This is a follow up to my recent post about the redesigned Beogram Commander remote control board, which now works in both (DC-motor) Beogr...
Tuesday, December 23, 2014
Beomaster 8000: Rebuilding and Testing of the Output Stages and First Power-Up
When I get a Beomaster 8000 of unknown provenance on my bench, the first thing is to check the main 10A fuse, and the color of the emitter resistors in the outputs (crispy brown indicates a burnout of the output transistors, the same can be said for an open circuit main fuse). The 8000 I am currently working on was negative on both counts, and so I proceeded to rebuild the output boards with new quality 105C electrolytic capacitors from name brand Japanese manufacturers (Nichicon, Rubycon et al.). I also put in 25-turn quality encapsulated Bourns trimmers to control the quiescent current of the outputs. This is probably the most important operation in any Beomaster 8000 restoration. Hot heat sinks during modest volume play, or, worst case, smoke coming from the back of the receiver are usually related to the failure of the often corroded original single turn 100 Ohm trimmers that were standard back then. These old trimmers also make it difficult to adjust the quiet current to its prescribed 18 mV equivalent voltage drop across the output resistors since they drift thermally and over time. 25-turn trimmers are much superior, and allow a precise adjustment that will stay put for a long time.
Here are the PCBs after updating them (the original look is shown in this post):
Below are the 25-turn trimmers before implantation in comparison with one of the original single turn trimmers. The 25ers need to be extended on the swiper lead to fit the old footprint. Also, it is important to orient them properly, so one gets clockwise increase of the quiescent current when turning the adjustment screw. If this is done reverse, then down the road another tech may get confused and accidentally fry the output transistors while adjusting the trimmers. Semiconductors behave exponentially, i.e. once the quiescent current threshold is crossed, the current will increase swiftly.
Once I had the new parts installed, I disconnected the power rails and the three plugs that are on the boards. Then I ran the PCBs with three external power supplies. While doing so, I adjusted the quiescent current to the correct 18 mV voltages at the test points:
Both stages took in the prescribed ~100 mV via each of the main ±54V rails, so all is good now in Output Ville!
Now that the outputs were certifiably stable, the first power-up of the Beomaster could be attempted without worries about producing 'magic smoke'. Before turning it on, I reconnected the clipping light control line to its proper place (see heat shrink on red wire below the indicator light that is on - someone just cut this wire for whatever amateurish reason...gotta love it when dilettantes find a screwdriver (or wire cutter for that matter) and wonder how best put it to use...;-)
I connected my two bench-speakers and then the Beomaster came on with a health double relay click, indicating that the start-up sequence worked properly and that the microcontrollers were doing their job as they should. I tuned to a station, and I got some decent FM reception of a local station, which played nicely on the speakers. Unfortunately the displays of this Beomaster are in need of fixing. They show the frequently observed missing segments issue:
I tested all the functions of the Beomaster, and it seems that everything is working except the remote control. None of the buttons provoked any response by the Beomaster.
I poked around a bit and finally realized that the dilettante did not only cut off the clipping wire, but also did some strange experiments with the remote receiver. Here is a picture of a part of the power supply board which shows some 'additions', notably a Sprague capacitor soldered to D25 in the output of the TDA 4050 IR preamp, which has no obvious function. Furthermore, pin 1 on P48, which carries the remote signal is bent upwards, i.e. the signal is effectively disconnected from the microcontroller remote input. No surprise that the remote does not work...I will check this out when I rebuild the power supply. Let's hope that the remote just works when disconnecting the cap and reconnecting the output to the connector.