C18Q1 Midi guitar

What's Hot
2»

Comments

  • RolandRoland Frets: 8590
    edited January 2018
    This afternoon is the first time in more than two weeks that I’ve had an opportunity to work on the challenge. I wanted to work out where to drill the bridge plate for piezo wires. Following my comment in another thread I measured the existing saddle positions, and soon realised that the node point on a Tele saddle isn’t as obvious as it is on a tunomatic. To get a decent measurement for the piezo saddles I swapped them for the top and bottom E strings on an existing guitar, and corrected their intonation:



    What do we find?
    1. The node points on the two saddle types are at different positions from the saddle front. The difference is 1.8mm, which is over 3 complete turns of the adjustment screw.
    2. The Graphtech saddles use American threads for saddle retention and height adjustment. The Graphtech’s don’t have much height adjustment, and I don’t have any longer spares with the correct thread. Thankfully there’s a compromise height which works.
    3. There’s a tone difference with the new saddles, but not enough to worry about.
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    Change of plan. After a little thought I'm going to put the midi gear into this existing guitar, rather than a new build. With everything else I've got going on a new build is increasingly unlikely to get completed before the end of March.

    So the midi stuff will go in like this. I'll cut a groove under the bridge for the piezo pickup wires and drill from that into the control cavity, rather than route the delicate wires alongside the magnetic pickup cables. The cavity will need enlarging to take the midi circuit board and midi volume. The 13 pin socket will go on the edge, near the end pin. Any lower and it will get in the way of my thigh when playing. The battery will go in its own box behind the bridge, with a hole drilled through to the control cavity.



    Sorry @Andyjr1515. This means that the jigs will get documented in another thread

    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • Andyjr1515Andyjr1515 Frets: 3127
    No problem @Roland.  It'll be an interesting project to watch anyway :)
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    Today I started by drilling holes in the bridge plate to take the piezo wires. The wires come out of the Graphtech plug using a pin to release the plastic trigger (arrowed in red).



    The pair of wires side by side measure 1.5mm, and a single metal terminator will go through a 1.5mm hole. To get the second terminator through alongside the first wire through the hole needs to be 2.0mm.

    Having pontificated in this and another thread about where to position the holes I discovered that they would be too close to the string-through holes. I ended putting them just behind the front edge of each saddle.

    After that excitement I drew up the outline of the enlarged control cavity, and cut out a paper template. Extending the cavity will be the first non-reversible activity, closely followed by cutting a hole for the 13 pin socket. Now this is a bound guitar, and I’m thinking about redoing the binding, including rerouting the ledge. This will be much more difficult with another socket hole in the side of the guitar. So I put it aside while I decide on the binding, and dressed the frets instead.

    Normally when I dress frets I take the strings off and let the neck settle, adjusting it back to flat over a couple of days. As luck would have it this neck was dead straight within a few minutes of removing the strings. I seized the chance and levelled the frets, then spent a happy couple of hours reprofiling and polishing them.
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 1reaction image Wisdom
  • RolandRoland Frets: 8590
    This morning I sanded off the old finish to remove pick marks and other divots. I’m going to refinish and rebind the face of the guitar. I also cut the channel under the Bridge for the piezo wires. Yesterday I bought a Makita mini router to redo the binding. The body face is contoured, and I want to use the Makit bevel foot to follow the guitar top at varying angles when I recut the binding channel. I wanted to get some experience of the new router by using it to open the control cavity. The new router is light weight, and ideal for the shallow channel.

    Don't forget to post the photos 


    I’ve changed my mind about how the piezo wires will get into the control cavity. To get the channel sufficiently deep in the body I’ve had to come into the pickup cavity, and then a new hole from there into the control cavity. This is where my new pillar drill is useful.


    ... interested in your thought process on jigs ...
    For the extended control cavity I was tempted to route it out with the aid of a few ad hoc fences. As things turned out I’m glad that I didn’t. First I drew the cavity outline on paper, cut it out, offered it up the the guitar, and checked the positions of the pots, sockets and circuit board. Then I cut a mask from 9mm ply on my scroll saw.



    Using the piece I’d cut out I drew a larger outline for the cavity cover, and cut that out too. The blank will be used as a template for the cover, and the hole as a routing template.



    Here’s where it went wrong. 

    To start with I drilled out the extra area with a Forstener bit to make life easier for the mini router. With my main router I’d just have piled in, and let the router do all the work. The new router worked well at enlarging the cavity. Then I reset it to sweep across the cavity floor. Halfway through the bit came loose, and cut too deep into the floor. I caught the problem, turned off and hoovered the cavity so that I could check for damage. Not too bad. I can bring the whole surface around the pots to this level, and insert a ply or metal platform to strengthen the wood and support the pots. Cleaned the shank. Set the router up again ... and the bit came loose again. This time it came within 0.5mm of cutting completely through. Thought about it for a while, packed the hole with plastic wood, and gave up for the day. I would have liked to use something stronger, but cyano-acrylic or wood hardener would seep through and inhibit the new coat of wood stain.
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    Damage recovery: Firstly the face of the guitar was restained blue, follower by cyano acrylic glue to rebind the damaged wood fibres, and strengthen the damaged area. The Plastic wood filler was then levelled with a router. This time I used my old router, which I have greater faith in. More superglue to fill and strengthen around the plastic wood because it can pull away from the wood as it shrinks as it cures.

    Then back to work. Binding channel recut. This is what I bought the Makita for. I’ll give it 8 out of 10. It’s light, and easy to guide. With the bevel base locked at 90 degrees it does a good job. Much better than my old heavy router. Around the forearm champfer I stopped and adjusted the base angle every two inches, which gave a smooth transition. The bevel base holds the router bit at a constant depth, but I forgot to allow for the width of the bit. As the angle increases the bit bites slightly lower. Of course there’s a three inch section where I didn’t notice this. It’s on the top edge, exactly where I’ll see it as I’m playing. There are a couple of choices: recut and use a deeper binding, redo the champfer to match the channel, or fill the gap with plastic wood. On any other guitar I’d recut or re-champfer. This one is the Recyclecaster, made from a builders’ plank I found in a skip, so it’s going to get the gap filler treatment.

    The original binding was put on with cyano acrylic glue. It sticks too quickly for my liking, and any excess can prevent wood taking colour and finish. @WezV, the fount of all guitar building knowledge, suggested plastic cement. The main ingredient is methyl chloride, or methylene chloride. It is used to make plastic models, and works by melting the plastic which then welds together. In this case I’m hoping that it will weld the plastic binding to the wood.

    Gratuitous picture. As often happens, this doesn’t pick up the true colour, which is dark turquoise.



    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • Andyjr1515Andyjr1515 Frets: 3127
    It's a little concerning that the bit loosened twice...  Did you ever bottom the root cause?
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    Not yet. My suspicion is that the collet on the new router is sensitive to the position of the bit. However I’ve not devoted time to testing that theory.
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • Andyjr1515Andyjr1515 Frets: 3127
    edited February 2018
    Roland said:
    Not yet. My suspicion is that the collet on the new router is sensitive to the position of the bit. However I’ve not devoted time to testing that theory.
    This would bother me, Roland.  The position of the bit - provided it is inserted at least up to the (usually) etched insert lines on the router bit shank - shouldn't make a difference.   Loose bits on routers is something that should never happen and it's something to urgently investigate if it does happen more than once.  They spin round far too fast to risk them being anywhere but fixed firmly in the collet. 

    I think either the bits are not inserted deep enough (which might be what you are saying), or the collet is faulty or it simply it hasn't been tightened up enough.  I say the latter because I find my Bosch palm router quite difficult to tighten properly using the lock pin - and Bosch don't supply a second (slim) spanner even though they do provide the spanner slots in the spindle.  I find that I'm having to put the pressure on keeping the spring-loaded pin engaged and not where the pressure should be - ie stopping the circular body from rotating.  It's one of the flaws of the available palm routers - particularly if a second slim spanner isn't supplied.

    Are you using two spanners or the pin?  If you are using the pin, try the two spanners (easier with the motor unit out of the base) and see if you are able to get a bit more purchase on it.

    If you are already using two spanners, then it would be worth taking the collet off and checking there's no swarf or dirt in the slots keeping it from clamping properly. 
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    The second time the collet was definitely tight as it would go. That’s the first thing I checked. I’ve used the same and other bits in the same collet since, and I’ve used the same bit in my other router. This tends to rule out a problem with the bit, or with the collet in general. What I haven’t done yet is to set up a depth test to see whether there’s a minimum length of shank which the collet will work with. 
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • WezVWezV Frets: 16543
    Inserting the bit too far can result in it coming loose, if it’s trying to grip to the transition between shaft and bit
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    If anything the shank was too far out.
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • Sounds like you have it all in hand @Roland :)
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    Saturday was final assembly day. To my surprise it all pieced together and worked first time. Here’s the bridge going in. Thanks to the marvel of photography the chrome bridge looks black. You can see where the piezo cables will drop into the channels in the body under the bridge, and the copper flex which earths the bridge.



    Here it is back together, and waiting for a chrome knob for the synth volume.


    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 1reaction image Wow! 0reaction image Wisdom
  • WezVWezV Frets: 16543
    Good work
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • Yes - looking good, @Roland ;
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    I’m going to need some help.

    Installing the saddles, electronics, and 13 pin socket was the easy bit. Getting everything working properly is a lot harder. My previous experience of midi guitar has been using a Parker Fly. First with external and internal GK-3 pickups, and later using the Parker piezos saddles into a Graphtech Hexpander. Tracking was a lot better with piezos: better note registration, and very few ghost notes. 

    This guitar tracks very well. If anything the Graphtech piezo signal is too strong, stronger than the Parker. There are no missed or late notes, and no ghost notes either.

    The problem is that some notes continue to ring, even when the string has been muted. As far as I can tell it’s either triggering from an inaudible signal, or there’s an electronic problem preventing the note from terminating.

    Anyone got any ideas?
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • m_cm_c Frets: 1211
    Anyway to connect up an oscilloscope to see what's the piezo signals are producing?
    If nothing else, it should tell you if it's some kind of ghost triggering, or something further down the signal chain.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    I’m pretty sure that the Graphtech saddles are part of the problem. I have the same Graphtech circuit boards in a Parker Fly, driven by Parker piezos. The GR-33 tells me that the Graphtech saddles are producing a higher voltage signal than the Parker saddles. With the Parker I have the tracking set to Accelerated Finger, whereas with this guitar it has to be set to Normal.

    I’m wondering whether a high pass filter on each piezo channel, set below the natural frequency of the open string, would reduce the noise.
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
  • RolandRoland Frets: 8590
    Incidentally, the Makita router has gone back.Yesterday afternoon there was a very dramatic puff of smoke. The workshop still smells of burned out wiring.
    Tree recycler, and guitarist with  https://www.undercoversband.com/.
    0reaction image LOL 0reaction image Wow! 0reaction image Wisdom
Sign In or Register to comment.