|
|
Post by 4real on Aug 4, 2013 16:45:13 GMT -5
I should add..... Pete, nice discussion! I like the way you think, and the way you and John come at things from different angles, then hash it out with lots of "oh hey, what about this idea?" give and take. My money is on you coming up with a trail-blazer of an amp mod that more than a few souls will want to copy. If you can, try to take sound samples before and after, using the same guitar, strings, cable, etc. Since this is an audio experience, you'll need to mic the speaker(s), so take notes on the placement of that, and if you run it through any kind of pre-amp or eq, then jot down those settings too, please. Anything to compare apples with other apples, instead of oranges.  Good luck!  Thanks Sum, much appreciated... That's the benefit of places like GN2, coming up with ideas and solutions and stimulating ideas for projects...I think this will be a cost effective and interesting project and one that might benefit many, especially with the direction many are thinking with modelling and piezos and even acoustics, and a useful addition...we'll have to see how it all works out, but hopeful on this one... John, will ahve to take that on board and probably need a bit of a diagram as before to make sense. The caps, are they going to need to be large voltages to cope with the power and such in this application? When I get time I will do a bit of a shopping list perhaps and see if the values are appropriate before ordering as with the cost of shipping, I'm not going to be wanting to order more than once, very ahrd to just pick up 'a part' here and there. But excellent work there, and seems to be on thr right track, the extreme high end is not necessary and only going to be noise anyway I suspect so a little roll off up there is good and letting the 12" handle the low and mids should improve things I feel, especially since I've EQ'ed much of that out. On the testing I've done though, this amp is single channel so can't use the stereo thing without a mixer, mostly I did these tests with the straight mag pup that does not ahve any controls on it. I was wondering if the pickup might perform better if there was a 500k or soemthing across it to simulate there being a volume pot on it or something. This is why a mixer is important though, as these different pup systems will require separate EQing and treatment for them to sound at their best and work well together. The test recording was a bit different as it was done to separate tracks to the computer... |
|
|
|
Post by JohnH on Aug 7, 2013 15:30:29 GMT -5
Just thought Id let you know that I have been doing some spreadsheets on the train this week, to really get a handle on these tweeters and how they might interact with other components. I wanted to understand the power dissipation better, plus frequency response, db sensitivity and overall impedance of the circuit. Ill post more soon, but all the versions discussed so far still seem like good safe options, subject only to listening to them.
The maths for these things I once understood briefly when i was 19, and forgot immediately after the exam. Then ive picked it up again in recent years looking at guitar circuits, but if I dont practice it for a few months, I forget it again. So this is quite a nice problem to tackle.
|
|
|
|
Post by 4real on Aug 7, 2013 17:39:40 GMT -5
I forgot to mention the switch type. I used standard toggles for robustness. But current isn't much with the series resistor there, and would be fine for any switch. But in fact my big toggles were not so a good choice since theres not enough flowing to keep them well connected - they need a spray now and again. If I did it again, Id use standard sized slide switches, which you often see on the back of amps and cabs doing impedance switching, the contacts are self cleaning, plus they are quite cheap. 'Nuther thing occurs to me, about whether the piezos are series wired as I did, or parallel: Mine were being connected to a 16Ohm speaker, and I picked the wiring to suit a volume balance that I liked. A 16Ohm speaker gets a voltage swing of 25V if there is 40W power. For an 8 Ohm speaker at 40w, there would only be 18V swing, so my arrangement of tweeters, picking up on that voltage, would be putting out less treble, by 3db. So a parallel wiring of tweeters might be better, which could boost the output by 6db, then adjust with series resistor. Id guess that 50 to 100 ohms might be good for series resistance with a paralleled pair and an 8 ohm cab. 50 would give full range with roll off above 15kHz, 100 would roll off at 8khz. On my series arrangement, I might raise the output a bit by dropping the 270 ohms to about 150 or 180, for a higher roll-off frequency J Thanks a lot John, glad it's only eating into your train time and enjoying it. As for the rheostat...a quick look this morning I see a couple of options...on ebay for instance various like these... www.ebay.com.au/itm/50W-Watt-200-Ohm-Ceramic-Wire-Wound-Variable-Resistor-Rheostat/150961738820?rt=nc&_trksid=p2047675.m1851&_trkparms=aid%3D222002%26algo%3DSIC.FIT%26ao%3D1%26asc%3D261%26meid%3D408203958800543059%26pid%3D100005%26prg%3D1088%26rk%3D1%26rkt%3D5%26sd%3D380530365388%26  Or something like this (right), $4 delivered with a knob and perhaps a little smaller... www.ebay.com.au/itm/4-7K-ohm-5-Wire-Wound-Rotary-Potentiometer-Panel-Pot-Single-Turn-w-Knob-Cap-/400517606602?pt=AU_B_I_Electrical_Test_Equipment&hash=item5d40b5accaAfter all, some of these things are for ovens and the like, so I'd not want to put in something 'silly' big... Less the $10 delivered from asia with a range of resistances...as an example if wanting to go 'variable'. That was a 'quick search' but I think there may also be ones with built in switches, though I suspect a separate switch would be advantageous for a full off as there is likely a 'sweet spot' for the amp and the guitar and my ears too, that would be tricky to predict if one were turning them to 'off' and seeking to find that 'spot' again. Otherwise, fixed would be ok if it were right and save a bit I guess. That extreme high end is perhaps a good idea to 'roll off' as a lot wont be heard and what is might be shrill or only be noise or something, or perhaps I dont know LOL. As for placement of the tweeters in the cab, it there any reason to put the tweeters 'above' the speaker, the easiest placement seems to be either side of the 12" speaker ie o--(0)--o ? The baffle is slightly angled back...I think they can be angled as you did but would be neater to be screwed in flush, does that matter? I've not opened the amp up for a while, but there may be a way of mounting the switches and resistors and such internally. If you look at the pic of the amp, there is an angled area under the controls with a bit of a 'grill' in the metal and an area under the tone controls or perhaps near the power on switch that could be utilized and perhaps the wiring carried out inside with the speaker leads to the circuit and then using separate leads to the tweeters, keeping the components safe and out of harms way and the inside of the open back cab clean and neat. Not sure if the rheostat is a bit big for this though....hmmm...alternatively there may well be room on the rear panel which might be better anyway...next to the line in and outs, ext speaker and such. I'm not sure how much it mattered, but just realised that last time I 'tested' the amp, it had the 15" ported closed bass box it sits on connected. I also had rolled the 'bass' knob down a bit on the amp with the eq. I used to use that box with this amp back in the 80's in bands and everything, it is very loud for 60watts and the added box moved more air and lifted the amp like a small 'stack'...but very heavy to carry and not likely to want to lug up and down stairs or anything! For a switch, was thinking of a stout power switch perhaps, similar to what you used, as the insall of drilling a simple hole would be easier than trying to cut a square slide switch hole, especially with a steel chassis... While I am 'at it' and modifying...I might rep[lace teh volume control which can have a couple of 'dead spots' in it at times, though might be tricky as it has a pull siwtch for bright in it. I probably should have pulled it when I was testing to see if it had more highs that way...doh...And, looking at it now, possibly room to wire in an extra input jack under the original. It has a hi-low inputs...could one just wire another input socket to the one in there to allow two inputs...the mag and piezo sides, for something without a mixer? The wrinkle with this is that one is a passive mag and the other an active piezo, but the guitar piezo preamp has vol, bass and treble on the side....hmm, perhaps the impedances would not allow this kind of thing simply. Plugging into the low and hi inputs, you hardly got much piezo signal it seemed through the low at the same time... If using a mixer, I might go to the line in, in mono I guess with a suitable combined lead...a mixer is a necessary thing as it is not just the signal in, butthe eq and such too...but there mihgt be some advantages to being 'able' to plug in multiple leads into the amp without it at times... Ok...keep at it john and others if they have thoughts about the options for this kind of thing... ... Meanwhile, being playing a lot more guitar lately and developing a 'sound and style' of sorts, particularly in my 'harmony' which contains a lot of 'second intervels and use of open strings and so unusual chord forms...enjoying that process... On guitar repairs, my GF is trying to re-learn guitar and brought her old cheap nylon string around...with green and broken strings. A new set of string and a slight lowering of the saddle did wonders, but would not hold in tune that well...so just installed a new set of cheap tuners ($6 delivered) and that has really done wonders and it actually plays and sounds really good. Might have to look at my classical down the track, but just shows what one might actually get out of a guitar that would be otherwise garage sale material...testing it now, it is still intune from more than a day, so that's good! |
|
|
|
Post by 4real on Aug 8, 2013 6:03:48 GMT -5
Ah..... I was going through soem photos and found this one on photobucket of my Roland on top of it's bass box by the front door....  |
|
|
|
Post by JohnH on Aug 8, 2013 15:32:18 GMT -5
I can see from that photo why you may be needing to move out. id be worried about that house until the foundations are releveled. Meantime, its probably a good idea to put a chair in front of the amp. On the piezo design, if you want a pot the sort of values that are useful are looking useful are in th e100 to 500 Ohm range. Herse a deal where you get 2 for $6, inc postage: www.ebay.com.au/itm/2-Pcs-WXD3-13-2W-100-ohm-Multi-Turn-Wire-Wound-Potentiometers-Pots-/330839262057?pt=AU_B_I_Electrical_Test_Equipment&hash=item4d078e8369searching ebay with the following 2W 100 ohm wire wound potentiometer brings up several versions of this in the industrial section. power rating is 2W, and whats important is how the power is distributed to the track so when the wiper is near one end, the dissiaption is concetrated in the active section. ie, if only 10% of the track is active, theres only 10% of the rating. But the numbers seem to work ok, with a couple of fixed resistors and caps. more later. J |
|
|
|
Post by 4real on Aug 8, 2013 15:48:06 GMT -5
Cool find john, I'd been avoiding those 'multiturn' pots I guess but certainly cheap enough if the supply some fine tuning to the project and dont burn out...looking forward to working this out and giving it a go...
Yes, well I've been trapped in here for a couple of weeks as the car hit a roo and took out the radiator and there's not a shop for miles...plus gale force winds straight from the Antarctic with rain and the house is like a tree house and shaking...brrr...
Still, getting some guitar playing in and learning new stuff too...
The 'moving' is really going to be a pain I can tell you, have been given plenty of notice and going through 'stuff' around the house and preparing to 'look' for a new place to live, hopefully close, but it's not a good feeling to have to move...looking for all kinds of distractions!
|
|
|
|
Post by JohnH on Aug 10, 2013 1:02:21 GMT -5
Heres a few different ways to wire up two piezo tweeters. These arrangements would go across the connections to the main speaker, which is assumed to be an 8 Ohm speaker running up to 60W.  Types A and B use a 100 Ohm wire wound pot to give some level control, and also a series cap to keep most of the low frequencies away from the piezo. A or B could also be made with all fixed resistors. Doubling the cap adds about a db to the output but with less rejection of low frequencies. C is simple parallel wiring with a resistor, and it will be the loudest. Varying the R value changes the drop off of upper frequencies, but no reduction is low frquency signals (which as we know, is not really a big problem with piezos but may be a nicety, to compare with A or B) D is simple series wiring, and it will be basicly 6db quieter than C, then again with treble roll off controlled by resistor value The db values are relative to two piezos in parallel across the output with no other parts (0db), so C with the lowest series resistor is closest to this. plots for A and C at max: Plot A  Plot c  Based on those and a packet of parts, I reckon youd be able to experimenst and find something that sounds good. John |
|
|
|
Post by 4real on Aug 10, 2013 17:19:54 GMT -5
Thanks john, A and B look like to be the most flexible and so when have time will put a parts list together and order enough to do some experiments with those values...I've seen some 'single turn' rheostats that come in 3x and perhaps will go for something like that with the anticipation of perhaps repeating this project with the fender or making an extension 'tweeter box' rather than to modify the amps some day.
The loss of a dB here or thare is of little concern as the roland at least is plenty loud enough to compensate, the fender too (at 40w) is loud enough to create a lot of feedback with an acoustic guitar if driven at volume...looking for the quality over the quantity of dBs as you can imagine.
I will be wanting to add a switch so that should make it possible to do some kind of recording with and without perhaps...if the moving does not interfere too much with proceedings...
Thank you for all this hard work...it wont be wasted!
|
|
|
|
Post by JohnH on Aug 10, 2013 21:58:51 GMT -5
Most welcome. I also intend to use this work to adjust my system, since it has helped me to understand more about whats going on.
i know that absolute db's are not in short supply with what you might use this for, but the relative differences in each circuit are important, because with a given amount of total volume, be it high or low, as detrmined by the main speaker, the db levels shown in the tables tell you which arrangements will sound more or less prominent in the overall mix. And you wont know which is best until you hear it.
But for comparison, my current piezo wiring is series type D, with a 270 Ohm resistor. When i run this with two main speakers which are 8 Ohms net, the theory says that at 10khz, the tweeters are at -9db relative to the other schemes on the chart (all relative, nothing to do with absolute db's). This would be equivalent to your 8 ohm speaker.
My perception of this setting is that i can hear it with acoustic, its quite nice, but Id like a bit more.
When i run it with just a 16Ohm main speaker, theres 40% more voltage coming out at a given volume level and the tweeters do 3db more relative to the main, so I'm at -6db at 10khz on the tweeters, again on this comparitive basis. This is how i selected it, and it sounds good for guitar, but its not an even full range response, it has a drop off of high treble relative to a standard full range PA speaker.
But one feature that you have on the Roland is a graphic eq, though it would be nice t find a setting where you could keep this at about neutral.
hen you get around to buy parts, some spares so you can double up capacitances etc might be useful. when Ive done excercises like this before, trying to work out some theory about how to make a circuit sound good, Ive often found that, althouh the theory is very helpful, the actual component values that end up sounding best to me can be a factor of two different.
cheers
J
|
|
|
|
Post by 4real on Aug 10, 2013 23:00:37 GMT -5
Ah ok...so relative volume to the main speaker...
so, what do you think of an adjustable system with wire wound pots, too much?
I'd intend to add a 'switch' to turn it off...
So...what 'parts list' would you suggest?
I agree that some experiementation to get the best results is likely to be required and untill I've heard them and with this application, it's kind of going to be hard to tell. Some of my anxieties may well be unfounded too, the only way is to 'hear' it I suspect. The adjustability is that the tweeters may well be a bit 'shrill' or smehting, or try and deal with frequencies too low for their capacity and cause distorions or something.
The EQ is a good feature at times, I suspect that the whole pickup system may well require quite a bit of EQing too....I am even considering an additional internal mic and adapting the lead to run three channels (it is a 4 conductor cable and so could be split again to handle it neatly)...
|
|
|
|
Post by JohnH on Aug 10, 2013 23:22:20 GMT -5
I reckon the adjustable schemes could be great, worth a try at least.
For a parts list, capable of testing most options A to D, without spended too much, Id get one each of each resistor on the chart, with an extra at 22 Ohms. All 2w min, so maybe 5W if you cant get 2W. Get another 100 if you cant get 47 (eg, they arrnt at Jaycar). The caps need to be not electrolytic, 0.47uF = 470nF is easy to find, and get a couple of these so you can try two in parallel, and also 330nF and 220nF if you might want to use them to reduce response instead of 470nF. The 100Ohm variable I pointed too of 2W is the minimum power rating Id use, if you can find one with more rating its a bit more secure.
Switch should not be a toggle. Mine have become unreliable and need frequent cleaning, due to not enough going through them. Slide switches are easy enough to drill for, if you can make two small holes then elarge with a small file to end up with a rectangular hole.
|
|
|
|
Post by 4real on Aug 12, 2013 7:37:54 GMT -5
You know John, I am thinking the idea of a rotary switch (perhaps with an 'off' selection might be the way to go, switching in various resistor values to get a range of roll off and the like, to get perhaps all or some of these options or just to alter the roll offs and such might be the way to go. A simple signle countrol and safe wattage resistors and likely cheaper as I ahve been looking around and rotary controls and all that...what do you think and could devise?
|
|
|
|
Post by JohnH on Aug 12, 2013 15:22:50 GMT -5
I thinks thats a very good idea. A standard rotary would be a very suitable type of switch i believe. How about a 2 pole 6 way, to give 5 active positions plus off? There could be a max setting, then a couple each of reductions either to the full range, or by rolling off the top frequencies. Each setting could be worked out as an even step down in terms of db's, and power rating could be very securely worked out.
Cool, Ill think more about it
J
|
|
|
|
Post by 4real on Aug 12, 2013 17:43:26 GMT -5
I thinks thats a very good idea. A standard rotary would be a very suitable type of switch i believe. How about a 2 pole 6 way, to give 5 active positions plus off? There could be a max setting, then a couple each of reductions either to the full range, or by rolling off the top frequencies. Each setting could be worked out as an even step down in terms of db's, and power rating could be very securely worked out. Cool, Ill think more about it J  |
|
|
|
Post by JohnH on Aug 13, 2013 7:50:58 GMT -5
Here is a bit of background research pertinent to the design of this tweeter switching system. Given that there will be discreet increments of tweeter output, rather than a continous range, its necesseary to decide on what is a suitable increment in db's between steps. Too great a step and you could have the choice between too much and not enough treble, with no just right setting. Too small a step ansd there may not be enoug s teps to cover a desired range. Also, what realy are the frequencies that need to be taken care if in an acoustic guitar sound, and which are not relevant. This clip will drive you mad if you listen to it more than twice. I know because I have... www.soundclick.com/bands/page_songInfo.cfm?bandID=674084&songID=12454064Its the same clip of strumming on my Maton, recorded direct to pc using its piexo pickup. The first group expores increments in treble content: 1. Full range 2. Everything above 5khz reduced by 30 db - ie virtually eliminated, an extreme case of what a normal guitar speaker might sound like, with no tweeter 3. >5 kHz reduced by 12db - this was the greatest amount of reduction that still offered some noticeable treble more than case 2 4. >5khz reduced 9db 5. >5kHz reduced 6db 6. >5khz rdeduced 3 db 7. >5 khz reduced 1.5db 8. Full range This series seemed to show tha 3db is generally a good step to have, not too much as a single step, except the final one 6 to 7 to 8 (max) needed half that step or it was quite a prominent jump in treble content. Also, my impression is that the 12 db reduction(3) is probably not needed, being not very much different from the full reduction(2) After those is a pause, then the same clip again, this time testing frequency significance. It goes: 9. Full range, 10. all >15khz reduced 30 db 11. >10khz reduced 30db 12. >7 khz reduced 30db To my ears, this shows that there is nothing much of conceren > 15 khz, but some significance at >10khz and very much so >7 and >5 khz. So, probably allowing roll off > 15 khz is fine. What is your impression of all that? cheers John |
|
|
|
Post by 4real on Aug 13, 2013 14:21:15 GMT -5
That's a great little experiement...still listening and likely too early in the morning...listening with monitors and headphones... My first thoughts are 'how realistic is this really I guess. I might not be 'getting this' but this proposed control is of the tweeters which are adding in the higher frequencies, over and above those of the inbuilt 12" speaker (and possibly a bass box 15 extension) and so, although the tweeters may be rejecting some midrange, they will still be present in the main speakers. From what little I ahve read, the tweeters will struggle with the midrange if it sees too much and possibly create distortions in the tweeters and so of some advantage to filter out, but that still lets those frequencies through in the 12"...or am i drifting down the wrong path? It is 4am...hmmm The other is that the sample is pretty 'low' down the guitar neck and strummed, where as a lot of what I play these days is 'plucked' and multiple single lines and all over the neck and quite a few harmonics. This is a decent part of the kind of music I am playing these days and am looking for that high end to be able to bring that 'tinsel' out adequately. On the other hand, one does not want hiss and noise or high end squeal or 'holes' in the spectrum, the guitar is pretty much a midrange instrument...but with a lot of high end transients with the acoustic. But a valuable experiement to understand the effect of the proposed control nontheless...it does 'highlight 'the problem' that generally guitar amplifiers are pretty 'boxy' and not capable of bringing out the best of the acoustic (or a range of other things like synths and effects and perhaps even clean electric at times) and that this kind of 'mod' could well address that. One application that I've had trouble with in the past is when using drum machines which can sound really 'boxy' through guitar amps and be a reason to ahve this kind of extended frequecies...or just playing music through the things where you would ant that kind of fidelity...say if one were playing along with a backing track and blasting it all out of a single gutiar amplifier. ... So, yep you are right, it can drive one mad, so with three listens, half asleep...ok, one more time... The 'full range' seems 'best' but, a little roll off gives a bit of a smoother response and could be less 'harsh' on the ears...so advantageous to roll off a bit so in preference reverse from 8 back to 2. However, there may be reasonsthat I or someone might want to add 'just a touch' of high fidelity to an electric guitar I would imagine to provide more 'definition' say to distorted tones or something. On the extreme high end, my concern would be more the potential for noise, hiss and squeal and such... But you may well want to control that extreme high end so people don't get their ears ripped off by excessive treble or the guitar sound 'thin'. I note hat with my very basic 'eq' experiementing, I boosted above 7khz (the highest band on the graphic), boosted the treble control a notch and the bass a bit, the graphic pulled down the 1.2 and 2.5khz bands and rolled a little off the midrange control as well. #12 reall demonstrates a 'notched effect' which I don't find attractive, but I wonder how these kinds of things actually sound when combined with the normal guitar speaker... And, I just have a feeling that the kind of 'throw' of the tweeters may be quite different from that of the speaker and so, what would the effect actually sound like 'in a room' in 3D? Perhaps you have more of an impression having already a box so equipped as to how it sounds 'in person'. Regardless, I think your impressions of that may well be similar. Some control allows some flexibility and at least with a switch, I think things would be more reliable and 'safer' and cheaper to impliment and be a lot 'faster' to get to and 'set and forget' the best of those selections. There are not too many advatages to having little effect other than, one might want that to just add a little more 'sparkle' to an electric say to brighten an amp...but it needs to be enough that it makes a noticable difference. The low end is a bit more subtle as I anticipate this to be compensated for bythe higher end ofthe main speaker, not a reduction overall in those frequencies. I dare say a lot can be done with EQ's on and into the amplifier as well. This is the kind of thing I am hopeing to experiement with quite a bit, but the amp needs to be able to reproduce these ranges of sounds which is the point of this modification. Given the idea that 'multiple pickup systems' will always sound better than one, as put forward by JG and others earlier in teh thread, I hope to learn to maximise that effect and EQ the pup systems to best produce various elements of the sound. The acoustic and guitars generally ahve a big 'harmoonic burst' at the start of the notes, after that they settle down quite a bit to a relatively 'smooth' tail end of the notes. If you roll off the attack of any guitar note, say with a 'volume swell', you find the guitar's sound to be 'relatively smooth, at times quite 'flutey' or closer to a sine wave as it tails off. There is also the use of compression which I think will be necessary, but with multiple pickups, there will be the opportunity to have little compression on one system to bring out that 'dynamic burst' while allowing say a mag which has a more even tone to be more 'compressed' and create a warmer and fuller 'tail' end and hold high end frequencies a bit more. I may even add soemthing like a mic into the guitar to pickup some of the more woody elements of the guitar at yet another level and compressed quite a bit to avoid feedback problems. Combined with that strategy, the use of different levels of reverb effects on each channel to allow 'presence' as well as 'depth'... As far as the graphic on the amp, I imagine that this could be really useful, if running more 'flat' is possible, to target problem feedback frequencies maybe. ... Anyway...yep roll off above 15Khz...perhaps not the extreme hi end, but as this is somehting that others might want to do and their needs might well be quite a bit different to mine, do you think that this is advantageous, or evn if this 'notch' would be so stark considering there will be the speaker as well running. ... How many 'steps' are you considering...5 or 6...one could do more I suppose if the switch could be single throw, but that seems like a few chunky resistors to house, many of which may not be used I am guessing once the 'sweet setting' if found. Certially 'off' should be one and so eliminating the separate 'switch' and likely in increments 'up' to 'full range' with no or little attenuation. The second highest to full range probably the 'high end roll off' and the others have thios and variations of midrange 'cut'...I suspect the prefered setting will be one of these depending on how the amp sounds as a whole with the speaker in it and if there are problems with distortions or other things related to driving midrange through the tweeters, this reportedly is the difference between a quality 'natural' sound out of the tweeters and something 'harsh'...or so I am to understand...and hence the desire to have a control to be able to 'find' the appropriate balance for the whole system and application and the 'real life' room effect (if there is one?! LOL)... Thanks again for the work and sorry for the 'essay' but hope that helps... ... Been a bit 'busy' lately, finally did my 'tax' yesterday and the election moved forward so best to get that out of the way, been sorting through a lot of stuff in anticipation of moving as well. Weather has been shocking and cold and especially windy and rain...so been playing a lot more guitar as well, amazing how much better one gest with regular practice. On a rare cold but sunny day recently though, I went round tothe ocean side of the island to see the sea...    |
|
|
|
Post by JohnH on Aug 13, 2013 15:29:48 GMT -5
OK thanks. I should clarify that the tests were not really to demonstrate how a given set up will actually sound at this stage, but just to find out what size steps in the range will be most useful and what atepa they should be to give a good transition from 'all off' to max. Enough info there to work on though, and your comments were helpful.
The actual sound will still be subject to test, but hopefully with some design, the switched system can be worked out so that somewhere near the upper settings, there will be a sweet spot to be discovered.
Im now thinking there should be just one range from max to off, and each should give a good response, in proportion to the setting, up to 15Khz. Above that, there will still be output, and different settings may vary it but its not essential to the sound and could roll off. These simple RC filter only have one or two parameters however, so there are some constraints to the exact shape of the curve.
I thought about using a sound sample using more of the neck, but I have in mind that the very high frequencies are not really about what notes are actually being played. Even the top fret on high e is only about 1200hz. i think the sounds being considered here are mainly percussive transients, or very high harmonics filtered by the response of the guitar body. But if there are more such tests, maybe we could use a dry recording from your actual playing and guitar. However, think thats all for now.
cheers
John
|
|
|
|
Post by 4real on Aug 13, 2013 18:53:24 GMT -5
Yes, teh 'magic' in this is in the very subtle details of the upper transients...play a nice complex cord with open strings and stick your head near the sound hole and you will hear a lot of complexity in how all those notes 'ring' together....thats the kind of thing that is 'lacking' in the guitar amp that the gutiar produces in real life and through amplification could do more and...well, sitting behind the instrument, you don't really get to hear what it sounds like to others and it is no where near loud enough.
Further, what I hope to explore and bring out in the amplified approach is even more of this with layers of pickup sources and light compressions and reverbs. I've always loved those sounds as they react against each other and the subtle dissonaces in the upper harmonic overtones, even in basic chord structures.
And lately, I've been exploring that aspect in arranging and such, trying to have a clear bass and melody but supporting the inner voicings or contrapuntal bits in the middle with a lot of notes a semitone or tone ringing together....the end results, especially in the upper harmoncis, is this ringing beauty behind the fundamentals that makes this kind of music 'work'...it's not 'just the notes' as in the fundamentals as, withut lyrics and other instruments where the guitar is supposed to be keeping in it's own midrange 'space' and keeping out of the way of things like cymbals and the like, it needs to be able to produce those 'sounds'
If this works out with the roland and when I get my fender amp fixed, I may want to repeat it on that for it's 'warmth'...I would mess with the impedances and give the tube amp a problem? I thought I could perhaps test it by connectiong the tweeters and control from this amp to the fender and see how it sounds. Another interesting 'idea' is to make a separate box, perhaps even incorporate a 'tweeter box' as a part of an amp stand like yours john and so avoid modifying the amp itself other than adding a socket or something from the speaker itself...perhaps even work out a way to include the mixer into it some how as well...perhaps some kind of 'drawer'...if I recall the amp stand maker of yous offered 'rack mounting' effects capability...lots of ideas come to mind.
Main thing I am aiming for is that high end 'clarity' and that is somewhat lost with one speaker trying to produce all those sounds and so 'distorted' or just 'lost'...I think we are on the way to doing that...
|
|
|
|
Post by JohnH on Aug 16, 2013 21:39:31 GMT -5
This is what Ive worked out for a rotary controller of two horn piezo tweeters:  There are 5 steps of volume, plus off, and the piezo response is concenrated in the region 5 to 15khz. They are protected from lower frequencies. Overall, the maximum output is about 2.5db down from what would result from just wiring them up directly with no attempt at cross over. My hunch is that this will be just about right, at settings 1 or 2 or 3, but needs testing of course. This is to go onto an 8Ohm speaker system. With 16 ohms, the tweeters will sound relatively 3db louder, so the tweeters might be turned down another notch. I was quite careful to judge the power dissipation on each component, basing the assessment as if the amp was running flat-out with equal power in each octave, and then added a factor of two. Also, the overall impedance of the speakers is not affected significantly by this circuit. It only adds load to the amp at high frequencies, where normal speakers have a higher impedance. There is always a resisatnce between the piezos and the amp. I'm reasonably confident that this can be connected to any amp, SS or tube, with no change in settings. ie, Id be happy to put this on my amps, but others should do so at their own risk! Here's the frequency response for the tweeters at each setting:  |
|
|
|
Post by 4real on Aug 17, 2013 3:00:54 GMT -5
Thanks John...will order parts in the near future and give it a go when I get a day off...
I really appreciate this and I think it gives great flexibility and a neat single control to the switching...
A quick look at the caps a different day, do they need to be able to take a particular wattage? Some I saw tyoical in crossovers were a bit expensive, like $8 each?
I'll let you know if I am having trouble locating parts from jaycar most likely. I might even spring for a slightly dearer horn, similar in type but are perhaps a little less 'directional' that I saw in their catalgue at $15 each...
|
|
|
|
Post by JohnH on Aug 17, 2013 7:19:14 GMT -5
The caps are quite low values compared to normal tweeters, so they are standard types. Jaycar does the green caps and the yellow MKT ones. Both of these have a 100V rating. Its voltage rather than power that needs to be rated in caps, since a capacitor does not actually dissipate any power. These basic versions are fine, since the highest voltages around the speakers are less than half of the 100V ratings. Normal dynamic tweeters need larger cap values, so getting these with the voltage ratings may get expensive. I showed two caps in parallel, since they may be easier to find, but it could be a single 1000nF cap. But also, you might like to try it with just one 470nF cap. These piezos seem to possibly have a bit of a peak around 5khz and that might ease it down a bit by raising the low end roll off frequency - needs experimenting with a careful ear.
The resistors are all available in the 5W wire wound range. jaycar dont have a 2W series, but there are 1W carbon film types available. Two 1W's of double the value in parallel would work well. eg, two 47 Ohm in parallel, to make 23.5 Ohm, would sub for a 22 Ohm. Two 68's will do for the 33, two 27s in parallel for the 15.
The design has been based around the best guess of the capacitance of those particular tweeters. So if you want to get a difefrent type, could you send a link? some have the same driver unit with a larger horn, but others have them with a high capacitance value which would change the response.
J
|
|
|
|
Post by 4real on Aug 17, 2013 16:50:01 GMT -5
The caps are quite low values compared to normal tweeters, so they are standard types. Jaycar does the green caps and the yellow MKT ones. Both of these have a 100V rating. Its voltage rather than power that needs to be rated in caps, since a capacitor does not actually dissipate any power. These basic versions are fine, since the highest voltages around the speakers are less than half of the 100V ratings. Normal dynamic tweeters need larger cap values, so getting these with the voltage ratings may get expensive. I showed two caps in parallel, since they may be easier to find, but it could be a single 1000nF cap. But also, you might like to try it with just one 470nF cap. These piezos seem to possibly have a bit of a peak around 5khz and that might ease it down a bit by raising the low end roll off frequency - needs experimenting with a careful ear. The resistors are all available in the 5W wire wound range. jaycar dont have a 2W series, but there are 1W carbon film types available. Two 1W's of double the value in parallel would work well. eg, two 47 Ohm in parallel, to make 23.5 Ohm, would sub for a 22 Ohm. Two 68's will do for the 33, two 27s in parallel for the 15. The design has been based around the best guess of the capacitance of those particular tweeters. So if you want to get a difefrent type, could you send a link? some have the same driver unit with a larger horn, but others have them with a high capacitance value which would change the response. J Ok....I saw the alternative model in a 2007 printed catalogue but don't seem to be offering them any more anyway so no matter. They appeared the same but with different shaped 'flute' for wider dispersion...they appear to have the smae specs and driver...but the cheaper ones will do as want only to buy in one go all the parts. Tweeters... www.jaycar.com.au/productView.asp?ID=CT1930So...for the switch... www.jaycar.com.au/productView.asp?ID=SR1212two pole, 6 position rotary Resistors... 5 watt 56 ohm wire wound www.jaycar.com.au/productView.asp?ID=RR32684x 22ohm... www.jaycar.com.au/productView.asp?ID=RR32581x 15 Ohm... www.jaycar.com.au/productView.asp?ID=RR32541x33 Ohm www.jaycar.com.au/productView.asp?ID=RR3262Suitable speaker wire... Capacitors... 2x 470nF www.jaycar.com.au/productView.asp?ID=RG5165Sorry, did not see the voltage capacity in the printed catelogue for some reason, sot that's fine... Am thinking of mounting all the components on a board to keep the connections safe and mounting it all perhaps on an aluminium plate inside the back...or perhaps inside the amp itself...testing first and a bit of an explore inside will decide. Would a 'switch' to cut off one of the caps be a worth while addition or the difference going to be very subtle? Anyway, seems like an affordable and simple shopping list, the wire wound resistors will be bigger, but less hassle than doubling up the carbons, but thanks forthe suggestion there... |
|
|
|
Post by JohnH on Aug 17, 2013 19:06:20 GMT -5
That seems like the right shopping list.
Im still thinking that you will start by lashing it up without the switch, just to check that the max level and basic tone seems feasible. Thats when you could try the one or two caps, then I think there is likley to be a choice that you can make that you dont need to switch for.
Actually, there is another slightly better way to test raising the low cut-off frequency, and that is by halving the value of the 56 ohm resistor, eg by getting two of them and running them in parallel for 28 Ohms. This would be expected to raise the lower -3db frequency from 3khz to 4.5khz, but without much loss of max level. Its still ok for overall impedance seen by the amp. So maybe get another 56 just for grins?
I think I remember seeing those other horn tweeters when I was shopping for these in 2009, and they did look like they had the same driver, but not for sale now.
J
|
|
|
|
Post by gumbo on Aug 17, 2013 20:32:13 GMT -5
4r.. ..could possibly do some interesting things with a couple of these ...  (eBay#) 320513147596 |
|
|
|
Post by 4real on Aug 17, 2013 23:03:06 GMT -5
thanks gumbo... www.ebay.com.au/itm/DEFOND-DP4T-SLIDE-SWITCHES-8A-250VAC-2-PCS-/3205131475964 pos slide switches. I have a 'nibbler' so I could make relatively neat holes to mount them, but not as neat as a single hole for a rotary control as with this scheme...so 'off' and 5 variations in a single knob...must check I ahve a suitable control knob, think I'll have a few that will suit. Thanks heaps john, few things on but will get around to ordering soon. I suspect in time I might make this project twice. Once built into this amp which will hoepfully make it more versitile and perhaps again in it's own box and mount a socket from the speaker in the fender rather than modify itshould I wish to add tweeters, perhaps even made into a 'stand' for that amp in some way... |
|
|
|
Post by gumbo on Aug 18, 2013 0:54:36 GMT -5
Yeah.. ...I was thinking 4-times-4... which (with a bit of forethought) could give you OFF + 12 combinations.. ...I figured that would keep you busy while it's raining...  ..anyway it's raining here now, so that's probably why I'VE got too much time on my hands as well.....  |
|
|
|
Post by 4real on Aug 18, 2013 20:21:36 GMT -5
Yeah.. ...I was thinking 4-times-4... which (with a bit of forethought) could give you OFF + 12 combinations.. ...I figured that would keep you busy while it's raining... ..anyway it's raining here now, so that's probably why I'VE got too much time on my hands as well..... I see what you are getting at and they are an interesting find for something else I dare say. For this I suspect there will be an setting that is just right or small adjustments based on a room or something, but not necessarilysomething one needs infinite control or a lot of 'settings' to make it better or one will need to adjust. So, simple is likely best, with the flexibility enough to find that 'sweet spot'. I feel the same way about guitars these days, the guitar is really going to have a few nice sounds and it's great to have perhaps a few extra settings, but there is a point in which more options just takes away from the functionality. Of the things brainstormed, I think the rotary option and fixed resistors is the best solution. It might be something to cosider if one were to make this project into a separate 'tweeter box' and consider using with different amps though, or if one were to add a crossover and a midrange horn as well, who knows. I might even consider something like that down the track some day if this works out well. ... And the weather...tell me about it...it's been very extreme lately...rain hail thunder but the worst has been this constant chilling high winds...i's been blowing a gale for some days now day and night and is expected to continue for the forseeable forcast...and living high on aa clifftop by the sea...the whole house is shaking and making a racket and I've been getting cabin fever, especially combined with not having a car for a few weeks until recently...brrr |
|
|
|
Post by 4real on Aug 19, 2013 4:25:35 GMT -5
Ok...placed order...$29.35 + $7 post AU, if people are interested, all parts available at Jaycar as above links...seems like a very cost effective project...will work out installation options when parts arrive and the weather clears enough to cut a few holes...thanks John...
|
|
|
|
Post by 4real on Aug 20, 2013 0:04:59 GMT -5
Leading in the mixer is this one... www.zzounds.com/item--BEHX1204USB The faders, mute and programable effects being large factors... A$255 at 'soundcorp' at the moment... |
|
|
|
Post by JohnH on Aug 20, 2013 15:37:29 GMT -5
That mixer looks like a lot of fun. I cant think of anything sensible that you couldnt do with it.
The picture is full size on my screen, so it makes me want to reach out and twiddle with it.
And if your signal is sent via this mixer on its way to the new piezo thingo, then those high 12khz treble pots on each channel might be a very good place to tweak the piezo sound.
|
|
