This goes especially to anatech. But any electronic pro could help:
I'm repairing a Bryston B4 amp and also having trouble by finding a suitable replacement for the BR6521 and BR6522 power transistors.
In a previous post I found that "2N5631 and 2N6031 are a good pair (anatech)" ( -> same as NTE60 and NTE61).
The Bryston had beed repaired ones before, and MJ802's were replacing all of the BR6521 transistors on one channel. MJ802 corresponds to NTE181, normally used as pair with the NTE180.
What do you think might work best: NTE60/NTE61 or NTE180/NTE181?
Thanks in advance for your reply.
PhantomBox
I'm repairing a Bryston B4 amp and also having trouble by finding a suitable replacement for the BR6521 and BR6522 power transistors.
In a previous post I found that "2N5631 and 2N6031 are a good pair (anatech)" ( -> same as NTE60 and NTE61).
The Bryston had beed repaired ones before, and MJ802's were replacing all of the BR6521 transistors on one channel. MJ802 corresponds to NTE181, normally used as pair with the NTE180.
What do you think might work best: NTE60/NTE61 or NTE180/NTE181?
Thanks in advance for your reply.
PhantomBox
I would presume that you mean "4B", not "B4"
To venture a wild guess, I would presume that they are custom-selected, custom-marked (ergo the "BR" for "BRyston" prefix). Likely, Bryston has the factory select for fairly close parameter tolerances that yields more consistent match between PNP and NPN devices.
The obvious first pass is to replace "BR" with "2N", "MPS", or one of the Japanese transistor prefixes. It doesn't seem to yield anything useful.
You could look at Toshiba and On Semi for their latest audio power transistors and give those a try. You can only guess at match of devices.
Ya know, you could order some from Bryston. I hear they have very good support, and shipping around the world shouldn't be too hard. Try sending them an email. The amp might even be still under warranty.
To venture a wild guess, I would presume that they are custom-selected, custom-marked (ergo the "BR" for "BRyston" prefix). Likely, Bryston has the factory select for fairly close parameter tolerances that yields more consistent match between PNP and NPN devices.
The obvious first pass is to replace "BR" with "2N", "MPS", or one of the Japanese transistor prefixes. It doesn't seem to yield anything useful.
You could look at Toshiba and On Semi for their latest audio power transistors and give those a try. You can only guess at match of devices.
Ya know, you could order some from Bryston. I hear they have very good support, and shipping around the world shouldn't be too hard. Try sending them an email. The amp might even be still under warranty.
Thanks BrianL.
Non of the Japanese prefixes gives me any result (just tried out any combination that I know of).
And asking Bryston for assistance doesn't help me either, I have to get the amp fixed for yesterday.
MJ802 (replacment of a BR5621, the only clue that I have) is a transistor equivalent to NTE181, and the number I can get at a local supplier for that transistor is 2N5882. Counterpart to that would be NTE180, equivalent to 2N5880, I assume.
As "anatech" mentioned in an earlier post, he uses 2N5631 for BR5621 and 2N6031 for BR5622 (equivalent to NTE60 and NTE61, numbers that actually are available to me), I would like his opinion on what pair he'd prefer: 2N5880/5881 or NTE60/61?
Thanks again,
PhantomBox
Non of the Japanese prefixes gives me any result (just tried out any combination that I know of).
And asking Bryston for assistance doesn't help me either, I have to get the amp fixed for yesterday.
MJ802 (replacment of a BR5621, the only clue that I have) is a transistor equivalent to NTE181, and the number I can get at a local supplier for that transistor is 2N5882. Counterpart to that would be NTE180, equivalent to 2N5880, I assume.
As "anatech" mentioned in an earlier post, he uses 2N5631 for BR5621 and 2N6031 for BR5622 (equivalent to NTE60 and NTE61, numbers that actually are available to me), I would like his opinion on what pair he'd prefer: 2N5880/5881 or NTE60/61?
Thanks again,
PhantomBox
Hi PhantomBox,
NTE will not cut it. Sorry. You need to buy real transistors with actual numbers on them.
Go for MJ21193 and MJ21194, they are not too fast. MJ15022 and MJ15023 would work also, but might oscillate in a stacked output circuit. I think I have used them in the past, but you should stick with the slower parts.
Bryston parts from back then were not matched very well. The numbers were simply to help keep their circuit a secret. I had to trace the 4B out back then as they refused to release any information. The only specialty part was the TO-220 drivers that they messed up on and required extended leads to reach the socket. That one gave me a laugh. Imagine how much money that added to the cost that was avoidable. Well, they had their custom part and the other only looked custom.
-Chris
NTE will not cut it. Sorry. You need to buy real transistors with actual numbers on them.
Go for MJ21193 and MJ21194, they are not too fast. MJ15022 and MJ15023 would work also, but might oscillate in a stacked output circuit. I think I have used them in the past, but you should stick with the slower parts.
Bryston parts from back then were not matched very well. The numbers were simply to help keep their circuit a secret. I had to trace the 4B out back then as they refused to release any information. The only specialty part was the TO-220 drivers that they messed up on and required extended leads to reach the socket. That one gave me a laugh. Imagine how much money that added to the cost that was avoidable. Well, they had their custom part and the other only looked custom.
-Chris
Thanks an awefull lot, Chris.
I don't know much about how to interprete the datasheet specs for a specific transistor, but I can try to locate these particular numbers.
By the way: you don't happen to have (or know where to get) the schematics of a 1987 Bryston 4B amp? That must be the 1st issue, or very close to that. I found a whole lot of Bryston amp schematics on their homepage, but the mainboard plans are all '88 and up (I just realized that).
Best regards,
PhantomBox
I don't know much about how to interprete the datasheet specs for a specific transistor, but I can try to locate these particular numbers.
By the way: you don't happen to have (or know where to get) the schematics of a 1987 Bryston 4B amp? That must be the 1st issue, or very close to that. I found a whole lot of Bryston amp schematics on their homepage, but the mainboard plans are all '88 and up (I just realized that).
Best regards,
PhantomBox
Hi PhantomBox,
They should be pretty close. Possibly some component changes, but the basic circuit should be the same. The current drawings should be usable, especially if you mark down the differences.
-Chris
Darn! I fired my hand drawn schematic into the garbage years ago after Bryston put these diagrams up.
They should be pretty close. Possibly some component changes, but the basic circuit should be the same. The current drawings should be usable, especially if you mark down the differences.
-Chris
Hi Todd,
Bryston doesn't bug me anymore, and haven't for years. I was only providing background and some factual info for the question posed. I understand that the SST amps are very good and I'll accept that.
Your choice of 2N3773 and 2N6609 are good ones. I used to use those until 2N6609 became scarce. They seemed to be available again after about 5 years of being scarce. They would be the better choice as the amp was designed and stabilized with those parts to begin with. My concerns with transistor choice was more centered around speed and oscillations.
I'm glad you got a laugh out of my view on early Bryston product. You ought to hear me when I am pissed with them!
Do you still have an old 4B or two? I have some ideas that might really make them sound much better. This is without changing the actual circuitry. If changes were allowed, I would concentrate on the bias circuit for reliability and to allow good sound soon after turn on. Inrush current limiting would be another worthwhile addition.
-Chris
-Chris
Bryston doesn't bug me anymore, and haven't for years. I was only providing background and some factual info for the question posed. I understand that the SST amps are very good and I'll accept that.
Your choice of 2N3773 and 2N6609 are good ones. I used to use those until 2N6609 became scarce. They seemed to be available again after about 5 years of being scarce. They would be the better choice as the amp was designed and stabilized with those parts to begin with. My concerns with transistor choice was more centered around speed and oscillations.
I'm glad you got a laugh out of my view on early Bryston product. You ought to hear me when I am pissed with them!
Do you still have an old 4B or two? I have some ideas that might really make them sound much better. This is without changing the actual circuitry. If changes were allowed, I would concentrate on the bias circuit for reliability and to allow good sound soon after turn on. Inrush current limiting would be another worthwhile addition.
-Chris
-Chris
Hi Chris,
Yes, I still have 2 of them (or 3 I thought, but I can't find the 3rd one). But I'm just using their hardware (case, power supply, heatsinks) for building some of the great DIY amp designs being produced here. Quasi, probably a DX amp, and my own higher voltage bastardization of a Leach amp. Anything else I can find that would enjoy +-75v rails.
The Bryston PC boards are just gathering oxide.
..Todd
Yes, I still have 2 of them (or 3 I thought, but I can't find the 3rd one). But I'm just using their hardware (case, power supply, heatsinks) for building some of the great DIY amp designs being produced here. Quasi, probably a DX amp, and my own higher voltage bastardization of a Leach amp. Anything else I can find that would enjoy +-75v rails.
The Bryston PC boards are just gathering oxide.
..Todd
anatech said:
Sound improvements without changing the circuitry? Not by peeling the plastic off the reservoir caps I hope, or demagnetizing my CDs, or $85 per gram magic potions for bathing my rhodium binding posts ...
I think your intentions for the bias circuit are good. Would one of those aforementioned changes be to thermally compensate the bias more quickly?
Do those big EI transformers really create the need for inrush limiting? I thought only toroids needed that consideration. Do tell!
..Todd
Hi Todd,
Sorry, been "away", I'm not ignoring you.
No, but I'll do some of that if you pay me to. Perhaps we'll send them to Ring Audio? 
Looking at the PCBs on my minds eye (I've seen many), there are many places to improve the sound quality. Some things would need redesign, but I'll bet I can really improve the sound quality of these.
-Chris
Sorry, been "away", I'm not ignoring you.
Yes.
LOL!
No, but I'll do some of that if you pay me to.
Perhaps we'll send them to Ring Audio? Exactly. That and the side benefits of good sound at turn on and lower operating temperatures. This is a major draw back of teh original design.
Yes, they do. Also, your power switch will last much, much longer.
Looking at the PCBs on my minds eye (I've seen many), there are many places to improve the sound quality. Some things would need redesign, but I'll bet I can really improve the sound quality of these.
-Chris
The Bias control could definately be improved. Looking at the service manual these 4B's drift considerably.
If you had some good suggestions I could be tempted to try them. If soem one designed PCB's I woudl definately be up for a pair, this Bryston 4B is rediculously heavy (22Kg) and it puts me off working with it much.
I got a Brston 4b of a popular auction site recently. I recapped it and replaced the rectifiers. The capactiors across the rectifiers was explaoded on both channels. (Was a film capacitor too.)
Once channel is working perfectly and they sound fine on my test speakers.
Bias is way too high on one channel. The Bias pot I replaced and I adjusting it and it seems to make not difference. That channel is getting very hot quiet quickly, 10-20 mins and its too hot to touch. I have yet to connect the scope to the output, to check for oscilations but the transistors look original. Though I do hear an thermin like sound as I power it up via a soft start circuit.
As the sound is fine on both channels, and within spec DC ofset, on both channels I expect the transistors are in OK form.
Any suggestions why the Bias should be excecive (4X normal) and the bais trim makes no difference and yest the sound it OK?
Replacement of blown output transistors
I realize this post is 11 years old, none the less; there is a tremendous amount of information this was not addressed. Knowing these items will help you find and replacement transistors in any amplifier, or any circuit application. These are some of the steps an engineer would follow in the design of an electronic circuit:
First: Determine and remove all defective silicon device. With no load on the output, power the amplifier and measure the power supply voltage. If possible disconnect the power supply and take your measurement, the prior method is acceptable, unless you are a novice!
I have never seen inside a Bryston amplifier, I do have a complete library of there amplifier circuit schematics.
1. Knowing the power supply voltage will give you the ability to calculate the basic output power, not taking into consideration of minor losses, like voltage drop across the transistors during the full-on condition and the drop across Emitter resistors, or in this case Colleertor resistors. The Bryston 4B has a power supply of +-80-VDC or 160-VDC total, across a totem pole design. A Totem pole design has 4 sets of transistors in series, this will be important later in your analysis. With 160-VDC power supply, what would be the maximum AC Voltage before clipping at the output? To convert a DC value to the equivalent AC value is E/2.828, or 160/2.828 = 56.58 Vrms = E2/R = 56.58Sqared/8 = 400.1 watts, no doubt this is a 400-watt amplifier into an 8-ohm load. When selecting a suitable replacement transistor in this case, to get 400 watts RMS you need to have 4 times the amount of transistor power ability. In the 4B, 80 Vpeak Sq/8 x 2 = 1600-Watts p-p. This is due to the peak-to-peak power of the sin wave. We need a transistor that can handle a minimum of 200-Watts x 8 transistors = 1600-Watts.
2. A transistor secondary Brake Down Voltage, this is the voltage when the transistor will go to a none-destructive short circuit, Primary Breakdown is when the device has catastrophically failed. The device needs to have a breakdown voltage of at least 20% above the maximum voltage the device will operate under. Now back to the totem pole, the reason for this is, in 1988 there where few devices that could handle a 160 volt supply rail and when you did find one, it was not cost-effective, also the majority of high voltage devices at that time were suitable for flyback power supplies, not audio! The answer was, instead of having a two-transistor push-pull configuration, engineers went to a 4 transistor totem to lower the required voltage of the output transistors. Instead of a 200-volt device, you can use 8 200-Watt,100-volt devices. The bottom two transistors of the design are current sinks whos current output ane determent by the voltage applied to the emitters of the current sources, by the above complimentary voltage followers.
3. No matter what you do, besides using 200-V devices, on turn on it is possible for the amplifier to become unstable and for a brief moment, one transistor can experience the full 160 volts across it and it will take out all eight devices. Clearly Bryston figured this out, as they no longer use this design.
4. In 1988 we only had the 2n3773, an excellent device but it had no compliment. 150-watts @ 140-V.
5. The final thing you MUST do is bata match your devices, hears why the minimum Hfe (current Gain) is 15 of a 2N3773 but it can be 60 or more. Let us assume you are using 4 transistors, 2 have a bata of 20, 1-15, and the other 60. The two devices of 20 will operate together, the device that has a bata of 15 will not keep up and cause the two devices of 20 to work harder, making up for the bata of 15, now the last transistor with a bata of 60 is going to try to handle the full load, because it will give more current gain than the total of the other three devices. the amount of current than the other devices, under full power this amplifier will fail. If the devices have good consistency you can get by with buying 50% more than you need, if the devices have a poor yield you will need to buy at least twice as many devices than you need, so you might get the number of matched devices you required to repair the amp correctly, to factory spec
I hope this helps people trying to figure out power. My background is 47 years in the professional music business as Electronics Engineer. If you have a question you think I can answer, please do not hesitate to ask. It is my turn to give back!
ra-tot@hotmail.com
I realize this post is 11 years old, none the less; there is a tremendous amount of information this was not addressed. Knowing these items will help you find and replacement transistors in any amplifier, or any circuit application. These are some of the steps an engineer would follow in the design of an electronic circuit:
First: Determine and remove all defective silicon device. With no load on the output, power the amplifier and measure the power supply voltage. If possible disconnect the power supply and take your measurement, the prior method is acceptable, unless you are a novice!
I have never seen inside a Bryston amplifier, I do have a complete library of there amplifier circuit schematics.
1. Knowing the power supply voltage will give you the ability to calculate the basic output power, not taking into consideration of minor losses, like voltage drop across the transistors during the full-on condition and the drop across Emitter resistors, or in this case Colleertor resistors. The Bryston 4B has a power supply of +-80-VDC or 160-VDC total, across a totem pole design. A Totem pole design has 4 sets of transistors in series, this will be important later in your analysis. With 160-VDC power supply, what would be the maximum AC Voltage before clipping at the output? To convert a DC value to the equivalent AC value is E/2.828, or 160/2.828 = 56.58 Vrms = E2/R = 56.58Sqared/8 = 400.1 watts, no doubt this is a 400-watt amplifier into an 8-ohm load. When selecting a suitable replacement transistor in this case, to get 400 watts RMS you need to have 4 times the amount of transistor power ability. In the 4B, 80 Vpeak Sq/8 x 2 = 1600-Watts p-p. This is due to the peak-to-peak power of the sin wave. We need a transistor that can handle a minimum of 200-Watts x 8 transistors = 1600-Watts.
2. A transistor secondary Brake Down Voltage, this is the voltage when the transistor will go to a none-destructive short circuit, Primary Breakdown is when the device has catastrophically failed. The device needs to have a breakdown voltage of at least 20% above the maximum voltage the device will operate under. Now back to the totem pole, the reason for this is, in 1988 there where few devices that could handle a 160 volt supply rail and when you did find one, it was not cost-effective, also the majority of high voltage devices at that time were suitable for flyback power supplies, not audio! The answer was, instead of having a two-transistor push-pull configuration, engineers went to a 4 transistor totem to lower the required voltage of the output transistors. Instead of a 200-volt device, you can use 8 200-Watt,100-volt devices. The bottom two transistors of the design are current sinks whos current output ane determent by the voltage applied to the emitters of the current sources, by the above complimentary voltage followers.
3. No matter what you do, besides using 200-V devices, on turn on it is possible for the amplifier to become unstable and for a brief moment, one transistor can experience the full 160 volts across it and it will take out all eight devices. Clearly Bryston figured this out, as they no longer use this design.
4. In 1988 we only had the 2n3773, an excellent device but it had no compliment. 150-watts @ 140-V.
5. The final thing you MUST do is bata match your devices, hears why the minimum Hfe (current Gain) is 15 of a 2N3773 but it can be 60 or more. Let us assume you are using 4 transistors, 2 have a bata of 20, 1-15, and the other 60. The two devices of 20 will operate together, the device that has a bata of 15 will not keep up and cause the two devices of 20 to work harder, making up for the bata of 15, now the last transistor with a bata of 60 is going to try to handle the full load, because it will give more current gain than the total of the other three devices. the amount of current than the other devices, under full power this amplifier will fail. If the devices have good consistency you can get by with buying 50% more than you need, if the devices have a poor yield you will need to buy at least twice as many devices than you need, so you might get the number of matched devices you required to repair the amp correctly, to factory spec
I hope this helps people trying to figure out power. My background is 47 years in the professional music business as Electronics Engineer. If you have a question you think I can answer, please do not hesitate to ask. It is my turn to give back!
ra-tot@hotmail.com
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.