|
|
Post by ChrisK on Aug 1, 2006 17:35:52 GMT -5
??
I know it's discontinued, but....
I've got a chance to score a new one for $769 (their normal price is $950).
And a '65 Deluxe Reverb for $683.
|
|
|
|
Post by sumgai on Aug 2, 2006 1:40:05 GMT -5
Chris, Snatch the Deluxe Reverb! Don't ask questions, just do it. (Unless an in-store trial shows it to have real problems.) I've always felt that the Pro Reverb (50 watts, 2x12") is just a weak Twin Reverb. It uses the same speakers as the TR, the chassis is pretty much the same as the Super Reverb (but no mid on the Normal channel), the cabinet is almost as large (narrower due to one less control), and it weighs in at only a few pounds less than the TR - 59 versus 69 lbs. If I've gotta hump that kinda freight, I better be able to make some ears bleed! ;D There may be a tone consideration to which I don't usually pay any attention. The lower output, into the same speakers, may cause an earlier breakup, yielding a more pleasing tone at a lower volume. But that's a very small difference, IMHO..... still, beauty is in the ear of the beholder. If you like how it sounds, you can always recoup your investment, should you ever feel the need.  Just my 2 yuan worth.  sumgai |
|
|
|
Post by ChrisK on Aug 2, 2006 12:06:13 GMT -5
Thanks, sumgai.
The Pro Reverb actually has 1x12". The 1/4 power switch changes the B+ voltage doubler to a straight rectifier and reconfigures the bias (50/12 Watts).
The Pro Twin is what you're thinking of perchance (2x12", 100/25 Watts) and is of no interest to me.
I've become a big fan of the low power, full distortion amps. I have the VOX AD50VT (w/ a Celestion), a Mark IV that will do 2x6L6 in triode mode, and the Epi Valve Junior.
Fender has developed the Princeton Recording amp that has a transconductance stage to allow full plate distortion at variable power (a la' the AD50VT). U.S. Patent 6,816,009. uspto.gov.
If'n ya gots an interest, I can email the pdf to you (all uspto files are stored as tiffs and are quite large) since I poured it as such.
The '65 Deluxe Reverb just seems too simple for the money. Four input jacks, yikes! I'd have to modify the first channel for an additional gain stage or something.
|
|
|
|
Post by sumgai on Aug 2, 2006 15:43:12 GMT -5
Chris, Somehow I had the impression that when you said "Pro Reverb", you meant as in a very close facsimile of the original, which was as I described it. What you're referring to now, your 'refined' description, is much different, obviously. Sounds to me like if it's light enough, yet can keep up with the demands of a small to medium venue (and of course, in a recording studio), then the purchase of the thing should depend only on the tone you get.... everything else is compromise-able. OTOH, the '65 Deluxe is a re-issue, a good knock-off of the original. Those four inputs must be present for the "classic" look and feel. Of course they are useless, but so what? They didn't add significantly to the overall cost, so they're a non-worry, IMHO. As for adding an overdrive or something, just do as I did when I had one of these back in '63 (d@mn, why wasn't I born with clairvoyance, and the ability to predict the future?  ?)..... just hook the output of V1b to the input of V2a. That was too strong for me, so I added a 82KΩ current limiting resistor in front of V2a's grid, and that sounded pretty darn good, if I do say so myself. (I also removed the 6.8kΩ resistor from the bottom of each tone stack, essentially emulating the Mid control being set to 1. Sounded even better to my ears, back then.) Value for return on investment in terms of how sweet an amp sounds is something I'm not qualified to estimate. Hell, I even have trouble making up my mind over my own gear purchases!  ;D sumgai |
|
|
|
Post by UnklMickey on Aug 2, 2006 18:45:52 GMT -5
... Those four inputs must be present for the "classic" look and feel. Of course they are useless, but so what? ... hey Sumgai, let me run this up the flagpole and see if anyone salutes it. extraneous input jacks on a Deluxe Reverb........hmmm....... what can we do with that? what if one of the inputs on the "normal" channel is converted to an output? the output from v1b is available at the tip, the switched portion of the jack carries the signal through the normal path, unless a cable is plugged in. (the reason for this will become apparent later!) so, for your built-in overdrive circuit, you plug the guitar into the unaltered input of the normal channel. the output of the normal channel goes through a patch cable to one input on the "vibrato" channel (that has been modified to taste for knocking the signal down to a managable level.) so far it's just like your built-in OD, but.... if you make a special A-B switch, you can stomp between OD, and straight into the other (unaltered) input of the "vibrato" channel. (the special part of this A-B switch ....... it needs to short whichever amp input isn't being used at the time) now here's the kicker.... unplug all that stuff and plug directly into the unaltered input of the vibrato channel. get a cable with a phono plug on one end and a 1/4" phone plug on the other end. plug the phono end into the output of the reverb pan. the phone end plugs into the unaltered input of the normal channel. so now we have a Deluxe Reverb, with "reverb deluxe". www.schematicheaven.com/mods/dualchreverbmod.htmwhatcha think? |
|
|
|
Post by sumgai on Aug 3, 2006 14:38:21 GMT -5
unk, I liked your first mod, the part about using an A/B box to get both OD and straight sound from the two channels, even though I can certainly do that without converting any jack to "output". Where does that come into play? (I know, you said 'it will become apparent later', but that "later" never came for me.) However, the second part, about the reverb thingie...... that left me scratching my head. So I went to the site, and looked around, and found this: Now your reverb knob is a "dwell" control. And the Volume knob on the Normal channel controlls the overall reverb volume. The new EQ adjustments you now have, give you control over your reverb much like that of the Fender Reverb Unit. Oh really? Come again, please? ( Edit: changed V4 stage nomenclature to reflect updated schematic, per ChrisK's post below.) Weeeeelllllllll. Let me see if I got this straight...... if you reroute the reverb pan's output away from V4a (the Reverb recovery stage), don't you also remove the Reverb control from the picture? So how is it that said control is all of a sudden capable of changing the Dwell? (Particularly when it was more of an "Amount" or "Intensity" type of control in the first place.) More to the point, when you run the pan's signal into the Normal channel, the signal will come to the junction where the two channels are combined (the summing point, if you will), and simple phase cancellation will cause a large deficit in one's tone and volume! Much as I hate to pee in schematicheaven's Cheerios, they are way off the mark on this one. I'd recommend it only to those who like tearing their hair out by the roots! Neat idea, unk, but it's back to the CAD screen! Now, as to my favorite mod for quick and dirty OD. If you're not gonna use the Reverb anyways........ pull the two cables out and tuck them under the pan on the cab's floor. Hook a resistor to two RCA plugs, and insert it into the two RCA jacks (Reverb In and Out). Now you've got all the OD you could ever want, complete with footswitch control for On/Off, and an Intensity control on the front panel. Start with 220KΩ, and go up for softer bluesy tones, or lower to get metal-like icepicks in your ears. (No, this isn't my original idea. It's all over the 'net, and it's anyone's guess who was the first to do this. I recommend that one ignores any claim of being "the original", it very likely predates all those who sell mod kits today.  )) sumgai |
|
|
|
Post by UnklMickey on Aug 3, 2006 16:04:41 GMT -5
Sumgai, i agree about the "dwell control". (what were they thinking?) but don't throw the baby out with the bath water. having the treble and bass controls for the reverb return are what seem to be the prize here. (although i don't know if that would be of any interest to you and the SG101 crowd.  ) i didn't follow up on it explicitly, but the "later" part was, since the connection from v1b's output is switched by the jack, it will allow for the normal channel to be used for the reverb return.so depending on your mood, you can configure for OD with standard reverb available. or by a different cabling setup, normal amplification with "reverb deluxe" (c'mon admit it. even if you don't see a need for t&b on the reverb, "Deluxe Reberb with reverb deluxe" did get a chuckle out of you.) EDIT:i think i recall doing that reverb overdrive back-in-the-day. a friend had one kernel of info on it and wanted me to do it to his amp. (he wanted it hard-wired) i used a pot, and wired it through the jacks. if i remember correctly, it actually had a "dead-spot" as you turned up the reverb control. then it started to get louder and distort, but wide open, it would oscillate. i had to switch the phase on his reverb transformer. only did the one, so i don't know if it was a wiring error, or if they were all like that. didn't seem to make any difference one way or the other, with the reverb pan connected for normal operation. then again, i was 14 at the time, so i don't remember all the details. unk |
|
|
|
Post by ChrisK on Aug 3, 2006 16:31:12 GMT -5
I do sort of see the "dwell" control (if it's actually dwell). V4A would have no input signal from the reverb pan, but since the grid is set to ground potential via a 220K resistor, the plate will present some complex impedance to the reverb level control via the coupling capacitor. This is then coupled to the channel summing node and will have "some" effect on the Vibrato (Note to Leo: !!!!Tremolo!!!!) channel.
The real issue that I see is that the schematic specified output level from the reverb pan is 1.4mVAC. The specified input signal level to the Normal channel is 37mVAC.
Also, each channel has a low and high signal level input jack (jacks ain't wired the same) that realizes the signal or the signal divided by 2.
Nowwwwwwwwwwwwww,
One could take the Normal channel and switch its output to one of the inputs on the Vibrato channel. A three-way toggle might give Normal, Vibrato, and Normal gain boost from Hades.
Let's see, just one input jack into the Normal channel. A three-way toggle in place of the other Normal channel input jack that either directs the signal thru the Normal channel, directs the signal thru the Vibrato channel, or decouples the Normal channel output from the summing node and redirects it into the Vibrato channel.
Hmmm, now there's two unused front panel jack holes. One might be a mid control for the Vibrato channel, and the other.....
Gee, I wonder what happens if one capacitively couples the Reverb footswitch disable point into the VibroTremolo disable point (and removes the 4.7uF cap).
Holy Toob Workbench Batman!
|
|
|
|
Post by UnklMickey on Aug 3, 2006 20:32:23 GMT -5
Chris,
i don't see a related 4.7uF cap on the old DR diagrams.
do you have a diagram for the reissue '65 DR, or am i missing something?
unk
|
|
|
|
Post by sumgai on Aug 4, 2006 0:44:34 GMT -5
Chris, Err, did you forget that V2b is feeding a healthy signal into the grid of V4b? That will override any static consideration of the 220KΩ resistor going to ground. After all, if there isn't any signal available from V2b, then there sure won't be any signal from V1 with reverb on it, now will there? ( Edit: changed V4 stage nomenclature to reflect updated schematic, per ChrisK's post below.) No problem. Look at V4a, and you'll see that the 12AX7 is wired to give about four times the gain as V1a or V2a. The sensitivity rating of which you speak is for full output - 37mVAC at the input jack will yield full rated output at the speaker(s). The reverb recovery stage has no such requirement. In fact, it is further burdoned with a 470KΩ resistor in series with the signal on its way to V4b. That's to compensate for the extra gain. Wha? Wazzat mean? What's yer aim here, pardner? (And like unk, I wanna know where you're finding that 4.7µf cap!) sumgai |
|
|
|
Post by sumgai on Aug 4, 2006 0:57:02 GMT -5
unk, Well yes, controlling the tone of such a large part of the signal is nice, I'm not gonna disagree with that one. But I still wanna know what's gonna happen when the signal hits the summing node (thanks, Chris). I say it will disintegrate into little tiny bits of BLAH!, and fart out due to phase cancellation. There are only two stages of amplification in the Normal channel, and there are three stages in the Vibrato channel. Since each stage reverses the signal phase, guess what? Can you say oops, boys and girls? Ah, that makes more sense. I had to think on it for several moments, but I finally got it.  As a rule of thumb, if a circuit oscillates only at higher gain settings, then it is not an issue of phase. But then again, no matter what one's age, if the problem is fixed (albeit with a kludge), then who's to know, right? ;D ( Edit: changed V4 stage nomenclature to reflect updated schematic, per ChrisK's post below.) My favorite method of increasing the apparent amount of reverb on these amps is to simply swap the 470KΩ and the 220KΩ resistors found in the plate circuit of V4a (and leading to the grid of V4b). This will pretty much double the voltage from one stage to the next. It won't sound any better, tone-wise, but when you get to 10, it sure sounds like 11! ;D sumgai |
|
|
|
Post by UnklMickey on Aug 4, 2006 1:40:33 GMT -5
....But I still wanna know what's gonna happen when the signal hits the summing node (thanks, Chris). I say it will disintegrate into little tiny bits of BLAH!, and fart out due to phase cancellation. There are only two stages of amplification in the Normal channel, and there are three stages in the Vibrato channel. Since each stage reverses the signal phase, guess what? Can you say oops, boys and girls? ... well, maybe i'm looking at this wrong (it's happened before, it'll happen again). but it looks like the output of the reverb pan originally went through 2 stages of gain (the reverb 1/2 of the 7025, then the third stage of the vibrato channel) before going to the P.I. with the "reverb deluxe" scheme it goes through 2 stages (the normal channel) then directly to the P.I. so there won't be a phase change. even if there was......... would the phase relationship of the dry signal, to the signal that's gone through the mechanical springs be so convoluted that it would be hardly recognizable anyway? regarding Chris's latest thought, i wonder if he's looking to "tremolize" the reverb return. seems more like the impedance of the the spring units output will be to low to allow much affect. OTOH, maybe he's thinking that the reverb signal will affect the tremolo oscillator. 'nuff speculation on that one. i'll just STFU and wait for him to bring us up to speed. unk |
|
|
|
Post by sumgai on Aug 4, 2006 10:30:56 GMT -5
unk,
(Edit: changed V4 stage nomenclature to reflect updated schematic, per ChrisK's post below.)
As I perceive it, there's an easy way to determine phase relationships (in this circuit). Since the original routing of the Reverb signal was sent to V4b in phase with V2b (the Vibrato channel's dry signal), then we need only compare it to whatever happens to the signal as it passes through the Normal channel. Under the new scenario, we are inverting the Reverb signal twice more (V1a and V1b), thus leaving it still in phase with V2b, right?
Well, V2b's dry signal is going to be reversed one more time (by V4b), but the Reverb signal is now not going to be reversed along with it. And therein lays the rub. And yes, the decaying signal will be interacting with the dry signal at 180°, thus giving us a very muddy, and probably thin, sound as a result.
All of this is easily evidenced without doing any mods to the amp. Simply run a "Y" cable into both channels (or a jumper from one channel to the other), and watch what happens. At some point in the mix, both volume controls will send approximately the same voltage to the summing node, and cancellation occurs. However, in our 'mod' scenario, we are sending the same signal (Reverb) to the Normal channel all the time (the input doesn't vary, except as a function of the Vibrato channel's volume control). While in the strictest sense the overall strength of the Reverb signal itself may not be cancelled out completely, it will be affected tone-wise, due to the harmonic interaction of the reverse-phased signals, if nothing else. That's all I'm saying here.
Hey, look at that clock - it's Miller time! ;D
sumgai
|
|
|
|
Post by ChrisK on Aug 4, 2006 10:45:35 GMT -5
www.fender.com/support/amp_schematics/pdfs/65_Deluxe_Reverb_Schematic.pdfI think that we're all looking at different schematics. Well maybe, traumatize the reverb. The reverb recovery stage (V4A) is driving the reverb level pot thru a 0.0033uF cap. This point is shorted to ground to disable. The Tremolo oscillator (V5A) is driving its grid thru 0.01uF caps. The node at C19 (4.7uF) is used to disable the tremolo thru a 10K shorted to ground. Because of the loading of this cap, removing it and connecting the reverb disable point to the tremolo disable point (with a blocking cap) seems like something stupid that I'd try just 'cuz I could. I'm not at all sure what will (or won't) happen. Like Lego blocks, they seem..... ...oh, shiny!
|
|
|
|
Post by sumgai on Aug 4, 2006 13:18:52 GMT -5
Chris, We sure are! I hadn't seen that particular one before. I note that it is almost identical in parts to the original AB763, with the majority of actual design changes taking place in the oscillator and power supply stages. Otherwise, it adds labels and voltages, which Fender did in the early years, but stopped doing when the circuit naming conventions were changed, and the schematics were redrawn with the new nomenclature (circa 1961). This new diagram is more "busy", but it adds some necessary troubleshooting info, which I can't help but like. Thanks for that particular tip. I've edited my earlier posts, reversing the V4a and b callouts, to reflect this updated information. sumgai |
|
|
|
Post by UnklMickey on Aug 4, 2006 17:04:39 GMT -5
...I think that we're all looking at different schematics... oh yeah! the AA763 and AB763 prints show a much different configuration for the reverb and tremolo switches. ...As I perceive it, there's an easy way to determine phase relationships (in this circuit). Since the original routing of the Reverb signal was sent to V4a in phase with V2b (the Vibrato channel's dry signal), then we need only compare it to whatever happens to the signal as it passes through the Normal channel. Under the new scenario, we are inverting the Reverb signal twice more (V1a and V1b), thus leaving it still in phase with V2b, right?
Well, V2b's dry signal is going to be reversed one more time (by V4a), but the Reverb signal is now not going to be reversed along with it. And therein lays the rub. And yes, the decaying signal will be interacting with the dry signal at 180°, thus giving us a very muddy, and probably thin, sound as a result.... Sumgai, i think you might be dropping a sine, when moving your point of reference. or losing track of your summing points. okay enough puns, let's track this down. to be sure we're using the same terminology: signal path for normal channel -- V1a, V1b, phase inverter. signal path for vibrato channel -- V2a, V2b, V4b, phase inverter. signal path for reverb signal -- V2a, V2b, V3a+b, reverb transformer, reverb pan, V4a, V4b, phase inverter. we'll track our phase polarities on the INPUT of each tube. we will always set the vibrato input as our reference point + we will call normal phase + , and inverted - so lets track the example of normal and vibrato through the input jacks to the input of the phase inverted. just as you stated if both inputs are +, the normal will be + at the input of the P.I. (it went through 2 inversions.) the vibrato will be - at the input of the P.I. (it went through 3 inversions.) since the two signals now have opposite signs as they are summed at the P.I, they will cancel. ********************************************************** conclusion #1: for the two signals to be in-phase at the input of the P.I.,
if a signal at vibrato input is +, the signal at the normal input must be -. ********************************************************** now lets track the vibrato and reverb signals. they both start the same at the vibrato input .....+ the dry signal reaching the input of V4b is +the input of V3a+b is the same.... + at the input of the transformer, the reverb signal is -. but what happens in the transformer, and the reverb pan? it's hard to track at that point, so let's instead determine what the result MUST be! remember we said the input of the V4b is +for the reverb signal to be in-phase with the vibrato channel at the input of V4b, the input of V4a MUST be - . so regardless of what happens with the reverb transformer and reverb pan, we KNOW.................................................. ********************************************************** conclusion #2 -- the OUTPUT of the reverb pan MUST be -********************************************************** since both the vibrato signal and the reverb signal are inverted by V4b........... ********************************************************** conclusion #3 the reverb signal must be - when it reaches the input of the P.I., just like the vibrato signal********************************************************** i know this seems like the long way 'round. but, here comes the payoff. read conclusions #2 and #3. now read conclusion #1. don't concern yourself with the relationship of the reverb signal to the signal at the input of V4b when routing the pan output through the normal channel. they don't sum there in this case. what is the phase relationship of the 2 signals at the P.I. (the first place they re-unite)? okay, workday is done, time to take off my tie. (so make mine, a mai-tai) unk |
|
|
|
Post by sumgai on Aug 4, 2006 19:01:38 GMT -5
unk, Nice analysis, I almost did the same thing, but it wouldn't have been as pretty as yours. EDIT This post is now null and void, and has been superceded by Reply #18. Please ignore all that has been struck through. I can't simply delete it, because if I did, then unklmickey's following response would look mighty ugly. 
I'm not referring to the ordinary summing node of the two channels at the grid of V6, I'm concerned with the "mini" summing node found at the grid of V4b, the dry and reverb signals. Luckily for our discussion, this node does not move due to our "modification".
If we accept that your Conclusion # 3 is correct, and if there is effectively no inversion across V1 (a and b), then indeed, the signal found at the summing node in question will be composed of a dry signal of +, and a reverb signal of -. Look at it like this: (As in your example, we're looking at the grid of each stage.)
Ordinary, unmodified version:
V2a .. +
V2b .. -
V3 .... +
V4a .. -
V4b .. +
As before, V4b is seeing V2b's output as +, and V4a's output as +. That's a good thing.
Modified version, as per instructions on schematicheaven:
V2a .. +
V2b .. -
V3 .... +
V1a .. -
V1b .. +
V4b .. - and + !!
Is this clear? V4b sees a + coming from the plate of V2b (dry), and a - coming from the plate of V1b (reverb). This is very easy to see on the diagram - we have taken away one inverting stage (V4a), and put in two inverting stages (V1a and b). All of our tone and volume problems will stem from this little factoid.
(As you noted, the small reverb transformer and the pan both invert the signal, so they're a wash, and needn't concern us any further.) BTW, I have given this some more thought. I have now come to the conclusion that this exercise is all fine and dandy, but just how much will this add to one's enjoyment of reverb? I think that simply amplifying the already reverb'ed signal to an even greater degree is not necessarily a good thing. Better would be to drive the pan itself a little bit harder, as would be the case in a true Dwell control, like that found in the 6G15 circuit of yore. <diatribe> There are many who simply accept reverb as it is delivered, and don't go any further. That works in many, if not most cases, but for me (and I suspect, for some other surf guitar players), you can have too much reverb if all it does is run on and on and on..... What we really want is what I call "fwip". That's the sound you make by 'palm-muting' the strings right at the bridge, while the reverb is turned up noticably. The more fwip, the sweeter the resultant tone. Don't believe me? Try Pipeline with no reverb. Ah, I see your hand lashing out at the reverb control, twisting it back up! ;D And to get the maximum fwip, use the softest pick you can manage to hang onto. My cymbal pick gets put down for surf instrumentals. </diatribe> sumgai |
|
|
|
Post by UnklMickey on Aug 4, 2006 19:33:01 GMT -5
...Is this clear? V4b sees a + coming from the plate of V2b (dry), and a - coming from the plate of V1b (reverb). This is very easy to see on the diagram - we have taken away one inverting stage (V4a), and put in two inverting stages (V1a and b). All of our tone and volume problems will stem from this little factoid. ... hey, now we're on the right track! this is where we differ on our analyses. i suggest that with the modification, V4b will see a + from the dry signal, but it won't cancel with the - of the reverb. why? ................................. because the reverb signal is never seen at the input of V4b with the modification! the first place that it is now re-introduced is at the P.I. (via the normal channel) and there, the dry and reverb signals are both - so how'd i do, is it your turn to buy the next round? BTW, loved the diatribe. not into the SG scene myself, but it's always great to learn about what creates certain tones and textures. and terms like "fwip"? hey, what's not to love about that! unk |
|
|
|
Post by sumgai on Aug 4, 2006 23:45:31 GMT -5
unk,
You said, way back when, that I must be losing track of my summing points. And indeed I did. So, without further ado, let's revisit the scene of the crime......
Ordinary, unmodified version:
V2a .. +
V2b .. -
V3 .... +
V4a .. -
V4b .. +
V6a .. -
Modified version, as per instructions on schematicheaven:
V2a .. +
V2b .. -
V3 .... +
V1a .. -
V1b .. +
V4b .. -
V6a .. - and - !!
OK, I should have worked that out before posting two messages ago. Then I would have realized that I was moving the "summing node". Obviously I was fixated on the Reverb signal, eh?
Previous post modifed to point to this correction.
sumgai
|
|
|
|
Post by UnklMickey on Aug 8, 2006 11:16:13 GMT -5
Hi Sumgai,
it looks like we're now on the same page regarding the phase of the signals.
first, i'd like to make a couple of statements.
since the reverb signal using the "mod" now skips BOTH V4a and V4b, the gain is probably similar using the normal channel.
granted, we would probably get a little more gain through the normal channel since it doesn't have the voltage divider (470k and 220k).
but we have some losses in the tone stack, so all things considered, we probably give up just a little.
bottom line: since it's both cheap and (plug-in) reversible, trying the basic "mod" out is probably time worth spent, for someone who might like to have T&B on the reverb return.
of course just ignore all that crap about using the existing reverb knob for dwell or anything else. (unless you just want to add some hum and noise)
now for more interesting thoughts.
whether or not one would do the "mod" it might be interesting to experiment with driving the springs harder.
i've been musing with that idea since you mentioned it.
so how to do it, with what's already (mostly) there?
the 2 halves of the 12AT7 are in parallel, so we have the capability to dissipate 3 watts in those plates.
i really doubt we really need all that.
so we might change to a 12AX7 and go from an amplification factor of 60 to 100. (might need to change the value of the cathode resistor too)
or maybe we can use half of the 12AT7 to give us lots more gain before it drives the second half that will still drive the transformer (need to exchange the primary leads to invert the signal too)
now the 1 meg resistor on the grid(s) of the reverb driver gets changed to a trimpot OR maybe the reverb control becomes a tandem pot, so that the drive and return are both controlled by the same knob.
we haven't yet thought about working with the size of the 3.3M // 10pF to affect the amount of regeneration (dwell?).
hey, don't mind me, or take any of this too seriously.
i'm just daydreaming out loud.
unk
|
|
|
|
Post by ChrisK on Aug 8, 2006 13:14:18 GMT -5
Hmmmm(ski), Dumb @ss question/observation. Since reverb is a mechanical delay thru the springy thingy, does electrical phase even/ever matter? The delay is fixed'ish (?mechanical?), the period of the waveforms vary with frequency, and phase is thence meaningless? ?? I can see where adding another gain stage will require addressing the phase issues of the "dwell" components 10pF||3M Ohm. But the propagation characteristics of the springy thingy...... Coffee break is over, I'll get back on my head now. |
|
|
|
Post by sumgai on Aug 8, 2006 13:52:05 GMT -5
unk, I've been through just about all of what you mentioned, having been in the surf guitar world for many a year. The output impedance of the two halves of the 12A(T||X)7 is matched to the reverb driver transformer. If you split that tube, you will need to replace the transformer, or come up with some other way to match the two items. The 3.3MΩ and 10pf parts are indeed a large loss on the dry signal, thus the need for V4b. By reducing the resistor size, you do garner some additional reverb regeneration, but you're probably never hear it - you're now driving V4b even harder with the dry signal, thus causing distortion even sooner than usual. Better would be to run a separate tap from V4a's plate back to the driver side, but don't forget to pay attention to the signal phase. I say that because of two things. Remember that an oscillator is merely an amplifier with positive feedback. If it's in phase, then it doesn't take much gain at all to start the circuit off on a real toot! If it's out of phase, then the regenerated signal is tonally affected. I've not done this, so I can't predict if the effect would be for the better or otherwise. The other thing is, supposing that you have conquered any tendency to break into oscillation, then what have you gained? Does the signal take so long to decay that you are into the next measure before it dies off? That's not an unrealistic expectation, we can do that now, digitally. And from that experience, I get the feeling that decay time can be overly long, to the point of hindering the music, not enhancing it. But that's just me, of course, and as the lawyers always make me say. YMMV! ;D A long time ago, I ran across a treatise of what a reverberated signal consisted of. It came about that the initial attack was interpreted by the pickup as "noise" - it was the sound of the string accelerating to full speed (defined as the time needed to reach 1 full excursion away from the zero point). That "noise" is seen by the reverb tank, and after being bounced around for awhile, it comes out as what I call "fwip". So the trick, as far as I'm concerned, is to get that 'noise' to appear to last longer. Since I can't change the string's properties, I'm forced to consider electrical (or mechanical) ways to enhance that first moment of the attack. This is indeed where regeneration may play a part. (Of course, the mechanical part is to modify the tank itself, something I'm not equipped to handle.) But how to keep the signal from running on and on? I'm thinking of inserting an expander network into that feedback loop, and essentially allowing only the initial attack to be sent back to the beginning. By manipulating the ADSR values, I should be able to reduce the overall feedback signal without affecting the first 10 to 20 msec. worth. Or so I hope, anyways. ;D Are you considering playing with this aspect of reverb? I don't have time this summer (I can't even get started on that pickup-placement project like I wanted to), it just isn't in the cards for the near future. I'd like to participate, even if only by posting negatives remarks about your results! j/k!  Now that we've gone way, way  , it's probably time to start a new thread, eh? sumgai |
|
|
|
Post by ChrisK on Aug 8, 2006 14:12:38 GMT -5
Bimp
(Jacques Clouseau)
|
|
|
|
Post by sumgai on Aug 8, 2006 14:27:32 GMT -5
He who posts while I'm composing, Nope, the electrical signal phase is still intact, and still important. What you're referring to would more closely approximate the harmonic relationships of each springy thingy transistion. It's a matter of the speed of electricity versus the speed of sound (and the harmonics relating thereto). Thankfully, the two transducers within the pan don't affect the electrical component of the signal (the "carrier"), only the sound portion. As for "delay", that's not what's really happening. The sound's decay property is being modified here, there is no delay. In fact, if the sound were to be delayed, it would come out as an echo, not as a reverberation. (The ear needs only about 1/12 to 1/10 of a second to perceive this effect.) And for the most part, decay is not "fixed". Indeed, that's what unk and I are discussing, ways to modify that property using what's already onboard the amp. The Dwell control on Fender's Reverb Unit is the classic example of varying the decay time of a signal. Regeneration is another, albeit less-often explored, avenue. sumgai |
|
|
|
Post by UnklMickey on Aug 8, 2006 14:33:44 GMT -5
i'm thinking not-so-much, and subscribe to the same thinking that led to your question/observation. fixed'ish is really applicable here. sending large pulses through the springs (KICK THE REVERB) will prolly affect the propagation time. but, i'm not prepared to make any hard and fast statements to that regard. that's just how it appears to me. ...I've been through just about all of what you mentioned, having been in the surf guitar world for many a year.... as i would have supposed! that's one of the reasons i voiced my musings. so you could tell me what might have merit, and what is total dookie. i doubt i'll ever actually follow up on any of these ideas. don't have a DR right now, and too many other toys are calling. (as are my creditors! J/K) but hey, if we're gonna go OT, i think this was a great place and great way to do it! N/K unk |
|
|
|
Post by ChrisK on Aug 8, 2006 16:34:15 GMT -5
Well, we have a parallel triode driving a transformer to the pan. We have a triode amplifying the output of the pan.
I assume that the pan has two coils in it. One is the drive coil to the springs and the other is the sense coil on the springs.
I further assume that the only input to output coupling in the pan is the mechanical spring(s).
Therefore, the speed of sound energy thru the spring(s) defines the delay and not the speed of electricity. If the speed of electricity was the coupling mechanism, it would have no effect since the transition time would be on the order of 0.2nSec. (about the clock period time on a 3 gHz Pentium processor).
Decay = 1/Dwell (They're reciprocal terms; decay is how fast something diminishes, dwell is how long it hangs around.) The 10pF||3M3 Ohm link is a regeneration path that feeds some of the pan output back into the input.
Based on the presumption that the pan is a mechanical device, there is a finite delay in the initial energy transition time thru the spring(s) and the mechanical echo effect of reflected energy in the spring(s) (just like a plucked string or taut clothesline).
The propagation time is relatively fast and won't appear as a discernible echo.
However, it is finite and unrelated to signal frequency (which defines relative phase shift), and therefore, the electrical phase of the drive to sense signals on the pan is immaterial.
??
|
|
|
|
Post by sumgai on Aug 9, 2006 2:15:59 GMT -5
Chris,
Seems we've been looking at this thing from opposing sides of the fence! ;D
Now that you've defined some terms, I'll easily agree with them, and simply state that your hypothesis are all correct. - I see no reason to "argue" against any of them. ;D
The unkster is correct, this too is crying out for some in-depth experimentation. Remind me of this come about T-day or so. I should be caught up by then, right? ;D
sumgai
|
|
|
|
Post by ChrisK on Aug 9, 2006 14:31:24 GMT -5
Well, I'm still thinking aboot the '65 DR RI for $683. If I get it, I'll be trying many things (including a possible conversion to point to point pegboard - or maybe I should just build one). I've given up on the Pro Reverb (pluses - lots of shiny knobs, minuses - heavy and half-power is wussy on 6L6's). And, I still gots to know aboot this one: www.epiphone.com/news.asp?NewsID=600 |
|
|
|
Post by UnklMickey on Aug 9, 2006 15:30:42 GMT -5
Well, I'm still thinking aboot the '65 DR RI for $683. If I get it, I'll be trying many things (including a possible conversion to point to point pegboard - or maybe I should just build one).... Chris, what does this have to do with reverb modifications! LSH... i think one deciding factor on converting to a turret board (i think turrets would be better than eyelet) would be if the tube sockets are already chassis mounted instead of on the printed board. if not, you'll have to do additional machining, the extra work might be the straw that broke the camel's back. OTOH, having an amp that looks like a DR reissue, with a well executed PtP circuit would be pretty trick. IMHO, it's not just the weight vs 2 6L6 limitation that turns me off on the Pro. 40w is still too much for a small amp, and 6V6s sound prettier than 6L6s anyway. so, if ya gots da bucks, and ya wanna make me (even more) jealous, buy the DR and make all the (stealth) mods. unk |
|
|
|
Post by ChrisK on Aug 10, 2006 13:09:42 GMT -5
Actually, I'm SERIOUSLY looking at Allen Amplification for their PTP KITS. www.allenamps.com/index.htmlThey all have external chassis jacks for re-bias. Many speaker choices. Great reviews on HC. The Sweet Spot, a Princeton Reverb type; 18 Watts 6V6, 25 Watts 6L6 Head - $1,300 Kit - $800 1x12 $1,500 Kit - $999 1x10 - $1,450 Kit - $900 1x12 - $1,500 Kit - $1,000 www.allenamps.com/sweetspot.htmlThe Accomplice, a Deluxe Reverb type; 22 Watts 6V6, 35 Watts 6L6 Head - $1,350 Kit - $849 1x12 $1,550 Kit - $999 2x10 - $1,600 Kit - $1,050 1x15 - $1,600 Kit - $1,050 www.allenamps.com/accomplice.htmlThe Encore, a mid 60's Vibrolux Reverb type; 35 Watts 6L6 Head - $1,500 Kit - $850 2x10 - $1,700 Kit - $1,150 1x15 -$1,750 Kit - $1,200 4x10 - $1,900 Kit - $1,350 www.allenamps.com/encore.htmlThe Old Flame, a mid 60's Super Reverb type; 25 Watts 6V6, 40 Watts 6L6 3 control reverb!!!!! Blackface/Tweed pre-amp voltage selection. Head - In revision Kit - $899 2x10 - In revision Kit - $1,100 1x15 - In revision Kit - $1,200 www.allenamps.com/oldflame.html!!!!!!My favorite!!!!!!!!!!! |
|
?)..... just hook the output of V1b to the input of V2a. That was too strong for me, so I added a 82KΩ current limiting resistor in front of V2a's grid, and that sounded pretty darn good, if I do say so myself. (I also removed the 6.8kΩ resistor from the bottom of each tone stack, essentially emulating the Mid control being set to 1. Sounded even better to my ears, back then.)
))
) 
