the tone cap myth...

[David Schwab]

I got some flak a couple years ago on another board because I changed a cathode bypass cap on the first stage in a Sunn 200S to a 300uf, and said there was a difference between that and the stock 250uf. By the numbers there should not be, but the effect that cap had on the sound of the amp was that the bottom end was just a little bit smoother and more present, and I liked it so I stuck with it. One person in particular was using all sorts of numbers and equations to prove I was wrong, but I could hear a difference, and so could others. The numbers still said there should not be a difference, but mostly everyone agreed that there was, though some could not hear a difference in their own tests. Since we all hear differently, that isn't surprising to me.

About 20 years ago I was doing some experiments with volume pots and EMG pickups. I had just put a set of EMG-J pickups in my Ibanez SoundGear 5 string bass, and I felt the pickups sounded better with 100K pots than the 25K they come with. This was based on the 100K pan pot in the Ibanez.

So I wired up a box with a rotary switch and a few different value volume controls, and switched back and forth and listened. The pickup was wired directly to this box. 100K was brighter and livelier than 25K or 50K.

In 1995 I got to meet some guys from EMG, including Doug Marhoffer, their chief engineer. I told him my findings, and he looked at me blankly and said "that doesn't make sense. It shouldn't be." But it was.

Just recently I learned that the output impedance of EMG pickups is not very low, about 10K, so obviously 25K is too low!

But on paper maybe it looked right.

[Possum]

...

Actually Doug was their sales manager, I was there when he was hired. He wasn't there very long. I don't think he was very effective there and was really glad to leave, he went to some big company and did well there...

[Mike Sulzer]

Mike,

I'm not discounting what you are saying here, but it seems like you are ignoring the fact that others can hear a difference between the different cap values when the pot is on 10, despite that the numbers say it shouldn't happen? Don't get caught up in the age old trap that a lot of engineers get caught in where they assume the math and numbers and test equipment explains everything. There are things related to sound that we can't measure and don't completely understand, and this is probably one of them.

Greg,

One of the oldest and most ridiculous things engineers used to say was: "This amp has less than 1% distortion and so it is essentially perfect, and you cannot hear the difference between it and reality." No one says that anymore for good reasons.

But in this case we have a change in frequency response with a lot of zeros before the first significant digit. One should not believe that anyone is hearing this without a true carefully conducted double blind test. Neither the listener, the guitarist, nor the person making the switch can know which state the instrument is in.

[Mike Sulzer]

About 20 years ago I was doing some experiments with volume pots and EMG pickups. I had just put a set of EMG-J pickups in my Ibanez SoundGear 5 string bass, and I felt the pickups sounded better with 100K pots than the 25K they come with. This was based on the 100K pan pot in the Ibanez.

Putting a tone control on a passive pickup (low Q resonant circuit) and on the output of an amplifier are two different situations. If you want the normal "series pot and cap" to work on the amp, you will need to make sure the amp has a high enough output impedance; this might mean putting a resistor in series with the output of an op amp. Then a larger value volume pot will give you just a bit more volume.

Do you remember "stereo saleman trick #1"? If you want someone to think that the more expensive of two essentially identical things is better, you play it just a little bit louder while telling the customer, "of course, we are using exactly the same volume." Then the louder one sounds, well, just as you said above: "brighter and livelier".

[dai h.]

is there anything in the actual physical manifestation of the circuit such as stray capacitance or whatever that could possibly provide an explanation for an audible difference?

as an aside I couldn't really think of anything in re: to the cap type but I tried measuring a 250k pot with an LCR meter and after getting some weird readings (which seemed too high to be accurate) I got a C reading of about 20pF from the wiper to the metal body which made me think of something I remember reading in (GP?) about how one of the Lynyrd Skynrd guitarists claimed that a Bourns pot in a guitar "gave a boost" when turned down (compared to a regular pot). It seemed that perhaps this could be explained if the pot was a plastic bodied pot without the stray C of the metal bodied pot (not as much loss). Also it seemed to make sense if one wanted to create a "no load" (or no additional loading) condition, it'd be best to disconnect the signal from the tone at the earliest possible point before any stray C in the circuit(in the gtr.).

[David Schwab]

Do you remember "stereo saleman trick #1"? If you want someone to think that the more expensive of two essentially identical things is better, you play it just a little bit louder while telling the customer, "of course, we are using exactly the same volume." Then the louder one sounds, well, just as you said above: "brighter and livelier".

My brother is a photographer, and he would always look at photo exhibits that had lots of sun sets and stuff (he always shot in B&W) and he would say; "make it good, or make it red!"

In the case of the EMG's, they didn't get much louder with the 100K pots, but the top end had more extension.

[Possum]

..

Well I'm glad that all the musicians here know that changing cap values in a guitar does change the tone, I was late to coming to that conclusion, players have known this for a long time as one post mentions, oh well. Unfortunately Mike you have only posed armchair engineer critiques with no basis in actual work. When something works and is known for years, the dismissal of one guy isn't going to debunk the whole thing, SK's frequency analysis charts aren't fabrications and my recordings were done identically, carefully controlled because those recordings weren't done to prove anything about tone caps, they were done to compare parts changes in the design of those pickups, they had to be identical recording setups and guitar, even the strings are the same. I have almost 100 recordings done from that project over about 4-5 months, the tone cap change was done only last week and both recordings were made one after the other, nothing changed. Believe what you will, I know what I did, who cares what is happening, it does work.

We've had alot of engineer types post here before, but they aren't pickup makers, never made a pickup, not even sure some of them own guitars. Credentials from one field don't qualify one in other fields just because they may be related. If you want your comments to carry any weight, tell us your experience with changing cap values in an electric guitar, what you did exactly and why you think you heard no difference. the others here have done this work and have used this knowledge to fine tune the tone of their pickups and is apparently common knowledge. Just saying something isn't so because you yourself haven't done the work or tried it carries no weight. I"ve been told things by engineer types here before who think because they know something about coils, think they can "improve" pickups by sitting in their minds and telling others what to do with no actual experience in the world of pickups. Unfortunately alot of what they've told me doesn't work in the real world or would make a pickup so hi-fi and shrill no one would want it. No one needs to do double blind tests on this little cap change trick because its stupid simple and gives one immediate tonal changes in their guitar. Instead of criticizing and acting as if you are some kind of authority on guitar electronics, maybe you actually ought to try it yourself, eh?

[Joe Gwinn]

As an aside, I couldn't really think of anything in re: to the cap type but I tried measuring a 250k pot with an LCR meter and after getting some weird readings (which seemed too high to be accurate) I got a C reading of about 20pF from the wiper to the metal body ...

Which make and model of LCR meter? Too much series resistance will definitely fool most LCR meters. We first saw this effect with pickups, which have a very large resistance for their inductance.

The quick-and-dirty test is the same: Wire a relatively pure (low resistance for its capacitance or inductance) component in series with a linear 50Kohm potentiometer and measure the capacitance or inductance of the series string while varying the potentiometer setting. If the capacitance or inductance reading varies much as the series resistance is increased, the LCR meter will not be capable of accurate measurements for like combinations of resistance and inductance or capacitance.

[dai h.]

it's this thing:

http://www.aandd.co.jp/adhome/products/sp/ad5827.html

(I think it's made in Taiwan for A&D and sold elsewhere under different brands.)

Too much series resistance will definitely fool most LCR meters. We first saw this effect with pickups, which have a very large resistance for their inductance.

I figured that's probably what it was. In hindsight I should probably have bought the Extech which I found being sold (here) some months after I bought mine (around USD$250? IIRC and including the extra software). Oh well.

[Joe Gwinn]

it's this thing:

http://www.aandd.co.jp/adhome/products/sp/ad5827.html

(I think it's made in Taiwan for A&D and sold elsewhere under different brands.)

It seems to use a different base instrument than the Extech. The specsheet values given in English didn't mention allowed range of Q values (ratio of reactance divided by resistance), which is a bad sign.

I figured that's probably what it was. In hindsight I should probably have bought the Extech which I found being sold (here) some months after I bought mine (around USD$250? IIRC and including the extra software). Oh well.

Well, I would make the test, and see what its limits are. Nothing to lose.

[salvarsan]

EMG op amps

In the case of the EMG's, they didn't get much louder with the 100K pots, but the top end had more extension.

The older EMG actives used 1st generation FET opamps, particularly the pre-RoHS LF442.

Overview:
http://www.national.com/mpf/LF/LF442.html
Datasheet (PDF):
http://www.national.com/ds/LF/LF442.pdf

It's has relatively low power consumption for good battery life but barely
enough drive for the recommended 10K load on them.

100K pots would seem to have been the trick, that and running them at 18 or 27 volts.

If battery life were less important, they could use opamps with better distortion
and drive figures; then you'd hear that pickup open up. It's also about money
since good audio opamps are expensive.

For comparison, the Burr-Brown OPA2604 that sounds good in audio use can
drain a 9volt battery in ~32 hours while the LF442 lasts at least 10 times that.
The respective prices are $2.40 and $0.59 in 1000 quantity.

-drh

[Mike Sulzer]

Unfortunately Mike you have only posed armchair engineer critiques with no basis in actual work.

Would you like to flesh out that statement with some facts? It is better to keep these discussions professional, leaving out the personal comments. If you think I am not doing that, tell me which of my statements are the problem.

SK's plots: The web page that he linked to has results for .047 caps. I did not see any comparison to .022 caps. He compared different types of .047 caps. My understanding of his verbal summary is that he could not hear much difference on 10 or 0. I see nothing in those plots that shows that a .022 and a .047 cap sound different on 10. His plots do not support your case.

My pickup experience: made my first pickups when I was in my teens in the early 60s. I do not do pickups professionally; I am a scientist and engineer; electric guitars are my hobby.

Before this discussion I had not thought much about the difference between the sound of .022 and .047 on 10. It never seemed worth thinking about given the huge isolation from the series resistor. I did try some casual testing last night. I pulled the vol pot out of a guitar and put a fixed resistor in the circuit. Put a switch in the hole and connected a .022 and a .047. Conclusion from last nights casual tests: I cannot hear the difference with a 500K resistor. Initial conclusion from last night: wow, listen to that difference! Why? After a while, I noticed that when I returned to picking after switching to one position, it was not quite the same position I used after switching the other way. Bias is very hard to avoid without a careful procedure.

When I get some time, I will replace that switch with a mini stereo jack. A bunch of plugs are needed, wired to connect either the tip or the sleeve to ground. Put the plugs into a plastic cup; shake, not stir. Grab one, record a numbered file, and put the plug in the same numbered bin of an egg carton. After doing a bunch, listen to the files for differences; then check the plugs to see which is which and record the results. Or send the files to someone else, but do not send then list until after they listen.

[Steve A.]

... Before this discussion I had not thought much about the difference between the sound of .022 and .047 on 10. It never seemed worth thinking about given the huge isolation from the series resistor...

Mike:

You keep insisting that the series resistance of a 250k to 500k tone pot wired as a variable resistor is "huge"- so large that the value and characteristics of the capacitor becomes insignificant.

With a pot wired as a 3 terminal potentiometer, it is the resistance to ground that reduces the signal level, not the series resistance. When you add in the third connection (to ground) you then have a voltage divider of sorts.

I had experimented with a 1M pot wired as a variable series resistor right after the input jack of a guitar amp about 10 years ago. I had expected to use it to fine-tune the strength of the signal from the guitar, but in practice it actually acted as a subtle tone control, shaving off some of the high frequencies when you set it all of the way up to 1M.

So in the common vernacular, a 250k to 500k series resistance to a passive pickup "ain't shit".

You have also suggested that you would not hear the difference between caps of different values if the tone control was set to 0. You obviously have not tried that since there is a very distinct audible difference going from 0.047uF to 0.022uF to 0.015uF (which Eric Clapton reportedly used in his LP's when playing with Cream to get that "woman" tone). With the tone control set to 10 the difference is more subtle, but when set to 0 the difference is very plain.

One more point: you are going to hear more of these subtle differences if you are using handwound boutique pickups. Most of the pickups from Seymour Duncan are over-potted, which reduces microphonics but it also reduces tone. You might say that they are the McDonalds of the pickup industry- the 800 pound gorilla- but nowhere near the quality of a fine 4 star restaurant.

As for your suggested test rig, when the rest of the tone caps I have ordered get here I plan to make a switch box that plugs in between the guitar and the amp, with a rotary switch to select between maybe 6 different caps and perhaps a switch to select between a 250k and 500k tone pot (hint: use a double ganged 250k pot and use a switch to add in the second gang for the 500k setting!)

As mentioned before it would be necessary to disconnect the tone cap(s) in your guitar by desoldering one of the leads. And this tone control would be global, unlike a LP, with normal or 50's wiring. So you would lose some of those subtleties.

For a totally random double-blind test you could remove the stop which would allow the switch to rotate continuously. Spin the knob around and see if you can guess the value!

But double-bind tests are not always necessary. Professional guitarists make equipment choices every day without having to make endless comparisons. Sometimes it is very obvious to them which guitar or box or amp sounds better- they are real pros and that is their field of expertise. Just like the professional winders on this forum, who happen to sell most of their pickups to professional guitarists- so you know that they have to sound good!

Steve Ahola (the professional amateur!)

[SteveJ]

But double-bind tests are not always necessary. Professional guitarists make equipment choices every day without having to make endless comparisons. Steve Ahola (the professional amateur!)

I agree with that Steve. You just have to listen. You have to train your self to hear the small differences and try not to be fooled by things. I went to a friends house a few weeks ago and another friend of ours had dropped of some gear to check out. There was a 2061X Marshall cab there and the guy was really digging it. He was ready to sell his cab. What was fooling him was that the Celestions were more efficient and therefore louder than his cab so he thought it was better. I sat him down and pointed out the differences and his cab actually sounded warmer and had a better break up. Point is that it's easy to be tricked but if you really listen you can pick out the differences. Like someone said earlier, the math and numbers don't always explain it all. I trust my ears and I have a few things laying around that were "the best" or "flavor of the week" that I don't really like.

[TD_Madden]

Absolutely. I can HEAR the difference.