Hi,
I'm having problems with my amp again and was hoping for a few pointers.
I started to get popping/clicking on the right channel.
I unplugged the speaker from that channel and finished my film.
When I checked it next day, both of the 6.3 fuses were blown.
Only other symptoms I can think is it occasionally used to click when powered up, but this has gone since I'm only using the left channel.
Any tips?
I'm happy to reward any advice that leads to a fix.
I'm an electrical engineer, but electronics has always been my weak point (well that and personality).
No oscilloscope, but decent multimeters.
The schematic is too large for forum upload, but I can send it to anyone.
Also if you search for rotel rb-991 technical manual it comes up in images.
this link may work
rotel rb-991 technical manual - Google Search
I'm having problems with my amp again and was hoping for a few pointers.
I started to get popping/clicking on the right channel.
I unplugged the speaker from that channel and finished my film.
When I checked it next day, both of the 6.3 fuses were blown.
Only other symptoms I can think is it occasionally used to click when powered up, but this has gone since I'm only using the left channel.
Any tips?
I'm happy to reward any advice that leads to a fix.
I'm an electrical engineer, but electronics has always been my weak point (well that and personality).
No oscilloscope, but decent multimeters.
The schematic is too large for forum upload, but I can send it to anyone.
Also if you search for rotel rb-991 technical manual it comes up in images.
this link may work
rotel rb-991 technical manual - Google Search
http://srv2.umlib.com/1129f1c84432965a0f8a682c99407786/Rotel-RB991-pwr-sm.pdf-3.png
An externally hosted image should be here but it was not working when we last tested it.
This may be easier for many here:
Rotel RB-991 - Manual - Stereo Power Amplifier - HiFi Engine
As only a preliminary suggestion - In the UK, a 6.3A fuse would also be fitted to the mains active input and both DC supply rails of each power amplifier or 5 fuses all the same rating. I assume you are referring to the DC supply rail fuses in the left channel only. It would seem likely with such a failure, that at least some pairs of power transistors would have shorted but their replacements will require a bit of expertise and professional matching of devices for a proper job. I'd also look into speaker or even speaker lead problems like shorts there, since it takes a lot of output noise/current to blow 6A fuses on a 200W amplifier.
Rotel RB-991 - Manual - Stereo Power Amplifier - HiFi Engine
As only a preliminary suggestion - In the UK, a 6.3A fuse would also be fitted to the mains active input and both DC supply rails of each power amplifier or 5 fuses all the same rating. I assume you are referring to the DC supply rail fuses in the left channel only. It would seem likely with such a failure, that at least some pairs of power transistors would have shorted but their replacements will require a bit of expertise and professional matching of devices for a proper job. I'd also look into speaker or even speaker lead problems like shorts there, since it takes a lot of output noise/current to blow 6A fuses on a 200W amplifier.
Hi,
Thanks for the tip/link.
I couldn't find a place that would let me share the link.
It was fuse 602/604.
So this is the +-73 V feed to the board.
When I design stuff I keep all AC/DC electrically separated; this amplifier stuff seems like some weird dark art to me!
They blew after the speakers were disconnected, so not speaker/speaker leads thankfully. I have shorted the leads before, it's quite exciting when that happens. All the output transistors from this channel self-disassembled and scattered themselves around the inside of the case...not easy to miss. Due to that, I modified the amp to have Neutrik Speakon connectors to prevent it happening again.
So apart from checking the transistors, where else should I look.
Potentially stupid question here, if I found a few dodgy transistors, would it be as simple as remove a few. Or do they need to be replaced? Is it just a maximum capacity thing or is there more too it?
Thanks again,
P
Thanks for the tip/link.
I couldn't find a place that would let me share the link.
It was fuse 602/604.
So this is the +-73 V feed to the board.
When I design stuff I keep all AC/DC electrically separated; this amplifier stuff seems like some weird dark art to me!
They blew after the speakers were disconnected, so not speaker/speaker leads thankfully. I have shorted the leads before, it's quite exciting when that happens. All the output transistors from this channel self-disassembled and scattered themselves around the inside of the case...not easy to miss. Due to that, I modified the amp to have Neutrik Speakon connectors to prevent it happening again.
So apart from checking the transistors, where else should I look.
Potentially stupid question here, if I found a few dodgy transistors, would it be as simple as remove a few. Or do they need to be replaced? Is it just a maximum capacity thing or is there more too it?
Thanks again,
P
Solution!
I had one of these in the workshop with same symptoms - blowing 6.3A fuses.
The fault was traced down to the glue that is applied in manufacture to the 1000mF 100v capacitors on the drive boards. The glue becomes conductive and tracks across the board. one 1N4148 diode had corroded and upsets the drive conditions.
Clean all the glue off the board and replace any damaged parts.
Perfect now 🙂
I had one of these in the workshop with same symptoms - blowing 6.3A fuses.
The fault was traced down to the glue that is applied in manufacture to the 1000mF 100v capacitors on the drive boards. The glue becomes conductive and tracks across the board. one 1N4148 diode had corroded and upsets the drive conditions.
Clean all the glue off the board and replace any damaged parts.
Perfect now 🙂
This amp is basically the same as the RB-1080. This amplifier has many design issues I can write a book about it 🙂.
Firstly, it is not stable at high frequencies and relies on the zobal network on the speaker terminals to attenuate the oscillations. If you have a high enough transient on the input it causes the output stage transistors to oscillate and cross conduct drawing a huge amount of current through the output stage transistors thus blowing the fuses. They fixed this issue in later designs by placing base resistors on each output stage transistors to isolate and dampen the stray resonant section using 22 ohm resistors, have a look at the RB-1582 amplifier circuit. They also added an extra zobal network.
Secondly, they do not use a dampened output inductor (inductor and resistor in parrallel) on the output of the amp to protect against capacitive loads that may be present on cables especially long run cables or cable that are badly designed. This again could cause instability.
Thirdly, these amps from Rotel, the VA stage (voltage amplifier) transistors mounted on small heat sinks on the amplifier PCB get very hot compared to the right channel. This is a main cable layout and positioning issue. They placed the mains cable inlet close to the left channel input without any shielding. The amplifier has high gain at this location and picks up the AC mains, amplifies it and feeds back through the VA and output stage back to the differential input stage. As a result, it over heats. If your differential pair input stage is not well matched, you may hear a light hum on the left channel. A fix for this is to use braided copper sleeving on the mains cable and ground this to the chassy.
Fourthly, the DC power cables and grounds are long as hell. This gives too much inductance as a result the high frequencies suffer and lack control due to improper dampening and lack of supply due to high impedance. This lead to a blended slush in the high end giving hearing fatigue due to odd order harmonics manifesting.
Fifthly, the input differential pairs where not matched well making odd order harmonics a real ear basher. Make sure these transistors are insulated and thermally coupled to each other otherwise your quiescent current will drift with the air temperature.
Sixth, mismatched input and feedback impedance (biasing current for differential amplifier) which leads to relatively large DC offset on the output (30 to 50mV). This may not be much of an issue, but good practice to minimize to reduce any drift issues in the differential amplifier.
I could go on and on 🙂 I know all these issues since I bought a second hand RB-1080 and saw all these nightmares. I corrected each issue step by step. The amp sounds completely different, fuller bass, well controlled mids and highs, do not suffer from hearing fatigue any more. Well I made significant changes its not longer the same.
Firstly, it is not stable at high frequencies and relies on the zobal network on the speaker terminals to attenuate the oscillations. If you have a high enough transient on the input it causes the output stage transistors to oscillate and cross conduct drawing a huge amount of current through the output stage transistors thus blowing the fuses. They fixed this issue in later designs by placing base resistors on each output stage transistors to isolate and dampen the stray resonant section using 22 ohm resistors, have a look at the RB-1582 amplifier circuit. They also added an extra zobal network.
Secondly, they do not use a dampened output inductor (inductor and resistor in parrallel) on the output of the amp to protect against capacitive loads that may be present on cables especially long run cables or cable that are badly designed. This again could cause instability.
Thirdly, these amps from Rotel, the VA stage (voltage amplifier) transistors mounted on small heat sinks on the amplifier PCB get very hot compared to the right channel. This is a main cable layout and positioning issue. They placed the mains cable inlet close to the left channel input without any shielding. The amplifier has high gain at this location and picks up the AC mains, amplifies it and feeds back through the VA and output stage back to the differential input stage. As a result, it over heats. If your differential pair input stage is not well matched, you may hear a light hum on the left channel. A fix for this is to use braided copper sleeving on the mains cable and ground this to the chassy.
Fourthly, the DC power cables and grounds are long as hell. This gives too much inductance as a result the high frequencies suffer and lack control due to improper dampening and lack of supply due to high impedance. This lead to a blended slush in the high end giving hearing fatigue due to odd order harmonics manifesting.
Fifthly, the input differential pairs where not matched well making odd order harmonics a real ear basher. Make sure these transistors are insulated and thermally coupled to each other otherwise your quiescent current will drift with the air temperature.
Sixth, mismatched input and feedback impedance (biasing current for differential amplifier) which leads to relatively large DC offset on the output (30 to 50mV). This may not be much of an issue, but good practice to minimize to reduce any drift issues in the differential amplifier.
I could go on and on 🙂 I know all these issues since I bought a second hand RB-1080 and saw all these nightmares. I corrected each issue step by step. The amp sounds completely different, fuller bass, well controlled mids and highs, do not suffer from hearing fatigue any more. Well I made significant changes its not longer the same.
Rotel has made some improvements on there newer products.
Ok, Rotel has made improvements to their newer models. The RB-1582MKII has made most of the corrections that were needed. If you have an older amplifier from Rotel that match the RB-1582MKII (RB-1582, RB-1080, RB-991), you can update your older model and follow what they have done on the RB-1582MKII.
Please note that only qualified technicians and engineers should attempt the modifications. The amplifiers should be tested on current limited supplies before supplying the actual supply to prevent damage.
Notability what to change is:
1. Placing a gate resistor of 4.7 ohms on each base of the output stage transistors to dampen the stray resonant circuit that exists in this part of the circuit (multiple paralleled transistors tied together). This is the main issue that blow the fuses.
2. Replacing the input bias resistor to match the feedback resistor, in most if not all cases, the feedback resistor is 15.4K, so replace the input bias resistor, which is 33K in most cases if not all, with a 15K or 15.4K resistor. This will reduce the output DC offset voltage.
3. Replacing and gain matching the differential stage transistors. The transistors (KSC1845-F and KSA992-F) shown in the RB-1582MKII are good replacements with higher gain. Its good practice to buy 25 to 100 transistors of each type and match them as close as possible (both differential pairs on each channel should have the same gains). Once placed on the PCB, they need to be thermally coupled to one another with some thermal compound and then thermally insulated (also insulate the leads of the transistor). Failing to do so will make the output stage bias setting drift with temperature). By gain matching, the DC offset will reduce further and the THD will also reduce.
4. Place a copper braided sleeve on the mains cable near the left channel input. This includes the loop with the fuse holder (include the fuse holder). The copper braiding must be grounded at the grounding point to the case/chassis. This usually fixes any small hum that might be present as well as making both channels run at the same temperature.
5. Replace the fuses to 8A slow blow type.
Some recommendations.
1. Replace the aluminum electrolytic capacitor with a film capacitor on the input.
2. Replace the aluminum electrolytic capacitor with a film capacitor on the feedback capacitor. (you may have to mount the capacitor off the PCB and route a twisted cable to the PCB since this capacitor tends to be large)
3. Replace the aluminum electrolytic capacitor with a film capacitor across the bias transistor, place this on the PCB rather on the transistor leads themselves.
4. Place a 10uF 50V film capacitor across the current source diodes.
5. Place a 1uF 50V/100V capacitor across the driver stage resistor (47 ohm resistor).
There are some further replacements and additions that can change the tonal quality of the amplifier by adjusting the power supply delivery parts, i.e the wiring and adding film capacitors to the rail supplies.
Ok, Rotel has made improvements to their newer models. The RB-1582MKII has made most of the corrections that were needed. If you have an older amplifier from Rotel that match the RB-1582MKII (RB-1582, RB-1080, RB-991), you can update your older model and follow what they have done on the RB-1582MKII.
Please note that only qualified technicians and engineers should attempt the modifications. The amplifiers should be tested on current limited supplies before supplying the actual supply to prevent damage.
Notability what to change is:
1. Placing a gate resistor of 4.7 ohms on each base of the output stage transistors to dampen the stray resonant circuit that exists in this part of the circuit (multiple paralleled transistors tied together). This is the main issue that blow the fuses.
2. Replacing the input bias resistor to match the feedback resistor, in most if not all cases, the feedback resistor is 15.4K, so replace the input bias resistor, which is 33K in most cases if not all, with a 15K or 15.4K resistor. This will reduce the output DC offset voltage.
3. Replacing and gain matching the differential stage transistors. The transistors (KSC1845-F and KSA992-F) shown in the RB-1582MKII are good replacements with higher gain. Its good practice to buy 25 to 100 transistors of each type and match them as close as possible (both differential pairs on each channel should have the same gains). Once placed on the PCB, they need to be thermally coupled to one another with some thermal compound and then thermally insulated (also insulate the leads of the transistor). Failing to do so will make the output stage bias setting drift with temperature). By gain matching, the DC offset will reduce further and the THD will also reduce.
4. Place a copper braided sleeve on the mains cable near the left channel input. This includes the loop with the fuse holder (include the fuse holder). The copper braiding must be grounded at the grounding point to the case/chassis. This usually fixes any small hum that might be present as well as making both channels run at the same temperature.
5. Replace the fuses to 8A slow blow type.
Some recommendations.
1. Replace the aluminum electrolytic capacitor with a film capacitor on the input.
2. Replace the aluminum electrolytic capacitor with a film capacitor on the feedback capacitor. (you may have to mount the capacitor off the PCB and route a twisted cable to the PCB since this capacitor tends to be large)
3. Replace the aluminum electrolytic capacitor with a film capacitor across the bias transistor, place this on the PCB rather on the transistor leads themselves.
4. Place a 10uF 50V film capacitor across the current source diodes.
5. Place a 1uF 50V/100V capacitor across the driver stage resistor (47 ohm resistor).
There are some further replacements and additions that can change the tonal quality of the amplifier by adjusting the power supply delivery parts, i.e the wiring and adding film capacitors to the rail supplies.
Last edited:
Thanks for the help. I think I might win an award for the slowest response ever!
The repair was a bit above me, however someone in the village is happy to help =)
We've removed the glue off the board, and checked all the transistors.
The 6.3A fuses still blow on power up.
I've checked everything else visually, nothing looks blown.
What else would you check?
The repair was a bit above me, however someone in the village is happy to help =)
We've removed the glue off the board, and checked all the transistors.
The 6.3A fuses still blow on power up.
I've checked everything else visually, nothing looks blown.
What else would you check?
I would check the power supply caps, if you haven't already. They do occasionally fail short and that'll blow the fuse and possibly the rectifier too. A shorted output stage would do it too, but you said you checked all the transistors.
Depending on the failure, you may be able to bring it up using a variac. Increase the mains voltage slowly until you get some voltage on the power supply but not so much that the fuses blow. Then measure the supply voltage of the amp at various spots. Compare the working channel with the dead one.
Tom
Depending on the failure, you may be able to bring it up using a variac. Increase the mains voltage slowly until you get some voltage on the power supply but not so much that the fuses blow. Then measure the supply voltage of the amp at various spots. Compare the working channel with the dead one.
Tom
As it's just the right channel board blowing the fuses, we've checked only that board.
The 100v 1000uF Caps test fine. I think they are c626 and C624
So immediately on power-up, no speakers or input connected, fuses f802 and f804 pop!
Is this the test you suggested?
The 100v 1000uF Caps test fine. I think they are c626 and C624
So immediately on power-up, no speakers or input connected, fuses f802 and f804 pop!
Is this the test you suggested?