Differences

This shows you the differences between two versions of the page.

--- projects:repair:beogram8000 [2018/01/20 01:27] – [B&O Beogram 8000 turntable] admin
+++ projects:repair:beogram8000 [2018/01/24 01:24] (current) – admin
@@ Line 1: / Line 1: @@
-====== B&O Beogram 8000 turntable ======
+====== B&O Beogram 8000 turntable repair ======
 Repair of a B&O Beogram 8000 turntable with following issue:
   * sudden speed increase, without any apparent reason. Originally about every 20 minutes, for a duration of 5 seconds.\\
   * Schematic diagram can be obtained free of charge [[https://www.vinylengine.com/library/bang-and-olufsen/beogram-8000.shtml|here]] (Free for personal use.)
-After measuring with an oscilloscope on some meaningful points in the circuit, there were no anomalies found. Signals looked as expected. Since the speed control circuit as a whole is in a closed loop, the cause could be anywhere as well within the loop. A better way to measure therefore was to place measuring probes before and after several functional blocks in the closed loop and see for each block how it behaves when speed increases suddenly. For this a Saleae capture device (Logic Pro 8) was used. With that the reason for the speed error was find:
+After measuring with an oscilloscope on some meaningful points in the circuit, there were no anomalies found. Signals looked as expected. Since the speed control circuit as a whole is in a closed loop, the cause could be anywhere as well within the loop. A better way to measure therefore was to place measuring probes before and after several functional blocks in the closed loop and see for each block how it behaves when speed increases suddenly. For this a [[https://www.saleae.com/|www.saleae.com]] capture device (Logic Pro 8) was used. With that the reason for the speed error was found:
 | {{ :projects:repair:beogram_8000_33.33_not_ok.png?direct&600 |}} |
@@ Line 11: / Line 11: @@
 | {{:projects:repair:beogram_8000_33.33_ok.png?direct&400|}} | {{:projects:repair:beogram_8000_33.33_ok2.png?direct&400|}} |
-|  Normal operation: Only SLOW UP/DOWN  |  Detail of signals  |
+|  Normal operation: Only SLOW UP/DOWN  |  Signal detail  |
 So to find the cause for this, the circuit which handles the speed sensor was analyzed:
 | {{ :projects:repair:beogram_8000_speed_sensor.jpg?direct&600 |}} |
-|  Ch. 1: OPE1 phototransistor, Ch. 2: Output of Opamp/Schmitt trigger  |
+|  CH1: OPE1 phototransistor, CH2: Output of opamp/schmitt trigger  |
-The B&O diagram from above is a bit different then what this turntable has: Phototransistor OPE2 is pulled up to 15v with resistor R47 (22k) and fed via resistor R48 (100k) to opamp IC1, which is configured as a 5v comparator (with a feedback of R49 (1M0). The output of the comparator is via a voltage divider (consisting of resistor R50 and R51) into the microcontroller. This is also different than what is designed in the circuit diagram from above. Maybe the 4013 schmitt-trigger was added in a later revision as a modulator for the speed signal.
+The B&O schematic diagram from above is a bit different then what this turntable has: Phototransistor OPE2 is pulled up to 15v with resistor R47 (22k) and fed via resistor R48 (100k) to opamp IC1, which is configured as a 5v comparator (with a feedback of R49 (1M0). The output of the comparator is fed via a voltage divider (consisting of resistor R50 and R51) into the microcontroller. This is also different than what is designed in the circuit diagram from above. Maybe the 4013 schmitt-trigger was added in a later revision as a modulator for the speed signal.
-With the limited triggering possibilities of the oscilloscope, measuring both signals for a long duration while staring at it and wait until one could see where the root cause was, was quite hard. Also, while investigating this, the cause did not appear a single time while observing. Still without repairing anything specific, it was difficult to understand that defects would be fixed somehow spontaneously. Due to this long duration, the Saleae analogue capturing device was not suitable either, because of the limited sampling time. Then, to find the root-cause (Or at least prove that the cause was not anymore present), a simple but effective Arduino Nano microcontroller as watchdog was taken. With some software it was able to guard both signals and indicate if one or both signals would fail and if speed was deviating from 33 1/3rpm.
+With the limited triggering possibilities of the oscilloscope, measuring both signals for a long duration while staring at it and wait until one could see where the root cause was, was quite hard. Also, while investigating this, the cause did not appear a single time while observing. Still without repairing anything specific, it was difficult to understand that defects would be fixed somehow spontaneously. Due to this long duration, the Saleae analogue capturing device was less suitable either, because of the limitations with maximum sampling time: If the error would not happen a single time within at least 24 hours, then the issue could be declared as resolved. Then, to find the root-cause (Or at least prove that the cause was not anymore present), an Arduino Nano microcontroller as watchdog was taken. With some software it could then guard both signals and indicate if one or both signals would fail and if speed was deviating from 33 1/3rpm.
 To reduce the influence of the Arduino measurement system, two opamps as voltage-followers were taken, fed with 20v to allow some margin above the 15v level to be measured:
@@ Line 27: / Line 27: @@
 Both input signals, Optical_sensor_P4.7 and IC1_p.14_pulse_out, are between 0v and 15v. The LM324 is here practical for interfacing, since for a single power supply, Vout can go to ground if sensing is near ground and this can then be easily converted to logic levels for the Arduino Nano. For this two cd4049 hex buffers in series were taken, only to create logic voltage levels with sharper edges.
-As the turntable always stops turning after about 30 minutes, transistor Q1 was selected to control the TURN-button automatically.
+As the turntable always stops turning after about 30 minutes, transistor Q1 was selected to 'hit' the TURN-button periodically.
 A 5v regulator was chosen to regulate 20v down to 5v for the Arduino Nano, if the Nano was not connected via USB to a computer. The 20v made it easy that only one power supply was required.
+The circuit was temporarily built on a breadboard:
+| {{:projects:repair:beogram_8000_circuit_breadboard.jpg?direct&600|}} |
+|  Interface circuit between Beogram 8000 and Arduino Nano  |
 ==== Monitoring speed with software ====
@@ Line 314: / Line 318: @@
 }
 </code>
-==== Root cause found ====
+==== Root cause ====
-After testing for several days, the error did not come back at all. Further analysis found that the most likely cause for the failing sensor is wrong alignment of the optical sensor unit (8005067). The sensor was too far in and with releasing the adjustment screw, it was possible to move it. Most likely, by touching the optical sensor unit, it had been shifted in the good direction, without knowing that. It solved the problem!
+After testing for several days, the error did not come back at all. Further analysis found that the most likely cause for the failing sensor was wrong alignment of the optical sensor unit (8005067). The sensor was too far in and with releasing the adjustment screw, it was possible to move it. Most likely, by touching the optical sensor unit, it had been shifted in the good direction. Without knowing that, it solved the problem!
 ==== References and further information ====
   * See about interrupts: [[http://www.gammon.com.au/interrupts]]
   * Variables to be declared volatile

Auditeon

User Tools

Site Tools

Differences

Page Tools