Hi jaycee,
Hi wg_ski,
Actually, your argument is on thin ice as far as device toughness is concerned. The improvements are not enough to really show an improvement in the number of devices needed for certain power levels.
-Chris
Yes, absolutely! Also the shape of the envelope is optimized. Marantz normally allows higher currents at higher output voltages instead of allowing high current at low output voltages.
Hi wg_ski,
Nothing wrong with a well executed V-I limiter at all. The alternative typically means fuses (bad solution) or relay (mosfet) solutions to disconnect the load. Not entirely successful as far as speakers and amplifiers are concerned. I have seen contacts welded together.
Actually, your argument is on thin ice as far as device toughness is concerned. The improvements are not enough to really show an improvement in the number of devices needed for certain power levels.
-Chris
VI limiters become unstable when connected to an inductive load, so I'm not a fan. But current limiting can be very subtle, and buys you precious milliseconds to pull a relay etc.
Perhaps a Marantz is a product for zealots like Range Rover and Jaguar, who's customers are willing to spend a lot of money on constant repairs for the sake of the mystique. Not me. When I said "over complicated", I was being polite. What I meant was "misguided". I understand them perfectly, but for amplifiers, more is not better because it adds stability problems and unnecessary bias drops that compromise the audio power vs heat. OK the class-A folks have no problem with that. Yes I repaired professional amps in the 80s, and I hated models that blew up repeatedly. As far as I'm concerned, it's just incompetent and irresponsible.
Perhaps a Marantz is a product for zealots like Range Rover and Jaguar, who's customers are willing to spend a lot of money on constant repairs for the sake of the mystique. Not me. When I said "over complicated", I was being polite. What I meant was "misguided". I understand them perfectly, but for amplifiers, more is not better because it adds stability problems and unnecessary bias drops that compromise the audio power vs heat. OK the class-A folks have no problem with that. Yes I repaired professional amps in the 80s, and I hated models that blew up repeatedly. As far as I'm concerned, it's just incompetent and irresponsible.
Hi steveu,
It's clear that we live in completely different worlds. I was lucky enough to be working doing Marantz warranty back then, and was authorized to do the 500 too. The Marantz product was tough, hard to kill and you really deserved it when you could.
The only repeat repairs were due to incompetent servicing, or due to an animal of a customer who blew everything up. You are way off base where Marantz is concerned. I still rebuild these products, and they continue to serve their owners well.
The Marantz VI limiters are stable. You can check the operation out using an oscilloscope and dummy loads. Like I said, you live in a completely different world. I'll say the same thing for McIntosh equipment. Well made, tough and they just keep on going with a little maintenance.
Everyone hated the "professional" amps that were glorified junk. But even the best amplifier out there is still made with parts, and they can fail at any point in time.
-Chris
It's clear that we live in completely different worlds. I was lucky enough to be working doing Marantz warranty back then, and was authorized to do the 500 too. The Marantz product was tough, hard to kill and you really deserved it when you could.
The only repeat repairs were due to incompetent servicing, or due to an animal of a customer who blew everything up. You are way off base where Marantz is concerned. I still rebuild these products, and they continue to serve their owners well.
The Marantz VI limiters are stable. You can check the operation out using an oscilloscope and dummy loads. Like I said, you live in a completely different world. I'll say the same thing for McIntosh equipment. Well made, tough and they just keep on going with a little maintenance.
Everyone hated the "professional" amps that were glorified junk. But even the best amplifier out there is still made with parts, and they can fail at any point in time.
-Chris
Old school 70’s and 80’s high fT devices were pretty pathetic - only handling full power out to 10 or maybe a whopping 20 volts. You couldn’t even make a reliable 50 watts out of a pair of TO-3s. Nowadays they have pushed the S/B point out to about 50 volts for triple diffused types. On par with the old epi-base and about 10x faster with a beta of more than 20 at 10 amps. At least they are useable on 50 volt rails and the mystery failures have pretty much disappeared. Doesn’t mean they usually put in enough outputs to handle heavy loads, but you do have to turn it up loud, use too many speakers, or short something to get them to die.
Hi wg_ski,
Well, not entirely true as we had types such as MJ15024 or MJ15003. I'm not going to bother looking up the actual physical construction. These parts have been around for a while now and are still good choices. Earlier 2N5631 type devices were also pretty good.
Now what you are talking about as far as handling their full power is known as second breakdown. New devices such as MJ21195 have increased the boundaries of the SOA, but only by a reasonable amount. Not night and day. For a part to handle it's full dissipation you are pretty much stuck with Mosfets. They have their own difficulties.
The fact is, if I replace a 2N5631 type pair of devices with the MJ21195 type parts, the amplifier might be a little tougher to blow up, but not by a noticeable amount in practice - as in real life. That example is moving from a popular 1970's device type into parts from this century, and that was the point you were making.
I still design with the old, ancient devices as long as they are still in production. I still have a few thousand MPSA06 and MPSA56 transistors that I doubt I will be able to use before I die. They may even be in production when I expire! Mind you, if the application calls for one of the latest parts, I'll design with those.
-Chris
Well, not entirely true as we had types such as MJ15024 or MJ15003. I'm not going to bother looking up the actual physical construction. These parts have been around for a while now and are still good choices. Earlier 2N5631 type devices were also pretty good.
Now what you are talking about as far as handling their full power is known as second breakdown. New devices such as MJ21195 have increased the boundaries of the SOA, but only by a reasonable amount. Not night and day. For a part to handle it's full dissipation you are pretty much stuck with Mosfets. They have their own difficulties.
The fact is, if I replace a 2N5631 type pair of devices with the MJ21195 type parts, the amplifier might be a little tougher to blow up, but not by a noticeable amount in practice - as in real life. That example is moving from a popular 1970's device type into parts from this century, and that was the point you were making.
I still design with the old, ancient devices as long as they are still in production. I still have a few thousand MPSA06 and MPSA56 transistors that I doubt I will be able to use before I die. They may even be in production when I expire! Mind you, if the application calls for one of the latest parts, I'll design with those.
-Chris
wg-ski and anatech
I think you both have good points. One of the best amps I heard (after a QC- 2n3055H design which turned out to be pretty poor anyway) was Bailey's "laboratory amplifier". That used 2N3716/2N3792 which had a second breakdown starting at 30V. You could certainly get 30W reliably, but you could also get 50W from a pair of 2N3055H's. I agree that a single pair of 3716/3792 MIGHT have given 50W but a reactive load could have popped them over the SOA limit. My first 50W 2N3055(H) amp was used school discos and the OP devices survived a "melted" transformer (one that used old style wax insulation) from overheating! I still have those and they still work in a second build of the amp.
Today, I would have no problem recommending MJ21193/4 (or 5/6) and the MJL3281A/MJL1302A or 2SC5200/2SA1943 etc. These latter devices seem to be equivalent to the Sanken devices (on paper- I haven't used Sanken at all) in that they are multiple epi, multiple emitter site, high frequency linear gain.
I think you both have good points. One of the best amps I heard (after a QC- 2n3055H design which turned out to be pretty poor anyway) was Bailey's "laboratory amplifier". That used 2N3716/2N3792 which had a second breakdown starting at 30V. You could certainly get 30W reliably, but you could also get 50W from a pair of 2N3055H's. I agree that a single pair of 3716/3792 MIGHT have given 50W but a reactive load could have popped them over the SOA limit. My first 50W 2N3055(H) amp was used school discos and the OP devices survived a "melted" transformer (one that used old style wax insulation) from overheating! I still have those and they still work in a second build of the amp.
Today, I would have no problem recommending MJ21193/4 (or 5/6) and the MJL3281A/MJL1302A or 2SC5200/2SA1943 etc. These latter devices seem to be equivalent to the Sanken devices (on paper- I haven't used Sanken at all) in that they are multiple epi, multiple emitter site, high frequency linear gain.
Last edited:
The 2N5631 is a lot tougher than it looks on paper. I’ve given them 50 volts at 5 amps on a heatsink in ice water for a few seconds and they take it just fine. If you look carefully, you’ll see that some of the characteristic curves in the data sheet are COPIES of the 2N3773 and MJ15003 graphs. What does that tell you? They are also 2 MHz devices, not 20.
Back in the day I used to put in 15024’s, D424’s, or even 5631’s in place of failed high fT jap types and they never failed again. Distortion probably ran higher, but you couldn’t tell at war volume. They have made improvements in high fT devices since then, and it all seemed to start about the first time I started seeing C3281’s. The problem NOW is they try to push more and more and more out of them. I’ve seen them used with 70 volt rails - in singles. That’s asking for trouble, even if you used a 15024.
Back in the day I used to put in 15024’s, D424’s, or even 5631’s in place of failed high fT jap types and they never failed again. Distortion probably ran higher, but you couldn’t tell at war volume. They have made improvements in high fT devices since then, and it all seemed to start about the first time I started seeing C3281’s. The problem NOW is they try to push more and more and more out of them. I’ve seen them used with 70 volt rails - in singles. That’s asking for trouble, even if you used a 15024.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.