News:

Follow us on Twitter for important announcements and outage notices.

Main Menu

Rhodesology: pickup wiring modification

Started by Rob A, November 18, 2010, 12:04:39 PM

Previous topic - Next topic

Rob A

This thread made me think about this. We all know that the sacred scriptures say that the pickup wiring pattern was changed to groups of three from groups of six in the early part of the 70's, in order to get a hotter signal. So it is written.

But Jim's comment
Quote from: jim on February 26, 2006, 06:54:56 PM
in my experience it's less noisy in sixes.
really raises an important point, and a technology issue.

Namely: the decision of parallel threes made good sense in 1971 or whenever, under the effective limitations of technology at that time. But how about 2010? Don't we have access to preamps that beat the socks off the 1971 stuff? (hint: yes we do) Would it possibly make more sense today to go back to parallel six configuration?

Well, this should be measurable. Assuming my esteemed colleague from Oz is correct, we should be able to measure an improvement. And it can be pretty easily arranged in a reversible way with some croc clips and wire. I have a feeling he's right, but seeing the numbers is a must.

So I propose a Rhodesological experiment: temporarily reverse the wiring mod described in the service manual by means of croc clips and short wires, then measure the change in noise floor that results. Assuming there's a benefit, then proceed to investigate the other tone changes that come about.

But then the thought occurred to me, why does it need to be all sixes or all threes? If it's beneficial, why not bridge some of the gaps and leave others open, say to get a free bass boost or whatever tone-shaping effect you want?

sean

#1
Oh Rob, I wish you never posted this.  It cost me hours of deep thought and ten minutes of experimentation.

If the quest was for less noise, I can't imagine that re-wiring the pickup rail would be my first remedy, but....

I studied the diagrams in the service manual:    http://www.fenderrhodes.com/org/manual/ch10.html#10-4
I had to think about this for a while, and those diagrams just weren't coming to life for me.  So I started my own diagrams, which you can see as a pdf at (dead URL link, sorry).

The first page shows the pickup wiring that most folks are familiar with: groups-of-three. 

The second page shows what it would take to restore the old-fashioned groups-of-six pickups.  The shocking truth is that you cannot create groups of six without cutting some wires.  This makes it a less-than-casual experiment.

However, I realized that I can create groups-of-nine pickups without cutting any wires (page 3 of the pdf).

When I thought about nine pickups in parallel, I realized that the pickups themselves would allow some of the signal to leak back around within each group.  The pickup that is generating signal is in parallel with eight other pickups that provide a return path for some of the signal... this will cause a decrease in signal strength at the outputs!  Yikes.  This I had to test.

So I created a group of nine pickups, and listened to it.  The group-of-nine pickups is significantly quieter than the groups-of-three pickups.  But it was still strong enough to work with my guitar amp.

So I expect that groups-of-six pickups will not be as loud as groups-of-three, but still useable.  It should be no problem with a high-headroom modern preamp. 

Page 4 of the pdf file shows how the 73-note harp would have been wired in the late 1960s.  If anyone has a solid date for the transition to groups of three, please speak up.  Page 5 of the pdf file shows that the service manual forgets to mention that you would have to install six jumpers to convert an old groups-of-six harp to a groups-of-three harp.

The last page of the pdf file simply shows the two ways that the tonebars were numbered.  Early pianos (through 1973 at least) were numbered from 1 to 73, and the tonebars were not stamped with a note name, just the number.  (I left the notename in the diagram so that I would not get lost.)  Later pianos had tonebars stamped with the notename and the tonebar number.  The numbering changed from 1-73 to 8-80.  I assume that the notenames were added at the same time that the numbering changed, but I have no evidence.  Anyone know when the transitions were made?

Converting a modern harp to groups-of-six would take a full evening with the soldering iron, and is guaranteed to make a mess of the harp wiring, so it probably isn't something I will ever do.

Sean





sean


But what about the more important consideraton:  TONE!

A harp wired with pickups in groups-of-six would have less end-to-end DC resistance, and also less inductance.  This might certainly have an effect of the overall tone character of the piano.  If it allowed for a less-tubby bass, that would be great!  Maybe there would be a little more brightness in the upper midrange.

That might make the ordeal of re-wiring the harp worthwhile.  Maybe not... my guitar amp has tone controls.


I am making the assumption that the pickups from the sixties were approximately the same as the later pickups.  I believe this is a reasonable assumption.  Yep, there were green-wire pickups, longer mounting tang on early red-wire pickups, and later pickups with white tape; but the overall number of windings was probably within the same range.  I find huge variability within pickups and harps of later production anyway.


If we approximate the DC resistance of a single pickup as 180 Ohms, then the whole harp resistance on a groups-of-three 73-pickup harp should measure about 1425 Ohms.  The same harp wired as groups-of-six would be only 355.7 Ohms.

The inductance would drop in a similar proportion.

These wiring schemes are all hum-bucking, and the drastic drop in resistance and inductance might actually make a difference in EMI-induced noise. 

Sean

Rob A

Well, I have a 1970 harp that is intact and all original to my knowledge. It's wired in sixes:



Note the zig-zag pattern, and note also the insulating tape in between two layers of alternating wires.

As I was writing the original post I wondered if this may also be behind the love that people seem to show for the older generation of tines. I just skipped mentioning it because I don't want to upset people.

Emielskey

does anyone know how the ek10 is wired? series of one?

change

Cormac Long

As far as I know even CBS didn't know how it was wired :-)
Regards,
   Cormac

Forum Administrator
admin@ep-forum.com

Twitter LinkedIn

Cormac Long

Quote from: sean on November 21, 2010, 08:18:31 PM
Early pianos (through 1973 at least) were numbered from 1 to 73, and the tonebars were not stamped with a note name, just the number.  (I left the notename in the diagram so that I would not get lost.)  Later pianos had tonebars stamped with the notename and the tonebar number.  The numbering changed from 1-73 to 8-80.  I assume that the notenames were added at the same time that the numbering changed, but I have no evidence.  Anyone know when the transitions were made?

My 1975 Mk1-88 tonebars are numbered with the lower 7 as 0 and the remaining as 1-81 but do NOT have the note names. So the 88-key/73-key common numbering model did exist at some stage without note names. Of course it could been introduced with note names and later changed to just numbers.. and then changed back!.
Regards,
   Cormac

Forum Administrator
admin@ep-forum.com

Twitter LinkedIn

sean


Rob - That is one sick harp!  Does it have a Turbojet part number in the corner?  (Not that I have a list of TBJ numbers to compare it to, but....)  Well, I wonder if that criss-cross pattern was a prototype, purposeful production work, or just some 1970's psychodelic soldering experiment.  It would have made replacing pickups even more painful... nearly impossible without breaking the tape.

It works, and it bucks hum, but it would have one unique feature that might definitely affect the tone of the piano:

As a single tine swinging up and down would generate current in its pickup, the same motion of that same tine would generate a small-amplitude signal in the neighboring pickups.  Since the criss-cross wiring puts the neighboring pickups in the opposite polarity, their small-amplitude signal will partially cancel the main signal.  This effect would be subtle, and I couldn't predict if it would be nice-sounding, or objectionable, or inconsequential.

Very curious.

Anyone else own or ever seen a harp like Rob's?




Emielskey

Sean, I think the EK-10 is wired the same way as rob's example? I mean the phase canceling is the same way you described... I guess.
change

Rob A

So at this point I've done a lot of reading, refreshing my understanding the trivia i learned in school about electromagnetics and AC circuits. which, by the way, was pretty short on applications and long on theory.

Anyhow, I've arrived at one conclusion that is applicable here which I will now expound upon.

You can't really analyze the harp electronics adequately without also having a model of the input that it is connected to. So, us talking in isolation about rewiring the harp to attain various lofty goals is only a part of the story. The thing you plug it in to plays a significant role.

I'm going to represent the harp as an AC voltage source in series with some resistance, and some inductance, and in parallel with some capacitance. Recalling that it's a series/parallel array, as we group bigger groups of pickups, the DC resistance and inductance will drop, while the capacitance will increase. For smaller groups of pickups, DC resistance and inductance increase, while capacitance decreases.

So is capacitance bad? Is inductance good?

Those questions can't really have answers unless and until you know something about your next stage's input impedance. At least this is the conclusion at which I arrived. If the input impedance of your next stage is adequately high, then even significant changes in the pickup configuration won't really matter a whole lot. Because in that case the load presented to the AC voltage source by that input is small. It won't pull a significant current, so the voltage dropped by the harp subsystem is low, the signal is strong.

If the input impedance is low on the other hand, the input loads the harp, draws more current, and the effect of the capacitance and inductance on the tone (and of the resistance on the volume) is much more significant.

So, the designers of the time made a compromise that was aimed at getting the best result across a variety of input impedance situations that were likely in the real world at that time. After all, Fender knew about amps.

I'm out of time right here, but I will post shortly about the practical consequences of this conclusion.

sean


Emiel -

EK-10 service manual:  http://www.fenderrhodes.com/pdf/mark3-service-manual.pdf

According to the EK-10 service manual, they put shields between the pickups.  You can see them in the photos you linked.
I assume this was to prevent the adjacent pickups from generating enough signal to trigger ghost notes from the electronic voices.

The first page of the manual tells how the signals from the pickups were grouped into the ribbon cables:  
Ribbon cable 1 carries  signals from all E's and Fs.
Ribbon cable 2 carries  signals from all F#'s and G's.
Ribbon cable 3 carries  signals from all G#'s and A's.
Ribbon cable 4 carries  signals from all A#'s and B's.
Ribbon cable 5 carries  signals from all C's and C#'s.
Ribbon cable 6 carries  signals from all D's and D#'s.

There is a paragraph on page 11 that describes how the individual pickup signals are routed:  "The signal is sent through mixing resistors R4[01-12] to [transistor] Q4 or Q5.  The amplified signal is sent [via ribbon cable connector C/P2] to the op amps on the mixer on the namerail."  On the mixer PCB that is on the namerail, the sound from the pickups is sent through [op amp] U3A/B (page 17).  (Then on page 12... "The signal from each of the pickups is also sent to an active rectifier and amplifier..." to create the envelope for the electronic voices.)

The result of all this means that each individual pickup is connected individually.  Dead pickups do not affect any other note.

Remarkably, I think this setup is still humbucking!  If you look at the schematic on page 24, the pickup signal that is coming in to connector P1 on pin 7 and pin 1 are from adjacent pickups.  As long as these two pickups are wired with one hot signal from the front and the other hot signal from the back, you will get EMI noise cancellation.  

If you look the second photo, you will notice that the purple and grey wires feeding the ribbon connector come from adjacent pickups (the fourth and fifth pickups from the right in the photo).  These two pickups are indeed wired with opposite polarity!  Cool.

Here is the photo again:  

Sean


Rob A

Practical consequences...

So you may be asking yourself what I meant by "adequately high" impedance above. Basically, as pointed out in line, that's high enough to not load the pickup array. In practice I think that 100x the impedance of the pickup array is adequately high. Maybe less than that, but for sure 100x is adequate.

"High" impedance for our purposes let's consider to be in the kohms range, while low impedance will typically be in the hundreds of ohms.

So the DC resistance numbers for our pickup arrays as listed above (1400 ohms for the threes, 350 ohms for the sixes) are at very best on the low end of the high impedance range, or low impedance in the case of the sixes configuration.

When you hook a low impedance source (harp) to a high-impedance input (amp, channel strip, stompbox, preamp, etc.), one more thing happens that you have to worry about. The guitar cable's capacitance may become a significant factor in tone (never for the better, sadly). It damps the higher frequencies by providing a capacitive path to ground for the signal. So use the shortest and best cable you can obtain for this.

This helps explain a lot of posts from people complaining about low level and dull response from their stage pianos. I am comfortable with the assumption that those people would be users of low impedance amp inputs in general. The usual forum advice is to get a preamp, and this is sound advice, but not for the obvious reason (gain). It's sound advice because one of the preamp's classic functions is to provide a buffering stage before the amp input to avoid loading your signal source.

Anyhow, enough of this impedance techno-babble. Back to regrouping pickups, and to tone. My conclusions are that

a) regrouping pickups will probably have a negligible effect on tone for any of us driving an adequately high impedance input (a hundred kohms or better as noted above)
b) noise effects are yet to be determined, however we've learned that harp wiring varied considerably over time (maybe the wiring pattern rather than the parallel group size better explains jim's observation about noise)
c) low impedance amp inputs are problematic for stage piano owners
d) ideal cables exhibit the lowest capacitance, and in general use the shortest cord that reaches your input.

It would be interesting to jumper all the gaps and do 1 parallel group of 73 pickups. Nominal DC resistance between 2 and 3 ohms.

Rob A

Bah, humbuck.

I am wrestling with sean's use of the term here somewhat. In contrast to a single-coil pickup, a humbucker (usually) has two windings in series (so that signal adds constructively) in magnetic fields of opposite polarity (so that noise adds destructively). I'm possibly missing the point of sean's post:

Quote from: sean on November 21, 2010, 08:48:31 PMThese wiring schemes are all hum-bucking, and the drastic drop in resistance and inductance might actually make a difference in EMI-induced noise. 

It sure seems like 73 single coil pickups to me.

Rob A

It also occurs to me that I ought to cite some credible sources for some of the baloney I'm peddling above.

Rane has an amazing library online of free application notes on all kinds of audio topics. For years I've used this as a resource. In particular their note 126 is useful to us:
http://www.rane.com/note126.html
Especially the "cable as low-pass filter" heading, which covers a lot of our ground nicely.

High-level discussion of series/parallel guitar pickup wiring here:
http://www.guitarnuts.com/wiring/serpar.php

Detailed pickup physics discussion here:
http://www.buildyourguitar.com/resources/lemme/index.htm

Interesting blab about loading, impedance and cables for guitars:
http://www.harmonycentral.com/docs/DOC-1183


sean



Humbucking is guitarist slang for electro-magnetic-interference (EMI) self-cancellation.   The EMI is commonly generated by the AC mains acting as an antenna, creating hum primarily at multiples of 60Hz in the USA (50Hz in far-away lands).

In a humbucking guitar pickup, two coils are used.  The coils are assembled on separate bobbins, and then connected together such that their windings are oriented in opposite directions.  For instance, one coil is wound counter-clockwise from start to finish, and the other is wound clockwise from start to finish.  Connect both "starts" to the tip/hot, and both "finishes" to sleeve/ground. 

If these two coils are sitting side by side (or even within a few feet of each other), then an electric or magnetic field at audio frequencies will induce a similar current in both coils, but since their windings are in opposite directions, the induced current will cancel. 

In a guitar pickup, you want the movement of the string to create a current that does NOT cancel itself, so you insert the permanent magnets such that the magnets around one coil are polarized in the opposite direction than the magnets influencing the other coil.  So the flipped-polarity coil also has a flipped-polarity magnet. 

The arrangement of the permanent magnets has no influence on the noise generated or cancelled by external EMI, only the relationship between the windings of the wire determines the humbucking.



On a Rhodes, all the windings are made in the same direction on the bobbin.  To make them humbucking, they need to be connected such that half the pickups are "front-to-back" and half of them are "back-to-front".  The EMI noise generated in the coils travels in the same direction in every pickup, but half of the noise signal arrives at the front terminals, and the other half of the noise arrives at the back terminal.  In each pickup group, the noise arriving at the bus wires cancels.

Since the 73-key piano has an odd number of pickups, I guess this humbucking is not theoretically perfect.

Sean

P.S. - I doubt that the factory inserted the magnets into the pickups with any concern for north-south magnetic polarity, but I have never checked.



Emielskey

Thanks for the research on the EK-10, sean!

I am restoring a mark 5 right now, "from the ground up". So rewiring was a consideration for me. But reading the thread I think I will go with the original wiring for now..

Emiel

Quote from: sean on November 23, 2010, 01:15:42 PM

Emiel -

EK-10 service manual:  http://www.fenderrhodes.com/pdf/mark3-service-manual.pdf

According to the EK-10 service manual, they put shields between the pickups.  You can see them in the photos you linked.
I assume this was to prevent the adjacent pickups from generating enough signal to trigger ghost notes from the electronic voices.

The first page of the manual tells how the signals from the pickups were grouped into the ribbon cables:  
Ribbon cable 1 carries  signals from all E's and Fs.
Ribbon cable 2 carries  signals from all F#'s and G's.
Ribbon cable 3 carries  signals from all G#'s and A's.
Ribbon cable 4 carries  signals from all A#'s and B's.
Ribbon cable 5 carries  signals from all C's and C#'s.
Ribbon cable 6 carries  signals from all D's and D#'s.

There is a paragraph on page 11 that describes how the individual pickup signals are routed:  "The signal is sent through mixing resistors R4[01-12] to [transistor] Q4 or Q5.  The amplified signal is sent [via ribbon cable connector C/P2] to the op amps on the mixer on the namerail."  On the mixer PCB that is on the namerail, the sound from the pickups is sent through [op amp] U3A/B (page 17).  (Then on page 12... "The signal from each of the pickups is also sent to an active rectifier and amplifier..." to create the envelope for the electronic voices.)

The result of all this means that each individual pickup is connected individually.  Dead pickups do not affect any other note.

Remarkably, I think this setup is still humbucking!  If you look at the schematic on page 24, the pickup signal that is coming in to connector P1 on pin 7 and pin 1 are from adjacent pickups.  As long as these two pickups are wired with one hot signal from the front and the other hot signal from the back, you will get EMI noise cancellation.  

If you look the second photo, you will notice that the purple and grey wires feeding the ribbon connector come from adjacent pickups (the fourth and fifth pickups from the right in the photo).  These two pickups are indeed wired with opposite polarity!  Cool.

Here is the photo again:  

Sean


change

Ben Bove

On a completely different side note, note names on tonebars only showed up on Mark IIs in my recollection.  Don't know about the '79 transition point but every 1980 that I remember seeing had note names.  79 Mk1s didn't seem to.

Anyhow, thought I'd poke the fire on a highly technical thread.  :)
Retro Rentals
Vintage Music Gear

http://www.retrorentals.net
(818) 806-9606
info@retrorentals.net

FB: https://www.facebook.com/retrorentals.net/
IG: @RetroRentalsNet

Silvestre

Quote from: Ben Bove on July 07, 2011, 04:53:12 PM
On a completely different side note, note names on tonebars only showed up on Mark IIs in my recollection.  Don't know about the '79 transition point but every 1980 that I remember seeing had note names.  79 Mk1s didn't seem to.

I saw a 1979 Mark I Stage Rhodes two weeks ago that was from about the 40th week and had note names and numbers on its tonebars.

I'm going to review some photographs, to see if I had one that could serve as an example.

Silvestre

Still searching for pics of the piano I was talking about... but here there're some from another 79 Rhodes with names and numbers on the tone bars.

Silvestre

Ok... this other Rhodes is the one I saw in person 16 days ago.

Here you have a few more pics: https://www.milanuncios.com/pianos/rhodes-mark-i-stage-piano-248871751.htm

You couldn't tell from that pics, but, as I said, there were also names and numbers on that piano's tone bars

pnoboy

I previously posted about the harp's frequency response vs. the input impedance of the device it's connected to.  FWIW, the stage piano is loaded by the 10k volume pot that Fender chose.  That value definitely cuts down on the output voltage and also makes a noticeable difference in the sound of the piano.  Most guitar amps have about 1 meg input impedance, so connect the harp directly to your amp to see how different it sounds when bypassing the stage piano's controls.  You'll be a bit surprised at how different it sounds.

http://ep-forum.com/smf/index.php?topic=8960.0