MIMF

Does any one have an understanding of the function of the slanting "harmonic bar" on Ramirez guitars. Is it possible that it cuts the brightness and adds more low end ?

I think it is also about tightening up the treble side to enhance the higher frequencies.

More what Barry said. Bass comes pretty easily in a classical guitar. It's the treble that usually needs assistance by stiffening the area between the bridge wings and the sound hole. It's not particularly a Ramirez thing though.

Here is the deal, a friend crushed the top of his Ramirez and had someone quite proficient replace the top, the workman ship is flawless but it is not Ramirez bracing. The sound is good and complex but lacking a strong low end and tends to be overly bright in comparison to the original top. As there are a few wolfs I have thought of adding some wood at the nodal points to reduce this, working thru the sound hole. I am just wondering if anyone has experience of adding a brace as the 'harmonic bar " after the fact and if it encouraged the low end.

It seems to me that adding braces will tend to stiffen the top, but you are looking for low end frequencies, so probably shaving some existing braces would be more productive to that end.

Michael, Thank you,My own experience in having success in getting more low end has been in thinning around the edges of the top. I agree intuitively that thinning braces sounds as a possible solution, to adding more low, but also adding more wolf. My concern is that I have seen a great many 'improved' instruments with thinned braces and sunken tops. Obviously it is in knowing how much to take off. When I read about the harmonic bar, it seems though stiffer, it has the possibility of adding richness, more fundamental. I realize it is an odd question as there are so many variables.

I've not done much investigation into Ramirez guitars because I've never really liked their sound. However, numerous makers use similar slanting lower transverse braces, including Fleta, whose guitars I have investigated more thoroughly.

In a nutshell, the slanted transverse brace makes the cross dipole mode of vibration asymmetric, whereas typical Torres style bracing results in a symmetrical cross dipole. Depending on exactly how you lay out your bracing, you can create a strongly radiating asymmetric cross dipole (a diagonal dipole, if you like) with either positive or negative phasing with respect to the main monopole, which is the guitar's main sound radiator. It is therefore possible to add another strong peak in the guitar's frequency response curve (generally a good thing) at a frequency not much above the main monopole frequency to help shape the sound of the instrument and broaden the gain-bandwidth product in the 200-300Hz range, giving more power and volume in that frequency range, if you get it right. The Ramirez 1A models that I've heard seem to be examples of getting it wrong, which is quite easy to do, because with the diagonal brace it is quite easy to just lock up the whole soundboard. That, I think, is why opinions tend to be polarised on the effectiveness of diagonal harmonic bars. Get them right and they work well, otherwise they're pretty disastrous. Details, analysis etc. in the usual place.

Trevor, thank-you that is quite informative. Oddly enough, as I am primarily a violin shop focusing on historic instruments and typically only work on guitars over 100 yrs. old, I have had quite a few Ramirez come thru mostly 60's and 70's, 1A's and 2 A's. I prefer not to make a judgment on the timbre only that it is distinctive and to my ears relatively consistent. Your transverse bar analysis seems quite sound. Sorry, pun not intended. Isn't it always a matter of getting the braces in just the right place with just the right dimensions.

Stephen Bacon wrote:...Isn't it always a matter of getting the braces in just the right place with just the right dimensions...

...and with just the right material properties.

Trevor wrote:
"Depending on exactly how you lay out your bracing, you can create a strongly radiating asymmetric cross dipole (a diagonal dipole, if you like) with either positive or negative phasing with respect to the main monopole, which is the guitar's main sound radiator."

I've done some measurements of the sound output of of a symmetrically braced classical in the low range, and, as usual, it is both complicated and interesting. I was working on a 'port' experiment, and plotted the sound level of the guitar at different frequencies, 5 Hz apart, at eight spots around it. The guitar was hanging from the head in my 'semi-anechoic' closet, with the guitar and driver mounted on a turn table, so I could rotate the whole rig without altering the relationship to the 'room'. I compared the same guitar with and without ports, to find out what the port did. What follows is a discussion of the un-ported box: suffice to say here that the port superimposes it's own asymmetry.

First, you have to remember that the 'bass reflex couple' has a pretty long tail into the higher frequencies, and at the 'cross dipole' pitch of most classical guitars the monopole response will still be quite strong. Also, keep in mind that a mode will change phase relative the drive force as you go through its resonant frequency: if you drive the bridge off center one side of the dipole will be in phase with the drive force, while the other side is out of phase,while the monopole will retain the same relationship with the driver through the range, since you're above it's resonant pitch. In the dipole range you're well above the monopole pitch, and it will always be out of phase with the driver. What happens, then, is that, as you approach the dipole pitch a pair of 'lobes' will develop in the output, with, say, the bass side of the top radiating more sound than the treble. As you go through the resonance the lobes will switch rather suddenly. It seems to me that the _fact_ of the asymmetry is more important than the _direction_: you could slant the bar downward on the bass side and get the same result, only 'flipped'. I'm not convinced that would make a difference anybody could hear.

It's possible for an asymmetric dipole to have a net monopole radiation by itself: if the 'big' side has the same amplitude as the 'small' side, it will naturally put out more sound. Offhand I'd think that the 'big' side would need to be a bit lighter than the 'small' one for this to happen, but that's probably simplistic; most likely you end up splitting the difference. I'm not sure how you'd drive the dipole without getting the tail of the monopole in with it.

I also wonder what the cost is: there don't seem to be many free lunches around. I used to slant the waist bar in my classicals, and made the fan asymmetric too. The cross dipoles were noticeably asymmetric, but I was not happy with the treble tone. In conversations with another maker who'd pursued the 'Kasha' ideas for a while, he pointed out that by the time he got the bracing asymmetric enough to work more or less the way it was 'supposed to', the monopole was so highly damped that you ended up losing a lot of sound. Since I've gone to using symmetric fans I've gotten a better treble sound, and find no lack in the low end. This has not been, in any way, a 'controlled' experiment, and there's a crying need for that.

Alan Carruth wrote:...he pointed out that by the time he got the bracing asymmetric enough to work more or less the way it was 'supposed to', the monopole was so highly damped that you ended up losing a lot of sound...

That's the "locking up" of the soundboard I mentioned.

Suppose you have created an asymmetric diagonal dipole and not locked up the monopole (can be done) then you have one of Christensen's "simple sound sources" which is the net output of that sound radiator. This may be either positively or negatively phased with respect to the monopole. Assuming the dipole has a resonant frequency above the monopole frequency (usually the case) when the sound fields of the two radiators add, and they are of the same phase, there is a deep valley in the frequency response curve between the two peaks corresponding to the resonant frequencies of the monopole and the asymmetric dipole. If they are not the same phase, there is a "saddle" in the frequency response curve (and differences higher up the frequency range, too). That difference is audible. (Well, it is to me!) I show examples of the modelling and the frequency response curves of various guitars in the book, but way too much material to include here, unfortunately.

Just a question. Ramirez III who introduced and consistently used, to my knowledge, this new (and bold) type of diagonal "crossing" strut almost at the same time he pionnered the use of WRC for tops, maybe was also tempted to fight a tendancy to correct a little bit the usually too generous, and not extremely refined character, of many large WRC topped guitars?
BTW, the use of a slanted bar, but located closer from the bridge area, and not crossing the waist bar, was a very standard feature of a large majority of (you say ladder braced, I believe?) early romantic guitars.. Italian, French, German... Even some baroque guitars had this, in a modest way however. All this spruce topped of course.

Very interesting discussion, this. I might be experiencing some of what has been mentioned above. I use an angled treble bar, and have found that I get more of the sound I'm after when I loosen up the top below the bridge, as well as moving the closing bars away from the bridge and flattening out the angle between them.

I wonder if someone has measured properly the thicknesses (including distribution) of a Ramirez III "post war" large WRC model. I do not remember having seen that info before. The short paper visible on internet (by Gruhn, Gruen or a name close to that) is silent on this point.

Hi Brad,
Maybe you would be just as well off not using the slanted bar and keeping a little more wood in the top. Classic Torres!

Clay - What fun would that be? Besides, I've had really good results with the treble cut-off bar. I never really liked the sound I got with a symmetrical bracing plan.