around it and trying for the loose contacts (although the problem must be in a fried / faulty component) I really suceeded to revive it for few minutes, than it went silent again. The accident stoped me from doing my "research" for some time. I disasembled the amp and stored the parts with a plan to use the cabinet, the speaker and the chassis for something different. These are the pictures of empty cabinet and the faulty PCB (some experts my even be able to locate the fault from the photo): https://lh3.googleusercontent.com/-b-dzkXjzsOU/U5dw-3vFBII/AAAAAAAAEH8/lT2lutcqsmM/w1534-h863-no/DSC_0123.jpg, https://lh6.googleusercontent.com/-HpiSCMxVUNk/U5dw-PchzRI/AAAAAAAAEHE/0i8iEyvXIYI/w1534-h863-no/DSC_0121.jpg, https://lh6.googleusercontent.com/-HdvVZVYdDro/U5dw_LuMbII/AAAAAAAAEHc/okIiK9Uaaa0/w485-h862-no/DSC_0124.jpg).
Some other time I bought a 40W Kemo power amp module, I thouht about using it as an amp for an active monitor, for my uncle's tractor "radio" or someting like that. I was kind of interested, how such thing works. Few months ago I began experimenting with it and got quite a "clean" playback if I connected it to my cellphone. (Kemo module - https://lh3.googleusercontent.com/-h-Y_0aBf6uU/U5dxAvp-O6I/AAAAAAAAEH4/OwVlcOeSaqw/w1534-h863-no/DSC_0127.jpg)
Also few months ago I assembled (just on breadboard) my first LDO PSU for the tube heaters. (When I made my first TT projects, it was with suplied and presented PSU, which uses a LM7812 fixed regulator. Because I wanted to regulate the heater for octal lamps like EL-34, which use 6,3V and (hot)~1,5A, LM7812 was not suitable. Because of high currents involved (which even at smaller voltage dropouts create a significant heat and so demanding cooling solutions for the regulating IC, I went into LDO (low dropout) adjustable regulators. First I read about LM108* class, but bigger ICs were pretty expensive. Then, when I stopped to be afraid of more than three pins on the component  , I found Micrel 29502 adjustable regulator and bought a few on eBay. I assembled a design on bredaboard and it worked.  - https://lh4.googleusercontent.com/-dF3D_KD9sM4/U5d6yMPNaLI/AAAAAAAAELA/TaGk1stQBYs/w332-h288-no/micrel.jpg (probably I will write a separate post about that PSU, as it is quite good (well, not really cheap), but I also have some problems)
With a functioning PSU design and a poweramp module there were two things yet missing for the whole amp: trafo and a preamp. I took a trafo out of non used halogen lamp, which had standard 11,8VAC output and was rated for 3,95A. Now, when I use the amp, this trafo gets pretty, pretty hot and I am not sure about the reason for that. It surely wasn't designed for such a task, but trafo is a trafo I presume (if there is somebody able telling me what is happening I would be happy to know; the thing I saw was that at the input clips of the trafo I think on the primary side there is something like a small resistor - maybe a thermistor for the softer start of the lamp (which has insanely low resistence when cold as far as I know) - if it is a thermistor, everything is pretty clear - you can se the schematic sign on the photo - https://lh6.googleusercontent.com/-iQfP9o92a3k/U5dxB0OeaoI/AAAAAAAAELk/Ib_tLBZhWiY/w1534-h863-no/DSC_0129.jpg). I mounted the trafo outside the chassis (there was not enough room inside + trafos can be quite a noisy devices).
For the preamp I decided to use a TT kit Vanilla Overdrive (which I never assembled before - https://www.tube-town.net/cms/userfiles/media/lov-vanillaoverdrive/lov-vanilla-schem.jpg). Surely I needed to change some things. I changed footswitches to toggle switches. First I thought about using another switch to bypass the transistor, but then I would have to drill additional hole in the chassis and I just didn't feel like. The main change was, that I moved the volume control so that it controls both channels (changed P3 to fixed 100k resistor and used a 1M audio pot instead of R20). Because of space considerations I also used much smaller caps (low voltage Wima)  than in the kit (I'm not sure, how this affected the sound - later about that).
I decided to combine PTP and PCB methods (I am not very good and precise at drilling so it would be hard to match the pcb to already drilled holes for the pots and switches; also I really didn't find a usable 90° PCB mounted switch. So I made a PCB with toner transfer method. There was a nasty emergency during the etching process, which caused me to leave it in the liquid for quite a long time - the traces are "eaten away" in many places (I am still not sure if it wouldn't have been better I made a new plate at the time). I drilled the holes in a hurry, so I destroyed a lot of pads with my 1mm drill (I bought a 08 and 05 drills later and figured out that there are only few components, which have their leads thicker). So the process of soldering was a hardwork, with a lot of improvisation and possible cold joints. Until I started using PCB rivets -they solve the problem with destroyed pads or discontinued traces excellently. It is quite a costly solution, but great, specially for later resoldering and you can easily attach components to both sides of the board - soldering the tube sockets to the bottom of the PCB was kid's play) Still I am not really sure about the quality of the solder joints. When the plate was finished and populated I started working on "PTP" part of preamp: input jack, first cathode bypass switch (combined with LED), drive pot, tone pot with a smaller PCB for two caps and a resistor, switch for "channels" (practicaly bypassing all of the preamp or not), master pot and output jack. Then I tested the device and there was nothing happenning. When using external PSU, there were at least LEDs lit, but no sound coming through at the "non bypass channel" setting and quite poor at the bypass channel. I was desperate, but then I found out: 1) there was a cold joint at the enable pin of the regulator 2) I used the wrong resistor for "programming" the output voltage (which caused sudden jump of output voltage from minimum to maximum, when I reached the end of regulation pot) 3) (I am deeply ashamed) I made mistake when drilling the holes for the tube socked and so rotated the socket for one pin (pin 1 of the tube went to socket 2 - what do you think, could that ruin a component; specially the transistor? what could be the consequences of such a mess? ) Anyway, luckily I used the rivets, so there was really not much effort needed to remove a socket with a desoldering pump). When I corrected all three mistakes there was still much noise on the line, but the system was working.
Next stage was additional drilling of the chassis. It was not really a masterpiece and I had many problems with mounting holes later.
Then I began to lay cabling for power supply: from the inlet (with the fuse1) to the switch and than to trafo, from the trafo to second fuse and than to the PCB / PSU. Then I mounted poweramp module with the cooler on the outside of the chassis (my clumsy drilling, yuck!) Mounting the PCB in the chassis was more work than it shoud be, because of inaccurate drilling (I know now it is best to drill holes, if you dont own a CNC, with an unpopulated PCB). The last task was connecting power to the PCB (I twisted the wires), connecting the output jack (swithed)  to the power amp and conecting all the grounding wires to the "star" ground point. Then final assembling.
The thing works and is quite impressive on "bypassed" channel (nice rounded tube sound with a bit of overdrive at higher volume), but is noisy as hell at "non bypassed" setting. It also has a nice "reverb" effect, which gives fantastic "space" to the sound. The sound on "non bypassed" is really not dependent on "drive" setting and goes to terryfing cracks when volume dial is over 12 o'clock. The overdriven signal is othewise nice, smooth and full. But gets practically lost in hissing, static and cracking. I intend to ask a friend with degree in el. engineering for help, but it could be:
a) cold solder joints
b) wrong use or lack of shielding (I think I overdid, but I am not sure)
c) damaged components, specially transistor, from the misoriented socket
d) mistake in grounding design (I think I overdid there also ...)
e) another collosal mistake in the style of already mentioned (and corrected) mistakes ...
Do you see anything in the pics? Any Ideas?
Al in all, when on "bypass" channell, the amp is very noiseless, but still: what to make it even quieter?

pics: https://plus.google.com/u/0/photos/102300033002717016211/albums/6023407147810329601

By friend's advice I removed transistor part form the circuit (connected drive pot directly to the V1/6). There is still some noise, but all in all it is much better. So this was the main problem. For now, I will leave the project just as it is.

https://plus.google.com/102300033002717016211/posts/ghiNBCwveh2
So here is "renewed Yahamaha", the "dIRT EDITION". Yes, you guessed, it is noisy as hell. Maybe someday I will make some order in this chassis. For now it was just a practical experiment, how Alembic bass preamp works (I really wanted to build it for a long time!) - with TT Vpump all this became much more simpler, because I only needed some generic components to build it, even PT was ordinary lighting 12VAC (50VA). It works great and is pretty loud. Master pot behaves funny in a way, it makes much less volume difference than expected. Tone strack works fabulously (but I becoma aware of HV present in it only in the process of buliding - I would better use some heathrinking from now on)). In front of Alembic there is a bypassable TT Banana booster circuit. With pots set up carefully it can serve as a nice overdrive effect. But sure enough this project gets quickly in the area of extremely ugly saturation /overdrive, where the sound decays. BUt with the correctly set-up pots both the clean and overdriven signal are not only satisfying. When you roll to 10 o'clock all the gain pots, switch to the bridge pickup and tune the tone commands, this monster weeps so convincingly that somebody would adopt it in a moment ... But it is mine!

Specs / build parameters:
- old fried amp's chassis, cab and speaker
- removed previously installed TT Vanilla circuit (look at the first few pics), because only bypass channel worked (I was thinking of troubleshooting it, but then the idea came to me to use long planned Alembic circuit - and TT Banana, which satisfied me as a stompbox)
- (unsuccessfully repainted the chassis few times, then began experimenting with "browning"
- PSU: IEC mains input with the fuse >> power switch >> PT 12VAC 50VA >> "micrelator" - a LDO linear regulated PSU board (with ST schottky diodes (I needed low dropout for normal function) and 10mF reservoir cap), set at 12,6V >> a) goes to LED, heaters, power amp module and banana plate b) TT Vpump - which provides plate voltage for the alembic
- banana booster with kick option, bypassable with a switch - first pot from the input jack - replaced Elkos with tantals
- Alembic tube bass preamp - schematic from the web with sweep and bright switches. In the schematic there is a 22uF cap drawn very close to the tube (looks like it is there more for "eating" the noise form the circuit than stopping the rippple), so I used it, paralled with another 2,2uF poly cap. Cap overkill again!
     Both made in "point to point" manner, in a messy way - i is probably better to change the socket with all the attached components than to desolder them.
- as the Endstufe Kemo 40W amp module (used previously) > speaker exit on the bottom of the chassis - to use also another speaker easily
- grounding with solid wire (an also all non screened signal wire) - the resistance value is not so different, but I like that the wire stays in place even without regular attachment (stranded wires tend to wobble around in chassis and cause random parasitics in the process.)

Problems (and what I learned):
- PTP in this way is impossible to troubleshoot. At least I should have made the leads to the ground perpedicular to the socket (not to make a sort of a tent)
- shielded cables are a nice trap for shorting screen to the signal line - better a little longer unshielded end than dead silence due to short circuit
- with a choke used in TT vpump and my multimeter I cannot determine if it is shorted, because the multimeter leads have higher resistance than the coil
- Don't stuff too much in a chassis - don't be greedy - I thought I learned that previously, but again the same problem ...  If I used factory made PCB it would have beeen different, but for the home production it is impossible
- things attached to the side of a chassis are hard to troubleshoot
- rust, if conditioned, can be a protective film for iron

Unsolved misteries:
- It was all working great, then I put the cable in Line out jack, and when I plugged it out, all died on me ->> could be a short between spk output and this jack
- I repaired the Vpump and it all worked, then I attached it (with protective washer, silicon shield and everything), and micrelator switched off due to schort circuit - where was it? Now I put a heatsink on a IRF741 and attached it trough the hole - and it all works great)
- how did the 68 resistor lead come in touch with DC lead a cm away

All in all, nice one.
It is really great to play it.
Maybe in some time I will get to make some more order in that chassis (and maybe reduce some noise) ...

With the experience in troubleshooting this project I also could have got enough knowledge to repair the "original" yamaha circuit. Besides from passives just two ICs basically - probably POweramp is fried.
But there is probably no chance I would put it back in. I wil find a box for it. And a speaker.