Rhodes

Junk Rhodes Contain Lots of Valuable Parts

Above are some Rhodes pianos being stripped of useful parts. This photo was taken just two or so months after I started to learn Rhodes repair. I was determined to clean up the Rhodes area by breaking down the "parts" units into actual parts and tossing out the cases. I eventually succeeded in this, but it took a few years!

There is already quite a wealth of info out there about servicing Rhodes. If I can think of some things that are worth reporting, I'll enter them here.

I do highly recommend the complete grommet and screw change for the tone bar assemblies. This really improves the your voicing work, both in overall timbre and in stability. It was always a bummer to get a phone call or email that the customer drove the Rhodes all the way home just to find 1 key out of whack from the transportation. Ever since I've started recommending the grommet changes, I really haven't had that issue any more.

This was an odd job where I was replacing the majority of the pickups, so I actually removed the pickup board from the harp. It is not uncommon for large numbers of pickups to die. The tiny wires can corrode and just turn to dust. If the Rhodes lived in a corrosive environment (salty sea air comes to mind) it might be a candidate for massive pickup failure. Check for greenish colors where the coils start and stop. If you get a break on the coil's end, you can usually just unwind the coil a little and reattach it. If the break is at the coil start, you have to rewind it to save THAT coil. Usually I just replace it since Rhodes pickups are easy to get.

Rhodes Fifty Four

The compact Fifty Four is my favorite amongst the models. Pretty rare to see one.

Keybed Repair

Ouch. A busted keybed. The symptom was hammers sticking up in the middle. This is from stuff busting through the bottom of the case. The screw for the legs can cause damage like this. I think it was much worse looking at first. This is probably after I got the wood pieces back together. The far right is how disassembled the Rhodes has to be to work on this part.

Suitcase Breakout Adapters

I've made some breakout boxes for the 4 pin and 5 pin Rhodes preamps. These allow me to service the keyboard without the suitcase. The 4 pin has a 2.1mm dc jack that I supply +25V from my bench power supply through. The 5 pin has banana jacks for +/-15V connection from the same bench power supply.

Currently the 2 signals are just broken out to 2 1/4" jacks. It would also be nice to sum them to 1 jack, perhaps a switch to toggle between: 1 <-> both <-> 2.

These are passive devices. Powered versions would be nice to make one day, but the bench supply does the job so well (including displaying the current draw) that I probably never will.

Volume Pot Oddity

I came across a stock looking Stage 73 Mk I that was in for a basic tune up. In my thorough examination, I noticed that both controls were the same CTS 50k reverse audio pots, and they had identical date codes: 137-7406, which translates to "CTS made this during the 6th week of 1974."

I am tempted to believe this was a factory error since the two pots are near identical and did not appear to have been touched since installation.

What is the effect of changing the pot value?

The volume pot should be a 10k audio (regular, not reverse) pot. So what is the effect of putting a 50k reverse audio in its place?

One effect is that full output is reached around "4" on the 0-10 legend. From "4-10" there is no perceivable change in volume. From "0-4" there is a rapid change in volume. Both the reverse audio taper and the increase in total resistance aid in creating this effect. A normal audio taper should spread out the perceivable change in volume to be fairly even throughout the 0-10 rotation. Increasing the maximum resistance in this configuration can also ruin the rotation.

A second effect is that the filter equation for the "Bass Boost" control shifts downward, and thus rendered the "Bass" control useless. To verify this, I used a resistor substitution box and placed 12.5k across the volume pot (12.5k || 50k = 10k). As predicted, the bass control worked again because shifting the output load impedance downward shifts the filter center upward. The responsible party there is the 47nF (or 0.047µF, if you prefer) capacitor. With a 10k load, the normal high pass cut off frequency is 338.6Hz. With the higher 50k load, this shifts downward to a cut off frequency of 67.7Hz. At that frequency, only the lowest notes can be affected by the filter, and even then it wasn't much of an effect.

Of course, if you don't use the controls, and just want maximum output, maximum bass, then the odd ball pot doesn't matter and doesn't need to be changed. The only problem here is difficulty in turning down the volume or bass.

Rhodes Stage Schematics

There doesn't seem to be an official schematic for the Rhodes Stage model's passive controls. Above are the notes I have on them. Note that it switches from a low pass filter on the early models to a high pass filter on the later models. The value of the capacitor, the value and taper of the tone/bass pot, and the connections, all change between the two designs. The volume control is setup the same way on both. The early ones will have very large (physically) 1µF caps. Often they are color coded. The bands, from the top, will be brown, then black, then green, and it doesn't matter beyond that (those bands indicate tolerance and max voltage which don't matter here).

They sound completely different when using the tone controls. The early version is just like a guitar tone control. All the high end gets cut, leaving the overall volume of the upper registers severely diminished. In the low regions, you get a soft and round bass timbre. Just like a guitar pickup, the Rhodes uses wire coils wrapped around magnets to capture the sound of the tines. These pickups are natural low pass filters by themselves, so a precise calculation of the filter cutoff is not simple to calculate (you need the specs, especially the inductance and output impedance, of the total pickup system). A Stage 88 from 1973 I had on hand measured 854mH, and 2.1kΩ dc resistance. Using this handy RLC calculator, I got a cutoff frequency of 172Hz, which is between the E and F below middle C. So everything near and above those notes gets dulled and diminished, and any harmonics from the bass register is subject to the filtering too. (Extra note: using the pickup specs above, the coil's inductance and the 10k volume pot create a low pass filter by themselves with a cutoff of 1.8kHz. The last 13 treble pickups, to avoid this issue, are connected in parallel to the rest, with a series 4.7n capacitor. Ah, but the tone control in parallel with the volume means there's 5k at the end... that shifts the low pass cutoff to below 1k... yikes. If you have this early version, consider lifting the ground connection on the tone pot, if it's there.)

The later version is designed as a high pass filter. If the pickup specs are ignored, the RC cutoff frequency is about 338Hz, which is between the E and F above middle C (exact measurements of the Rs, Cs, and Ls in the real circuit may shift this up or down a little, or a lot, see previous entry "Wrong Volume Pot"). As the Bass Boost control (as the later tone control is called) gets turned down, the lower half of the keyboard loses its deep bass, but will retain the upper harmonics so you can still hear the bass notes, just with a "thin" timbre that can be an interesting effect. If you add in the pickup's natural low pass filter to the equation, the overall RLC circuit is a bandpass filter centered right in the midrange.