ARP ProSoloist analog monophonic synthesizer

main-pic

Last Update 03-08-2015


I usually prefer tweakable synthesizers with a full panel of knobs - so how did this preset organ accessory sneak into my arsenal?  A musician friend offered it to me for cheap (this was before the analog synth revival of the 1990s).  When I played it I was not terribly excited about the static sound of a single oscillator and very little tweakability.  I'm a Moog diehard but for some reason I liked the sound of this deceptively simple preset synthesizer.  Personally I avoid ARPs because the slidepots on the control panel have a history of going bad and there are no new replacements available (almost every website that offers them are actually scavenged from ARP carcasses and refurbished but have a limited life as they are NOT new).  I didn't have an ARP in my collection, the slidepots all worked, so I bought it.

Mind you, I was a late bloomer on progressive rock and did not know the association to any keyboard artists such as Tony Banks of Genesis.  The fuzz guitars, reeds (bassoon, oboe, etc), trumpets, trombone, and violin/cello are very nice for an analog synth. The traditional sounds have an intriguing organic timbre to them.  Since it only plays one note at a time I don't find much use for monophonic piano/banjo/harpsichord/acoustic/steel guitar presets.  What won me over was the pressure touch sensor - great controller!  A flick of the switch changed the sound instantly - remember that this was well before patch programmability became a reality.  Back then the keyboardist had to manually change the synthesizer controls to change the sound, which took some time (especially on a dark stage) and it was not likely to be the exact same sound from show to show.  ARP targeted the ProSoloist to the musician who wanted instant sound changes and did not want to learn analog synthesis.  There are thirty presets - bassoon, english horn, oboe, clarinet, flute, tuba, trombone, french horn, trumpet, cello, violin, bass, piano, banjo, fuzz guitar 1, buzz bassoon, sax, space reed, telstar, song whistle, noze, pulsar, comic wow, mute trumpet, steel guitar, harpsichord, space bass, steel drum, country guitar, and fuzz guitar 2.  Some of the presets are goofy sounds (comic wow, noze, telstar) that were popular in that era.

The ProSoloist was an upgraded product for the Soloist which had half the presets and largely analog preset control.  The Soloist was produced only from 1971 to 1972 and was originally marketed as an organ accessory with its small low profile case designed to sit on top of an organ console and its controls on the front of the instrument under the keys.  The Soloist was unstable and fragile with none of the ROM digital preset control/logic switching of its later derivative, and all the filters are modules encased in rock hard epoxy.  It was drifted out of tune - in fact a Steely Dan song was cut using the Soloist and every time Donald Fagen was stacking multiple tracks using the ARP he had to keep re-tuning it and re-playing the parts because its pitch drifted too much.  This infuriated him so much that when he finally completed the multi-tracking, he threw his Soloist down the stairwell and took out his rage by jumping up and down on it.  The band completed the destruction by spreading an accelerant on it and setting it on fire.  The charred unrecognizable remnants still remain as a decoration in that studio.  Probably the sole module worth salvaging from the Soloist (but not later models) is the 4012 "moog ladder" filter module, made famous in the ARP 2600 until Moog slammed the patent infringement hammer forcing ARP to abandon the 4012 module.

Steely Dan aside, no doubt ARP had been hearing from organ customers about the unreliable Soloist.  ARP also realized that the product had to have a complete makeover to appeal to the professional musicians.  The ProSoloist replaced much of the analog preset control with digital control, expanded the presets, and offered much more stable and reliable tuning.  Many keyboard artists of the 1970s embraced the quick-change-artist ProSoloist - Tony Banks, Vangelis, Jon Lord, Patrick Moraz, Steve Walsh (Kansas), Joe Zawinul, Gary Numan, Tangerine Dream, Renaissance, and Wings are just a few.  The engineer in me (not the kind that drives the choo-choo) was curious about this so I got a hold of Tim Smith (RIP) who was a certified ARP technician and acquired a service manual as well as an extra set of slidepots.

The architecture of the ProSoloist is fascinating.  It is the usual VCO->VCF->VCA architecture but provides some processing and routing not seen in non-modular analog synthesizers (and unusual for a preset synth).  The VCO generates a ramp and core pulse waves with widths of 1/64, 1/9, 1/14, 2/11, 1/2 (square) or any combination thereof.  Filtering is not limited to just a lowpass VCF - the ProSoloist has banks of fixed bandpass ("resonators" in ARP speak) and highpass filters.  You can use the fixed filters in series or in parallel with the VCF.  The ramp and/or pulse wave can be processed through any of four highpass filters which then go to the VCF.  The pulse wave can be routed through any of ten bandpass filters which can be routed through the VCF or directly to the VCA.  The ProSoloist packs an AR and ADSR that can be routed to either/both VCF and VCA.  This is a pretty interesting system that has its obvious origins in the ARP 2500 and 2600 modular synthesizers.  I even adopted some of its techniques to my sound design skills.

Could a panel of controls be hacked into this thing?  Sure, but you will not have continuous variable control and are limited to the VERY discrete settings.  For instance, the attack times for the envelope generator is set by fixed resistors which are enabled using the open collector outputs of the ROM ICs.  Back then ROMs were very expensive and only available in low densities, which limited the number of control lines.  As a consequence there are only 3-4 discrete attack times available.  Likewise for all other control parameters.  The effort just wouldn't be worth the very limited gains.

I later learned from Tim Smith that the 4034 VCF filter module is a 24dB/oct transistor ladder configuration, copied from the patented Moog ladder filters.  That explained why its sound appealed to me.  When ARP got a nastygram from Moog's patent lawyers, they replaced all their infringing filter modules (4012, 4034, 4035) with their own designs.  The Pro/DGX model 2730 and 2733 with all metal case, orange pushbuttons, and orange panel graphics replaced the 4034 with the 4075 VCF module which was ARP's own design.  They are claimed to not sound as good due to the lack of the transistor ladder VCF, but those 4075 filters did contain a circuit error that limited their maximum frequency cutoff to a dull 12Khz, and the correction can be found on the internet.  Note that the transition Pro/DGX model 2720 with wood endcaps, white/green/red panel graphics, and black pushbuttons replacing the toggle switches on the ProSoloist did use the 4034 module. The Pro/DGX is the exact same architecture and preset library.  A year after the model 2733 Pro/DGX appeared, the Prophet-5 was released boasting complete variable programmability which made the preset synthesizer obsolete.  But the Pro/DGX soldiered on until ARP was liquidated in 1981.

ARP employed some real maverick designers.  The ProSoloist was designed by Jeremy Hill and used many novel circuits that - save the Pro/DGX - were never used in another ARP product, or any analog synth for that matter.  The US patent 3,930,429 is a fascinating study of this system (and a good supplement to the service manual).  One part of the pitch stability solution is the keyboard controller which is a monophonic digital scanning system, not far from the Emu/Oberheim polyphonic scanning system that became the industry de facto standard (if only...).  The scanning system generated digital words that described the pressed key, eliminating the analog drift of the usual current driver/resistor string circuit.  This digital word is a binary coding numerical system (BCD) with an octave code of two bits to indicate the octave (four octaves to cover the 37 note range) and a key code of four bits to indicate any of 12 keys within the octave (C, C#, D, D#,... B).  The key codes drive a discrete DAC whose R-2R string is comprised of a custom resistor array in an integrated package.  The octave code also drives the DAC but only as an error compensation for the oscillator.  Thus the DAC only provides the pitch CV for a one octave range for the 12 keys in the octave, from ~5V to ~9V.

Many people see this DAC and automatically class the ProSoloist as a DCO synth.  It is not surprising because there is so much confusion over this VCO.  In actuality, the oscillator is a charged cap core with a linear voltage to exponential current converter controlling the charge time of the cap and the frequency of the oscillator.  It's the same core system in other VCOs!  The pitch is continuously variable (this does include pitch bend and vibrato) therefore the ProSoloist is a VCO synth.  The oscillator include a glide processor (log response) that is continuously smooth.

But it deviates from the usual design from here.  The VCO is essentially a closed loop error correction system implementing an error correction block with the DAC CV on one input and the output driving the oscillator, whose output is driving an error detection system consisting of a frequency-to-voltage converter and a pulse train integrator generating the error voltage to the other input of the error correction block which completes the closed loop.  From classic opamp control loop theory, the error correction block constantly adjusts the output (and the oscillator frequency) to keep the voltages at the inputs the same, and the RC component of the lead/lag network in the integrator is selected for optimum settling time and stability of the closed loop.  It is very rare to see advanced control system theory applied to an analog synthesizer in a closed loop system this sophisicated.  I wasn't kidding when I said that ARP had some real maverick designers on their staff.

The oscillator has a four octave range (three octave keyboard with high C comprising the sole key in the fourth octave) and actually generates a frequency that is 7-8 times higher than the fundamental pitch.  The reason for this will be explained later in the waveshaping circuit.  The error detection system only needs to operate over a one octave range, which makes the circuit much simpler and more reliable.  This is the same reason why the DAC only covers a one octave range. So how does it generate the four octave range of the oscillator...?!?  The answer is in the octave bits from the keyboard scanner, which also control a set of divide-by-two blocks in the error detection system.  As you navigate between different octaves, the dividers divide the oscillator square wave TTL signal by two (which is halving the frequency) or by four or by eight depending on the octave.  This divided signal drives the frequency-to voltage converter, thus forcing the error detection system to generate the error voltage necessary for the correct oscillator frequency.  That's how the DAC and error detection system operate over a one octave range.  It's an oddball but very novel circuit that works REALLY well.  So well that the VCO is rock steady in tune on cold power up (no warm up time required), and I never had to adjust the master tuning since I acquired it in 1994!!!

The primary reason why conventional MIDI-to-CV converters will not work on the ProSoloist is because the circuit is not the tradition CV input architecture.  There's only two analog summing points where the CV from the converter could be applied and neither will work because the converter is missing the octave control bits that are crucial to the operation of the closed loop system.  If one wanted to make the ProSoloist a MIDI controller, there is a -1volt/oct source tapped from the oscillator for VCF keyboard tracking...

Now we move to the waveshaping circuits.  The oscillator output is a TTL-compatible square wave generating a frequency that is 7-8 times higher than the fundamental frequency of the audio sound.  This TTL signal goes through a set of divide-by-two dividers for the manual octave control on the left hand panel of the synth.  There's another set of dividers for preset octave control, because you want a different range with a bass preset versus a flute.  Then the final signal is fed into a serial configuration of six dividers whose 2n divisions are summed to produce a 64 step sawtooth waveform.  The divided signals also are logically summed to generate core pulse widths of 1/64, 1/9, 1/14, 2/11, and 1/2 (square).  While any combination of these pulse widths can be summed for complex waveshapes, the pulse width is not continuously variable.  There is a dynamic pulse generator (used in the fuzz guitars, sax, and bass presets) that is discrete PWM with the ADSR modulating a traditional ramp-to-pulse converter and the discrete stepping is slightly audible in the fuzz guitar presets but it never stopped the prog rockers from exploiting the preset.  The sawtooth and pulse waveforms provide the basic signals to accommodate the preset voices.

The con of this VCO design is that it uses some customized components (the R-2R resistor integrated package for the DAC), PWM is not continuously variable, is more part intensive than most VCO circuits, and a more complex troubleshooting effort for the tech who doesn't have a complete understanding of the circuit (you need the patent for the full description).  The pros are 1) no $$$ tempco resistor, 2) rock steady in tune at power up, and 3) extremely stable pitch which is a very important feature for the market this synth was targeted at.

There's only two ways to retrofit this synth with MIDI: either use solid state relays across all 37 keyboard contacts, or convert the MIDI note on message to separate key codes and octave codes compatible with the VCO system along with a gate/trigger generator.  The former would be easier.

We've already covered the filter processing.  There are fifteen preset select switches above the keyboard to select the sound.  But the ProSoloist has thirty presets...  Ah, but notice that each preset switch has a label above (upper) and below (lower).  To the far right is another switch that selects the sound associated with the upper or lower label.  Now, this system is less than intuitive.  Because they are toggle switches, the far most right switch that is in the DOWN position is the preset that is active.  The toggle switches can get dirty contacts with age and this can make the preset selection system malfunction, which can result in an embarassing performance when the spotlight is pointed at you for your fuzz guitar solo and you are playing the comic wow preset (oops).  The Pro/DGX replaced the toggle switches with momentary buttons and LEDs which made it a little more user friendly.

So how does the ProSoloist change sounds so fast?  Sprinkled throughout the circuits are ROM ICs which use open collector outputs to enable and disable parameter controls for the VCO, the filter processing system, the audio routing, the EG control routing, even the delayed LFO with auto-vibrato.  The preset selector circuit generates an address which is simply tied directly to the address pins of the ROMs.  The concept of open collector outputs is when they are enabled they can short a control point to ground, and when they are disabled they are a high impedance open circuit which can be tied to a voltage other than the Vcc pin (within reason).  For example, this allows the ROMs to mute an audio routing by shorting it to ground, and when the ROM output pin is off the high impedance ensures that the audio routing is not loaded down.

Probably the nicest feature of the ProSoloist is the expression system.  There is a pressure touch sensor under the keyboard and a set of switches on the left of the panel to select what the pressure signal controls.  The options are pitch bend (up only), wow (which sweeps VCF cutoff with resonance at maximum), brilliance (VCF cutoff), volume, vibrato, and "growl" (some would argue it sounds more like "gargle").  "Growl" is effective on brass presets, especially trombone.  Fading in vibrato with the touch sensor is very cool.  A slidepot varies the sensitivity of the sensor and there is a nice deadband built in so that fast aggressive playing doesn't trigger the aftertouch.  Also included is a "repeat" feature that re-triggers the EG at the rate of the LFO, useful for the percussive presets (and handy for troubleshooting faults!).  The expression system is the best feature of the ProSoloist, and it is the only axe where I can configure a manual vibrato using pitch bend and volume only by varying pressure with my finger.  I cannot duplicate that expression on any of my other gear with pressure sensors.  Some of the presets (like the cello/violin) have delayed auto-vibrato.

If you ever have to service one of these, some words of caution.  Be REAL CAREFUL removing the bottom cover, the keyboard assembly, and some circuit boards as there are hidden washers used as spacers in some mounting bolts.  It is a royal PITA to re-assemble some of these washers without having them fall inside, and the last thing you want in an electronic device is a loose metal washer!  The circuit boards are interconnected with ribbon cables as well as some single wire points so disassembly or troubleshooting a single board can be a challenge.  If you are removing components (none of the ICs are socketed) it is very easy to lift PC board traces (Grrr...).  I had a fault with the dynamic pulse waveshaper in my unit, and with several potential failure points I made d@mn sure the suspect circuits were working to minimize the risk of lifting PC board traces.  These things are not service friendly!  The brilliance slidepot in my unit ultimately died - I tried to restore it but the resistive element went bad and it was unsalvageable.  Since there are no new replacements available in low quantities (they have to be an EXACT FIT in the PC board holes) and the other slidepots would not be far behind, I replaced ALL the slidepots with rotary pots. One of the wood endcheeks separated from the metal frame which required new mounting brackets and new screw holes.  Also try to avoid removing any locktited nuts from integrated bolts as they may shear the bolt right off the panel.

With the right effects - especially digital reverb - the orchestral sounds of the ProSoloist can sound real authentic.  These things are not just for prog rock.  While the presets are a mixed bag of traditional and goofy sounds, you don't buy a ProSoloist for tweaking your own sounds - other than VCF cutoff there is no variability at all.  It's a preset synthesizer.  This limitation and the lack of any MIDI retrofit has kept the ProSoloist from reaching the vintage price craze of today's market.  But back in the heyday of analog synthesizers the ProSoloist was enough of a contender that it spawned me-too products like the Moog Satellite/Minitmoog (read the spelling, NOT Minimoog), Yamaha SY-1 and SY-2, and similar preset synths from Roland and Korg.  But none of them had the expression or sound of the ProSoloist and it remains the only preset synthesizer with "vintage" status.

contact info

Home