The Gibson Varitone

[wolf]

This is probably the umpteenth topic to discuss this circuit. With the recent death of BB King, it made me think that he was quite a fan of using this circuit.
Basically what I want to know is what type of circuit this is.
I think it is a bass-cut treble-cut circuit, which is usually (and incorrectly) referred to as a "mid-boost" circuit.

[b4nj0]

I think you're about right. As I see it, it's a series resonant LC circuit that will exhibit low impedance (and thus act as a band pass filter) at the resonant frequency of the two lumped constants. The bandwidth will depend upon the "Q" of the circuit. Changing one (or both normally) of the values will change the resonant frequency and thus the band pass frequency. The higher and lower frequencies will encounter a greater impedance and will thus become attenuated. It is not a mid boost but it might give that impression. I am guilty of extrapolating from RF here, so if I'm out of my tree, I apologise and hand it over to the gurus!

Edit;
Seeing Newey's reply made me realise that I should have written that because it is connected to ground, the effect is as Newey says, a notch filter.

e&oe...

[newey]

I think it is a bass-cut treble-cut circuit, which is usually (and incorrectly) referred to as a "mid-boost" circuit

Just the opposite. A circuit which allowed the midrange through, but cut the highs and lows, would be a "gate" filter, also called a "band-pass" filter (since it allows a certain band of frequencies to pass through to the output).

But the Varitone is actually a "notch" filter, also called a "band cut" or "band stop" filter. It cuts the mids out, leaving just the highs and lows. Turning the Varitone knob changes the frequency range that is cut, widening or narrowing the "notch".

[sumgai]

Remembering that that filter itself is actually a bandpass jobbie, and that it is shunted, we are left with the characteristics of a notch filter at the output jack. IOW, the Varitone circuit itself acts to pass only certain frequencies between a low cutoff point and a high cutoff point. In reality, those passing frequencies are sent to ground. All the left-over frequencies that weren't grounded get to move on to the output jack.


Now, in some research for this reply, I was reminded of the following link I had found some time ago. Pay attention, this guy is the Real-Meal-Deal:

www.ampsnaxes.com/wp-content/uploads/2013/12/varitonepaper.pdf

Particularly, his statements on the first and last pages are exceptionally poignant - he disqualifies himself as a so-called "expert", and in between those disclaimers, he reduces several semesters of Electrical Engineering School to a handful of paragraphs and diagrams. And IMO, he does it correctly. I'm figuring this guy and our own JohnH are two peas in a pod.

I'd also like to call attention to his discussion on so-called "proof" postings - what we commonly call "before and after" sound files. He channels me exactly, though perhaps with a tad more acerbity.

HTH







sumgai

[JohnH]

The way Gibson show it on their old hand-drawn diagrams, I reckon there are 3 key elements:

1. When in the five active positions, the direct route from pickup to output is via a 100k resistor in series. This raises the impedance at the output, which gives the inductor/cap combination an easier job of shunting some mid frequencies to ground. It will also lower the general output level, and kiss goodbye to the usual high treble peak caused by cable/pickup interaction.

2. The inductor is in series with the selected cap, all between output and ground. The interesting characteristic of a perfect capacitor and inductor in series is that, at the frequency where their impedances are equal magnitude, they actually add up to zero net impedance because they are fully out of phase. Its like adding equal '+' impedance to '-' impedance to make 'zero' impedance at that frequency. The result is a very strong notch filter, with a depth determined by the simple resistance of the inductor coil. Changing the caps with the inductor staying constant will change the notch frequency.

3. A less significant part: Each cap has a full-time connecttion from hot to ground via a 10M resistor. This is to keep static charges from building up to avoid crackles when switching. But with 5 caps so connected, the effect when bypassed is like an extra 2M load on the signal. Still not much but enough to take a bit of edge off even in the 'off' position.

I think I can use GuitarFreak to show these effects. It has a 'Varitone' setting for cap and inductor. Although it doesnt have a specific component to represent the 100k, I think the bass-cut circuit can be pressed into use, without bass-cut cap. Will post when done.

[wolf]

I might as well attach my varitone graphic:


I thought about the circuit too simplistically.
As you can see, it has capacitors going to ground which (supposedly) would be a treble cut.
Then the inductor going to ground would (supposedly) be a bass-cut.
I didn't take into account that the capacitors and inductor are wired to each other and would behave a lot differently.

So, anyway, this circuit behaves as a notch filter. So, is this what is usually referred to as a mid-cut circuit or a mid-scoop?

[newey]

Yes, also called a "mid scoop" or cut.

[ChristoMephisto]

Depending on how you arrange them, you can also make a low pass or high pass. Much like a RC circuit can do both.

[JohnH]

Here is what I have worked out, using close to the Gibson values 0.001, 0.003, 0.022, 0.033, 0.22 uF, with the 1.5H inductor and 100k resistor:



It looks like the values could do with being spread out more evenly.

[ashcatlt]

wolf - Your pic is not exactly the same as what I'm seeing in the link that sg posted, nor does it agree with JohnH's description. I don't think your version works quite the same.

I really don't like all this talk about "shunting to ground". Ground is not a destination! It's better to say that certain frequencies complete their circuit (back to the pickup) more easily via this path than through the load of the amp, but it's even more accurate to say that this is a frequency-dependent voltage divider in which the "bottom resistor" is very small for midrange frequencies.

[ChristoMephisto]

Can't remember where I saw it, but the old Gibson used 7H inductors.

Isn't there a LRC calculator for this?

[b4nj0]

It's a tuned circuit and as such it doesn't matter what size the inductor is so long as together with the selected capacitors, it resonates at the same frequency. (The "calculator" is f=reciprocal of 2*Pi*√(L*C).)

The larger inductor mentioned (and resultant necessary different values of capacitors) will affect the "Q" and therefore the bandwidth of the pass band for any given switch position. (Which in this application is connecting a portion of the signal with what I feel comfortable in thinking of as "ground" (even though I agree it's not.)

I have no trouble in thinking of "ground" or "earth" in the context of my house wiring, even though I know that technically, it is not at that potential (at least with PME it's not). For me it's the same with a guitar's "ground"- If all the common points were strapped to true ground, wouldn't the circuit would work just the same? (albeit with safety considerations and potentially unwanted ground loops.)

Too, there are no electronics in a passively wired guitar (ie, the vast majority) but I constantly read about "electronics" in guitars. It doesn't hinder my understanding of points being made.

I think that we need to distinguish between what a tuned circuit does and what it accomplishes when we use it as a "Varitone". I fell into this "trap" (no pun intended) with my first posting. For instance, we could connect a series tuned circuit in series with the signal and it would no longer act as a notch filter.

Again, I'm coming from an RF viewpoint here, so if I've lost the plot I hold my hands up! Thanks go out to 'gai for that excellent .pdf.

e&oe...

[wolf]

ashcatlt
The graphic that I posted was something that I posted here many years ago.
When you say, it doesn't look like the varitone in sumgai's link, which graphic do you mean?
Figure 1, 8, 12 or 15?
I just figured we all should get together and decide on just one circuit.

[ashcatlt]

Jun 6, 2015 1:01:26 GMT -5 wolf said:

ashcatlt
The graphic that I posted was something that I posted here many years ago.
When you say, it doesn't look like the varitone in sumgai's link, which graphic do you mean?
Figure 1, 8, 12 or 15?
I just figured we all should get together and decide on just one circuit.

It doesn't really match any of them. It's closest to fig1, but with a couple of very important differences. The biggest problem is that the 100K resistor needs to be in series with the pickup before this whole thing rather than sort of in series with this structure which is presumed to be parallel to the pickup except that the resistor is shorted by a wire so not actually doing anything. Also, the connection to the tone pot is significantly different.

Edit - Honestly, I'm not so much worried about the pot. You've adapted this for a 5-way rotary, eliminated the "normal Tone" function, and made the whole thing variable. That's all fine with me. But as JohnH said, that 100K resistor is pretty important. For the middle frequencies that this thing affects (before the pickup's reactance becomes significant) this resistor is the bulk of the "top resistor" in the frequency dependent voltage divider that makes this work. If it's not there, then this thing will react quite a bit differently, and might not be very satisfying. You might think to just remove the wire that's shorting it in your diagram, but that makes the 100K into the bulk of the "bottom resistor", and basically turns the whole thing most of the way off. The spot where it should go is not depicted on your diagram at all.

[JohnH]

I think there are a few circuits that have been called 'Varitone', but it seem like this is the most classic one from Gibson:

That's the one that I made plots for above.  The things about it are:

• It is right at the end of a standard Gibson HH circuit with 2V & 2T controls

• The control of the Varitone is just by switching caps, no extra pot is part of it

• The 100k resistor is active as the top half of a voltage divider when selecting the main settings, then bypassed when it is off

• But the caps are always loading the output through the 10M drain resistors. I checked that out, it makes about a -0.7db difference in high treble, not a big deal, but the effect is as if the normal pots were around 450k instead of 500k

Not having built one, but I didn't find the plotted results looked like it is something that I would want on one of my guitars! Too much loss of level in most settings. I think there is more potential either in having a pot in series with the caps and choke, with no 100k resistor, or make the 100k variable in which case it can double as a bass cut control. Also, I think a larger choke value may be more interesting, you can buy a 3H rather than a 1.5H, and the caps need to be selected to give a better spread of effects.

[YouXE]

Hello, everybody!

This is my first message. I've been reading your (very interesting) posts for a long time, but now I've decided to write and share something with you.

Stellartone has released a new circuit of passive tone shaping. They are well known for their great ToneStyler: a very well made variable low-pass filter, with many options for bass and guitar. But recently they offer a new product: the Vari-ToneStyler.
Here you have a nice demo of it:  (the classic tones first, the mid-scooped later). It sounds very good! I tried to investigate if there's something especial in their design, and as you can see they use a 15H choke. Not a 1.5H, a 15H choke.
So I ask them why that inductor value. Here is their answer:

"The original Gibson varitones from the late 1950s used 15 Henry chokes. These were surplus radio parts, and were inexpensive. As will be explained in greater detail below, this value and part was likely selected by Gibson based upon trial & error testing, plus ease of installation. If a smaller and cheaper part would have worked equally, that certainly would have been procured by Gibson.
In classic Gibson style, the published service schematics for the ES-345 and ES-355 guitars deliberately mis-labled the value of the chosen audio frequency chokes, at 1.5 H The inserted decimal point was intended to confuse competitors and copycats, but not authorized service centers who would be obtaining the correct parts from Gibson anyway. Decades later, a Gibson draftsman, unaware of the little joke, wrote the correct 15H value and an RCA part number on some varitone schematics.
If you look at a photo of the original 15 Henry chokes, they are large and heavy... in fact, a pair was used for the earliest stereo varitones.
In comparison, a choke with a 1.5 Henry value only measures about 1/2" in length, width, and depth. For this reason, cheap aftermarket varitones using this insufficient choke value perform very poorly, and produce several adverse tonal defects.
The worst varitone instructions and kits found on the internet use a tiny impedance-matching transformer sold by Mouser electronics, intended for use inside microphones, and inside 1/4" to XLR adapters. These tiny transformers only measure about 0.8 Henries (800 mH), and are totally incapable of producing the desired mid-notch "V" filter shape.
The opportunistic marketers of these kits were duped by Gibson's mis-labeled specifications. To make matters worse, both the buyers and sellers are unable to distinguish proper vs improper audio performance. Compared to an original 1958 Gibson Varitone, the audio filtering result is just awful, which is why the kits are rarely completed, except by gullible or tone-deaf victims.
The sub-standard quality of varitones made after about 1963 is the primary reason for their continuing bad reputation. Once the original 15 Henry surplus chokes were no longer available, Gibson resorted to winding pickup wire around sewing machine bobbins, with a 1/4" steel bolt inserted though the center to create a magnetic "core". These are very poor substitutes for the originals, for two reasons: first, the inductance value is only about 2.5 H, and second, the DC resistance value is way too high, at about +3k Ohms. Just like the cheesy aftermarket kits, this wrong inductance value produces poor results.... which is why Gibson's BluesHawk and Epiphone's Lucille models have been unsuccessful.
Enough background.... here's the practical information:
By experimenting with audio frequency chokes with inductance values ranging between 1 Henry and 60 Henries, it's immediately apparent that the best-quality audio filtering, and the most distinctive "V" shaped mid-notch, is produced by the higher, rather than the lower inductor values. A-B performance testing confirms that low-value inductors, such as in Gibson's practical joke, produce very inferior results. Conveniently, values above 20 Henries produce diminishing returns - a further increase in size and weight does not produce a proportionate audio-quality benefit. So, 15H is the winning value.
If larger-value chokes with a low DC resistance were compact, they could be used inside of musical instruments. But, the physical requirements to create higher inductance means larger metal cores and armature wire gauges. Chokes above 15 Henries are much larger and heavier than a typical humbucker pickup, which creates installation problems.
For this reason, a balance between proper audio performance and practical size has been achieved in both Gibson's1958 surplus radio choke, and the 2016 Stellartone choke, as shown in these product photos.
Bottom line: a small choke performs poorly, a medium choke does not perform as well, but a large choke does not install easily. The nominal 15 Henry rating is long-proven to be effective in a mid-notch filter. A choke below 3 Henries will never sound right, despite any reader's attempts to use SPICE circuit simulations to plot the same "V" notch. This is not a matter of debate; just listen for yourself!
Both the original 1950's surplus choke and Stellartone's new hand-wired choke are potted to avoid microphonics, and use mu-metal cases for superior background noise reduction. The Stellartone choke also offers an improvement upon the obsolete Gibson version: instead of only one ferrite core and winding, our choke module has two sets, wired out of phase, to create the well-known and desirable "humbucker effect"... canceling out RFI, EMI and AC mains hum, which otherwise might be received by the varitone circuit.
I hope this adequately explains why SIZE MATTERS when choosing the best audio frequency choke for a passive mid-notch filter. Thanks for your interest! Best regards, D. Campbell, Stellartone USA."

"There's really no alternative to a properly-designed and constructed varitone. Audio filters featuring a $3 impedance-matching transformer are not worthy of consideration, testing, or installation. Think of the audio performance of a PAF humbucker pickup, vs some copper wire wrapped around a pencil. Both items will detect string vibrations, but they are certainly far from equivalent items!
In this scenario, the "joke varitone" with the 800 mH choke will make your PAF humbucker sound exactly like a coil of wire wrapped around a pencil! See what I did there? Emoticón wink Best regards, D. Campbell Stellartone USA."

What do you think about this? I've tried to simulate the effect of the circuit with GuitarFreak software, and there seems to be something that makes the difference: less loss of high frequencies.
Maybe JohnH can confirm this, as he controls way much better the parameters.

[reTrEaD]

Feb 4, 2016 13:26:33 GMT -5 YouXE said:

Hello, everybody!

This is my first message. I've been reading your (very interesting) posts for a long time, but now I've decided to write and share something with you.

Hello, YouXE. I'm sure newey will stop by to give you the official Guitarnutz welcome with all the colored text and bells and whistles but for now, Welcome to Guitarnuts2.


1) mid-scooped
Actually, it goes beyond that. The Varitone is a notch filter.

2) Stellartone
I don't see any pricing via that link but I suspect it won't be cheap.

3) The original Gibson varitones from the late 1950s used 15 Henry chokes.
I'm not sure about this one. It seems plausible. The big difference between using a 15H choke vs 1.5H would be the center frequency of the notches. The lists below are the calculated frequencies in Hz for each capacitor value listed in the original Varitone diagram.

15H choke
1299
750
411
237
88

1.5H choke
4109
2373
1299
750
277

4) DC resistance value
The higher the resistance in this network, the less sharp the notch. It would seem the lowest resistance would be most desirable. If too sharp, one could add a resistor to broaden the notch.


5) two sets, wired out of phase, to create the well-known and desirable "humbucker effect"
I wouldn't expect a choke to be nearly as susceptible to hum as a pickup, especially if the choke had shielding around it. But this is still a smart move.


6) I've tried to simulate the effect of the circuit with GuitarFreak software, and there seems to be something that makes the difference
Unless you know the resistance of the chokes involved, the main difference you're likely to see via modeling would be the relative position of the notches.  Imho, resistance will play a significant role.

[JohnH]

Interesting and I can believe it...

Varitone circuits usually consist of a capacitor and inductor in series. If these were 'perfect', then at the critical frequency, their impedances are equal and opposite, adding up to zero. So they make a very deep notch in the response as they shunt the signal to ground at this frequency.

Lets say we have a 15H inductor and a 4.7nF cap in series, then the notch frequency is about 600hz, which is in the range of interest for a tonal effect. We can also match this notch frequency using the 1.5H inductor, with a 47nF cap. Different caps can vary the notch-filter frequency.

In practice, the depth of the notch is limited by 'imperfections', mainly the inductor resistance. But in any case, it is a very deep cut at this frequency.

So both sets if values can make the same notch frequency. But I think the most significant differences are revealed by considering other frequencies. Take 2000hz for example. Ignoring coil resistance, the series impedance of the 15H & 4.7nF combo at 2khz is about 170k, which is high enough to have not too much effect, so preserving the tone in this region as well as notching-out at 600hz.

But at 2000hz, the 1.5H/47nF design has impedance 17k. That will destroy the tone at this higher frequency, and similarly across the broad range of tones that we want from the guitar.

So, I can fully believe that Gibson used 15H in their original Varitone! Although Ive never tried one, the numbers make the smaller values look very unappealing. I think that with 1.5H, or better about 3H, a coil and capacitor series circuit is best used with a pot also in series to control it. Then it can make a broad 'mid scoop' but not totally wreck the tone.

GuitarFreak plots to follow later...

[ChristoMephisto]

I've been looking into making a varitone with a concentric pot, which I may not do, but I've come across a lot of info on the net if you can filter through it.

It should be a RLC circuit instead. If you have just a LC, it creates a very narrow notch, you can't have a perfect LC because you will encounter some form of resistance either in wiring or component leads.

The resistor is needed to control the width of the notch, or called Q, for the Quality. It's more of an engineering term than a musical one. The Q is the bandwitdth that is effected. A low Q will be wide, and the same with a high Q having a narrow Q.
Here's a Bandwidth Calculator you can use once you know your Q.
www.sengpielaudio.com/calculator-cutoffFrequencies.htm
So with the last example, a 15H and a 4.7nf, and add a 100k like the typical Gibson values.
The center frequency is 599hz, the Q is .56, the upper and lower cut off is 268 to 1340 hz.
Change the values to 1.5H, 47nf cap and a 10k resistor, you may get the same. I've heard the inductor needs to be large enough for it to work, but not sure about that tho.
Sounds like it would cut out a lot, but the varitone is not a series RLC circuit, which does work as mid cut circut.
Looking at the schems, and other gibson varitones, it's a resistor parallel to the capacitor that's in series with a inductor.
This link has different LC circuits with SPLICE pics and codes.
www.allaboutcircuits.com/textbook/alternating-current/chpt-6/resonance-series-parallel-circuits/
From what I'm seeing, the resistor in parallel to the cap in series with the inductor has a frequency peak instead of a notch.
If you want to try a 15H, you can put two 42TL017 in series. They are 7.5H on the primary side.

[YouXE]

Thanks for your feedback. Anxiously waiting for John's plots and comments... Meanwhile, I would like to share some notes (if I am wrong, please let me know):

1) The inductance of the choke is important. With the same resonant frequency (adjusting the capacitor value to match it), a higher inductance value makes the impedance at that frequency higher, so the gain reduction at that point of resonance is lower with an e.g. 15H choke than with a 1.5H one? Theorically, no. At resonant frequency, impedance equals resistance, because reactance (the other part) is zero.

2) Above the resonant frequency, the 15H inductor gives a larger impedance to the filter than the 1.5H, so it keeps more of the tone at the higher registers.

3) But what about the DC resistance of the chokes? Calculating some combinations with this RLC in series calculator, I get:
    - 1.5H and 4kOhm choke (e.g. 42TL021) with 0.047uF capacitor:
      4kOhm of impedance at 599Hz, 17kOhm of impedance at 2kHz, 37kOhm of impedance at 4kHz
    - 1.5H and 40kOhm choke (fictitious) with 0.047uF capacitor:
      40kOhm of impedance at 599Hz, 46kOhm of impedance at 2kHz, 54kOhm of impedance at 4kHz
    - 15H and 40kOhm choke (e.g. 2x42TL017 in series) with 0.0047uF capacitor:
      40kOhm of impedance at 599Hz, 176kOhm of impedance at 2kHz, 371kOhm of impedance at 4kHz
    - 15H and 4kOhm choke (fictitious) with 0.0047uF capacitor:
      4kOhm of impedance at 599Hz, 172kOhm of impedance at 2kHz, 369kOhm of impedance at 4kHz

The results greatly confirm the Johnh's estimations. The example DC resistance values make a relevant difference in the results at the resonant frequency, but their role is secondary/marginal at higher frequencies, especially when interacting with an inductance of 15H.
I got the values of the Xicon's chokes from this datasheet, plus the indications of ChristoMephisto (thank you for your suggestion). It seems to be common knowledge that the 42TL021 is 1.5H at primary, and I've also read the 7.5H value of the 42TL017 in some forums, but is it easy to calculate those inductances only with the information in that sheet? I would like to be sure that the 42TL017 is 7.5H.

[YouXE]

Oh, no! What a big fail!
I confounded/changed the impendances and resistances of the datasheet! And the indicated impedances are at 1kHz!
If the resistance of the chokes are as listed in the document (between ~150 and ~700 Ohms), I think they are even less relevant.
Maybe the people who calculated those inductances got the values based on the assumption that the 42TL021 is 1.5H, so: If the 42TL021 gives an impedance of 4kOhm at 1kHz and the 42TL017 gives 20kOhm, then the inductance of the latter is 5 times the inductance of the former --> 1.5*5=7.5

[JohnH]

Here are some GuitarFreak plots, using 15H and 1.5H inductors.

These are just electrical response, I switched off pickup position and string vibration features in order to concentrate on the matter at hand.

The plots represent a PAF humbucker with 500k Pots. The Varitone is represented by an inductor, capacitor resistor all in series from hot to ground. I picked 1.5H with 47nF and 15H with 4.7nF, to keep the notch frequencies the same, then resistors of 1k, 10k and 100k to vary the depth. The value of 1k is about the dc resistance of the 15H inductors which seem to be available at Mouser and elsewhere.

The 15H lines are thicker, and dark blue is 'no varitone'. The bigger inductor value does look better, and if the caps were varied, the notch would sweep up the scale.

[helmuth]

On different websites I found three different notch frequency figures:

A) 1850 / 1050 / 650 / 350 / 140 Hz
B) 1950 / 1100 / 620 / 360 / 120 Hz
C) 1875 / 1090 / 650 / 350 / 130 Hz

I think there will not be great differences in sounds between all these.

With the capacitors (1 nF, 3.3 nF, 10 nF, 33 nF, 220 nF) the inductance be calculated. It varies between about 6 and 8 Henries. For my Varitone replicas (called "ToneVari") I use 7 Henries.

[perfboardpatcher]

Feb 14, 2016 17:05:19 GMT -5 YouXE said:

The results greatly confirm the Johnh's estimations. The example DC resistance values make a relevant difference in the results at the resonant frequency, but their role is secondary/marginal at higher frequencies, especially when interacting with an inductance of 15H.
I got the values of the Xicon's chokes from this datasheet, plus the indications of ChristoMephisto (thank you for your suggestion). It seems to be common knowledge that the 42TL021 is 1.5H at primary, and I've also read the 7.5H value of the 42TL017 in some forums, but is it easy to calculate those inductances only with the information in that sheet? I would like to be sure that the 42TL017 is 7.5H.

Hi YouXE,


I did some calculations,

Let's presume impedance is 20k at 1000 Hz.
L = X_L /(2 x pi x f) = 20000 / (2 x pi x 1000) = 3.18 Henry

If we take 3.18 Henry when we calculate the low corner frequency we get:
f = R / (2 x pi x L) =693 /(2 x pi x 3.18) = 34.7 Hz
That's already pretty low, so it's not very likely that L is 7.5 or 15 Henry.

I looked further on the internet and found these 2 from Xicon:
42TM117-RC Z 50k primary 1k2
42TM025-RC Z 50k primary 700
If we apply the same calculations we get:
L = X_L /(2 x pi x f) = 50000 / (2 x pi x 1000) = 8.0 Henry
That's more like it!

42TM117-RC


Paul

[YouXE]

Feb 17, 2016 14:59:58 GMT -5 perfboardpatcher said:

Feb 14, 2016 17:05:19 GMT -5 YouXE said:

The results greatly confirm the Johnh's estimations. The example DC resistance values make a relevant difference in the results at the resonant frequency, but their role is secondary/marginal at higher frequencies, especially when interacting with an inductance of 15H.
I got the values of the Xicon's chokes from this datasheet, plus the indications of ChristoMephisto (thank you for your suggestion). It seems to be common knowledge that the 42TL021 is 1.5H at primary, and I've also read the 7.5H value of the 42TL017 in some forums, but is it easy to calculate those inductances only with the information in that sheet? I would like to be sure that the 42TL017 is 7.5H.

Hi YouXE,


I did some calculations,

Let's presume impedance is 20k at 1000 Hz.
L = X_L /(2 x pi x f) = 20000 / (2 x pi x 1000) = 3.18 Henry

If we take 3.18 Henry when we calculate the low corner frequency we get:
f = R / (2 x pi x L) =693 /(2 x pi x 3.18) = 34.7 Hz
That's already pretty low, so it's not very likely that L is 7.5 or 15 Henry.

I looked further on the internet and found these 2 from Xicon:
42TM117-RC Z 50k primary 1k2
42TM025-RC Z 50k primary 700
If we apply the same calculations we get:
L = X_L /(2 x pi x f) = 50000 / (2 x pi x 1000) = 8.0 Henry
That's more like it!

42TM117-RC


Paul

Oh, thank you very much, Paul.
Now I am even more confused about the inductance values of those chokes. Because the 42TL021 has been known as a 1.5H, but with those formulas (supposing 1kHz is the measurment frequency) the value for the 42TL021 would be around 0.650H.
Could anybody measure them with an accurate tool and tell us what are the results?

[vinnie1971]

Feb 6, 2016 8:47:37 GMT -5 reTrEaD said:

Feb 4, 2016 13:26:33 GMT -5 YouXE said:

Hello, everybody!

This is my first message. I've been reading your (very interesting) posts for a long time, but now I've decided to write and share something with you.

Hello, YouXE. I'm sure newey will stop by to give you the official Guitarnutz welcome with all the colored text and bells and whistles but for now, Welcome to Guitarnuts2.


1) mid-scooped
Actually, it goes beyond that. The Varitone is a notch filter.

2) Stellartone
I don't see any pricing via that link but I suspect it won't be cheap.

3) The original Gibson varitones from the late 1950s used 15 Henry chokes.
I'm not sure about this one. It seems plausible. The big difference between using a 15H choke vs 1.5H would be the center frequency of the notches. The lists below are the calculated frequencies in Hz for each capacitor value listed in the original Varitone diagram.

15H choke
1299
750
411
237
88

1.5H choke
4109
2373
1299
750
277

4) DC resistance value
The higher the resistance in this network, the less sharp the notch. It would seem the lowest resistance would be most desirable. If too sharp, one could add a resistor to broaden the notch.


5) two sets, wired out of phase, to create the well-known and desirable "humbucker effect"
I wouldn't expect a choke to be nearly as susceptible to hum as a pickup, especially if the choke had shielding around it. But this is still a smart move.


6) I've tried to simulate the effect of the circuit with GuitarFreak software, and there seems to be something that makes the difference
Unless you know the resistance of the chokes involved, the main difference you're likely to see via modeling would be the relative position of the notches.  Imho, resistance will play a significant role.

Interesting, but is it true? Not so sure.
I have looked st the graph to help me select my next capacitor values, but I can say I have made passive tone controls using the Xicon 1.5 H inductor. I looked at videos of the original Gibson 345 using .22,.033, .01,.003 and .001. At first I thought how can point .22 eve be usable but I see it in the graph and I understand that Hz is not a linear scale. So I ended up bunching mine a bit closer together for the big caps as its on a push pull to activate the inductor.
The results are very good, i don't get that thin tone a .22 cap would get, but I do get close to the .033 with the .047 and .022 using the 1.5 H inductor, the tone is in a similar ball park. I also have 0.01 which sounds very close to the videos of the 0.01. Beyond that I got .0047 which is s bit warmer and woody compared to the .003 and .0022 is a bit brighter. I also have .001 but, it's a log scale so the difference between that and 0.0022 is barely noticeable and it's both get close to wide open. I have this varitone hooked up to a tone pot and that's the tone pot for the guitar, I dial a tone by selecting a cap, my tone pot determines the roll of and push pull determines if it's an LC or RLC circuit.
I am so happy I am making 2 more varitones for 2 other guitars. I will be using the Gibson values for one and custom values for another to get a different spread.
I have not yet wrapped my brain around the maths, but I can say that the results of using 15 or 1.5 Henries but I can attest that 1.5 works just fine, it sounds like the proper varitone but mines has slightly different notch frequencies.
This has been a good discussion, I only looked to see if I could get some better ideas for spreading my capacitors but looking at the graph I have selected a good range.

[vinnie1971]

Feb 17, 2016 5:49:48 GMT -5 helmuth said:

On different websites I found three different notch frequency figures:

A) 1850 / 1050 / 650 / 350 / 140 Hz
B) 1950 / 1100 / 620 / 360 / 120 Hz
C) 1875 / 1090 / 650 / 350 / 130 Hz

I think there will not be great differences in sounds between all these.

With the capacitors (1 nF, 3.3 nF, 10 nF, 33 nF, 220 nF) the inductance be calculated. It varies between about 6 and 8 Henries. For my Varitone replicas (called "ToneVari") I use 7 Henries.

In reality does that actually translate to much sonic difference. I would be interested to hear it.
I have spare 1.5H inductors, if I want to make a 3H inductor do they need to be connected in series with each other?
I might just try a bigger one though ...

Just in this subject I know stellartone are making some unsubstantiated claims that sound like urban legend ( is they guy who did the Gibson design even alive today to validate it?) and Rothstein and others who make RLC tone controls all use the Xicon 1.5H inductor that I have been using. There are a lot of videos of these controls and they all seem to work as expected.

[perfboardpatcher]

But if 42TL021-RC's primary is 1.5H then Xicon has an awkward way of writing it down in their spec sheet.
While in that case the impedance of 4K corresponds to a frequency of 425Hz the mentioned impedance variation concerns a frequency of 1k.

Makes no sense to me.

[reTrEaD]

Feb 18, 2016 15:14:42 GMT -5 perfboardpatcher said:

But if 42TL021-RC's primary is 1.5H then Xicon has an awkward way of writing it down in their spec sheet.
While in that case the impedance of 4K corresponds to a frequency of 425Hz the mentioned impedance variation concerns a frequency of 1k.

Makes no sense to me.

Hello, Paul.

It makes no sense because you've made some assumptions that aren't valid. The inductance of a transformer isn't directly related to its impedance'.

The quoted impedance is what's referred to as 'reflected impedance'. If you load the secondary as specified, the primary will look like the rated impedance to whatever circuit is it connected to.

The impedance ratio of a transformer is the square of the turns ratio.
In theory, we should be able to take the square of the turns ratio and multiply that times the load impedance and get the primary impedance. In practice, there are losses due to internal resistance and coupling inefficiencies. But in general the calculations are close, within reasonable parameters. We certainly don't expect a transformer with a shorted secondary to look like a zero ohm load on the primary. And an unloaded secondary won't look like an infinite impedance on the primary.

Side note: I was surprised to see such a variance from theoretical calculations (using turns ratio) to actual specs on the 42TM117-RC. Several other transformers listed held much truer to calculated values.

Bottom line: Don't try to determine inductance of a transformer used as an inductor by using its rated impedance. It will lead you astray.

[sumgai]

Any other oldtimers here get shivers down their spine, as they read reTrEaD's post above?  I swear, he was channeling Unklmickey!!

(And no, he wasn't reincarnating ChrisK, or else he would've just said "Turns are, and ratio is", and left the rest of us scratching our heads....)