Starting out on J45

[Trevor Gore]

All in a 2m x 3m space.

2m x 3m!!??

You was lucky. When I were a lad...

[Steve.Toscano]

All in a 2m x 3m space.

2m x 3m!!??

You was lucky. When I were a lad...

Yep, it was a garden shed. Later i upgraded to a single car garage space which wasnt all that much bigger.
When you were a lad ..... were you making in a water tank or something?

[Trevor Gore]

[Richardl]

Hi All. Well, I'm the proud owner of a square 3/8 inch hole. All has turned out well. I visited a friend with some heavy equipment. Of course, the collars supplied with the mortise kit didn't fit either drill press he had so turned up some that did fit one of them. Then he gave me the drill press . It's an old Richardson E39, weighs a tonne and is extremely heavy duty and runs perfectly true with no slop at all - they don't make em like that any more. Cut the mortise no sweat. Just goes to show though, the little handy-man/person drill presses aren't up to this sort of punishment. So, I'm going to bed a happy camper

Cheers
Richard

[Trevor Gore]

That's good to hear. I don't think I've ever come across a piece of stationary manufacturing equipment that would function better for being lighter!

[Richardl]

It's very easy to over-build a J45. The originals were made with less-than-best wood and retailed for $45 (i.e. cheap) compared to say the J200 which was originally $200. So if you choose good materials it's easy to end up with everything too stiff.

Having just re-read Section 7.1, I can see that it is not the clearest bit of writing! The dimensions given (20 x 10mm) are essentially back brace dimensions and the procedure described is primarily the back brace making procedure. You need to modify the dimensions to the X-braced plan you're using. The plan from the book can use brace blanks starting at ~15 x 7mm so you can finish at ~1/2" high at the X and a width of 1/4" (~13mm x 6.3mm), Section 11.2.2. The better your jigs and shop practices, the closer you can start to the finished brace dimensions, but the same procedure as for making back braces can be used, whilst leaving the un-gabled bit for the X as described in Section 7.1. These are the dimensions we use to get the resonant frequencies in the right target zone without risking undue distortion. Most published plans (usually taken from factory made guitars) are way over-braced when good materials and assembly practices are used.

There is no need to notch finger braces or face braces into the X braces. If you make good joints with fresh surfaces and glue with Titebond, the wood will fail long before the glue (at normal temperatures).

The bridge plate is beveled or rounded on the exposed top and bottom edges (Section 11.2.2.2) and is glued down first and then acts to locate the X-braces accurately when they are glued down. The X-braces are tightly butted up to the bridge plate edges, to which they are glued.

If you follow the procedures and plans, you should be pretty right. A good many people have been pleased with the results. However, if you do find anything else that requires clarification, just ask here.

Hi Trevor

I'm in the midst of making my second J45. I liked the first but thought I could improve. Quite a lot! And since I have all the jigs...

I have re-read the bracing bit and just hoped to clarify the back bracing please. The text seems to indicate that the bracing doesn't matter too much in a non live-back guitar and some folk make very heavy bracing. The text seems to indicate 10mm x 20mm gabled and the plans indicate that too except for the box referring to the J45 which indicates 6.3 mm thick x 20 mm high gabled. I've asked about the bracing previously but more in relation to the soundboard which I think I have sorted. So, the question is does the back bracing matter in this case as its a standard ladder back? And would a stronger 10 mm thick bracing be the better option?
Thanks
Richard

[Trevor Gore]

In Section 11.2.2.13 I talk about the details of J45 back bracing. The "real" J45 has a live back and the live back is important for the reasons given in that section. So if you want to improve over your previous J45 I'd recommend going for a live back and the details of how to do that are also discussed in that section. The 4 bar bracing system works for a live back, but you have more control of the pitching of the resonant frequencies if you use the 3 bar "star" bracing pattern. If you want a non-live back, 4 bars at 10 x 20 and a slightly over-thickness back will do that.

[Richardl]

Thanks Trevor.

I must have read that bit last time and gone with the traditional 4 transverse braces of 6.3 wide on 2.8 mm thick mahogany - which, I hadn't understood, gives a 'live' back after all - a contribution to the sound from back vibration - which I take it is better and richer in complexity as long as (in layman's terms) it doesn't clash with the soundboard. Must say, to my ear it sounds good, but i don't have a means of measuring the frequency response. Have I generally understood that right?

So, the potential improvement you are suggesting is a more controlled response from the back using the radial braces and 3 x gabled transverse braces, the lower one scalloped.

The plan shows the radial bracing layout. It indicates a J45 traditional (BUT does that mean 4 transverse braces as described in 11.2.2.13) has a 6.3 mm "B" dimension, as I used previously. Does that mean that the radially braced design with a scalloped lower bout brace is made from 10 mm wide x 20 mm tall (in the middle of the board) spruce (the lower brace flat area reduced to 13 mm high - in the middle of the board)? That seems correct from the photo in Fig. 11-20.

Cheers
Richard

[Trevor Gore]

I must have read that bit last time and gone with the traditional 4 transverse braces of 6.3 wide on 2.8 mm thick mahogany - which, I hadn't understood, gives a 'live' back after all - a contribution to the sound from back vibration - which I take it is better and richer in complexity as long as (in layman's terms) it doesn't clash with the soundboard. Must say, to my ear it sounds good, but i don't have a means of measuring the frequency response. Have I generally understood that right?

Using the 4 brace scheme and a random thickness back does not necessarily give a live back, though clearly that happened occasionally with the originals. If you want to be certain about a live back, you have to design and make it that way. If you can't measure the frequency response, you won't know what you have, so the first step would be to set yourself up with Visual Analyser, otherwise the rest is moot.

[Richardl]

Hi Trevor

thanks for your reply. Having just had a quick look at 4.4.9 (must admit I haven't picked up the first book in a while) it appears the answer to my question about radial brace dimensions is: it's 10 mm thick and 20 mm high - the "standard solution that has worked well on many guitars".

I understand that if you want to be certain about performance, one might need to invest in ways to analyse the response, as you might in any industrial lab. However, as a weekend self-taught woodworker, I haven't got a frequency analyser at my disposal and am not too sure where to access one. Personally, I'm still grappling with the difficulties of making a decent structurally-sound box (that plays in-tune) with a limited array of hand tools. So, maybe, if a live back can add unwanted embellishment of the sound, a non-live back is the better option for the novice? My rosewood back is still a 2.9 mm thick flat panel, so if I use it as is, from what I understand from this discussion, it will contribute little to the top's frequency response curve, irrespective of the bracing. Is that correct? Alternatively, where would you find a frequency response analyser - outside a physics lab? Cheers

[Richardl]

Hi Trevor

thanks for your reply. Having just had a quick look at 4.4.9 (must admit I haven't picked up the first book in a while) it appears the answer to my question about radial brace dimensions is: it's 10 mm thick and 20 mm high - the "standard solution that has worked well on many guitars".

I understand that if you want to be certain about performance, one might need to invest in ways to analyse the response, as you might in any industrial lab. However, as a weekend self-taught woodworker, I haven't got a frequency analyser at my disposal and am not too sure where to access one. Personally, I'm still grappling with the difficulties of making a decent structurally-sound box (that plays in-tune) with a limited array of hand tools. So, maybe, if a live back can add unwanted embellishment of the sound, a non-live back is the better option for the novice? My rosewood back is still a 2.9 mm thick flat panel, so if I use it as is, from what I understand from this discussion, it will contribute little to the top's frequency response curve, irrespective of the bracing. Is that correct? Alternatively, where would you find a frequency response analyser - outside a physics lab? Cheers

OK, I have just discovered Appendix A in the first book. I'll have to think how I can set this up at home with a microphone and will check out the software suggested.
Cheers
Richard

[Richardl]

Hi again. I've downloaded Audacity and SpectraPlus (realtime ed). I managed to get a spectra of sorts using SpectraPlus, after tapping the bridge, but it doesn't have the controls discussed in Appendix 1 - can't even adjust from a log scale of frequency to linear. Can't make head nor tail of Audacity for this purpose although it looks great for manipulating recorded tracks.
https://www.flickr.com/photos/143232438 ... ed-public/

I feel like I'm trying to learn to play the trumpet by a combination of Braille and semaphore. Something that isn't necessarily going to end well. I'll give up for the moment at this point. Are there any you tube tutorials on generating relevant frequency response curves?
cheers
Richard

[Trevor Gore]

Best just to download Visual Analyser (for free) and follow the instructions in the Appendix.

If you're using a Mac, you'll need an emulator type program so you can run Windows apps (Parallels or Wine seem to be the popular ones) or buy a cheap Windows PC.

If you need further instruction, you can get it from here: Guitar Analysis and Testing with Trevor Gore

[Richardl]

Hi Trevor
I downloaded Visual analyser - it seems far better for this application. I also found your instructions for it's use that you had posted earlier to someone else, so managed to set it up. I generated this response from my J45 using a small external omnidirectional microphone used for calibration of a Stereo amp (may not be the right sort, but is all I have). The first major peaks are at 93 and 195 Hz, the latter seem to be higher than that shown in 4.7-2, although I don't profess to understand the significance of that yet, unless the higher frequency says the top is too rigid. The signal looks noisy.

https://www.flickr.com/photos/143232438@N05/?

If it looks like I'm vaguely on the right track, I'll start to look at the boards. However, I have already planed these close to thickness and joined them so can't apply the footnote on p 11-7 to determine the target thickness which brings me back to the conclusion that perhaps I'm safer making a guitar with a non-live back and concentrating on getting the woodwork right (if I have already gone too far to tune the boards properly).
Cheers
Richard

[Ed Haney]

I hope I'm not hijacking this thread, but I'm building a guitar and I've been referring to the J-45 plans as a reference. My client wants a Martin dread shape and I've found it easy to insert the J-45's ladder and "star" braces design into the dread's back. I have a good handle on thicknessing the top and back according to the methods and calcs in the "Book", so that is not an issue. I also made good E measurements on my brace wood so that I know what it is. And I have mass support braces glued on the rims. But i want to home in on the back design without just using pat brace dimensions for the back (and the top braces too.)

- But I do not see any reference anywhere in the Book that tells me what Young's Modulus value was assumed for the J-45 back bracing plan. Should I assume the Book used a value of E = 10GPa which is the average for Sitka spruce given by the Book and is also my experience? Or did I miss it and the Book gives the E value assumed for the plans?

- If the Sitka I'm using is, say, 8.77GPa, is there a fully fleshed out example somewhere that I can review to increase the dimensions of the J-45 back bracing to equal what the plan was shooting for? If an example entails using Equ. 4.4-2 (pg 4-38 Design Book) and the parallel axis theorem, then don't we need the J-45 plan back thickness and E value to complete the design with the bracing?

- Same questions as above except for the top brace design.

- Independent of the J-45, is there a complete start-to-finish EXAMPLE anywhere showing all the calcs for bracing design of a steel string top (medium size or J-45 or Martin Dread)? Unless I have missed it, the bracing design is not as detailed in the Book as the top and back thicknessing. Unfortunately, I need more hand-holding for calculating bracing designs. (The falcate design is given in detail but not with calculations. Adding the CF reduced some need for calcs, I assume.)

If someone could point me in the right direction to answer these questions I would really appreciate it. I had signed up for a workshop with Trevor in July in the US that has just been cancelled. These were areas I needed help.

Thanks,
Ed

[Ed Haney]

With enough time and effort I was able to figure out the spreadsheet formulas for calculating the soundboard bracing system.

I still wonder what value of E was used for J-45 in the Book.

[Trevor Gore]

I still wonder what value of E was used for J-45 in the Book.

10GPa. Don't go less than that, but stiffer is OK as you can always shave a bit off the tops of the two lower bout braces.

[Ed Haney]

Thank you, Trevor.

[Ed Haney]

As mentioned above, my spreadsheet for flexural rigidity is working fine on the example in the Book.

I ran the spreadsheet on the J45 plan and got a Flexural rigidity of 58.6 Nm^2. If I understand the Book correctly, 50 Nm^2 is a good target for steel string guitars with light gauge (12's) strings. Correct?

Below are the bottom line numbers from my spreadsheet.


"Flexural Ridigity EI
Component" "Flexural
Ridigity
Nm^2"

Soundboard EI 11.29

Triangle Brace EI 2.01

Triangle Brace EI 2.01

Triangle on Gable EI 16.63

Rectangle on Gable EI 4.99

Triangle on Gable EI 16.63

Rectangle on Gable EI 4.99

TOTAL SYSTEM EI 58.55

What EI for the J45 are other folks seeing?

[WilliamDavidReynolds]

I do all assembly at 45% RH.

The sides can be problematic because they tend to be heat and damp tortured and then ignored. They are often assembled at unknown RH and, mostly, people seem to get away with it. I tend to bend the sides early in the process and give them a week to normalise at 45% before I cut to length and glue in the end blocks, then linings and side stiffeners. The more you move away from "standard" woods (EIR, mahogany, blackwood) toward the less used local woods, the more I'd want to assemble everything at 45% RH, not just the top and back.

Section 11.2.2.2 states New Guinea Rosewood for transverse brace, I looked for it on Wood Data base, for the properties, weight/density/ym,, etc, it’s not there. Is there another name for this wood?
I have a lot of wood that could be used here rather than spruce if it has similar qualities to the New Guinea
Like perfectly tight quarter sawn African mahogany(Khaya)
Or ,,,

[GregHolmberg]

Section 11.2.2.2 states New Guinea Rosewood for transverse brace, I looked for it on Wood Data base, for the properties, weight/density/ym,, etc, it’s not there. Is there another name for this wood?

I searched for it on wikipedia and found this: Pterocarpus indicus, commonly called Narra (wood database).

You can see a summary table of the data from wood-database here: tonewood (wikipedia). From the table, a similar wood might be Sapele or White Ash.

Greg

[WilliamDavidReynolds]

Section 11.2.2.2 states New Guinea Rosewood for transverse brace, I looked for it on Wood Data base, for the properties, weight/density/ym,, etc, it’s not there. Is there another name for this wood?

I searched for it on wikipedia and found this: Pterocarpus indicus, commonly called Narra (wood database).




You can see a summary table of the data from wood-database here: tonewood (wikipedia). From the table, a similar wood might be Sapele or White Ash.

Greg

Thanks, as soon as I posted this a googled it and saw Narra, I shoulda updated the post or deleted my comment. Haha I’ll tell you this, it’s nice to interact with you sir. I just work on my prototype archtop bass guitars and now a new line of 34’ radius flat tops, all alone and no one to shop talk luthiery with. I’m yacking at my wife and she displays appeasement but generally isn’t into it,,,, yet! She was hyped on the math though, she did her own spreadsheet version of the 4.5-7 equations and we were comparing our outputs. Romantic nerds