Featured Post

Beogram Commander Remote Control: Maybe This is the Final Version!..;-)

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...

Wednesday, February 28, 2018

Beomaster 8000: Evaluating the Audio Performance

This Beomaster 8000 receiver has been playing music in my lab since I finished the recap and output amplifier adjustments. The sound is great as it always is with these receivers but human ears are subjective and we like to measure some key values with test equipment to have some actual numbers to compare with. For amplifiers we typically use a sine wave test input voltage and measure the output of a preamplifier or power amplifier to check the total harmonic distortion (THD) and sound to noise ratio (SNR).

For this Beomaster I set my test up to check some input voltages at some key frequencies and measure the output amplifier at the speaker outputs.

For the speaker output measurement I connected a pair of fixed 8Ω loads. Obviously the fixed resistive load isn't exactly like a real speaker system load but by using the fixed 8Ω load I should always have a good measurement value I can compare other amplifier results to.

Here are my dummy speaker loads. They are power resistors connected to a very large heat sink. When I drive the output amplifier up to its rated power output (100 Watts into and 8Ω load in the case of the Beomaster 8000) the load resistors will get quite warm.






















I am going to measure the THD and SNR values with my QuantAsylum QA400 audio analyzer. Since the high speaker output voltages would damage the analyzer inputs I use a low noise differential probe to lower the signal to a level the analyzer will accept.

This picture shows the Beomaster 8000 left and right speaker outputs connect to the dummy 8Ω resistor loads where the differential probe is attached and the signal connected to the QA400 inputs.






















For the test inputs I want to use a constant sine wave of 1Vrms at frequencies from 20Hz to 20KHz.
I use the signal generator of an HP8903 audio analyzer for the test input signal and check it with an oscilloscope as I route the test signal to the Beomaster 8000 Tape 1 (TP1) inputs.





























I started with a 1KHz signal as that is a common frequency for a lot of the audio specs manufacturers publish. It is also in the middle of the frequency range on the log scale.

It should be noted that before starting the measurements I adjusted the Beomaster 8000 source input levels to set the left and right channel levels as close as possible to each other. Those input level adjustments are on the left side of the Beomaster cabinet.



Using the Beomaster volume control I increased the volume of TP1 while monitoring the output with the QA400 analyzer and the Beomaster clipping lamp.

With the 1Vrms input signal I got to the clipping point when the Beomaster reached 5.9 on the volume indicator. I backed off that volume to 5.8 and measured between 28Vrms and 29Vrms at the speaker load. That corresponds to about 100 Watts of output.

Here is what the QA400 analyzer measured.











































































The THD levels are very good. Both channels are comfortably less that 0.05%.
For the SNR value I notice that B&O specifies their Beomaster 8000 SNR value as "A Weighted" and should be >77dB for the Tape input. The frequency and output level isn't specified but I am doing my measurements at the maximum rated output level so I turned the A-Weighting on in the QA400 and got these measurements (at 1KHz).



The SNR values are below the expected 77dB for the Beomaster and the left channel THD went up some. It is still below 0.05% but I thought it was odd that it moved and the right channel didn't move much.

Checking other frequencies (400Hz, 10KHz and 15KHz) resulted in the left channel being within THD specs but not as good as the right channel.

These results are not bad but I would like to see closer values for the left and right channels during this testing as I am using the same input signal and the load resistors are identical. A common culprit in the amplifier performance for the Beomaster 8000 are the OpAmps used in the preamplifier board and tone control board. Those are the LF353N, TL072CP and uAF772TC OpAmps. The audio source signals go through these OpAmps so their performance directly affects the Beomaster amplifier performance.

In the case of changing out the OpAmps it is best to change them all (left and right channel). That way all of the signal paths are using new OpAmps that should be pretty much identical.

When replacing the OpAmps I am also going to install 8-pin sockets for the integrated circuits. Here is the preamplifier board with its original OpAmps (seven total)

 Here are the sockets for the new OpAmps.

Here are the new TI LF353N OpAmps installed.























Next is the Tone Control & Filter board. This Beomaster 8000 unit has five OpAmps to be replaced on this board. Note that some Beomaster 8000 units have a six.  This picture shows the five OpAmps replaced.






























The OpAmp identified as 4IC6 is for the Beomaster volume control circuit. Whenever that OpAmp is changed it is very likely that the volume control circuit offset adjustment will be necessary to be performed.






















Here is the circuit schematic.
























The section outlined in red shows the left channel trimmer for the volume control offset adjustment.
You immediately know if the adjustment is necessary because you will hear audible clicks in the related channel's speaker as you turn the volume up or down. Adjusting out the offset removes those click sounds.






















Now that the OpAmps are changed out and the volume control offset is readjusted I remeasured the amplifier outputs.

The THD and SNR (A-Weighted) both improved plus the left and right channels are close to the same.







































I have done a couple Beomaster 8000 restorations where I replaced the OpAmps with a higher performance TI OPA2134A (SoundPlus) device. However, in measuring the performance with that OpAmp versus a new TI LF343N OpAmp I could not see any measurable differences with the analyzer. Since the OPA2134A costs four times that of a new LF343N I recommend using the latter.

As a final check I used the QA400 impulse stimulus and measurement system to make a frequency response measurement of the Beomaster. This test also uses the same test setup. The difference is the Tape 1 stimulus is from the QA400. I have to admit I am not 100% comfortable with this test yet but it is worth looking at and recording for future reference with other Beomaster unts. The results look pretty good from 20Hz to 15KHz. However, I was hoping the 20KHz level would be a little higher. As it is it measures around -0.8 dB instead of the -0.5dB.







































I will finish closing the Beomaster 8000 cabinet back up then do some actual audio component tests using the Beogram 8000 and a Beocord 9000. That will make sure all of the playback and recording features work as well as the Beomaster 8000 remote control.


Friday, February 23, 2018

Beogram 4000: Restoration of the Keypad Cluster

After replacing the electrolytic capacitors and restoring the AC motor of the Beogram 4000 that I am currently restoring, it was time to rebuild the keypad cluster. The PCBs below the keypad house the control center of the Beogram, which is a unique early digital control system based on TTL logic chips. This logic system comes to its operational conclusions largely based on inputs caused by the many mechanical switches throughout the turntable. The keypad contains 8 of them enabling user interaction with the Beogram.
The keypad is held in place by a single screw, which was missing in this Beogram:
An indication that 'human interaction' had taken place earlier (which was confirmed once I looked at the switches on the PCBs - see below). I took the pad out and opened it up:
The upper board contains some of the logic chips and the light bulbs that illuminate the position indicator and the RPM trimmers. The lower PCB is populated with the eight switches for the keypad and one more logic IC. As usual, the switch terminals were heavily corroded:
I removed the board from the keypad, which is necessary for extracting the switch terminals:
This shows the side of the board that houses the switch actuators. The small green, white and grey 'plungers' are actuated by the keys on the keypad, which pushes the switch terminals on the other side of the board making or breaking the associated contact. Note the location of the single grey plunger top left. This is the wrong location for this particular one. It needs to be in the center of the bottom row, since this particular switch is a break switch. This incorrect installation immediately explained the malfunction of the arm lowering circuit that I noticed after the restoration of the main PCB. The green plungers are longer, and therefore this switch, which is responsible for lowering the arm, was permanently open (actuated). This shows the three switch types in comparison:
The grey one is the shortest. It is used for break switches. The green one is used for make switches and the white one is for two-pole make/break switches (the << and >> keys, which have slow and fast functionality, depending on how hard the keys are pressed).
I removed the switch terminals:
After removal of the oxide layer with 2000 grit sand paper, I coated the terminals with a gold layer:
and then soldered them back into place:
The final step was to replace the light bulbs with LEDs. The position indicator scale lights were replaced with custom designed LED boards (available to other B&O enthusiasts), each containing two red-green LEDs tuned to yield an incandescent-like sheen. The RPM trimmer back light bulbs were replaced with standard red LEDs and current limiting resistors:
This shows the LEDs in action after installation:
After that it was time to put the keypad back together. Unfortunately, the center key was not attached to the keypad, i.e. I needed to reinsert it. This can be difficult and there is a danger to scratch the other keys while doing it. For this reason I used 3D printed tools that I developed earlier, which make this process much easier. They allow pre-bending of the spring that holds the pad in place and that allows it to bounce back after pressing it:
Once the spring is bent up, it is fairly easy to get the pad on it. Careful removal of the printed taps releases the spring holding the key in place. And this shows the pad installed:
A test revealed that all keys are now working properly. On to gold coating the remaining switches below the carriage and in the arm lowering mechanism.













Thursday, February 22, 2018

Beomaster 6000 (2702) restoration: mounting the power supply into the main chassis

After my disaster with the electrostatic discharge that fried several transistors in the output stages (click here), it was now time to put everything back into the main chassis of the Beomaster 6000. I had not touched them yet since the dismantling, so it all looked kind of dirty !





And after cleaning







And yes, I like to keep the stickers & lettering on it to keep the look & feel of vintage...

The 4 rubber feet had there best time, so I replaced them with 4 spare better looking ones.




The transformer went in first after adding a new power cord to the on/off switch. Then the voltage selector was mounted with the 2 screws. The small metal frame with the bridge rectifier and the stabiliser transistor are fixed. The larger clear metal frame was put in place, the on/off switch mounted and the plastic safety cap plugged on top.




The plastic holder with the two main fuses was screwed on again, but the copper foil did not stick anymore to the chassis. Not sure what type of glue they used because it's supposed to be electrical conductive. So, I used some duck tape to keep it in place and connected.



As mentioned in one of my earlier posts, some of the later Beomaster 6000's have an extra fuse (4A slow) mounted in the 27V AC connection from the transformer to the power supply PCB as an extra protection.  


I decided to add one also in the same spot. I used a encapsulated one versus the "open" one as seen in the picture above.

I needed to modify the metal bridge, that holds the power capacitor, a bit and drill a hole in it.





The transformer, main bridge rectifier, voltage selector, main fuse holder, on/off switch, power caps and new fuse are now in place.