Fullscreen HTML5Med Res (??MB)HTML4

Modifying the Studio 200 power amplifier Back in February 1988 we published the Studio 200 power amplifier which has proved to he a very reliable and consistent per£ormer. In the intervening period a number of readers have written to us describing modifications they have made to "audibly improve" the performance. Here are the modifications made by one reader. By LEO SIMPSON Since it was published in February 1988, our Studio 200 power amplifier has proved itself to be very trouble free and many hundreds have been put together without problems. Constructors have been particularly delighted with its very low noise. In that respect it is a lot quieter than many The Studio 200 power amplifier was actually developed from this module which was published in the December 1987 issue. This circuit pioneered the use of PTC thermistors for amplifier and loudspeaker protection. 100 SILICON CHIP commercial power amplifiers, regardless of their cost. But just as there exists a small specialist market for those well heeled people who want commercial amplifiers to sound better, so there is also a temptation among keen audio enthusiasts to modify the Studio 200. Up till now, we have been reluctant to encourage the idea of modifications because of the considerable risk involved. First of all, there is the risk that uninformed modifications may lead to a costly failure in the amplifier. This would be particularly frustrating if the amplifier had performed perfectly up until it was modified. And then there is the risk that the amplifier may sound no better or may sound worse, after the mods have been made. We have also been reluctant to publish readers' modifications to our amplifier designs because, without exception, such modifications have not been accompanied with any analysis or objective measurements to justify the work. This is not to say that the design of Studio 200 is absolutely right in every way. As with any commerical design, the Studio 200 amplifier is a tradeoff between cost and performance. At the time we produced it, we believed the Studio 200 to be a very good compromise. We still do. But ... In publishing the accompanying modifications we are in no way endorsing them. We do not even agree with them. Why then, are we publishing them? To stimulate debate. Here then, is the reader's letter: In relatively recent times I have replaced op amps to gain a small audible improvement in various consumer audio products. I was not happy with the apparent loss of depth information in NE5534 preamps compared to an earlier discrete design I had used, so I built an experimental preamp using an LM394, BC559 and LH0002 buffer in each channel (I believe in low output impedance). I repeatedly became aware of a phenomenon whereby after a certain amount of time I realised that the soundstage had considerably improved and the sound was ,--_ _ _ _ _ _____ _ _ _ _ __ _ _ __ _ _ _ _ ____,.,__ _ _ _ __ _ _ _ _ _ _ _ v+ 5A r lN;:_..s 22k 1'/, 22k 1% OUTPUT .,. -;- * 04 47k 1% 2xBF469 2.2k 1% 2.2k 1% 5A 100() 1°/, ,..___ _ _ _ _ _ _ _ ____,.,__ _ _ _ _ _....,_ _ _ _ _+ - - ---<11-- - - -- ..--------------v+ ......----v- 0.1+ RIGHT 0.22 * HFE MATCHED TO 10% 240VAC 0.22 10k 1W V- RIGHT N V+ LEFT 0.22 10k 1W 0.22 10k 1W . V- LEFT Fig.1: the modified circuit of the Studio 200 amplifier uses current sources for the second differential pair instead of the now tried and proven current mirror arrangement which is very common in op amp ICs. Note that quite a few more components are employed and that the amplifier has a gain of 23 down to DC. "sweeter". Each time I noticed the change, I read all DC voltages and on every occasion, the potential difference between the collectors of the input differential pair was less than 5mV! Unfortunately the improvement slowly disappeared (dependent on the surrounding temperature or ventilation of the preamplifier). Several versions of the preamp later on, the BC559 was replaced with an imported 2N3811A dual transistor (at greater than $20 each) and the LH0002 with a discrete version using selected BC559 and BC549 transistors. The unused half of the 2N3811A was given an identical collector load to the other half and its collector voltage used to control a shunt tr ansistor across the reference LED of the adjustable input pair current source to maintain optimum conditions. This arrangement also resulted in ± 5mV offset from the DC coupled preamplifier. I fluked obtaining a beautiful dual 18-position rotary switch in a Jaycar switch pack and fitted 1 % metal film resistors to it, to make a precision matched dual volume control. Surprisingly, the sound was less harsh also. More recently, I decided to try similar circuitry in a 100 watt/ channel amplifier based on the Studio 200. The amplifier (as it now stands) is not suitable for the average home constructor as it requires some parts selection and the dual PNP input transistor. A competent designer could overcome most of the drawbacks however. The audible effect of the input pair collectors being adjusted to less than 5mV difference is at least as audible as in the preamplifier and the amplifier audibly outperformed my reference " Tillbrook 6000" which uses matched pairs throughout, 1 % resistors and polystyrene capacitors in lieu of ceramics wherever possible. The JANUARY19 90 101 +4!1.5V t 01 1N4002 2700 02 1N4002 f 48.1V 0.5V :_J B 01 BC556 C ~ .,. 18k 33.3V et,. 07 B0139 .012 --1 B l 06 BF470 C 5A 0.1i B E 0.1 INPUT 22k B 1k ... VR1 5000 47 6VW:r .,. 04 B l ! 2xBF469 C B 48.3V 2.2k ~ :-i --1 :__J 1.9V 2.2k 1.9V -49 .5V O.llSTUDIO 200 STEREO POWER AMPLIFIER SC1-1-288 Fig.2: this is the circuit of the Studio 200 as originally published in the February 1988 issue of SILICON CHIP, although minus the power supply. Its signal to noise ratio was close to - 120dB with respect to full power. That's very quiet. equipment used for auditioning was a much modified Marantz CD65 player, the previously mentioned preamplifier and B&W 801s in an almost ideal listening room. I am not sure what the improvement is caused by, as I expect the distortion would already be vanishingly low. Perhaps supply rejection is the answer? I feel that this phenomenon is worthy of further investigation, particularly as most designers do not fully take into account the loading effect of the succeeding base emitter junctions across the collector resistors of the input pair. The loading transistors must track each other thermally if the input pair is not to be unbalanced. Alternatively, if only half the input pair is loaded, an appropriate value resistor should be fitted to the collector of the other input transistor. Food for thought? (A.K., Chatswood, NSW). Potential problems Readers should compare the accompanying circuit diagram (Fig.1) with that for the Studio 200 (Fig.2) - see also SILICON CHIP, February 1988 - to see how extensive the modifications are. We do not propose to analyse the effect of these modifications as we would prefer Notice to Subscribers Are you about to change your address? Please advise us by the middle of the month by filling in the form on your address label and posting (or faxing) it to us. If you have any queries about your subscription, call us or write to us at: SILICON CHIP, PO Box 139, Collaroy Beach, NSW 2097. Telephone (02) 982 3935. Fax (02) 982 9553. 102 SILICON CHIP readers to make comments. Briefly though, we see a number of potential problems in the modifications. First, the use of separate bridge rectifiers and filter capacitors for each channel, while superficially an advantage from the point of view of better channel separation, may cause circulating currents in the earth returns. This could well lead to audible distortion or low frequency instability. Second, we can't see how the use of a LED (light emitting diode) will lead to improved power supply rejection. At a current of 2mA, a LED will typically have around the same dynamic resistance as a 1N4002 power diode which was originally specified for the Studio 200. In fact, at 2 milliamps, the dynamic resistance of 1N914 or 1N4148 small signal diodes is a good deal better. There are many other ramifications of the proposed modifications. For the time being we will not comment on these. We invite readers to have their say. ~