Damp

After Second Bass I decided to update my amp. My trusty 200W Class B amp does a great job as a practice amp when the 40W valve amp isn't going to be heard, but I felt there was "space" at the top for something else.
After a previous place of work got into financial difficulty, I found myself the proud owner of some industrial PSUs. These were destined for large Automated Test Equipment installations and were 800W units giving 48V at around 16A each. The idea of putting them to use in a bridge amplifier felt like my duty in these modern times of recycling...
The original intention was to design a Class D amp and put the tricky Class D modulator stuff inside an FPGA. I used an Altera Cyclone 4 device running at 200MHz, but the trade-off between simple PWM resolution and speed didn't look promising. I made a low level output stage and the results sounded (to my ears) like a badly tuned radio. I expect this is officially known as intermodulation distortion, but it was never going to cut the mustard.
To combat this fundamental problem the world of DSP promises much but when it comes down to details most is hidden away as proprietary information. The best chance of success was to use a sigma-delta topology which has similarities to SACD. After blowing a few sets of MOSFETs up I eventually had to resort to the tried and tested International Rectifier IRS2092 (now owned by Infineon). This may not be perfect, but with careful attention to layout (especially the dc offset protection monitor in my case!) and avoiding the usual grounding pitfalls it has performed admirably, albeit always making a weird weeeep noise when it is switched on.

The preamp has two separate inputs A and B, a compressor and a parametric EQ stage before the volume control. The controls were constrained by only having five excellent solid aluminium knobs that really needed to be used. The six controls are Gain A, Gain B, EQ frequency, EQ boost/cut, master volume and compressor level. Seeing as six controls and five knobs don't align properly, the compressor gain has to share the same knob as Gain B using a dual gang pot. Most of the time I use Channel A and on the rare occasions I need both channels, the compressor can be easily switched out if it misbehaves. The other switch mutes the output from the master volume which is handy at gigs to ensure no nasty noises build up between sets without having to twiddle anything else.

The Digital Amplifier working name was shortened to D-Amp and eventually just "damp". Maybe not as good as moist, but considerably better than fetid or stagnant.
I tried a resistive power test once it was built. I cobbled together some wirewound resistors to give 8ohms and then plunged them into water to keep them cool. Electricity - water - safety - blah blah - don't try it at home etc. I got to over 400W before the resistors burnt out! Before they let go, the amplifier was definitely happier at higher frequencies such as 1kHz and not so keen on 20Hz. This may be down to the amount of reservoir storage capacitance available, or the power supplies hitting their 20A current limits earlier than I'd hoped, but with real bass guitar signals it has so far belted out nothing but a good solid wall of sound. Or eaten watts and poohed bass as I have heard it put ;-)

The front is smoked perspex with a drilled aluminium panel behind. This allows the control fixings to be concealed, the (very noisy) PSU fans to be vented front and rear discreetly and the opportunity to put some eye-candy LEDs on the front around the master volume control. The ring shaped WS2812B LED board was sourced from a popular online auction site along with a simple controller. I was so pleased with these LEDs it made me ponder on another use for them...

Orange Overload

 

After the two weeks was up, the lacquer was lightly rubbed down with wet'n'dry paper, rubbing compound and T-cut which gave a pretty good high-gloss finish. There's a few not-so-good areas but from any reasonable distance it looks great. The bright orange is so intense that the camera gets saturated. According to Kia-Ora, it's too orangey for crows, and apparently too orangey for Nikon's too!

 

 

 

The jack socket doubles up as a strap button. An extra hole was drilled to allow the E string bridge to be earthed. As the nut is aluminium, it should earth all of them in a roundabout way.

 

There are no plates over control cavities on this guitar as it would spoil the organic curves. The pickup selector switch is fitted underneath the pickup and accessible from the rear (the bagel!). No other controls are needed - it's very minimalist...

 

The pickup was then fitted to (a) pickup the string vibrations and (b) cover the dodgy woodwork and shoddy soldering.

 

Now for the artistic shots. Chewie got me some orange Neon strings for my birthday, so this was the time to whip'em out. They really are the icing on the cake! 

 

You can't beat a good angle shot...

 

Much better than straight photos...

 

The white strap matches the fretboard.

 

I'm getting the hang of not overloading the camera now. It's a case of standing further away and using the zoom lens. Simple really, s'pose.

 

 

Here it is, hanging together. It still hasn't snapped.

 

After a bit of settling in, the strings seem to stay in tune for at least a few hours. I'm pleased that the stainless steel re-inforcement has worked well as the strings can be tuned quite independently. I'm still getting used to the Delano Xtender pickup as I cannot decide on the best switch position. There is plenty of top end, but without sounding lacking anywhere else. It's a good place to start, and I'll have to give the 4x10 cab full beans with it soon ;-)

War paint

Painting is tedious.

Sanding is tedious.

Sanding down paintwork doesn't figure very highly in my list of favourite things to do, but it is a necessary step along the way. The water-causing-cracks-problem behind me, I took the opportunity to add a bagel-like shaped feature behind the pickup so that I can add a pickup switch. This three position switch will give humbucker-series / single-coil / humbucker-parallel. This tonal variation may prove very useful seeing as there will be no other controls.

I did a bit of reading up on t'internet and white undercoat is recommended for fluorescent paintwork. 

 

I then realised, after refinishing the body a second time, that the headstock is a different colour due to the white undercoat.

Painting is tedious.

The headstock was also re-finished to make it match. Both parts were then lacquered with a polyurethane based coationg that promised to be heavy duty and suitable for alloy wheels. It smelt just like a tin of wood varnish, but it does seem tougher than the original acrylic lacquer. Actually I was very pleased that the white undercoat did make a difference as the orange colour is now much brighter. Every cloud has an orange lining!

Another two week wait to allow the lacquer to harden and then it is ready to be flattened down again. This time I used the bare minimum of water with the 1200 grit paper, 2500 grit followed by rubbing compound and finally T-cut. It looked pretty good in most places, but I noticed a few areas where I had rubbed through the lacquer and it was flaking at the edges. D'Oh!

Painting is tedious.

I knew the flaking had to be dealt with before it started to spread everywhere, so the surface was roughened up slightly and a few more coats of PU lacquer were applied.

I now need to wait for another two weeks because "Painting is tedious".

Unless you're painting bagels...

 

The Learning Curve

I put on four or five coats of bright orange spray paint and allowed it to dry for a few days. I then lightly sprayed two coats of glitter paint to give the flat orange colour some added sparkle before putting on another half-a-dozen coats of clear acrylic lacquer. The can said allow two weeks before using rubbing compound! It was looking really good, but it was a long wait before I could do anything else to it.

I was starting to get quite excited about nearly completing the bodywork. All this waiting for solvents to evaporate is quite tedious, but a necessary evil. During this fortnight I took the opportunity to see how the pickup fitted. I had to enlarge the hole in the centre to allow some space for the cables, and drill three fixing holes.

As the lacquer had dried to a bobbly finish, I got the wet'n'dry 1200 grade out to get the surface smooth again. I rinsed it down and set to work with 2500 grade. Definitely promising. Another rinse and out came the rubbing compound. A good dash of elbow grease later (rinse!) and I'm onto an old bottle of T-cut which will provide the final polish. Wow! It looked really shiny, but it has got a bit wet throughout the process. I was so pleased with the result, which was multiplied by the fact that I wouldn't be messing about with sandpaper, paint & polish any more. I shook the drips off and brought it back indoors to dry out as the garage gets a bit damp this time of year (especially when I'm throwing water about).

The next morning I noticed the central heating had stopped working. I did a bit of prodding about, and eventually it became apparent that the previously very noisy pump was now really really quiet. This is not a good sign. I went to work and took a detour on the way home to get a new C/H pump. All I had to do was isolate the valves each side of the pump and swap'em over. Of course, like any good soap opera, the ending can always be predicted. The valves were seized onto the pump, so the system had to be drained and the valves replaced. The new valves were fitted and then it became evident that they were shorter than the old ones. I looked online for 22mm / 1.5" BSP valves but no-one specifies how long they are! So the next day it was an early start off to B&Q for 7am, get some 22mm yorkshire style elbows and get the blowtorch on the go. Warmth by 8:45!

After the central heating escapade, my attention returned to the bass body, so I picked it up to admire my good work and congratulate myself on a cracking job when I did indeed notice some cracks appearing. I was understandably annoyed by this. The water I used to rinse the paintwork had soaked into the wood grain where it was exposed (neck joint, jack, pickup and screw holes) and caused it to swell in places. Capilliary action works incredibly well on maple (so it would seem) and there was little choice but to let it finish drying out and have to start remedial work. Gutted!

To minimise the risks of the same thing happening twice, I'm attacking the problem on different fronts. The bare wood has been given a diluted coating of PVA to seal it and the cracks have been filled with wood glue. I ought to spray the neck pocket with a coat of paint or two. The majority of initial sanding (400 grade?) will be done dry. If I have to resort to wet sanding later, then I'll use a damp cloth to remove most of the residue. Of course, once I get to the painting, I'll still have to wait two weeks for the lacquer to harden. I'm so impatient!

Sanding Equals Dust

Sanding is quite dull. You take some paper with miniature rocks stuck to one side and rub it a few times against your piece of tree. This enhances the look of the wood inversely proportional to the ache in your arms. In the meantime everything gets a liberal coating of dust.

 

Some of the larger gaps had to be built up with polyester car-body filler and smaller holes with knifing putty. Both of these products are quite uncontrollable (in my hands), but I threw in a bit more sanding to compensate.

 

 

If wood dust isn't sufficient, sanding through layers of paint and filler gives several different colours of dust. I tried light brown, white and two or three different shades of grey. Not fifty though - it's the wrong sort of curves.

 

 

Perhaps a camouflage paint job would have been an idea? This shows the high-spots and low spots left by the straight grinder. Then I did some sanding.

 

 

A lot of the time it is difficult to see any improvement. Whilst waiting for signs of progress it is worth doing some sanding to pass the time...

 

After a bit more sanding, most blemishes are blended in fairly well. This process could carry on almost indefinitely, but I'm not a stickler for absolute perfection. A few more coats of Filler / Primer and I reckon it's nearly ready for the top coat to be applied.

 

  

 Strangely, the filler primer is a light brown / orange sort of colour. Looks promising...

 

 

I think six bolts should be enough to keep the neck on.

More bodywork

Creating the neck pocket took a while. The neck was carefully measured and the slot cut deliberately undersize with a saw. This was then whittled away with a chisel until the slot was *just* large enough to accept the neck heel. Six holes were drilled in the back so that the neck could be fixed securely. After that, it was time to assemble the basic bits to see if it would snap as soon as some string tension was wound in.

Fortunately, apart from an awful 20mm action height, it survived. I had aligned the bridge and neck etc so that the strings were centralised. As can be seen from the photo, this wasn't terribly successful. The string tension was sufficient to compress the 'C' shaped body and start to close it up a little. The E string was nearly off the f-board, whilst the G had more than enough space. The body bending is something I will just have to live with, so I allowed it to settle for a week before moving the bridge mounting holes down a few millimetres. I also took the time to remove some body wood from under the bridge to get the action down to about 3mm. Once refitted, the whole experience was more civilised...

This is the point where it could be played (acoustically) if only it fitted on my lap. It really needs a strap, but that will be one of the finishing touches. The other point is that the body alone weighs 3.7kg without any hardware! It feels much heavier than expected. The stainless steel reinforcement won't be seen once finished so I'll can try to convince people that it's filled with concrete...

Curves

Now that the whole will-it-snap possibility was sorted out, it was time to get rid of the blocky look. The maple body needed to be shaped which required the removal of quite a lot of wood. An orbital sander would not cut the mustard here. Some new toys had to be found - enter a DWT Swiss GS06-27LV Straight Grinder and a couple of Foredom Typhoon carbide burrs.

The body and neck were separated once more and the grinder was used to give the body a roundish cross section. Goggles, ear plugs & dust masks were the order of the day. The bridge and neck pocket areas have to remain flat and blocky, so the trick is to try and blend them in.

 

The rough shaped body is a great improvement over the slab look, but the process creates a huge amount of really fine sawdust. I made a dust castle in its own lunar landscape...

 



 

After the grinder had done the bulk of the work, there are several gouges and grooves which usually result from me using power tools. This time they were in the wood and not me, which I take as a good sign. Hand sanding with 60 grit paper is the first stage. It needs to look right at the end of this! I think it's going to be a long job, and I will also need to use some filler in places too. I foresee a lot more sanding ahead. Bored now...

 

 

Nuts

Due to the springiness of the stainless steel and the maple, the middle section needs a bit of clamping to keep it in the right place. Whilst the body parts are being glued together (not now Igor...) it seemed like a good time to have a go at the nut that should be on the neck.

 The original nut was a broken creamy-coloured plastic item that didn't inspire much confidence. The garage was invaded to see what was about and eventually I chanced upon an old PentiumII heatsink assembly. This was an old one - before they decided that everything gets a fan. The heatsink assembly consisted of a flat aluminium transfer plate and a separate finned thing. The black anodised transfer plate was 3mm thick and appeared to be a harder grade of aluminium than the regular stuff that often crops up. An original fixing hole would provide acces to the truss rod - great!

The aluminium was sawn and filed into shape whilst being protected from the vice jaws by small pieces of card. I marked out the centres of the slots and made them 3/8" apart. Once the corners were rounded off slightly, the nut was able to be used as a guide for the chiselling that was required on the wooden nut-holder!

Chiselling the neck was done as carefully as possible to avoid messing up the impressive paint job. I don't like the idea of just gluing a nut on without some decent foundations. When a note is fretted it probably makes virtually no difference, but for an open string it must have an effect.

Time to inspect the gluing of the middle section before the top layer is put on. It is looking pretty rough, but the revised pickup shape looks much more circular.

Here is the nut in place on the neck. It needs a bit of a tidy up round the edges, but nothing major. With the benefit of hindsight, I can see that the string grooves don't look equally spaced, and that maybe I should have used a fixed distance between the strings instead of their centres. In other words the string separation should be further for the fatter E string. Smeg! I might have to whittle the grooves a bit somewhen.

The neck to body joint will be the next thing to tackle...