Theo Verelst Audio Page

Lately I've been able to start building some of the equipment I had in mind,  to find out what works and what doesn't (thus far everything according to my thoughts), and to present working prototypes to generate funding and/or further business activities.

This page has been visited times

Data Sheets

These sheets give an idea of the applied power amp part, pretty good figures (definately for a PA system). sheet 1 sheet 2 sheet 3 sheet 4


The basement of gallery 'Donkersloot' in Amsterdam has been experiencing the creation of the following seemingly unorderly collection of electronics parts:

Another picture showing a broadband only, open baffle speaker design and a cardboard/mbf experimental frontplate with breadboards containing an oscilator and lin-log converters:

In fact an additional tweeter is shown (the little cone in the broadband has an irritably directional high response), the broadband is in the same position as in the other pictures. In fact the bass response was quite acceptable (30Hz resonance), albeith not very tight.

19 inch rack

Below a picture of a 3 unit  19' rack (about $50) prepared for power Amp (Torroid supply tranformer, rack-cooled rectifier, heat sinks, large jack parts, volume control) and basic synthesizer (9 pots, 2x12V supply), and more prototype circuits (space for various vertically mounted Euro-cards with double sided accessability).

The rack now hold space space 10 pots with enough room to access them comfortably without twisted fingers, and obviously has a power switch. Currently, the pots are screwed directly in the front panel, which requires knobs with a rim that covers the nut, which I currently didn't afford. A alu U-profile behind the frontplate could solve this as well. The advantage of this design is that everything excpet the Euro power cord connector only need simple round holes, so produciton with low-cost tools is fairly easy.

Power amp

I've made various proto's with varying power rating and quality (basically some TIP31/32 based circuits), and by far the most effective optioin is the TDA7294 that I've applied here. About the size of two or three match boxes and a heat sink gives about 80Watt RMS continuous power with very decent distortion figures, and this is limited by the supply voltage in this case, because the IC can consume about 50Watts (given sufficient cooling), allowing about 150W continuous sine power, theoretically. In bridge, 200 Watts with very nice figures should be very feasible, I consider this for in-cabinet driving of the woofer in the speaker described here, as part of a three way actively seperated powered monitor.

This amp costs about $8 + maybe another $5 for some trivial parts, and a heat sink, and a heavy power supply (about $50 dollars for 150 Watts or so, non-industrial prices).

The Amp's current supply transformer and rectifier + Elco's is shown in the upper left corner below, 2x26 Volts, 3.3 Amp, with room to grow.


Speaker cabinet

The cabinet I currently use for testing is about 40 x 40 x 60 (from memory), and contains a 12' low cost woofer, an 8' broadband, and a Philips dome tweeter (because I had one). The woofer costs about $40 and is rated for 300 Watts (Continous ?), 34 Hz resonance, fairly tigh suspension. The broadband has very loose suspension, is rated at 50Watts Max (it has an excusion of about 1.5 cm when I drive it at about 40Watts...) and the tweeter sounds ok, but not very exciting or accurate. It has worked for a whole night at about 60W total drive power, including some times of heavily clipped signal at even higher power, and didn't breakdown.

Here, I've used a 4.7uF capacitance as a very basic seperation filter (cutoff about 4kHz) for the tweeter, and the other speakers are simply put in parallel (all are 8 Ohms), limiting the total power.

Overall, it sounds OK, I used it with moderately high volumes in a tiled basement, and it cranks out very clear, distinguised, non-boomy bass lines, fairly clear mid, and high as described above. Pleasurable listening experience, I need to try it in a 'living room' situation, and preferably stereo fashion to judge it better. Certainly beats all the consumer hifi equipment I've recently heard.

Oh, I should add it currently 'pumps', that is at higher power you can probably blow out a candle with the air leaking out some holes. Some silicone-kit will solve this, and the resulting bass response, assuming I can indeed drive the woofer at about three times current power (which allready is pretty loud, even playing music with non-forgiving basses), the bass end is definately up to acting as a PA system. Considering this design, including wood and amp cost me under $150 (non-industrial proces), and not too much work, that could be interesting.

The tweeter should preferably be replaced by a higher-power and preferably a bit more clear version, the broadband ($20) could act as a squaker.

I've currently made all speakers seperately driveable, and used a single 1.5 sq mm cable jack-ed into the 19 inch rack.

A set of wheels completes the prototype. Now some paint, or preferably a rugged coating (plastic, maybe metal edges). The mbf itself looks ok though.

Considering the speakers are real low cost, I think with better damping material, an air-tight cabinet, and active filtering / seperate amps, it will sound even better.


The same enclosure that holds the power amp will hold a variation of the circuitry of the mini-moog analog synthesizer. A variation, because is used different parts, redesigned some parts to fit my current desires, and because some parts are obsolete or more expensive. See below for the rationale behind the current prototypes.

I've included detailed schematics of both the VCF, the VCO, and the exponential converter, and even an accurate SPICE electronic analysis in my pages on synth design.


The filter circuitry is spun around a 4-th order balanced ladder filter, where each stage utilizes the BE-resistance of its two transistors as a variable resistance, by changing the emittor current. This circuit is pretty much an exact copy of the minimoog's circuit, which is famous for its sound. The rest is a difference (instrumentation) amplifier, input amplification and output buffering.

The above picture shows the integrated 2x12V supply, fed through a simple AC wall wart.

Thus top view contains a legenda to the parts of the filter prototype.

The 64k question: does it work ?

Yes, it does. It sounds pleasing, fat, versatile, has at least 20-20k Hz cutoff range, doesn't badly noise or hum (depending on distorion figures, this is open for some more measurements), even with a lot of resonance, and self-oscilates nicely, except that in the lower frequency range, a lot of feedback is needed fot unity loop gain, just as in the original circuit. I've done an early (breadboarded) prototype where I worked around this by a pole in the feedback loop, which made things better, but I stuck with the original circuit for this prototype board, which is intended to hold the other circuits as well to form a self-contained euro-card minimoog. The essentials, that is.

Currently the exponential converter described on my SPICE pages (which only has some parts scaled as compared to the original) currently is still on breadboard, because I consider using the same cicuit as is used in the VCO for the filter, because it is probably more accurate over a larger range, and has a intrinsic input zero correction (minus an opamp offset voltage).

Judging by listening to easy to percieve octave intervals, the volt per octave accuracy is sufficient.

Feeding audio signals through this filter and fiddling with the cutoff and resonance controls makes for very strong and good sounding filter wah's and oumphs, and the resonance can accurately cut out single harmonics, e.g. from a saw tooth. Fast changes in the cutof control give good chorussy and rich synth sounds.


For testing purposes, I've used some simple 74HC14 schmitt triggers with RCD networks to generate some saw and triange waves, and I've accurately simulated the VCO design which has a different comparator but for the major part is the same as the original circuit, and it should pose mo problems.

The rest

The VCA, EG's and some other circuits such as audio mixers are still to be built, but I've built several types of all of them in the past, so they should pose no problem, except that the moog circuits are quite straightforward, so the supply decoupling and audio signal path seperation (to prevent disturbance) are probably important.