| Welcome to my dusty corner :-) |
|
| Welcome to my dusty corner :-) |
|
My Hobby: DIY Audio An old amplifier FET amplifier model Hybrid valve-FET amp |
An old amplifierBeing a student in 1989 I bought a brand new reel-type tape recorder as a pile of scrap-metal. It was very cheap because it had been broken right from assembly-line and didn't pass factory quality control. It contained a powerful and reasonably-good amplifier. I fixed the device by replacing a handful of transistors. Since CD and PC era it has been working well as an amplifier. Last summer I decided to look inside it and replace dried electrolytic capacitors, old-fashioned transistors, etc. The main idea was to get rid of quasi-complementary output stage and solder complement modern transistors there. This is the original schematic from 80-th:  After replacing electrolytic capacitors (and adding film-type capacitors in parallel), I've replaced old 3-MHz KT815, KT502 and KT503 and unsafe KT315/KT361 transistors by modern ones. I have had to cut and rewire a couple of PCB paths to transform the output stage into a complementary follower. I'd chosen reasonably cheap pairs of output BJTs: 2SC3182N and 2SC1216N. They were packed into TO-3P packages, but fitted well over the old TO-3 places. Even now I do not know who'd made those beasts and they are still under suspicion. Experimenting with standing current, I've burnt one pair, but it was my fault: the protection was locking my efforts and I've switched off VT13 and VT14. To continue the experiment I have had to activate the protection again and decrease its sensitivity by adding diodes VD9A and VD10A. This is the resulting schematic:  Then it sounded better than original version, but not with my Creative Audigy 4 sound card as a signal source. The problem was: high frequency oscillation and noise when ultra-sonic signal was connected to the input. My Creative Audigy 4 is not equipped with low-pass output filter, so its spectrum is very noisy. Simple low-pass RC filter R1-C3 does not eliminate the oscillation and I had no idea how to suppress it. My next idea was to remove R8 and replace BJT VT7 by low-power MOSFET. And there were no need for such high standing current through VT17 and VT18, so I intended to increase R49 and R50 to 56-75 Ohms. But the only powerful resistors I had were 39 Ohms. Funny, but there was no need to change R7 resistance. The resulting DC offset became +20 mV for the right channel and -18 mV for the left one. I've been watching for the offsets after heating the amp for about 30 minutes and nothing have changed. Surprisingly, but the amp became more stable. Now it can stand Audigy4 spectrum and I have even decreased C5 capacity to 56 pF. I like the new sound I have! This is my current schematic:  Recently I've modeled the original and the current schematics. This is the resulting THD level of the original schematic:  And this is the resulting THD level of the current schematic:  Comment this |