ETI 4600 Owner Jeff Jones' Page
|Loading Photo & caption.|
I first became
interested in synthesizers when a friend loaned me a record of
Pictures at an Exhibition performed by Isao Tomita. I was amazed at
the range of sounds that were generated electronically. I then
started looking into synthesizer construction and found the
Practical Electronics Minisonic project. This was a simple and
affordable synthesizer to build and so undertook the construction.
The synthesizer worked and all the basic modules of a synthesizer
were there but it was poor at holding tune - it could go out of tune
by the time I had finished playing a song. Then Maplin started
advertising the kit for the ETI3600 and 4600 synthesizers and after
a lot of thought I decided to order one at a cost of £460. I think
the year was 1976 and the synthesizer took about 6 months to
I had a number of problems with the initial project, many I found were to do with errors in the original construction manual. Fortunately a letter appeared in the Maplin magazine written by someone who was also building the synth and asking to share experiences in the construction. About six people replied and all correspondance was circulated around the group (I still have some of the original letters). This proved invaluable in sorting out problems and completing the project. Some of the problems which I had with the original construction were:
Keyboard Controller - unable to tune to correct frequency, found out that series resistors were putting adjustment potentiometers out of operating range, replacement with different value resistors solved problem. Key Direct and Key Modulate outputs would not track together. Did not find a solution to this with original design and so just used Key Direct to control oscillators.
Keyboard - key bounce sometimes generated random notes and it had to be played carefully ie lifting off one note before hitting the next to ensure each key sounded.
Power Supply - occasionally tripped out, especially when bass on equalizer was turned up. Did not find a solution to this with original design and so did not turn bass up!
Many modules did not work first time - in nearly every case this was due to a blown CMOS which are very sensitive to static electricity and failed despite using earth straps to protect them. Fortunately sockets were used for them and simply replacing them usually solved the problem.
Audio breakthrough - when the VCA's are fully attenuating there should be no sound on the output, but rich harmonic sounds could be clearly heard when it should have been silent. The original manual specified normal hookup wire to connect the audio signal paths - it should have been screened wire. This was specified in the later 5600.
VCF - no resonance control.
Most amusing problem - I was playing one evening and after I had finished I heard exactly what I had been playing coming from next door - he had been able to pick up the sound on his FM radio so recorded it and played it back to me! It took a lot of work to solve this one but I eventually found that a small capacitor across the power supply pins of IC5 of each oscillator stopped it transmitting (and stopped my performances being broadcast).
Once constructed and debugged the synth has been absolutely superb. It has been totally reliable and has not gone wrong once despite being stored in the garage for five years. Also the tuning has been incredibly stable (a rare phenomenon for synthesizers). I was able to start a multitrack recording, forget about it, recommence it three months afterwards without needing to alter the tuning. Additionally it stays in tune no matter how long it is turned on. I replaced one VCF with the 3600 VCF (a common modification I understand) to obtain a resonance control. This dramatically improved the range of sounds available.
After a number of years Maplin brought out the 3800 and 5600 synths. I decided to use circuits from the 5600 to overcome some of the remaining problems. I therefore purchased the parts for the binary keyboard controller and power supply kits. Kids and wife got in the way of construction of these modules and it was about 10 years later that they were constructed and installed. The binary keyboard controller had none of the problems of the original and the changeover keyboard contacts make it easy to play. The power supply is very stable.
The only other things I have done to improve the sound is to attach a Zoom sound multiprocessor to the output which reinforces the sounds superbly, and a graphic equalizer. There are a few small modifications which I may carry out in future:
phase locking the oscillators
adding voltage controlled mark space ratio to the square waves
construction of an LFO
I will keep readers of this web site notified of developments and look forward to sharing ideas with them.