Featured Post

Beogram 4002: Restoration of DC Motor Video Published - Check It Out!

By popular request (really, I got quite a few emails about this!...;-), I finally completed my Beogram DC motor restoration video! It demon...

Saturday, May 26, 2012

Beomaster 6000 Heat Sink Temperature

I recently fixed the output stage of a Beomaster 6000 and adjusted the quiet current to specifications. I thought it would be interesting to know what the corresponding heat sink temperatures are for future diagnostic purposes. Here is what I measured with a digital IR thermometer on the top of the 'cooling tower' in the back of the BM6000:

After ~1 hr of listening to the radio (volume set to 10) the right half of the cooling tower measured about 42-44C, while the left half measured 38-40C. The reason for this difference is that the right side houses the power supply, i.e. more heat is dissipated there. I expect that these numbers vary depending on the ambient temperature of the location. When I measured, it was about 20C.
If the BM6000 delivers temperatures significantly higher than the above values, I would think that the quiet current needs some adjustment.

Sunday, May 6, 2012

Beomaster 8000 Display Repair Marathon

****note: This post is 'historic'...Please, visit this page for my latest display repair procedure: http://beolover.blogspot.com/2012/09/beomaster-8000-display-repair-hopefully.html****

I am on a roll!!! Today, I did the displays of my second Beomaster 8000 in one setting. Desoldering all the displays at once, opening them up, putting the SMD LEDs in...it took a few hours, but they look very nice. My best result so far...I guess practice pays off...;-).

I made one further change to the procedure: I used "No Clean Flux Paste" from MG Chemicals. I realized that paste does not evaporate as quickly as the liquid flux does. This makes it possible to put flux on all locations at once and then put the SMD LEDs on in one setting. It is recommended to do them in sequence by orientation, i.e. do all vertical ones in one setting, then the horizontal ones. This speeds the process up considerably. Also reliability increases. This time I had no failed LEDs. They all withstood the toothpick stress test (yes, not only European banks get this test...;-). The "No Clean" aspect of the paste did not work. I probably put too much on there. Therefore, I washed the boards in ethanol with a toothbrush. This worked very nicely. Here some pics of the completed boards:

I like doing the balance display best now. I realized that the smaller pads in combination with the green solder mask stops the solder from spreading.  This makes it much more straight forward to get enough solder on the contact pads of the LEDs. It also prevents the solder from obstructing the bezel, which fits perfectly after the refurb without the need for Dremeling the slits to allow for the solder thickness...These pics show the balance display board:

Saturday, May 5, 2012

Beomaster 8000 Display Brightness Adjustment for New LEDs

This post describes what I did to adjust the brightness of my refurbished beomaster 8000 display units (see previous posts) to the brightness of the original displays (as compared side-by-side with my second 8000):

After studying the circuit diagram for the display PCB (#8) I tried to adjust the brightness of the display with trimmer R15. Unfortunately, even fully turning it down only reduces the display voltage from the originally prescribed 5V to about 4.5V (measured at TP 14). This did reduce the brightness a little, but by far not enough.

This led me to change R14 from 4.7k to 3.3k, and R16 to 1.5k. This pushed the voltage range of R15 down to about 3.6V-4.7V, while keeping the voltage divider intact (to not affect the "reset" circuit). Consequently, the supply voltage for the LEDs also dropped. This allowed a significant reduction of  the brightness. Adjusting the display voltage to 3.85V yielded a brightness similar to the original displays (maybe a bit higher, but I always felt the original displays are a bit low for daylight situations, i.e I decided to leave it at that; if desired the new voltage divider allows to go considerably lower with the intensity, but care needs to be taken to avoid dropping the keyboard strobe below ~2.7V). Here is a cutout from Diagram 3 from the service manual showing the discussed components:

It is interesting to note that R14/R15/R16 also control the strobe voltage for the keypad (measured across R24,27 or 30). Accordingly, my 3.85V setting at TP14 reduced the keypad strobe pulses to ~3.12V (from an original 4.2V). Here is a screen shot from my oscilloscope (after modification):

It turns out that this new lower voltage does not cause any trouble. The reason for this is that the 6500 microcontroller is based on TTL logic. TTL logic accepts HIGH signals down to 2.7V. Turning R15 fully low pulls the strobe below 2V, and the keyboard consequently stops working. This demonstrates that there is enough safety margin if the strobe is at 3.12V, i.e. no further modifications are necessary.

Thursday, May 3, 2012

Beomaster 8000 Display Restoration

It has been a while...I finally had some time to take on the display of my first  Beomaster 8000. Some segments were out and this had to be done.
I used the same soldering approach like with my Beomaster 6000 displays. Dry tack per Norman Mier and then a quick soldering at 340C with regular leaded solder after dousing the tacked LEDs (I used again the LUMEX SML-LX0603IW-TR model with 14 MCD) with no-residue liquid flux (Kester #951). The most significant difference to the BM6000 experience is that one has to take out the entire 3-display bar at once. This is not possible without using of a professional desoldering pump (I use a Hakko 808). Serious damage to the fragile boards could occur if one tried to do this with de-solder braid or a spring loaded pump. 

The volume, program selector and frequency displays are very similar to the Beomaster 6000 displays. This picture shows them after repair during the test run. A test run is highly recommended. In fact, it is a good idea to run them without the diffusors on and test them for mechanical stability. I did this with a toothpick. I pushed them relatively hard on all four sides. Indeed a couple of the 54 LEDs flickered during this test and had to be replaced. It appeared that these two were damaged during the soldering process (the plastic cap wiggled on the metal base - I guess they got a bit too hot). 

Another thing I did different was that I did not use the brightness-reducing window tint foils to adjust their luminance to the original spec. I rather modified the display board enabling a proper reduction of their current to achieve the right intensity. More about that later.

The balance display is differently constructed. The tracks are much thinner and there is less space for the new LEDs. This means that one needs to be pretty accurate with the positioning of the new LEDs to avoid having to significantly Dremel the bezels out to make room for misplaced SMD LEDs. This picture shows the balance display after disassembly with the original LEDs.

These two pics show the display after repair:

And here the final result after adjusting the brightness. I will report about this process in the next posting.