Idea for a dedicated Rhodes preamp, need help !

[PermO]

I have a wish for something that not really exists out there, just the way I would like it.

A dedicated Rhodes pre amp.

Some years ago I got this MK1 Rhodes piano, and since then I discovered not all DI's are made equal.
Never had any problems with bass or guitar, but the high output impedance and low level of a Rhodes piano really show the limits of many DI stages, also running it into pedals directly is not always a great succes, some sort off work, but it never sounds great.
The things I miss is this top end sparkle and details, it often results in a rather dull and flat sound.

I have build a Alembic tube DI with transformer (Edcor) balanced output that sounds great for recording my Rhodes piano, though I feel it lacks some low end punch and it does not drive pedals.
Also have a RNDI-S which does sound great, but needs a mic pre, and the trough output is not buffered so also no good for driving pedals in this case.

The Peterson preamp design is out there but also has the reputation of sounding rather dull.

I think the best thing I have heard was the DI input of the Aurora GTP1 that someone brought over, that's a sound I really like, it really brings out all the details and nuances, low end is tight and punchy, top end sparkles.
So, get a GTP1 ? .. I'd love one ! yeah but, that's slightly over budget, I don't need the mic pre and I would still have troubles driving pedals.

I think impedance is the key factor in this.

I don't know how to design a circuit like this, I only have a wish list.

Top notch sound quality like the Aurora
Clean, no EQ, no funny distortion or vibrato or anything
10M (?) High Z input
Balanced line level output
Unbalanced high Z buffered output for driving pedals
Single knob level control
No tubes, transformers or any exotic expensive parts
Small formfactor 9.5" 1 HE max, or small enough to build into (behind) Rhodes output panel.

Could it be a pedal ? .. I don't know, I've never heard anything really exiting running on 9V... So I guess a +15 / -15 powersupply would be needed for this circuit. Fet input stage ? opamp balanced line driver ?


So I wonder, do my wishes make sense ?
I'm probably not the only one who would love to have a dedicated Rhodes preamp like this.

I wouldn't know where to start, I don't have the design skills, also picking the RNDI and Aurora as an example I think I put "the design bar" quite high, I don't know if these units have something common in their designs, but I really like the sound quality of these units.

So, I'm just dropping this here and... let's see what happens.

 

[totoxraymond]

Bojan Z who is a famous piano player (at least around here) has a preamp like that. Though i thin it only outputs to his effects pedals and then we DI that to the console. (I didn't see him for a while)

I don't remember if it was a custom preamp or not...

I'll try to look if i can find more info.

Thomas

 

[rock soderstrom]

Exciting idea, I'm curious how this will develop.
I don't have a concrete circuit suggestion, but I would first think of an optimized Fender Blackface circuit with tubes.

Since you ruled that out, I think a JFet input stage simulating that would be a conceivable way to go. A very good address for this topic is runoffgroove.com , there are some circuits that could be suitable as a basis. ( Fetzer etc.)

Another idea would be to adapt the Hamptone JFet preamp, e.g. install 10M resistor and take the source capacitors out, then the circuit remains relatively long clean. After the first stage you could decouple for the pedals, with the second stage you could drive line inputs, through an impedance balancing or an IC stage.

Just some ideas for the start...

 

[PermO]

Jup, I like your ideas Rock, I was thinking in that direction.

I'm not against tubes and transformers at all, I have quite a lovely bunch here...

Transformers are cool, but they always have a sonic imprint, are rather expensive and sometimes are really hard to get, depending on where you live.

Tubes need high voltage and heaters, so, a dedicated transformer, I'd rather stay away from that also

So in this case I would like to keep it small and "economic" I also feel that if you want to bring a reasonable noisy, low level high impedance output to a workable level without breaking the bank a modern design would be the way to go.

I'd like to end up with a small PCB design using trough hole components that really anyone could build, I think almost every Rhodes / Studio owner would love to have a dedicated thingy like that.

Main focus should be on sound quality, not on "cheap" ...and I really feel this can be done, only nobody did it yet.

 

[rock soderstrom]

I'm not against tubes and transformers at all, I have quite a lovely bunch here...

True. Your Alembic DI is essentially a blackface preamp.

I agree with your perspective and I'm sure it can be done very well without tubes.

Your advantage is, you can tailor the preamp exactly to your instrument and your taste. No compromises

PS: I am sure there will be other suggestions and ideas.

 

[abbey road d enfer]

You could do worse than that...
Note the briged-T notch after the first stage, and the Fender-style tone stack that also creates a notch at about 600 Hz.

 

[PermO]

I think I found a nice output stage here; Balanced Transmitter and Receiver II

His stuff is always good !

I have build a Rod Elliott spring reverb driver and recovery amp that was also based on a 5532 and I was really impressed with the result.
So this is rather simple, unbalanced input, input impedance can easyly be set by changing the value of R102, 5532 is easyly available, no caps,... he asumes a stereo version and uses two opamps for that, in my case I can do this with just one 5532 I guess ?

I have these parts, so I won't need to order anything to breadboard this, also found a small 2x 15v~ 80mA transformer that could power this experiment.

P87B PCB version;

 

[abbey road d enfer]

Unfortunately it's the worst example here. It is known as "the Tascam problem". Stay away from it.

 

[PermO]

Ah, good to know, thanks for this advice,... I will

 

[Rossi]

I don't own a Rhodes, but I have a Hohner Pianet T. I usually record it with a super simple DI which is nothing but a single FET source follower with a 4:1 transformer (actually, a tiny Neutrik NTE4 backwards). It sounds absolutely brilliant. Simply because a) the input impedance is higher than usual (10 M) and b) I keep the cable very short (about 30 cm / 1 ft). As you may know, the cable capacitance plays a big part in the sound of (passive) magnetic pickups. The shorter the cable the higher the corner frequency of your pickup.

My super simple DI runs off a 9V battery. It has enough headroom for the Pianet T and most guitar/bass pickups. You do run into audible clipping with something like a high output distortion humbucker when you hit the strings really hard.

Obviously, a super simple DI like this won't meet all your requirements. It requires a mic preamp, and doesn't have a second buffered output. But your solution may be simpler than you think. I'd start with a FET source follower and, instead of the transformer, add a BJT emitter follower to avoid gain loss. That's very easy to breadboard. See if you like the sound and get enough level.

 

[L´Andratté]

But first of all why not the simplest form:
A FET opamp like the OPA134 with high input Z (like 10M), gain switch(shouldn´t need more than 20dB?) and enough output drive for the "pedal-out", accompanied by either a THAT balancing chip or a dual opamp in the usual "non-Tascam problem" configuration. (see D.Self´s small signal design for that)...
I love that Rhodes sound!

 

[igloo701]

It's not well known, but the Rhodes has a rather low impedance (around 1 kΩ). So it's perfectly clean with any passive DI and does not require anything special...

 

[maarja]

Unfortunately it's the worst example here. It is known as "the Tascam problem". Stay away from it.

What do you mean by "the Tascam problem?"

 

[JohnRoberts]

You could do worse than that...
Note the briged-T notch after the first stage, and the Fender-style tone stack that also creates a notch at about 600 Hz.

Back in the day there was a guy hot rodding those.. IIRC his mod was called "Dyno my Piano"... Basically a hi Z JFET preamp/buffer.

JR

 

[abbey road d enfer]

It's not well known, but the Rhodes has a rather low impedance (around 1 kΩ). So it's perfectly clean with any passive DI and does not require anything special...

Indeed.
The passive "preamp" is essentially a 10k potentiometer.
The active preamps have an impedance of about 30k.
However I think it tends to muffle treble too much. But no need for 10 Megohms, or even 1 Megohm. I believe 250k is an optimum value.
But still some sound-shaping is necessary.

 

[PermO]

Passive DI's make instruments sound like they are still in the box with the bubble wrap

So I followed Rossi's advise found some stuff on the internet and came up with this;
I thought it would be easy sourcing parts for this... well, not so... I came up with J109 and 2N5401 those were easyly available, cheap, and seem to be ok for low noise audio applications.
@ L'Andratté I Also put the OPA134 in my basket, but they did not offer any of the THAT chips.
(haven't ordered anything yet, just exploring)

I think I'll settle with a +15 / -15 V powersupply for this experiment as that will give me plenty options to wotk with any opamp design or transistorpair I guess.

So I think first I should build a nice and quiet powersupply, there's nothing to evaluate without a proper powersupply.
So first things first.

 

[Rossi]

As the pickup seems to be rather low Z, its ouput level is likely rather low. I'd start with a 9 V battery; that makes it very easy to breadboard. Also, you don't get hum when you place the box on top of your Rhodes. And you could power it from the same PSU you use for your stomp boxes.

I don't think you need a higher voltage with a gain of 1. If you do run into distortion, you could try two batteries. Battery voltage may be noisy, so don't forget to bypass with an electrolytic.

R2 and R3 seem rather small. I'd probably start with 3.3K and 2.7k, which is what Douglas Self recommends for a CFP emitter follower (with BJT input). You could actually use a BJT for an input Z of 1M. Although I'd probably stick with a FET for now and try different values for R1. See what sounds best to you.

 

[rackmonkey]

Take a look at MNATs version of the UREI 1108. PCB available through Hairball Audio for $15. Plenty of designs for 24V power supplies available. Or even 24V wall wart for a laptop computer could power it. You could probably get by with 18v for your application. Use a 1M pot i place of the LDR to adjust gain. (And you wouldn’t use the input /output transformers in your case).

MNATS 1108 Mic Pre

 

[PermO]

That's seems to be a nice pre, but not really what I am after.

The main goal is to experiment with low voltage opamp, transistor, FET stuff as I have never done that before, probably make some stupid mistakes and destroy some parts...
This does not really need to be finished as I have plenty options to record my Rhodes.
And it is all new to me, there's no tubes... so it will probably take a while before I have something satisfying.

I worked out a bit of a concept to start this with, I got a hammond box, open top, that will hold some prefboard with a basic 78xx 79xx powersupply, these are stable and pretty clean, I don't need a lot of current so that should work fine to power whatever will happen. And it will hold my input and output jacks in place, a six position Lorlin switch for gain stuff and I will have a breadboard inside to hold the circuits for easy component swapping.

Also found the THAT1646, that's a nifty little chip, I need to try that, so I bought it.
Found OPA134, some cheap JFETs .. and some very expensive ones that I did not buy.

 

[Whoops]

Make the input 10 mega and not 1 Mega like in the schematic