Somebody brought this up recently on (gasp!) a different forum (can't remember which one) but I couldn't find it again, so I figured I'd ask here...
The last couple of paragraphs concerning Strats is interesting, but I don't understand exactly how John is connecting the 20k resistors in the circuit, and how it would work...
Post by UnklMickey on Jan 11, 2006 13:24:23 GMT -5
seems pretty self explanatory to me:
"If you want a different sound from your Strat or Tele try putting about a 20k resistor between the hot lead of each pickup and the selector switch"
so, at the selector switch, disconnect the wire from the pickup, solder a 20k resistor to the lug of the switch, and connect the wire to the other end of the resistor. (do this 3 times, once for each pickup)
won't make much difference in the sound at 1,3,or5. but the coupling of the coils is reduced in 2 and 4.
i haven't tried this myself (and probably won't), but it seems to have some basis in theory.
unk
Last Edit: Jan 11, 2006 13:41:41 GMT -5 by UnklMickey
"Silence is golden, but duct tape is silver. " -- Steve Hopton
"some people say happiness is just a State of mind...................... i think it should be a whole freakin' Country!" -- unklmickey
Post by UnklMickey on Jan 11, 2006 14:13:22 GMT -5
i expect you might not even be able to detect the difference in volume.
i really can't say whether or not it would matter if all the pickups had a resistor in series, or if it would be better to have one just one, in series with the middle.
the point is, since the pickups are not directly connected together, there would be less interaction. that would (at least i think it would) change the tonality. but i don't know if that would be better or worse.
unk
"Silence is golden, but duct tape is silver. " -- Steve Hopton
"some people say happiness is just a State of mind...................... i think it should be a whole freakin' Country!" -- unklmickey
I was thinking that the resistor on just the middle pickup would make the most sense, but it was a little unclear to me on the original article. If I ever try this, I'll be sure to post...
I would advise caution in applying this concept. With a resistor (lets say 22k, to stick with standard values) on just one pup such as the middle, then that pup will be significantly reduced when combined with another, such as in positions 2 and 4.
You can picture each pup as a pure voltage source in series with an impedance. The impedance of the pup and that of the other, form effectively a voltage divider, reducing the output
Lets say you have a 6k impedance (lets just think of low frequencies – it rises at higher frequencies). the middle pup with the 22k resistor now has about 28k impedance. The other (say the bridge pup)still is at 6k.
the signal from the middle pup is now just:
6 / (28 + 6) = 18%
signal from other pup is:
28 / (28 + 6) = 82%
Hence it has thrown the balance strongly towards the bridge pup
This is actually a good principle for making a blender control, if you can vary the resistor.
How about having resistors on all pups?
This is better, but still be careful. if anyone tries this, make sure you can do an A/B test in some way to make sure the result is desirable. The risk is an undesirable reduction in top-end output.
To analyse this it is a bit more complicated, but also more interesting:
The response from the pickup is dependant on the inductance of the pickup, the resistance of the pickup and any other resistor, plus various capacitances of which the most significant is the guitar chord to the amp. I simulated this using pSpice, with this circuit:
This represents a single coil pickup with 6k of its own resistance and 2.5H impedance, a 500k volume pot, a 20’ cable with a capacitance of 800pF, feeding into a 1M input impedance. These are typical values for a single coil pup. The source is a nominal 1V over a frequency sweep, and we read the output reaching the amp at the marker on the right.
Here is the results, first with exactly that circuit and no resistor added:
Notice a peak in response around 3kHz frequency
Here it is, adding a 22k resistor, modelled by increasing the pup resistance to 6k + 22k = 28k:
That peak is now much less, which would result in a difference to the top end response.
Volume Pots
While we are about it, lets check out a different value for the volume pot. The runs above were with a 500k pot, but many people prefer a 250k pot with single coils. Here is the response, without the extra resistor:
It shows a reduced peak, but not by so much as the resistor in series caused.
Back to the original question
To get the possible benefit of having the extra decoupled resistors, but without the potential loss of treble, one solution is to put a single FET preamp within the guitar. I have come to the conclusion that this is the right thing to do to sort out cable capacitance, treble bleed, blending , driving other devices and a raft of other small problems, but that will be another thread.
Post by UnklMickey on Jan 12, 2006 16:19:49 GMT -5
John,
thanks for looking at this from a more analytical viewpoint.
i'll disagree with the "percentage" values. they may be true in terms of voltages, but to be more useful, they would have to be expressed in dB (logarithmic) terms.
none the less, your point about things being strongly biased toward the signal from the bridge pickup is well taken.
i see you are using pspice. what flavor is it, how much did it cost, and how do you like it?
unk
"Silence is golden, but duct tape is silver. " -- Steve Hopton
"some people say happiness is just a State of mind...................... i think it should be a whole freakin' Country!" -- unklmickey
Unk - yes those are just voltage %. I left out the leap to db = 10logV, for the sake of younger viewers, just to illustrate that there is a significant change in output when you add the resistor.
Im using the student edition of pSpice 9.1, which is free.
Here is a link from which 9.2.3 can be downloaded (I havnt tried it however): www.sss-mag.com/spice.html
It is great for any kind of passive circuit with R, C and L. I also use it for active circuits, although I do not have data for all the common transistor types. it is very useful for testing a principles, but i find that sometimes the best values for components turn out a bit different in real life.
Hi folks! I'm very pleased to find this community again (pretty well hidden, though! Any chance of a link on the original GN site?). The combined group expertise seems to be several notches higher than when I was last here (18 months ago, plus or minus?). I feel like a novice again!
John's post has pre-empted a couple of questions I was mulling over on the bus in preparation for my six-monthly rewiring of my long-suffering axe.
Q1) Imagine (I suggest a quick sketch):
a) OUTPUT to PUP2 to 125K RESISTOR to EARTH; b) "10 end" of a 250K LINEAR TAPER POT is wired to PUP2(hot) and "0 end"of POT is wired to PUP2(cold) (i.e. the pot is in paralell with pup2); c) WIPER of pot goes to PUP1(hot); d) PUP1(cold) goes to EARTH.
(If only I had a scanner I'd be able to express it in schematics!)
I reckon (and I think John's comments back this view although my in depth electronics know-how is a bit sketchy) that you have the following sounds at various pot positions:
10 -Pup1 on it's own 5 -Pup1 and Pup2 in paralell 0 -Pup1 and Pup2 in series
Presumeably turning 10 to 5 would fade in Pup2 in paralell, can't quite imagine what the effect would be turning 5 to 0.
What do you reckon? My thought was to include two pots wired thus with a tele switch in a strat with the following: 1) N 2) N/B (paralell) 3) B
In pos1 the Neck "tone" control would bring in the Mid pup as envisioned above; In pos3 the Bridge "tone" control would bring in the Mid pup as above. (So you can "set" each position independently). No "tone" control in pos2.
Q2) (This is the last one I promise!) What do you reckon the best way of wiring a pot to fade in a pup in series?
I reckon: a) EARTH to PUP2(cold); b) Linear 250K pot wired in paralell with PUP2; c) Wiper of pot wired to PUP1(cold); d) PUP1(hot) to output;
With a bit of luck PUP1 is always on and PUP2 fades in smoothly as the pot is twiddled.
Hi Hastings. I'm glad you found us - I know from seeing your input on the old board, that you will have a lot to offer here.
I have also been trying to figure out how to fade pickups in series in parallel. The ToneMonster2 on the schematics page has one version of a blender, which works best with a buffer circuit.
Here is my latest version for two humbuckers with series and parallel, which I have built into an old Cort LP. I have only recently finished this, and it makes a huge range of sounds. Here is the schematic as a taster:
It is along the lines discussed earlier in this thread, with a JFET buffer as the key to allowing those 56k seperating resistors to be introduced, without causing loss of signal. There are also some 'treble bleed' resistors and caps, which I added to give a smoother control and because it sounded better that way.
With an ounce of encouragement, I'll make a post of this design with wiring diagrams and photos too - we need more on the schematics board.
Hi Hastings. I'm glad you found us - I know from seeing your input on the old board, that you will have a lot to offer here.
Thanks for the compliment! Although, my undertstanding of arcane concepts like "impedance" and "frequency response" is about as good as my understanding of Inuit sub-dialects.