Fender rhodes peterson vibrato preamp


2011-03-06 2:58 am
Hello all,

I've got a 1967 peterson era rhodes preamp on hand.

Schematic: http://www.fenderrhodes.com/org/ch11/fig11-8.jpg

There is considerable hiss brought on by the circuit. Cutting the treble knob helps greatly (but no fun with treble cut!). When the vibrato circuit is active, the his is worse. Following these discoveries, I re-capped the board given the schematic values as well as those shown on this picture:


At this point I've only changed out the electrolytics. Swapping them had little effect on overall hiss, but the swap caused the vibrato to thump in the attack portion of waveform. There was no thumping before the re-cap. When I received this preamp I noticed that someone had altered the cap values. If you follow that picture, the 4x 47uf caps and 1x 16uf cap were all installed as 100uf.

I know there is a quick-fix for thumping vibrato by adding some RC networks to the oscillators, but I want to avoid this if possible. This changes the characteristics of the vibrato wave to a triangle and I think it will negatively impact the sound.

Before I go re-installing all those caps I was wondering if anyone had any input on the matter.

Last edited:
Hiss is not generally caused by filter caps. They fail and cause hum or unwanted DC leakage. Usually in a low voltage circuit like this my first suspects for hiss would be semiconductors. Changing the e-caps is probably a good idea though.

If the treble control drops the hiss, then the source is before that control, so the input stage. It only has two transistors. SCope it to see, or just replace them. Drawing says 2N3392, but most any nice low noise signal transistor should work fine. The circuit only runs on +25v, so most transistors cover that.
Low noise NPN signal transistors? My own go-to choice would probably be an MPS8097 or maybe 2N5088. And if there are any left in my drawer some MPSA18. I forget now what I used them in a lot, but I have a bunch of 2N5210 too. Those are old school numbers, there may be some more recent favorites, but I have large numbers of those in stock.

I suspect if you went to On Semi and pulled up the selector guide page for low noise TO92 you would find some other possibilities. It really isn't critical.

If transistor XYZ is quiet enough in a circuit, I see no reason to sub something else just because it has a lower noise figure. This isn't hifi.
Fender Rhodes do hiss. Always did, always will; we just didn't notice so much before digital. Its a function of being high impedance (all those series coils, like lots of guitar pickups) and inductive; the theoretical noise generated by a resistor at a given temperature goes up with the value of the resistance, and the amount of high frequencies goes up because the impedance rises with frequency.

My trick was a dynamic noise filter; a sort of noise gate with a very fast attack speed and quite fast delay, operating only on the high frequencies. It opens with the "ting" of the tine attack, and drops the hiss as soon as the sustain comes in, leaving a relatively sinusoidal decay (which is what a Rhodes produces anyway, unless you hit it hard enough for major nonlinearities in the tine/pickup position)

Using a ring of three circuit as a crossover you can easily sweep the frequency; invert the polarity of the bandpass output before mixing it with the LP and the gated HP (incidentally, if you put pots on the three signals this is a very interesting instrument equaliser).

I'm afraid I'm in the middle of moving and all my circuits and paperwork are sealed in cardboard boxes, but I'm sure noise gate circuits will have improved in the twenty years since I built it, anyway.
Nothing prevents you from trying those NTE parts, but always remember the NTE cross refernce tells you which one of THEIR transistors they think is closest to the original, the NTE parts are NOT the originals. That is why one NTE number crosses to hundreds of industry parts, and not just one.

Yes they are noisy, but I think transistors have improved in the last 40 years, and for the whole dollar a pair of the newer types would cost, it is worthwhile to try to reduce it. But I don;t disagree, a noise gate might be the ticket.
Not noise gate; that chops off all the sustain. Dynamic filter; only the high frequencies brought down, Mine was a copy of the Audio and Design Scamp module, with a LED/photocell active device, thus there was a touch of delay on the attack; I always felt it would have been better with a totally solid state solution, but (having installed the unit into the Rhodes 88) never got round to it.

And we sold on the Rhodes over ten years ago (new proprietor loved the sound, but was quite upset about me having made the unit phantom powered. Had to add an external power supply).


2011-03-06 2:58 am
I think the NTE123ap made a modest improvement. Today I noticed that one channel has more hiss and gain than the other. I can get a handful of MPSA18 so perhaps it would be wise to replace the lot?

Chris - I should mention that my preamp is built into a separate enclosure for any input source.
I think the NTE123ap made a modest improvement. Today I noticed that one channel has more hiss and gain than the other. I can get a handful of MPSA18 so perhaps it would be wise to replace the lot?

Chris - I should mention that my preamp is built into a separate enclosure for any input source.

And a very good idea, too, as long as you go straight from the RCA and avoid all the fender passive controls. I was too recently back from touring, and wanted minimum wiring, rather than maximum flexibility.

More gain? What kind of circuit makes the gain dependent on the actives rather than the passives? (I can't access the schematics page from your first post, gives me a frowny face, perhaps because my ip address isn't in the states.
The only thing you've got after the volume control is emitter followers; they shouldn't hiss (no gain) and should give same signal in or out. And, when you switch the vibrato in, you're not adding any extra components, and you're not even starting the oscillator up (I'd be ready to accept the astable multivibrator moving the ground a bit, although I'd expect LF noise rather than hiss), and you're reducing the gain of one side or the other, never boosting it. When you switch the vibrato out of circuit all you're doing is turning the little light bulbs on full. I could understand a mismatched photocell, or light bulb for that matter, unbalancing the gain, but generating hiss?