Power Supply - - Replacing Blown IGBTs.
The PLX IGBTs are driven by an active, direct coupled integrated
circuit, rather than a gate drive transformer. IGBT or driver failure should be
rare (when correctly assembled) but when the IGBT’s blow, it usually damages
the following parts:
CHECKLIST AFTER BLOWN IGBT’S
Q96, Q97, (IGBT’s generally fail in pairs)
D78, D79, R358, R359, gate drive coupling comonents, check after removing blown IGBT’s.
Ul8, lR2110 high-side gate driver, Fault current when low-side IGBT shorts to upper rail. Such currents also typically damage the gate coupling parts noted above.
Ul9, 3525 controller, Blows from currents shorted thru Ul8, or possibly by overvoltage on the supply rail SOMETIMES Ul4, 556, powered from 5V output of 3525, which may fail high when 3525 fails.
RARELY Ul3, which has fairly high supply voltage ratings.
PROBABLE CAUSES OF MASSIVE IGBT FAULTS
SHORTS IN CONTROL CIRCUIT.
The parts operate well within their ratings and should hold up well in the field. The usual cause of failure is when both IGBT’s turn on at once, shorting Pri-Hi to Pri-Lo. This occurs when something causes the drive signal to one part to remain on when the other part is supposed to turn on. Shorts from solder or debris are one obvious cause.
SHORTS IN THE LOAD.
Although there is peak current shutdown, shorts in the power amplifier transistors or secondary-side supply components can cause currents to increase too quickly to prevent damage.
OVERVOLTAGE ON THE BIAS SUPPLY.
If the TOP-210 bias supply fails to operate, no harm occurs, the unit simply does not operate.
However, open circuit (missing part) in several key components can cause the Bias supply voltage to be much too high, This blows the 2110 and thus the IGBT’s.
QUICK TEST OF BIAS SUPPLY.
Ramp the AC voltage up slowly to 25% of regular voltage (30V for 12OV unit). If the bias supply is working normally, the green “power” LED should come on between 30 and 35Vz with its usual, steady “half-brighr start-up level. If the LED comes on at 2OV, or not until 5OV, or blinks, DO NOT RAISE VOLTAGE PAST 60V until you have measured the bias voltage. The switching will not start until you reach 9OV, so you can save the IGBT’s from blowing.
Confirm that bias voltage at Cl38 is 18-19V. Open or missing D63, 64, 65, 66, 67 or R349 will break the feedback to U16 and cause overvoltage.
D78, D79, R358, R359, gate drive coupling comonents, check after removing blown IGBT’s.
Ul8, lR2110 high-side gate driver, Fault current when low-side IGBT shorts to upper rail. Such currents also typically damage the gate coupling parts noted above.
Ul9, 3525 controller, Blows from currents shorted thru Ul8, or possibly by overvoltage on the supply rail SOMETIMES Ul4, 556, powered from 5V output of 3525, which may fail high when 3525 fails.
RARELY Ul3, which has fairly high supply voltage ratings.
PROBABLE CAUSES OF MASSIVE IGBT FAULTS
SHORTS IN CONTROL CIRCUIT.
The parts operate well within their ratings and should hold up well in the field. The usual cause of failure is when both IGBT’s turn on at once, shorting Pri-Hi to Pri-Lo. This occurs when something causes the drive signal to one part to remain on when the other part is supposed to turn on. Shorts from solder or debris are one obvious cause.
SHORTS IN THE LOAD.
Although there is peak current shutdown, shorts in the power amplifier transistors or secondary-side supply components can cause currents to increase too quickly to prevent damage.
OVERVOLTAGE ON THE BIAS SUPPLY.
If the TOP-210 bias supply fails to operate, no harm occurs, the unit simply does not operate.
However, open circuit (missing part) in several key components can cause the Bias supply voltage to be much too high, This blows the 2110 and thus the IGBT’s.
QUICK TEST OF BIAS SUPPLY.
Ramp the AC voltage up slowly to 25% of regular voltage (30V for 12OV unit). If the bias supply is working normally, the green “power” LED should come on between 30 and 35Vz with its usual, steady “half-brighr start-up level. If the LED comes on at 2OV, or not until 5OV, or blinks, DO NOT RAISE VOLTAGE PAST 60V until you have measured the bias voltage. The switching will not start until you reach 9OV, so you can save the IGBT’s from blowing.
Confirm that bias voltage at Cl38 is 18-19V. Open or missing D63, 64, 65, 66, 67 or R349 will break the feedback to U16 and cause overvoltage.
QUICK TEST OF BIAS SUPPLY.
Ramp the AC voltage up slowly to 25% of regular voltage (30V for 12OV unit). If the bias supply is working normally, the green “power” LED should come on between 30 and 35V, with its usual, steady “half-brighf’ start-up level.
CAUTION: if the LED comes on at 2OV, or not until 5OV, or blinks, DO NOT RAISE VOLTAGE PAST 60V until you have measured the bias voltage. The switching will not start until you reach9OV, so you can save the IGBT’s from blowing.
Confirm that bias voltage at Cl38 is 18-19V.
BIAS SUPPLY VOLTAGE MUCH TOO HIGH
D63,64,65,66,67 or R349 open or missing -- breaks feedback to U16
NO BIAS SUPPLY VOLTAGE
U16 missing or blown.
Tl missing, reversed, or open primary
D62 open or missing.
BIAS VOLTAGE ERRORS
The exact voltage is controlled by the feedback through D63, 64_ 65, 66, 67 and R349 as follows:
Cl38 is the “+18V’ rail with about 18.8V typical.
D63, 64, 65 each subtract a diode drop (0.7V) from Cl38.
Cl39 , is the “+16V” rail with about 16.6V typcial.
D66, a IOV zener diode , plus diode D67, subtract about 1 IV from +16.6V.
R349 subtracts about 0.5V, bringing the net voltage at Ul6, feedback pin 4, to about 5.lV.
Ul6 uses this feedback to adjust the “on” time at pin 5, in order toraise or lower the flyback voltage charging Cl38 and thus maintain regulation of the +16V and +18V supplies C142, R356, and R349 form a closed-loop stability circuit which prevents the regulated voltage from “hunting”.
Q99 and associated R374 reduce the voltage of the Bias supply by 33% when the AC voltage is turned off.
This prevents the Power LED from showing at half brightness after turn-off, since U16 continues to run from the main filters for some time after shut down R375 and 376 sense the output of Ul3:3, the “Loss of AC” comparator, and cause Q99 to turn on. If Q99 is shorted, the bias voltages will remain 33% low when AC is turned on.
Ramp the AC voltage up slowly to 25% of regular voltage (30V for 12OV unit). If the bias supply is working normally, the green “power” LED should come on between 30 and 35V, with its usual, steady “half-brighf’ start-up level.
CAUTION: if the LED comes on at 2OV, or not until 5OV, or blinks, DO NOT RAISE VOLTAGE PAST 60V until you have measured the bias voltage. The switching will not start until you reach9OV, so you can save the IGBT’s from blowing.
Confirm that bias voltage at Cl38 is 18-19V.
BIAS SUPPLY VOLTAGE MUCH TOO HIGH
D63,64,65,66,67 or R349 open or missing -- breaks feedback to U16
NO BIAS SUPPLY VOLTAGE
U16 missing or blown.
Tl missing, reversed, or open primary
D62 open or missing.
BIAS VOLTAGE ERRORS
The exact voltage is controlled by the feedback through D63, 64_ 65, 66, 67 and R349 as follows:
Cl38 is the “+18V’ rail with about 18.8V typical.
D63, 64, 65 each subtract a diode drop (0.7V) from Cl38.
Cl39 , is the “+16V” rail with about 16.6V typcial.
D66, a IOV zener diode , plus diode D67, subtract about 1 IV from +16.6V.
R349 subtracts about 0.5V, bringing the net voltage at Ul6, feedback pin 4, to about 5.lV.
Ul6 uses this feedback to adjust the “on” time at pin 5, in order toraise or lower the flyback voltage charging Cl38 and thus maintain regulation of the +16V and +18V supplies C142, R356, and R349 form a closed-loop stability circuit which prevents the regulated voltage from “hunting”.
Q99 and associated R374 reduce the voltage of the Bias supply by 33% when the AC voltage is turned off.
This prevents the Power LED from showing at half brightness after turn-off, since U16 continues to run from the main filters for some time after shut down R375 and 376 sense the output of Ul3:3, the “Loss of AC” comparator, and cause Q99 to turn on. If Q99 is shorted, the bias voltages will remain 33% low when AC is turned on.
REPLACING BLOWN TOP-210.
If U16 has blown, check T-l for continuity after removing Ul6. Its primary may be open. Pins I-2 It should measure about 15 ohms.
If U16 has blown, check T-l for continuity after removing Ul6. Its primary may be open. Pins I-2 It should measure about 15 ohms.
Replacing Blown Output Transistors
OUTPUT TRANSISTOR SHORTED
Short in one device tend to cause the opposing device to blow as well,
If an output transistor shorts:
Drive transistor will be shorted (Q26, Q27, Q71, Q72)
CXNne transistors will short in pairs (Q39 & Q40, Q36 & Q37, Q84 8 Q85, Q81 8, Q82)
The rest will short in fours (Q28, Q29, Q34 & Q35; Q73, Q74, Q79 8, Q80)
IGBT’s 8 their associated components may fail
Short in one device tend to cause the opposing device to blow as well,
If an output transistor shorts:
Drive transistor will be shorted (Q26, Q27, Q71, Q72)
CXNne transistors will short in pairs (Q39 & Q40, Q36 & Q37, Q84 8 Q85, Q81 8, Q82)
The rest will short in fours (Q28, Q29, Q34 & Q35; Q73, Q74, Q79 8, Q80)
IGBT’s 8 their associated components may fail
CHECK EMITTER AND BASE RESISTORS WHILE DEVICES ARE REMOVED,
Each output transistor has an associatated pair of 0.47 ohm resistors in parallel.
Each BANK of output transistors has a 15 ohm resistor from base to rail (emitter bus).
Each output transistor has an associatated pair of 0.47 ohm resistors in parallel.
Each BANK of output transistors has a 15 ohm resistor from base to rail (emitter bus).
Audio Outputs Troubleshooting Current Limit
WEAK CURRENT LIMIT -- PREMATURE CLIPPING.
The usual symptom of weak output current is premature clipping of one or more peaks of the audio voltage.
This could be caused by missing step, weak current limit, or dead output section.
PREMATURE CLIPPING at 60% VOLTAGE, SIMILAR AT ALL IMPEDANCES:
This points to a step problem (2402, 3002 only). See Step Troubleshooting.
If the amplifier reaches full voltage at 8 ohms, but prematurely clips at 4 ohms or 2 ohms, we can assume the step is OK but the output current is too low (see below).
NO OUTPUT AT ALL ON ONE POLARITY.
This indicates complete failure (open circuit) in the circuit leading to the dead output polarity. Check the series components in the current splitter for missing or open:
Ch 1: Positive, Ql9, R381, Q20, R70, DIO, D14. Negative: Q22, R382, Q21, R71, Dl I, D12
Ch 2: Positive, Q64, R383, Q65, Rl93, D35, D39. Negative: Q67, R384, Q66, Rl94, D36, D37
CONSTANT, PREMATURE CLIPPING, WORSE AT LOW IMPEDANCES.
First, check the clamping voltages on C21 (Ch I+), C22 (Ch I-), C56 (Ch 2+), C57 (Ch 2-), as shown in table below. At idle, all four voltages should all be similar. If one is out, check parts according to the
following table. The exact voltage varies with temperature. Look for the mismatching value on the weak cell.
A too-low voltage causes early clamping of that output section, as explained in the previous several pages. If the voltage is correct and current is still low, also check for missing - unsoldered output device, or emitter resistors.
The usual symptom of weak output current is premature clipping of one or more peaks of the audio voltage.
This could be caused by missing step, weak current limit, or dead output section.
PREMATURE CLIPPING at 60% VOLTAGE, SIMILAR AT ALL IMPEDANCES:
This points to a step problem (2402, 3002 only). See Step Troubleshooting.
If the amplifier reaches full voltage at 8 ohms, but prematurely clips at 4 ohms or 2 ohms, we can assume the step is OK but the output current is too low (see below).
NO OUTPUT AT ALL ON ONE POLARITY.
This indicates complete failure (open circuit) in the circuit leading to the dead output polarity. Check the series components in the current splitter for missing or open:
Ch 1: Positive, Ql9, R381, Q20, R70, DIO, D14. Negative: Q22, R382, Q21, R71, Dl I, D12
Ch 2: Positive, Q64, R383, Q65, Rl93, D35, D39. Negative: Q67, R384, Q66, Rl94, D36, D37
CONSTANT, PREMATURE CLIPPING, WORSE AT LOW IMPEDANCES.
First, check the clamping voltages on C21 (Ch I+), C22 (Ch I-), C56 (Ch 2+), C57 (Ch 2-), as shown in table below. At idle, all four voltages should all be similar. If one is out, check parts according to the
following table. The exact voltage varies with temperature. Look for the mismatching value on the weak cell.
A too-low voltage causes early clamping of that output section, as explained in the previous several pages. If the voltage is correct and current is still low, also check for missing - unsoldered output device, or emitter resistors.
Audio Power Stage, Current Limit Troubleshooting
CURRENT LIMITS WHICH COLLAPSE PREMATURELY.
An immediate collapse of ALL current limits at full power could be premature triggering of “Power Supply Cutback”, which is described in the section below. Cutback after several seconds of full power operation at two ohms is normal.
Cutback of one or more output sections, at full temperature, while approaching full power into two ohms is also normal. However, cutbacks at 4 ohms, or when cold into two ohms, indicate problems with the transistor power measuring circuitry.
An immediate collapse of ALL current limits at full power could be premature triggering of “Power Supply Cutback”, which is described in the section below. Cutback after several seconds of full power operation at two ohms is normal.
Cutback of one or more output sections, at full temperature, while approaching full power into two ohms is also normal. However, cutbacks at 4 ohms, or when cold into two ohms, indicate problems with the transistor power measuring circuitry.
CLAMPING VOLTAGES COLLAPSE TOO SOON.
TROUBLESHOOTING “POWER SUPPLY CUTBACK”.
As noted in the section on Power Supply, the amplifier’s current limit cuts back when necessary to protect the power supply. Because the Observed effect is a reduced output voltage, in response to prolonged operation above the long-term current limit, we commonly refer to this behavior as “power supply cutback”, but we must remember that it is actually amp/Tier current hmifhg in response to an overload signal sent from the power supply. Full power operation into 2 ohms (both channels) should produce a 50% cutback of current after several seconds.
If both channels of the amplifier fail to cut back after about 3 seconds, 2-ohms, both channels driven, the cutback signal is probably missing. CAUTION: Prolonged operation under these conditions could blow IGBT or burn out Cl44. Test for 6-10 seconds maximum.
Check the output (secondary side) pins of U17 (sh 4). Confirm presence of +6V on pin 5. Pin 4 should normally be at about OV, and go high (I-5V) after 3 seconds at full power, If Ul7-pin 4 does not go high, check U17 itself. If it appears OK, trace the circuitry driving U17 (PRIMARY SIDE, CAUTION). Check for continuity through L6:2 to Pri-Lo, check missing or open R343, D61, Q95, R347, all of which drive optocoupler U17. A short in R346 or Cl31 will also prevent drive to U17.
If Ul7-pin 4 goes high on schedule, and BOTH channels fail to cut back, trace voltage on“PS_OL” bus to R273 (sh 3), which connects to “MUTE+” bus. Continue tracing voltage on MUTE+ to Ql6 (sh I) and Q61 (sh 2). If only ONE channel fails to cut back, look for missing Ql6, R 65, Ql7 (sh I) or Q61, Rl88, Q62 (sh 2).
Cl31 controls the speed of cutback.. If missing, the amplifier current limits will enter cutback almost immediately at or above full power, 4 ohms.
As noted in the section on Power Supply, the amplifier’s current limit cuts back when necessary to protect the power supply. Because the Observed effect is a reduced output voltage, in response to prolonged operation above the long-term current limit, we commonly refer to this behavior as “power supply cutback”, but we must remember that it is actually amp/Tier current hmifhg in response to an overload signal sent from the power supply. Full power operation into 2 ohms (both channels) should produce a 50% cutback of current after several seconds.
If both channels of the amplifier fail to cut back after about 3 seconds, 2-ohms, both channels driven, the cutback signal is probably missing. CAUTION: Prolonged operation under these conditions could blow IGBT or burn out Cl44. Test for 6-10 seconds maximum.
Check the output (secondary side) pins of U17 (sh 4). Confirm presence of +6V on pin 5. Pin 4 should normally be at about OV, and go high (I-5V) after 3 seconds at full power, If Ul7-pin 4 does not go high, check U17 itself. If it appears OK, trace the circuitry driving U17 (PRIMARY SIDE, CAUTION). Check for continuity through L6:2 to Pri-Lo, check missing or open R343, D61, Q95, R347, all of which drive optocoupler U17. A short in R346 or Cl31 will also prevent drive to U17.
If Ul7-pin 4 goes high on schedule, and BOTH channels fail to cut back, trace voltage on“PS_OL” bus to R273 (sh 3), which connects to “MUTE+” bus. Continue tracing voltage on MUTE+ to Ql6 (sh I) and Q61 (sh 2). If only ONE channel fails to cut back, look for missing Ql6, R 65, Ql7 (sh I) or Q61, Rl88, Q62 (sh 2).
Cl31 controls the speed of cutback.. If missing, the amplifier current limits will enter cutback almost immediately at or above full power, 4 ohms.
SHORT CIRCUIT CURRENT DOESN’T CUT BACK.
CAUTION: DO NOT MAINTAIN A SHORTED LOAD IF CUTBACK FAILS TO OCCUR WITHIN 1 SECOND.
It will be necessary to measure the output current with a DC current probe, or by noting the voltage across a low value resistance with a DC scope, in order to determine which output cell is failing to cut back. Failure to cut back could indicate either lack of clamping, or lack of voltage cutback. Measure the voltage on the respective clamp capacitor. If the voltage decreases, but current limiting does not cut back, check the clamping transistor.
CAUTION: DO NOT MAINTAIN A SHORTED LOAD IF CUTBACK FAILS TO OCCUR WITHIN 1 SECOND.
It will be necessary to measure the output current with a DC current probe, or by noting the voltage across a low value resistance with a DC scope, in order to determine which output cell is failing to cut back. Failure to cut back could indicate either lack of clamping, or lack of voltage cutback. Measure the voltage on the respective clamp capacitor. If the voltage decreases, but current limiting does not cut back, check the clamping transistor.
Troubleshooting Thermal Tracking
MOUNTING PROBLEMS WITH IOK SENSING NTC.
The thermal sensing for fan and bias tracking depends on a 1 OK NTC which is mounted in a hole in the heat sink. The hole is filled with thermal grease to improve coupling. If the NTC is not straight while mounting the heat sink, it may short out against the side of its hole. It is mounted on a standoff which protrudes into the hole, so this should not occur if care is taken while installing the heatsink. If shorted to the heat sink, the amplifier output voltage is coupled to the NTC. If the short is to the grounded lead of the NTC, it may not damage anything. If to the other end, a large voltage is put across the NTC which will probably damage it.
SHORT FROM “LIVE” END OF NTC TO HEAT SINK:
Replace affected NTC
BE AWARE: This can short (relatively) quietly and then appear to be operating normally. Poor bias tracking can indicate this problem. Sometimes this short will not happen until the amplifier is driven past IV input.
NTC BENT OVER AND SHORTED TO DRIVER TRANSISTORS.
May touch Ql9, Q26, or Q64, Q71.
This causes severe overcurrent to the affected output cell, possibly damaging the parts in series with the shorted transistor. It may also blow the power supply
The thermal sensing for fan and bias tracking depends on a 1 OK NTC which is mounted in a hole in the heat sink. The hole is filled with thermal grease to improve coupling. If the NTC is not straight while mounting the heat sink, it may short out against the side of its hole. It is mounted on a standoff which protrudes into the hole, so this should not occur if care is taken while installing the heatsink. If shorted to the heat sink, the amplifier output voltage is coupled to the NTC. If the short is to the grounded lead of the NTC, it may not damage anything. If to the other end, a large voltage is put across the NTC which will probably damage it.
SHORT FROM “LIVE” END OF NTC TO HEAT SINK:
Replace affected NTC
BE AWARE: This can short (relatively) quietly and then appear to be operating normally. Poor bias tracking can indicate this problem. Sometimes this short will not happen until the amplifier is driven past IV input.
NTC BENT OVER AND SHORTED TO DRIVER TRANSISTORS.
May touch Ql9, Q26, or Q64, Q71.
This causes severe overcurrent to the affected output cell, possibly damaging the parts in series with the shorted transistor. It may also blow the power supply
Replace affected NTC, drive transistors, check components in
series with drive transistor
Ch 1: Ql9 shorted, check, R381, Q20, R70, DIO, D14.
Q26 shorted, check ALL outputs and opposing driver transistor on this channel.
Q26 shorted, check ALL outputs and opposing driver transistor on this channel.
Ch 2: Q64 shorted, check R383, Q65, Rl93, D35, D39.
Q71 shorted, check ALL outputs and opposing driver transistor on this channel.
Q71 shorted, check ALL outputs and opposing driver transistor on this channel.
Audio Output, Troubleshooting Stability Feedback
HIGH FREQUENCY OSCILLATIONS
SEVERE-DRAWS CURRENT-GROSS DISTORTION
C27 (62) missing or wrong, or series R367 (368)
C 25, 26 (60, 61) missing
Secondary filter capacitors missing or open (unlikely that ALL are defective).
SEVERE, BUT DOES NOT DRAW LARGE CURRENT
R22 (146) open.
MARGINAL -- MAY APPEAR ONLY AS EXCESS DISTORTION
Cl4 (49) missing
Cl 6 (50) missing or too large
C28 (63) missing
C 25 or 26 (60 or 61) missing or too large
Cl95 or 196 missing (input board).
EXCESSIVE OSCILLATION JUST BELOW CLIPPING, 2-4 ohms
Cl 7 (52) missing
NOTE: about 0.1% oscillation right below clipping at 2 ohms is normal.
EXCESSIVE SWITCHING INTERFERENCE.
Switching interference may LOOK like an instability, however it is at a much lower frequency (1lOkH.z) than most instabilities. It will be more visible at low frequencies (2OOHz) and at lower impedances.
Missing jumper at R224.
Missing Cl29, 134 on output board.
Grounds not connected to chassis at output board and front chassis mounting screw.
SEVERE-DRAWS CURRENT-GROSS DISTORTION
C27 (62) missing or wrong, or series R367 (368)
C 25, 26 (60, 61) missing
Secondary filter capacitors missing or open (unlikely that ALL are defective).
SEVERE, BUT DOES NOT DRAW LARGE CURRENT
R22 (146) open.
MARGINAL -- MAY APPEAR ONLY AS EXCESS DISTORTION
Cl4 (49) missing
Cl 6 (50) missing or too large
C28 (63) missing
C 25 or 26 (60 or 61) missing or too large
Cl95 or 196 missing (input board).
EXCESSIVE OSCILLATION JUST BELOW CLIPPING, 2-4 ohms
Cl 7 (52) missing
NOTE: about 0.1% oscillation right below clipping at 2 ohms is normal.
EXCESSIVE SWITCHING INTERFERENCE.
Switching interference may LOOK like an instability, however it is at a much lower frequency (1lOkH.z) than most instabilities. It will be more visible at low frequencies (2OOHz) and at lower impedances.
Missing jumper at R224.
Missing Cl29, 134 on output board.
Grounds not connected to chassis at output board and front chassis mounting screw.
FEEDBACK PROBLEMS: GAIN INCORRECT
Gain of output stage set by R23, 31 (147, 153)
Gain of Ch 1 volume control buffer stage set by RI 1, 16.
Gain of Ch 2 volume control buffer stage set by Rl37, 139. Make sure Q48 is turned on - grounding RI37 at Q48 should not affect gain. Check RI 32 (drives Q48).
Gain of balanced input is set by 4 matched resistors R9, 8, 12, 13 (129, 130, 135, 136). Confirm both sides of balanced input are working. Check R5, 6 (123, 124).
Gain of output stage set by R23, 31 (147, 153)
Gain of Ch 1 volume control buffer stage set by RI 1, 16.
Gain of Ch 2 volume control buffer stage set by Rl37, 139. Make sure Q48 is turned on - grounding RI37 at Q48 should not affect gain. Check RI 32 (drives Q48).
Gain of balanced input is set by 4 matched resistors R9, 8, 12, 13 (129, 130, 135, 136). Confirm both sides of balanced input are working. Check R5, 6 (123, 124).
Audio Output, Troubleshooting Clipping, Limiting
EXCESS STICKING (TOO MUCH DISTORTION DURING CLIP LIMITING)
Cl4 (49) much too large (also causes increased high frequency distortion).
R38 (161) missing.
R38-39 (161-162) have wrong values.
Q9, 10 (54, 55) missing
R34 (157) or R35 (158) missing
Q8 (53) missing.
Cl4 (49) much too large (also causes increased high frequency distortion).
R38 (161) missing.
R38-39 (161-162) have wrong values.
Q9, 10 (54, 55) missing
R34 (157) or R35 (158) missing
Q8 (53) missing.
CLIP LIMITING DOESN’T WORK:
BOTH CHANNELS:
Check U3 missing,
Check U3 supply voltages, +I 3VCL, -13VCL, on C73* 74
ONLY ONE CHANNEL BAD:
Probe output of U2 (7), pin 7 while clipping. If output exceeds 4V during clipping, check :
R38 (161) missing.
R38-39 (161-162) have wrong values.
Q9, 10 (54, 55) missing.
R34 (157) or R35 (158) missing
Q8 (53) missing
If output at pin 7 clamps at 3.5.4V as expected, check parts surrounding U3:
R32 (154) missing
Q7 (52) missing
R28 (151) missing
Q6 (51) missing
RI8 (141) missing
RI9 (142) missing
SW I:1 (1:lO) not making contact
Check each pin on U3
CLIP LIMITING OSCILLATES:
Cl 3 (48) missing.
R21, 27 (144, 150) missing.
EXCESSIVE STEP DISTORTION (STEP GLITCH)
Close scrutiny of the distortion trace, and scope probing of the switched waveform, will help determine the cause of excess step distortion, The step should switch when the output voltage is within IO-12 volts of its respective rail, This switching margin should be fairly constant from20- 20kHz. The switching event itself should be a fairly uniform up and down ramp, moving at about 25 volts/us, therefore taking about 2us to complete its transition.
Close scrutiny of the distortion trace, and scope probing of the switched waveform, will help determine the cause of excess step distortion, The step should switch when the output voltage is within IO-12 volts of its respective rail, This switching margin should be fairly constant from20- 20kHz. The switching event itself should be a fairly uniform up and down ramp, moving at about 25 volts/us, therefore taking about 2us to complete its transition.
Step Switching Too Close to the Rail:
This will cause increased step glitch, especially at low impedances. If present at all frequencies, check the reference voltages:
Negref: 17.5 volts above its intermediate rail: D88, R276-7-8.
PosRef: 20V below its intermediate rails: D87, R256, 257, D53.
This will cause increased step glitch, especially at low impedances. If present at all frequencies, check the reference voltages:
Negref: 17.5 volts above its intermediate rail: D88, R276-7-8.
PosRef: 20V below its intermediate rails: D87, R256, 257, D53.
Confirm correct values in output voltage divider: R48 loaded by
R49-50 (Rl71, and Rl72-173). If present only at high frequencies, check the
value of the speed up capacitor C20 (C55) in the output voltage divider, or
look for slow switching.
Step Chattering.
If the step repeatedly switches on and off, usually at a low frequency, it creates an oscillation burst which increases step glitch at low frequency. The tendency is usually worse at low impedances and low frequencies. 2-3 “false trials” at very low frequency, 2 ohms, is normal, but prolonged bursts of maximum frequency chattering may cause FET failure.
Positive step: check hysteresis resistor R66 (RI 89)
Negative step: check hysteresis resistor R69 (RI 92) and capacitor Cl 87 (Cl 93).
Slow or Fast Switching.
Slow switching reduces step glitch but puts more strain on the FET. Fast switching increases step glitch. The usable limit is17-27 volts/us, If both slopes are equally off speed, check the slope capacitors:
Positive, C30 (C65) and Negative, C29 (C64).
If only one slope is slow, check the resistors and buffer transistors:
Positive step: R78, 79, Dl5, Q30-31 (R201, 202, D40, Q7576)
Negative step: R83, 84, Dl7, Q32-33 (R206, 207, D42, Q77-78).
If the step repeatedly switches on and off, usually at a low frequency, it creates an oscillation burst which increases step glitch at low frequency. The tendency is usually worse at low impedances and low frequencies. 2-3 “false trials” at very low frequency, 2 ohms, is normal, but prolonged bursts of maximum frequency chattering may cause FET failure.
Positive step: check hysteresis resistor R66 (RI 89)
Negative step: check hysteresis resistor R69 (RI 92) and capacitor Cl 87 (Cl 93).
Slow or Fast Switching.
Slow switching reduces step glitch but puts more strain on the FET. Fast switching increases step glitch. The usable limit is17-27 volts/us, If both slopes are equally off speed, check the slope capacitors:
Positive, C30 (C65) and Negative, C29 (C64).
If only one slope is slow, check the resistors and buffer transistors:
Positive step: R78, 79, Dl5, Q30-31 (R201, 202, D40, Q7576)
Negative step: R83, 84, Dl7, Q32-33 (R206, 207, D42, Q77-78).
Step FET Oscillation.
Certain FET types oscillate at extremely high frequency while ramping up and down. This injects interference into the amplifier which increases the step glitch. Such problems are supposed to be found while (dis)qualifying specific FET types. If they crop up in production, Engineering needs to know.
STEP WON’T TURN ON (Premature Clipping)
If the step refuses to switch high, the amp will clip prematurely, at the intermediate rail, at any load. Make sure the clipping is not acually current cutback, usually evident only at 2 ohms. Probe the output voltage and intermediate rail voltages to confirm clip point and lack of step action. Trace the circuit from the step FET back via gate drive to drive circuit to locate cracks, missing part etc. Check DC power on step driver (14V on EACH positive step drivers, 12V on BOTH negative drivers). Check voltage of PosRef (20V below +65V rail) or Negref (17.5V above -65V rail). Look for severe mismatches of the comparator resistor ladders.
Certain FET types oscillate at extremely high frequency while ramping up and down. This injects interference into the amplifier which increases the step glitch. Such problems are supposed to be found while (dis)qualifying specific FET types. If they crop up in production, Engineering needs to know.
STEP WON’T TURN ON (Premature Clipping)
If the step refuses to switch high, the amp will clip prematurely, at the intermediate rail, at any load. Make sure the clipping is not acually current cutback, usually evident only at 2 ohms. Probe the output voltage and intermediate rail voltages to confirm clip point and lack of step action. Trace the circuit from the step FET back via gate drive to drive circuit to locate cracks, missing part etc. Check DC power on step driver (14V on EACH positive step drivers, 12V on BOTH negative drivers). Check voltage of PosRef (20V below +65V rail) or Negref (17.5V above -65V rail). Look for severe mismatches of the comparator resistor ladders.
STEP STUCK ON (Switched Rail Voltage Stuck On Full)
If the positive step is stuck on, (evidenced by permanent high voltage on switched rail) the FET is probably bad, since the positive gate drive cannot sustain DC turn-on due to the bootstrapping. If the negative step is stuck on, it could be a bad FET, or the gate drive circuit could be holding the FET on, which will easily be confirmed by measuring the gate voltage. Malfunctioning gate drive circuitry should be checked as noted above under “Won’t Turn On”.
REPEATED FET FAILURE.
Repeated failure of step FET’s is usually caused by failure to fully switch ON or OFF (lingering in the linear region). The actual failure usually occurs at 2 ohms, where the dissipation is highest. After replacing the FET, the step waveform should be monitored, starting at light load to avoid repeated failure, and advancing briefly to heavier loads while closely watching the waveform. You will need to use an isolated scope probe which allows voltage readings to be taken with respect to the intermediate rails, or to FET sources.
FET Does Not Fully Turn On:
Generally causes problems at low frequency, 2 ohms.
Confirm that the step FET remains fully on for the entire cycle (2OHz). If not, confirm weak gate drive and determine cause.
Weak positive drive: check voltage on C32 (C51) for 14V. Check C31 (C66), low RIO4 (R227),
missing D18 (Dl48). Check R78, Dl5, Q30 (R201, D40, Q75).
Weak negative drive: check voltage on C67, 12V. Check R83, Dl7, Q32 (R206, D42, Q77).
FET Turns On or Off Very Slowly:
Generally causes problems at high frequency, 2 ohms.
If both slopes are equally slow, check the slope capacitors:
Positive, C30 (035) and Negative, C29 (034).
If only one slope is slow, check the resistors and buffer transistors:
Positive step: R78, 79, Dl5, Q30-31 (R201, 202, D40, Q75-76)
Negative step: R83, 84, Dl7, Q32-33 (R206, 207, D42, Q77-78).
Severe Step Oscillation.
Generally observed at low frequency, low impedance, right at threshold.
Positive step: check hysteresis resistorR66 (RI 89)
Negative step: check hysteresis resistor R69 (RI 92) and capacitor Cl 87 (Cl 93)
Troubleshooting DC Fault Shutdown
NORMAL BEHAVIOR OF THE CIRCUIT.
Any amplifier fault which causes a non-symmetrical output, such as premature clipping of one polarity, a missing step, etc, may trigger DC fault shutdown. This indicates normal operation of the circuit
TRACING THE CAUSE OF FALSE TRIGGERS.
If amplifier is shutting down for no apparent cause, the source of the false signal must be found. Be sure the output is checked with a DC coupled scope in order to confirm absence of an actual DC offset. The circuit will trip on DC offsets exceeding about 4V.
The optocoupler’s input can be safely disabled by shorted Ul5, pins I-2 together. This will indicate if false triggering is before or after Ul5.
The output of UlO:l, pin 1 should be monitored. If it goes low during DC shutdown, it is sending the false signal.
SHUTDOWN OCCURS AS SOON AS SWITCHING STARTS.
Disable Ul5 as noted above, determine if there is a DC fault condition. CAUTION: use 50-ohm resistor in series with AC line to limit fault current in case of shorted outputs.
If amplifier output looks OK, check UlO:l output. If low, check voltage on pins 2 and 3 UlO:l, pin 2: should be zero (no signal)
UlO:l, pin 31 should be about 2V! set by R243, 244, 245.
Check R348 at Ul.5.
SHUTDOWN OCCURS ABOVE ABOUT 4V OUTPUT:
Q87, C7, R240 or D48 missing.
Confirm D48 is pulled low (-13V), holding Q87 on. If not, check RI 17, 118, Q42
NOTE: this control voltage responds to the Br Mono switch, pole 7.
Check R348 at Ul5.
Bad connection at step diodes (D21, D22, D46, D47)
Fan Speed Troubleshooting
FAN STUCK HIGH:
Q88 or Q89 failed
R30 or Rl55, Thermal Sense NTC, shorted to heatsink
R266 or R271 missing.
FAN DOESN’T RUN
Check fan voltage, should be 1 IV when cold, 29V when hot.
Voltage OK -- replace fan.
No Voltage: check Q91, 89, 90, R264 and 265 missing. Confirm that +/-15V is reaching circuit.
Q88 or Q89 failed
R30 or Rl55, Thermal Sense NTC, shorted to heatsink
R266 or R271 missing.
FAN DOESN’T RUN
Check fan voltage, should be 1 IV when cold, 29V when hot.
Voltage OK -- replace fan.
No Voltage: check Q91, 89, 90, R264 and 265 missing. Confirm that +/-15V is reaching circuit.
See the circuit diagram by previous post
