NIKKO NR-519 stereo receiver restoration project
These receivers were manufactured in late 70-ies, so it had at least 35 years on it when I got it. Cosmetically the unit was in pretty good shape, and a little scrubbing with office equipment cleaning fluid made it look pretty decent. I blew the dust from inside out with air compressor.
4 different variations were manufactured, according to service manual, for different markets. Besides different power supply, major differencies for the models are in FM tuner section. Some models had outputs for 1 pair of speakers, some for two. Mine has two - A pair and B pair.
I got the service manual from Analogalley.com. It is a good quality scan, although when printed out schematics is hard to read. No problem on a screen though.
The receiver is relatively small, and very straightforward in its construction. There are two main boards, basically one for tuner and another for power amplifier. Amp board also hosts power supplies and filter capacitors, while tuner board has phono preamp, input selector, and on its corner balance control potentiometer just before the signal jumps over to amplifier board. Both top and bottom covers are removable so that both main boards can be worked on without disassembling the entire unit. Front panel is also very easily removable without even taking any of the buttons off.
The tuner is built around Panasonic AN217P, but for stereo decoder they have chosen to use NEC uPC1161C instead of, for example, Panasonic AN362. Would be interesting to know why that choice was made. The tuner is permanently powered, and at volume turned to max you can hear the audio from tuner bleed through to output when input is switched to aux or phono.
The amplifier is also quite interesting. There is only one capacitor direcly in signal path from AUX and TAPE inputs all the way to speakers. The tone controls are wired to negative feedback path in a way that won't even try to understand. Potentiometers are all from Alps, and electrolytic capacitors are Marcon branded - certainly looks like high-quality piece from its era.
It does not have
Most of the electrolytic caps in the receiver were branded Marcon, at the time it appearantly used to be a good capacitor brand, Nippon Chemi-Con bought them back in 1995. I pulled couple of them from amp board and tested them, all were rading 20-35% above their rated capacitance, and several ohms ESR.
It was the first recap job for me, so I did some research for what type of capacitors to use. For that I watched audio gear restoration videos on Youtube, and scanned forum posts, then checked the data sheets if needed. I had no intention to buy audiophile-grade golden capacitors, but I wanted to do it reasonably right anyway.
Recapping is much easier if you have a desoldering station. I did not own one before, but wanted to have one for a while, so this project was a good excuse to buy one - bought a reasonable priced one SSD-15 desoldering station 33400100. I am sure there are fanicer and better ones available, but this one gets the job done good enough. It is not ideal, the biggest problem is that it has a fan that is very loud. The desoldering gun itself is ok, but getting the solder collecting container off for cleaning is a bit fiddly. I still think it is ok for the price, unless you are are planning to do desoldering daily, and all day long.
Forums suggest to go for polyester film in audio path for 2.2uF or lower values
Forums also suggest to replace tantalums with modern electrolytics, low leakage type in critical applications. Or multilayer ceramic.
NR-519 capacitor shopping list
Thanks to my own stupidity I managed to blow two transistors in power amp, and this forced me to try to find replacement parts. This is what I used, not sure how well these actually match the originals, I replaced Q707 and Q708 with KSC2316 (just to make the channels the same) and Q712 with BC640, only in blown channel because I did not order a replacement for its complement anyway. At least I cannot hear any difference between the channels, so that'll do.
Fixes and modifications
Recapping solved AM radio problem, phono preamp noise, and FM stereo sensitivity.
Automuting circuit in FM radio part was reacting very slowly, so in stereo mode you would miss most stations when tuning. The timing is determined with 47uF capacitor (C123) and 22k resistor (R136, which actually measured 22.6k). I made it to react slightly faster by replacing the resistor with 18k value. The stereo mode enabling had an intermittent problem that turned out to be bad solder joint on R135.
For replacing the 8V 0.3A bulbs I used LED bulbs, look for "T10 194 168 W5W COB 8 SMD LED CANBUS Silica Warm White License Light Bulb" on eBay. These are for 12V, and I was just barely able to get 11V out of the 8V AC voltage by using full-bridge rectifier made from 1N5819 schottky diodes, followed by 220uF capacitor. At that voltage these bulbs glow just borderline acceptable, but have a greenish tint to the light which I rather like. You may want to use something else if you want the dial brightly lit, or perhaps use these bulbs but pair them up in series and get power from main power rail which is over 30V.
For potentiometers and switches I just used Deoxit 100, and all of these still work well.
The Nikko power amp does not have any offset or bias adjustments, so that's that. All adjustment procedures in service manual are for RF part, and FM stereo decoder. When working on the receiver I did not have RF signal generators to do it properly and the radio reception did not seem to have any big problems, so I did not mess with the working system.
That said, this far I've touched few things for which the scope was all that was needed:
T102 which is for detecting FM signal strength to turn on stereo mode. This was still spot on.
HVR302 which is for setting stereo decoder VCO frequency to 76kHz. This was slightly off at about 77kHz.
The FM tuning scale pointer was about 500kHz off, for this I just tuned on a station of known frequency, loosened two screws that hold the tuning rope wheel to variable capacitor, and adjusted the tuning knob to put the pointer at correct frequency on the scale, then tightened the screws again.
Since then I've built a simple test signal generator (see rfgenerator project) and verified that the tuning scale is still a bit off, but I decided that it is not worth messing with.
Copyright © Madis Kaal 2000-