|
Post by JohnH on Sept 20, 2008 3:58:42 GMT -5
I want to show you a spreadsheet. I have no idea if anyone will be interested, but it plots the frequency response of a guitar circuit, with all the main variables adjustable, so you can see what happens when you change to a different pot value, or a new capacitor, and also what happens when you turn the knobs. Its here: EDIT 06/10/08: Actually, its now in the reference section, updated versions, see here: guitarnuts2.proboards45.com/index.cgi?action=display&board=reference&thread=3627This is how it should look: Its all interactive, change a value in a yellow cell to see the graph jump about. Also, there's some buttons on the side, which have macros assigned that write in the yellow cells. So you can click to see the difference with longer or shorter cable, different pickups and different treble bleed arrangements. The math is all down the page. I think its probably right (complex numbers), since it matches similar circuits modelled with spice analysis. The advantage of this one is that its all very instant, plus its helped me remember some stuff I used to almost know 30 years ago. Comments and suggestions are invited. John EDIT: Just a watchit for young players: These kind of analyses can only pick up the electrical aspects of a guitars tone, and don't know anything about coil spacing and position, timber qualities and general mojo. Also, it does not know about the relative overall output level of different pickups, ie that a typical Hb will have greater output than a single coil etc - all the analysis is done with the same basic level (reference 0dB) of input signal.
|
|
|
Post by newey on Sept 20, 2008 7:40:27 GMT -5
John-
Very Cool! +1
However, I get a message disabling the macros because of the lack of a security certificate. This requires lowering the security setting to use the macros, which Mr. Gates tells me is "not recommended". But I trust that your macros remain pristine and unhacked.
|
|
|
Post by andy on Sept 20, 2008 9:43:21 GMT -5
I have no idea if anyone will be interested I should imagine that lots of people here will be VERY interested! I try to steer clear of too much electronics stuff due to my sloppy soldering skills, but to have an idea of the effect of putting in a different value component, when searching for tonal perfection, could be invaluable! (Yes, pun intended, but it's still true).
|
|
|
Post by ashcatlt on Sept 20, 2008 10:37:22 GMT -5
That is awesome!
Considering the number of times I link to the static graphs you've created, I'm sure I'll be pointing this out fairly regularly. The one thing I see is that some of the values (specifically self-capacitance and inductance of the pickups) aren't always that easy to determine. Manufacturers are happy to tell us the DC resistance, but...
Thanks much for this.
Oh, and +1.
|
|
|
Post by JohnH on Sept 20, 2008 16:07:16 GMT -5
Thanks for the comments guys. Newey pointed out an issue with the macros being suppressed by the MS data police which are embedded in Excel. I think this is an issue with Excel 2007 and Im looking for a workaround that avoids the need to open the flood gates to other security problems. The sheet is written in Excel2007, and saved in compatibility mode for 97-2003. I'll keep at it. In the mean time, here is a good article on pickup properties: buildyourguitar.com/resources/lemme/index.htmand buildyourguitar.com/resources/lemme/table.htmcheers John
|
|
|
Post by ChrisK on Sept 22, 2008 15:28:15 GMT -5
I'm assuming the the frequency response effects of the magnet(s), material, and structure are also excluded within "general mojo". For only an analysis comprised of one pickup, the differences in output levels doesn't matter since all is relative (dB). A suggestion might be to also chart the phase angle (but then you'd have to explain it .) I would also suggest that you label the signal output and the signal return (the ground with a triangle thingy) otherwise you'll be forever explaining it. All in all a good effort. Thanks, it makes it easier for many to "tune about" I still cannot get my wife to use pSpice. Just last week she said that the car was "getting bouncy". I said that the resonant frequency of the suspension was coming more into play as the damping effect of the struts and steering damper were wearing. I suggested that she model it to better understand it. She strongly suggested that I get it fixed.
|
|
|
Post by D2o on Sept 22, 2008 16:01:44 GMT -5
Hi John,
I meant to offer my two cents as well. Thank you for this. I am excited about what the worksheet may be able to do, once it lets me do it.
It would be interesting to see how the various values interact. I always see these pspice charts and I don't really know what they mean. With your spreadsheet, I may be able to input certain "known flavours" (e.g. a Les Paul ... or a Stratocaster) and then see what tales I can exstratolate from there.
In the meantime, I look forward to gaining access to the Tales From the Unencrypted.
D2o
|
|
|
Post by warmstrat on Sept 22, 2008 16:04:41 GMT -5
I have the same problem with the macros.
Not REALLY a problem, as I trust JohnH, but... yeah. Y'all know where this is headed.
+1 again, I suppose.
|
|
|
Post by JohnH on Sept 23, 2008 1:00:21 GMT -5
Thanks for the further comments.
Chris – phase angle would not be too hard to graph, once I figure out properly how to extract it, (plus what it really means). I’m not sure what seeing it might lead to, but it could give some further insights into what is going on, so long as it does not obscure the response results.
On the macros, I’m sure they are clean, but it also is best not to lower your guard generally just to let one spreadsheet run. It seems to be Vista or Office 2007 which is being picky. I saw, on my home computer, an option in the Excel2007 security tab to allow macros to run if they are saved in ‘trusted locations’ on your machine. There is a list of where they are and you can define more. So maybe saving it to such a place would allow it to run without overall reducing security.
But all the macros do is to type numbers into the yellow cells by the diagram, which you can alsol do manually.
Cheers
John
|
|
|
Post by ashcatlt on Sept 23, 2008 14:21:10 GMT -5
I thought I might mention, in the name of total geekdom, that this shows the response of the filter circuit that is the guitar. It is not necessarily what you'd see if you did a spectrum analysis of a guitar actually being played. That would (obviously?) depend on who's playing what and how.
That dropoff seems pretty severe and very low, sitting around 3-4K. In comparison to the 20K bandwidth assumed for human ears it is. Sits right in what we might call the "high-mids". Keep in mind that the highest fundamental available to most of us is somewhere short of 1.2K, and most of what we do is much lower than this.
I'd submit that there really isn't anything but pick noise and EMI above the cutoff. Nothing you really want to hear anyway. In fact, for most stuff we tend to filter even lower (via speakers which are not made for anything near full range output) after the guitar output. The sound of most guitars directly out of a Hi-Z buffer or similar is just plain nasty!
|
|
|
Post by ChrisK on Sept 23, 2008 20:05:45 GMT -5
Well, there is some stuff up there. I always play each new guitar acquisition thru the same PA system to hear what it really sounds like. One can always get the tone out later, it's a lot harder to put in if'n it isn't there to begin with. Most guitar speakers are low end drivers. In a stereo we cheat and add midrange drivers and tweeters. We didn't have to with tube amps since they are crap in frequency response anyway(yeah, yeah, I know, I used to be a "glass audio" aficionado meself, back when I could hear). Well, at least the one's with "mojo". If a guitar was a piano, we'd only use less than half the keyboard anyway.
|
|
|
Post by JohnH on Sept 26, 2008 6:31:10 GMT -5
I added a few more bells and whistles, including a value for the height of the resonant peak (relative to 100Hz) in dB, a button for an on-board buffer (=much lower capacitance than a guitar cord - I like them!), plus a 'Hot' humbucker for comparison to the single coil and Vintage PAF Hb. This extra pickup has a higher inductance and resistance and is based on values I found for a Dimarzio Super Distortion. You can see the roll off in top end with the hotter pickups.
The first post now links to the updated file.
John
|
|
|
Post by D2o on Sept 26, 2008 9:36:32 GMT -5
Hi John, Thanks for continuing your work on it and for sharing it with us. I have been muddling my way around this manually, and I am appreciating it more and more. Sweet! +1 from me too. D2o
|
|
|
Post by D2o on Oct 3, 2008 12:37:47 GMT -5
Shouldn't this be sticky-ed?
D2o
|
|
|
Post by JohnH on Oct 3, 2008 16:45:40 GMT -5
newey will know what is best. I was thinking of having it ultimately in the reference section, but I posted it here first for review where more people could see it.
John
|
|
|
Post by ChrisK on Oct 3, 2008 18:30:22 GMT -5
JohnH,
Once you tell me it's final, I'll copy it, downsize it to just the spreadsheet post, and move it to the "Reference" section. I'll leave this thread in place.
Then after, as it gets revised (error repairs), please update that thread post. As it get re-versioned (updates in a non-backward compatible manner) please post anew the new version while leaving the current version(s) in place.
Please embed the version and revision visibly in each instance and file name (ie Vx.yy as in Version "x".Revision "y" and name x_yy.extension for the file name.
Of course, if I was doing it, it would be name x_yy_CCYYMMDD.extension. (CCentury/YYear/MMonth/DDay)
After 30 years of producing products with embedded and PC-based firmware.........
|
|
|
Post by newey on Oct 4, 2008 11:01:48 GMT -5
Chris is right, leave all our long-winded discussions in place here, and post the lean and mean version in references. I'll sticky it once it's there.
|
|
|
Post by JohnH on Oct 4, 2008 19:00:25 GMT -5
Chris and Newey - OK thats all fine. I'll make a self contained post for the reference section soon, and I'll start following a numbering convention as Chris recommends to track versions. I think in terms of what the spreadsheet covers, it's got enough, and for any more circuit configuration versatility, its better to go to a Spice analysis. What the spreadsheet wins on however, is the direct hand to eye link of the controls, so you can drive it on screen and instantly see the result. To that end, I'm now fiddling with 'scroll bars' and 'radio buttons'. Right now, I've got a volume and a tone pot slider working - "Holy dancing frequency response graphs Batman!" I think the ideal reference for this type of application is Duncans Tone Stack Calculator, although I'm a long way from being able to create such a bomb-proof app. On the other hand, if I keep this spreadsheet all open-book, others can adapt it. For the controls I'm also thinking of something to quickly demonstrate the difference of parallel and series pup combinations (which can be done by doubling or halving appropriate component values), also controls to reset everything to either a generic stock Fender or Gibson arrangement. In the meantime, in view of the shakeyness of my grasp on the underlying theory of combining impedances, some QA is required. The following two files are screen shots of the spreadsheet, with identical circuits analysed using 5Spice. I've tried to set up arrangements where all of the components have an effect. Its showing good agreement within about 0.01dB on all the points I've looked at. I've selected the peak values in the images, and they are listed in the spreadsheet and 5Spice graph. people.smartchat.net.au/~l_jhewitt/circuits/GF1_0check1.gifpeople.smartchat.net.au/~l_jhewitt/circuits/GF1_0check2.gifBTW its a long weekend here and its raining - otherwise I'd obviously be outside doing active Australian activities, as we do (yeah right!) John
|
|
|
Post by JohnH on Oct 6, 2008 1:30:50 GMT -5
|
|