Hi,
has anyone built the early version of LC-audio "The End"?
It was published in a swedish magazine "High Fidelity" 1989.
I built it in the early 90 but never got in right.
The amp runs and sounds very well until you try to connect/unconnect the audio-cable.
The amp then gets unstable and several transistors blows.
This has happend 2 times and this forum will be my last try
before I scrap the project.
Perhaps the wires from the caps are to long..?
see pictures:
http://members.chello.se/surf/Image01.jpg
http://members.chello.se/surf/Image02.jpg
Any ideas?
Thanks
/Jan
has anyone built the early version of LC-audio "The End"?
It was published in a swedish magazine "High Fidelity" 1989.
I built it in the early 90 but never got in right.
The amp runs and sounds very well until you try to connect/unconnect the audio-cable.
The amp then gets unstable and several transistors blows.
This has happend 2 times and this forum will be my last try
before I scrap the project.
Perhaps the wires from the caps are to long..?
see pictures:
http://members.chello.se/surf/Image01.jpg
http://members.chello.se/surf/Image02.jpg
Any ideas?
Thanks
/Jan
I have just built The end and have similar exsperiences.
I just blew one channel by connecting a preamp with a slight DC-offset at it's output. Never connect the cabels AFTER you have turned it on. It's very sensible to occilations.
First connect the cabel to the preamp. Then turn on the preamp - then the poweramp.
You may allso consider to install a turn-on-delay and DC protection circuit. Welleman has a good kit called M4700.
To protect the inputs of the poweramp from DC you can mount a 2,2uf bipolar cap before the input.
Groundig: It's very importent that you have only ONE ground in the cabinet. Every gnd should be fixed to this.
Hope this helps!
Peter
mail@peterlund.org
I just blew one channel by connecting a preamp with a slight DC-offset at it's output. Never connect the cabels AFTER you have turned it on. It's very sensible to occilations.
First connect the cabel to the preamp. Then turn on the preamp - then the poweramp.
You may allso consider to install a turn-on-delay and DC protection circuit. Welleman has a good kit called M4700.
To protect the inputs of the poweramp from DC you can mount a 2,2uf bipolar cap before the input.
Groundig: It's very importent that you have only ONE ground in the cabinet. Every gnd should be fixed to this.
Hope this helps!
Peter
mail@peterlund.org
Transistors fail both in powerstage and in differential stage
and everytime the amp burns it is different transistors
which fails. Sorry, but I have not registered which transistors
due to this.
I've heard from several builders that grounding is important, otherwise it gets unstable.
Signal ground and powersupply ground is connected to each other but I have not connected ground to cabinet. Nothing
in the building instruction says that. Should I connect ground
to cabinet anyway?
A dc protect circuit sound like an good idea.
How is sound affected when you put a cap before inputs?
and everytime the amp burns it is different transistors
which fails. Sorry, but I have not registered which transistors
due to this.
I've heard from several builders that grounding is important, otherwise it gets unstable.
Signal ground and powersupply ground is connected to each other but I have not connected ground to cabinet. Nothing
in the building instruction says that. Should I connect ground
to cabinet anyway?
A dc protect circuit sound like an good idea.
How is sound affected when you put a cap before inputs?
Missing starground can very well be the trouble you are in. Here is how I have done: Make the point between the caps in The PSU your Main ground.
From here let a wire go to the bottom of the cabinet and make a good connection.
Let the ground on the PCB's (the one near to the output) go here.
The inputs should NOT be connected to the cabinet - they should float. Input ground should go to the gnd at the input.
Even if fuses a non-linear component you may save some transistors by putting somthing like 5A fuses in both the + and - just before the PCB.
Also you good shielded wire form each inputconnector to the PCB's.
I can see nothing but air under the PCB's. Not good. Try to cover it up by somthing shielding.
This amp is very delicate to HF and occilation. Mostly because it misses one thing: Feedback!;-))
I will try to attach a picture of my amp soon for you to see.
Hope this helps.
From here let a wire go to the bottom of the cabinet and make a good connection.
Let the ground on the PCB's (the one near to the output) go here.
The inputs should NOT be connected to the cabinet - they should float. Input ground should go to the gnd at the input.
Even if fuses a non-linear component you may save some transistors by putting somthing like 5A fuses in both the + and - just before the PCB.
Also you good shielded wire form each inputconnector to the PCB's.
I can see nothing but air under the PCB's. Not good. Try to cover it up by somthing shielding.
This amp is very delicate to HF and occilation. Mostly because it misses one thing: Feedback!;-))
I will try to attach a picture of my amp soon for you to see.
Hope this helps.
Hi
If you build it wrong, it will burn every single time! It's like any other amplifier or DIY project of any kind.
However if you build it according to instructions, it will work every time. More than 1800 of this amplifier was built back in 1989 - 1993, where this version of The End was first published. Many of which still work fine to this day.
First the much newer Millennium XP, posted above has no resemblance to The End 1 or 2 (can't see which on above pics), so let's avoid mixing up these two amplifiers.
To make The End 1 or 2 stable, you must do as written in the magazine High Fidelity: Use the suggested transistors TIP 35C and TIP 36C for output devices, and TIP 41C / TIP 42C for drivers. Many amplifiers failed because people were attempting all kinds of other transistors, without taking further precautions to keep it stable.
Secondly: Connect a 100nF cap DIRECTLY from the power GND to the metal surface of the heatsink. This will cure any instability from poor grounding. The PG terminal is placed some 2 cm. from the heat sink on the original PCB's (which you are clearly not using). So 2 cm. (or less than 1 inch) is the length of wires you should keep on this cap from Power Gnd to the nearest of the screws holding a Power transistor. Just use a small and inexpensive MKT cap, nothing fancy.
If you build it wrong, it will burn every single time! It's like any other amplifier or DIY project of any kind.
However if you build it according to instructions, it will work every time. More than 1800 of this amplifier was built back in 1989 - 1993, where this version of The End was first published. Many of which still work fine to this day.
First the much newer Millennium XP, posted above has no resemblance to The End 1 or 2 (can't see which on above pics), so let's avoid mixing up these two amplifiers.
To make The End 1 or 2 stable, you must do as written in the magazine High Fidelity: Use the suggested transistors TIP 35C and TIP 36C for output devices, and TIP 41C / TIP 42C for drivers. Many amplifiers failed because people were attempting all kinds of other transistors, without taking further precautions to keep it stable.
Secondly: Connect a 100nF cap DIRECTLY from the power GND to the metal surface of the heatsink. This will cure any instability from poor grounding. The PG terminal is placed some 2 cm. from the heat sink on the original PCB's (which you are clearly not using). So 2 cm. (or less than 1 inch) is the length of wires you should keep on this cap from Power Gnd to the nearest of the screws holding a Power transistor. Just use a small and inexpensive MKT cap, nothing fancy.
Just wanted to add:
Good practice after a burnout (of any amplifier) is to change ALL transistors from the output stage - right back to the voltage amplifiers. This is because any transistor in the driver stage, output stage and BIAS control will suffer complete or partly, when the amplifier burns out.
Even if you use an ohm meter to check the health of a transistor, you can NOT trust the transistor to be OK.
In some cases you might get the amp to work for some time, but after a few days the amplifier will burn out again with no apparant reason. The reason is (as any repair professional in the world knows): the transistors were partly damaged at an earlier burnout.
To avoid this problem, simply change all output transistors, drivers, predivers and BIAS transistors IN ONE GO. Without connecting power in the middle of the projct ect.
If you don't follow this practice, i would say, it's simply not worth doing the repair job at all, and it applies to any kind of power amplifier out there.
Good practice after a burnout (of any amplifier) is to change ALL transistors from the output stage - right back to the voltage amplifiers. This is because any transistor in the driver stage, output stage and BIAS control will suffer complete or partly, when the amplifier burns out.
Even if you use an ohm meter to check the health of a transistor, you can NOT trust the transistor to be OK.
In some cases you might get the amp to work for some time, but after a few days the amplifier will burn out again with no apparant reason. The reason is (as any repair professional in the world knows): the transistors were partly damaged at an earlier burnout.
To avoid this problem, simply change all output transistors, drivers, predivers and BIAS transistors IN ONE GO. Without connecting power in the middle of the projct ect.
If you don't follow this practice, i would say, it's simply not worth doing the repair job at all, and it applies to any kind of power amplifier out there.
I was just trying t be funny....
This amp is based n the principle of no-feedback, which means that there is no global feedback from output to input as in most amps.
This amp is based n the principle of no-feedback, which means that there is no global feedback from output to input as in most amps.
If you rebuild the amp it would be a good idea to test it on a lower supply voltage (use a rheostat), that way if it does oscillate it will be less likely to destroy the output transistors.
I had a similar problem with this type of output stage (and no global feedback) oscillating at around 400kHz and cured it by adding small (10s of pF) 'miller' capacitors to the driver transistors T13 and T14 in this case.
I had a similar problem with this type of output stage (and no global feedback) oscillating at around 400kHz and cured it by adding small (10s of pF) 'miller' capacitors to the driver transistors T13 and T14 in this case.
Wrong schematics...
AFAIK this one is the MK1 version of "The End".
Look at the pictures in the beginning of this thread...
(I have built both a Mk2 and a Mk3 so I should recognize it)
😉
Lars Clausen said:
AFAIK this one is the MK1 version of "The End".
Look at the pictures in the beginning of this thread...
(I have built both a Mk2 and a Mk3 so I should recognize it)
😉
OK, the Mk1 is the same as Mk2 except for the input stage, and that should not be in scope of this problem.
The Miller capacity would be nessescary if you were using japanese drivers, especially the 2SA968 and 2SC2238 set. These are inherently unstable, and require a lot of nursing to keep from oscillating. In fact, even if you were to use them as just an emitter follower, they will be likely to oscillate (!).
The TIP's that are used in The End are slower, they will not cause you any problems.
The Miller capacity would be nessescary if you were using japanese drivers, especially the 2SA968 and 2SC2238 set. These are inherently unstable, and require a lot of nursing to keep from oscillating. In fact, even if you were to use them as just an emitter follower, they will be likely to oscillate (!).
The TIP's that are used in The End are slower, they will not cause you any problems.
From my experience
Hi,
I built the amplifier in both mk1 and 2 versions and i burned quite a few (3 😀) mk2's before i reread the article and found that i had forgotten to connect the low level ground to the star ground. After that the aplifier was stable as a rock. This was with the tip35 36 transistors. I had to retire it when my preamp gave up and i didn't have enough money to buy a new one.
Regards
Me
Hi,
I built the amplifier in both mk1 and 2 versions and i burned quite a few (3 😀) mk2's before i reread the article and found that i had forgotten to connect the low level ground to the star ground. After that the aplifier was stable as a rock. This was with the tip35 36 transistors. I had to retire it when my preamp gave up and i didn't have enough money to buy a new one.
Regards
Me
OK, we can play word puzzles 😉
According to my dictionary, the word 'inherently' can have two meanings. One is that the item 'inherently' is used to describe, has some property within, or from it's birth. The other meaning is something like 'when all is said, it is inherently ... ' (unstable).
Both meanings would support my point perfectly.
Check it out!
According to my dictionary, the word 'inherently' can have two meanings. One is that the item 'inherently' is used to describe, has some property within, or from it's birth. The other meaning is something like 'when all is said, it is inherently ... ' (unstable).
Both meanings would support my point perfectly.
Check it out!
Here we go again
Once again you are venturing in areas concerning responses from a manufacturer that belong in private Email rather than on the forum. The nature of these responses would better serve you since what seem to me to be very poor customer relations. This surely cannot be help to sales if these post are typical of what a customer can expect as a response to questions.
Concerning TIP transistor parts:
Usually these transistors are VERY loosely specified parts that fall out the specs for another intended part number. The specs often state minimums with no typical or maximum values for many of the parameters. Designing something as sensitive as a compound output stage with transistors whose characteristics vary so widely is begging for trouble. It sounds like many other people have had problems with this same issue. A responsible designer would pick a popular and tightly specified transistor and design a compensation network that assured stability and consistency instead of blaming the builders of the amps.
Once again you are venturing in areas concerning responses from a manufacturer that belong in private Email rather than on the forum. The nature of these responses would better serve you since what seem to me to be very poor customer relations. This surely cannot be help to sales if these post are typical of what a customer can expect as a response to questions.
Concerning TIP transistor parts:
Usually these transistors are VERY loosely specified parts that fall out the specs for another intended part number. The specs often state minimums with no typical or maximum values for many of the parameters. Designing something as sensitive as a compound output stage with transistors whose characteristics vary so widely is begging for trouble. It sounds like many other people have had problems with this same issue. A responsible designer would pick a popular and tightly specified transistor and design a compensation network that assured stability and consistency instead of blaming the builders of the amps.
Re: Here we go again
Fred,
I don't see his responses any different from what Nelson Pass is doing. This is acceptable behaviour conduct on this forum and I suggest you stop making further remarks of that type.
Peter Daniel
diy Audio Moderator
Fred Dieckmann said:
Fred,
I don't see his responses any different from what Nelson Pass is doing. This is acceptable behaviour conduct on this forum and I suggest you stop making further remarks of that type.
Peter Daniel
diy Audio Moderator
- Status
- Not open for further replies.