Have any of you considered doing a PCB layout for the Superamp using the TO-247 style transistors. I would really like to build the Superamp but would like to make use of some of the heatsinks that I already own. (They are similar to the ones Jens is using.)
I think it would be very similar to the board that Jens has designed for the Low TIM.
Thanks, Terry
I think it would be very similar to the board that Jens has designed for the Low TIM.
Thanks, Terry
1, I only want to know if all these Gerber files are of pcb only .
2,Next for protection what is the voltage across the emitter and base of T-10 & T-12 YOU taken into account to switch them on to shut down the drive to the driver trs .
2,Next for protection what is the voltage across the emitter and base of T-10 & T-12 YOU taken into account to switch them on to shut down the drive to the driver trs .
still4given
Now high voltage transistors are avilible and the normal Leach amp can give the output of the Leach super amp by increasing the supply voltage only , The ss\uper amp was designed many years ago when these output devices were not availible.
Now high voltage transistors are avilible and the normal Leach amp can give the output of the Leach super amp by increasing the supply voltage only , The ss\uper amp was designed many years ago when these output devices were not availible.
rajeev luthra said:
Hi rajeev,
Are you saying that by changing the transistors in the Low TIM, I would be able to use my 87VDC per rail transformer? That would be great. Can you give us more information about this?
Thanks, Terry
I would sagest MJL1302 and MJL3281. They have already been tested in this device and are known to work. Some people have also reported they sound better, and some people have reported that they don’t hear any difference. At any rate the specifications are much better. These are also the same transistors used in the new Mark Leveinson amps.
However for changing the transistors and razing the voltage you pay a price. First of all with higher voltage you will not be able to run them at as high a current, they are already power limited by the fact that they are plastic devices. This is because the insides get hotter. You will have to use more of them in parallel. There is a limit as to the max number you can use and adding more then you really need is not a good idea.
You will also have to reconfigure the safety circuit to activate at a lower current. Depending on your final DC rail voltage you’ll find a graph of the SOA (rail voltage v.s. max current) on the data sheet. Unfortunately Leach does not give any instructions on his web site as to how to do this. You will have to raze the value of the .33ohm resistors, but I don’t know by how much. Does anybody have any suggestions / formulas 😕
Leve
However for changing the transistors and razing the voltage you pay a price. First of all with higher voltage you will not be able to run them at as high a current, they are already power limited by the fact that they are plastic devices. This is because the insides get hotter. You will have to use more of them in parallel. There is a limit as to the max number you can use and adding more then you really need is not a good idea.
You will also have to reconfigure the safety circuit to activate at a lower current. Depending on your final DC rail voltage you’ll find a graph of the SOA (rail voltage v.s. max current) on the data sheet. Unfortunately Leach does not give any instructions on his web site as to how to do this. You will have to raze the value of the .33ohm resistors, but I don’t know by how much. Does anybody have any suggestions / formulas 😕
Leve
LBHajdu,
"However for changing the transistors and razing the voltage you pay a price."
Indeed. In purchase quantities of 100 from OnSemi:
MJL1302 $2.88 each Plastic case
MJ15003 $1.87 each Metal case
Orderable Part IC Max (A) VCEO(sus) Min (V) hFE Min
MJL1302A 15 230 60 *
hFE Max fT Min (MHz) PTM Max (W) Polarity Package
175 * 30 200 PNP TO264
Orderable Part IC Max (A) VCEO(sus) Min (V) hFE Min
MJ15003 20 140 25 *
hFE Max fT Min (MHz) PTM Max (W) Polarity Package
50 * 2 250 NPN TO3
Original quote:
"Originally posted by rajeev luthra
still4given
Now high voltage transistors are avilible and the normal Leach amp can give the output of the Leach super amp by increasing the supply voltage only , The ss\uper amp was designed many years ago when these output devices were not availible. "
"increasing the supply voltage only", Excuse me?
And more MJL1302A devices are required? How many?
"At any rate the specifications are much better." How's that again?
Can someone please explain how this is an advantage over a proven design?
With all due respect to "Mark Leveinson" amps and rajeev luthra/
LBHajdu, your reply is directed to a self acknowledged rookie trying to learn. Perhaps you could make your case more clear as to how any of this idea is an advantage.
BTW, Prof. Leach is probably not in the business of giving "any instructions on his web site" on how to futz up the Superamp or the LowTIM amp.
His biographical info here:
http://users.ece.gatech.edu/~mleach/bio.html
Prosit
"However for changing the transistors and razing the voltage you pay a price."
Indeed. In purchase quantities of 100 from OnSemi:
MJL1302 $2.88 each Plastic case
MJ15003 $1.87 each Metal case
Orderable Part IC Max (A) VCEO(sus) Min (V) hFE Min
MJL1302A 15 230 60 *
hFE Max fT Min (MHz) PTM Max (W) Polarity Package
175 * 30 200 PNP TO264
Orderable Part IC Max (A) VCEO(sus) Min (V) hFE Min
MJ15003 20 140 25 *
hFE Max fT Min (MHz) PTM Max (W) Polarity Package
50 * 2 250 NPN TO3
Original quote:
"Originally posted by rajeev luthra
still4given
Now high voltage transistors are avilible and the normal Leach amp can give the output of the Leach super amp by increasing the supply voltage only , The ss\uper amp was designed many years ago when these output devices were not availible. "
"increasing the supply voltage only", Excuse me?
And more MJL1302A devices are required? How many?
"At any rate the specifications are much better." How's that again?
Can someone please explain how this is an advantage over a proven design?
With all due respect to "Mark Leveinson" amps and rajeev luthra/
LBHajdu, your reply is directed to a self acknowledged rookie trying to learn. Perhaps you could make your case more clear as to how any of this idea is an advantage.
BTW, Prof. Leach is probably not in the business of giving "any instructions on his web site" on how to futz up the Superamp or the LowTIM amp.
His biographical info here:
http://users.ece.gatech.edu/~mleach/bio.html
Prosit
Hey Guys
There is so much information in this forum , a little effort to go through
some threads will really be of advantage and answer most of our
questions .
Terry
The answer is YES
In the super amp the transistors are used in series to increase their
working voltage , now we need not do that and use higher working
voltage devices in parallel as per requirement , thus making it possible
for the 120w version to work at higher voltages and deliver higher
wattage .
acenovelty
You can use these
MJL4281A/MJL4302L or 2SC-5200 / 2SA-1943 and some others also.
If you are interested in T0-3 devices the MJ15024/MJ15025 but now the
new devices are better .
LBHajdu
Please go through these threads , here there is some interesting text on
TO-3 and TO-264 devices , and in the thread "Leach clone, pretty good
looking" You will find circuit diagrams in Jens site with four and ten pairs
of output devices .
http://www.diyaudio.com/forums/showthread.php?s=&threadid=40184&
goto=nextnewest
http://www.diyaudio.com/forums/showthread.php?s=&threadid=46824&
perpage=10&pagenumber=1
http://www.delta-audio.com/Leach-Clone.htm
I am sure you will be satisfied after going through the above .
Yes the protection values are a little tricky that is why I had asked Jens
the following question
"what is the voltage across the emitter and base of T-10 & T-12 YOU
taken into account to switch them on to shut down the drive to the driver
trs ."
I hope Jens is reading and will reply .
There is so much information in this forum , a little effort to go through
some threads will really be of advantage and answer most of our
questions .
Terry
The answer is YES
In the super amp the transistors are used in series to increase their
working voltage , now we need not do that and use higher working
voltage devices in parallel as per requirement , thus making it possible
for the 120w version to work at higher voltages and deliver higher
wattage .
acenovelty
You can use these
MJL4281A/MJL4302L or 2SC-5200 / 2SA-1943 and some others also.
If you are interested in T0-3 devices the MJ15024/MJ15025 but now the
new devices are better .
LBHajdu
Please go through these threads , here there is some interesting text on
TO-3 and TO-264 devices , and in the thread "Leach clone, pretty good
looking" You will find circuit diagrams in Jens site with four and ten pairs
of output devices .
http://www.diyaudio.com/forums/showthread.php?s=&threadid=40184&
goto=nextnewest
http://www.diyaudio.com/forums/showthread.php?s=&threadid=46824&
perpage=10&pagenumber=1
http://www.delta-audio.com/Leach-Clone.htm
I am sure you will be satisfied after going through the above .
Yes the protection values are a little tricky that is why I had asked Jens
the following question
"what is the voltage across the emitter and base of T-10 & T-12 YOU
taken into account to switch them on to shut down the drive to the driver
trs ."
I hope Jens is reading and will reply .
acenovelty,
“Can someone please explain how this is an advantage over a proven design?” – who said anything about an advantage ?
The question where trying to answer is how can Terry use this amp with his 87VDC per rail transformer, I think the advantage to Terry is that his amp won’t explode. This forum documents at least two people using these transistors without any bad effects.
purchase quantities of 100 – I had in mide more like five or six per channel.
You have always got to try new avenues to improve the amp, otherwise how can progress be made?
Terry
There are some other things that come to mind:
There are also to resistors in the diff pair who’s values must be changed. Leach provides the formula on his web page. These set the bias for the diff pair.
And of course use all of Jens Rasmussen substitutes as it’s not just the output stage that has to be rated for the voltage.
Rajeev - is right, read the thread.
Rajeev,
I have asked Jen about the protection values, several pages it this thread ago, but got no reply. He probably figured it out by trial and error, which may be what we have to do. The other option is to go right to the source and ask leach himself.
“Can someone please explain how this is an advantage over a proven design?” – who said anything about an advantage ?
The question where trying to answer is how can Terry use this amp with his 87VDC per rail transformer, I think the advantage to Terry is that his amp won’t explode. This forum documents at least two people using these transistors without any bad effects.
purchase quantities of 100 – I had in mide more like five or six per channel.
You have always got to try new avenues to improve the amp, otherwise how can progress be made?
Terry
There are some other things that come to mind:
There are also to resistors in the diff pair who’s values must be changed. Leach provides the formula on his web page. These set the bias for the diff pair.
And of course use all of Jens Rasmussen substitutes as it’s not just the output stage that has to be rated for the voltage.
Rajeev - is right, read the thread.
Rajeev,
I have asked Jen about the protection values, several pages it this thread ago, but got no reply. He probably figured it out by trial and error, which may be what we have to do. The other option is to go right to the source and ask leach himself.
Actually I have had no time to sit down and derive the needed formula 😉
Please be patient, I will post a formula sometime…
Regarding output devises I like the MJL4281/MJL4302 pair
\Jens
Please be patient, I will post a formula sometime…
Regarding output devises I like the MJL4281/MJL4302 pair
\Jens
Have any of you respondents built even one of these amps successfully?
"purchase quantities of 100 – I had in mide more like five or six per channel" Which is it 5 or 6 or maybe 7, 8?
100 quantity purchases get you the best low volume discount from retail. Please apply the simple arithmetic you were supposed to learn in the 6th grade. Each MJL1302 costs $1.01 more than the documented working transistor. 8 are required/channel.
2X8=16
16X$1.87=?
LBHajdu/rajeev luthra,
Perhaps you have photos of your amplifier, working boards, or schematics or board layouts to back up your claims? Lots of talk, gentlemen. Show us your work.
" the new devices are better ." Care to present some specs or other tangible evidence to demonstrate how they might be better?
Oh yes, "the advantage to Terry is that his amp won’t explode." if he uses the TO3's mounted on a 1/4" 2"x2" equal angle Al bolted firmly to a flat back heatsink using standard components. My local hardware is asking $3.95 for a 48" section of this material.
Prosit
"purchase quantities of 100 – I had in mide more like five or six per channel" Which is it 5 or 6 or maybe 7, 8?
100 quantity purchases get you the best low volume discount from retail. Please apply the simple arithmetic you were supposed to learn in the 6th grade. Each MJL1302 costs $1.01 more than the documented working transistor. 8 are required/channel.
2X8=16
16X$1.87=?
LBHajdu/rajeev luthra,
Perhaps you have photos of your amplifier, working boards, or schematics or board layouts to back up your claims? Lots of talk, gentlemen. Show us your work.
" the new devices are better ." Care to present some specs or other tangible evidence to demonstrate how they might be better?
Oh yes, "the advantage to Terry is that his amp won’t explode." if he uses the TO3's mounted on a 1/4" 2"x2" equal angle Al bolted firmly to a flat back heatsink using standard components. My local hardware is asking $3.95 for a 48" section of this material.
Prosit
Terry,
I'm in the process of modifying a Leach to use an old Yamaha M-2 chassis and PS that runs 80-84volt rails depending on line voltage.
Don't forget to bump up the power rating of R13/14 to 1 W - you'll be dissipating close to .4 W with 87 volt rails. You'll probably have to mount it high off the board to get the leads to fit, but that's a good thing when it's hot.
The Q16 and Q17 will need to be changed MJE15032/33 or 34/35 for their highervoltage ratings. USe a decent size heat sink.
I'd suggest 4-6 pairs of output devices depending on how hard you want to work the amp. In addition to those already mentioned MJL21195/96 are an option. (I'm using the TO-3 version, MJ21195/96)
If you want to lower the current limit, lower the value of R37-40. Earlier version of the amp used 470 ohms instead of the current 680. Leach states this was done to avoid triggering the protection circuit at low load impedance.
To calculate where the protection circuit kicks in, use one leg: Q10, R37 and R28. Q10 will begin to conduct when its base reaches a little over .6V. R37 and R28 form a voltage divider, sensing the voltage across the emitter resistor (R45). So when the voltage across R28 reaches .6V, we'll be in protection. Calculate the voltage across R45 that will cause .6V across R28. then divide this by the value of R45 to get the current PER OUTPUT DEVICE PAIR at which the protection kicks in. I'll leave deriving the formula and the calculations to your education, unless you need help.
With this number you can go into the SOA curves to decide how many pairs of outputs you need to meet your power/ load requirements.
I'm in the process of modifying a Leach to use an old Yamaha M-2 chassis and PS that runs 80-84volt rails depending on line voltage.
Don't forget to bump up the power rating of R13/14 to 1 W - you'll be dissipating close to .4 W with 87 volt rails. You'll probably have to mount it high off the board to get the leads to fit, but that's a good thing when it's hot.
The Q16 and Q17 will need to be changed MJE15032/33 or 34/35 for their highervoltage ratings. USe a decent size heat sink.
I'd suggest 4-6 pairs of output devices depending on how hard you want to work the amp. In addition to those already mentioned MJL21195/96 are an option. (I'm using the TO-3 version, MJ21195/96)
If you want to lower the current limit, lower the value of R37-40. Earlier version of the amp used 470 ohms instead of the current 680. Leach states this was done to avoid triggering the protection circuit at low load impedance.
To calculate where the protection circuit kicks in, use one leg: Q10, R37 and R28. Q10 will begin to conduct when its base reaches a little over .6V. R37 and R28 form a voltage divider, sensing the voltage across the emitter resistor (R45). So when the voltage across R28 reaches .6V, we'll be in protection. Calculate the voltage across R45 that will cause .6V across R28. then divide this by the value of R45 to get the current PER OUTPUT DEVICE PAIR at which the protection kicks in. I'll leave deriving the formula and the calculations to your education, unless you need help.
With this number you can go into the SOA curves to decide how many pairs of outputs you need to meet your power/ load requirements.
Hi, further to BobEllis guide, the current limit and SOA will ensure that your output transistors survive a short term overload for a few sub-bass cycles and even a dead short for some seconds.
The dissipation of short circuit current times Vrail voltage amounts to enormous power and your sinks will heat up very quickly. This will lower your SOA due to Tc rising. How will you limit power dissipation for a long term current limit? A thermal switch? or some kind of electronic latch? to lock out the PSU or input stage until the problem is corrected?
regards Andrew T.
The dissipation of short circuit current times Vrail voltage amounts to enormous power and your sinks will heat up very quickly. This will lower your SOA due to Tc rising. How will you limit power dissipation for a long term current limit? A thermal switch? or some kind of electronic latch? to lock out the PSU or input stage until the problem is corrected?
regards Andrew T.
Hi Bob,
Thanks for the instructive post. A bit of that went over my head but I will try and study the circuit before I start asking questions. Could be I can figure it out by studying what each componant does that you mentioned.
I would like to know if you are working with the Low TIM or the Superamp design.
I already have 25 each of the MJE15032/33. They lists as, "8.0 Amperes Power Transistors Complementary Silicon, 250 volts, 50 watts."
The MJE15034/35 list as, "4.0 amperes Power Transistors Complementary Silicon, 350 volts, 50 watts."
Which would be better?
On Semi carries the MJE15034/35 so I could easily get those instead.
Thanks, Terry
Thanks for the instructive post. A bit of that went over my head but I will try and study the circuit before I start asking questions. Could be I can figure it out by studying what each componant does that you mentioned.
I would like to know if you are working with the Low TIM or the Superamp design.
I already have 25 each of the MJE15032/33. They lists as, "8.0 Amperes Power Transistors Complementary Silicon, 250 volts, 50 watts."
The MJE15034/35 list as, "4.0 amperes Power Transistors Complementary Silicon, 350 volts, 50 watts."
Which would be better?
On Semi carries the MJE15034/35 so I could easily get those instead.
Thanks, Terry
Terry,
It's a the Low TIM amp. Last time I looked, Prof. Leach had a decent explanation of how the protection circuit works on his website. Basically if the source resistor voltage gets too high, the protection transistors turn on and remove (short) the drive to the output stage.
Either will work - I am using the MJE15032/33 and since you already have them...
It's a the Low TIM amp. Last time I looked, Prof. Leach had a decent explanation of how the protection circuit works on his website. Basically if the source resistor voltage gets too high, the protection transistors turn on and remove (short) the drive to the output stage.
Either will work - I am using the MJE15032/33 and since you already have them...
HI BobEllis,
Can I correct your last post.
Leach's description for the protection transistor limits drive and thereby the current through the source resistors. I do not think it removes the drive. I think the drive limit will will be Vout + Vsat.
this will form a feedback loop from output to protction and the output will try to produce constant current into the load until the overcurrent condition is removed.
please correct me if I am wrong.
regards Andrew T.
Can I correct your last post.
Leach's description for the protection transistor limits drive and thereby the current through the source resistors. I do not think it removes the drive. I think the drive limit will will be Vout + Vsat.
this will form a feedback loop from output to protction and the output will try to produce constant current into the load until the overcurrent condition is removed.
please correct me if I am wrong.
regards Andrew T.
Andrew, you are correct. I oversimplified my description of the circuit's operation trying to help Terry get a handle on it. Thanks for keeping me on the straight and narrow 😉
Hi Bob,
I will say your discription was easier to follow. 😀
I wondering if it is better to be relying on the protection circuit than to have the room to spare by running things in series?
Is there more loss in power and effeciency with the Superamp method? I really like the design that Jens has come up with. Very smilar to the way the P101 goes together. Anyway, I have both types of transistors so I guess I'll keep an eye open and try to make some decisions when my copper clad boards get here.
Thanks, Terry
I will say your discription was easier to follow. 😀
I wondering if it is better to be relying on the protection circuit than to have the room to spare by running things in series?
Is there more loss in power and effeciency with the Superamp method? I really like the design that Jens has come up with. Very smilar to the way the P101 goes together. Anyway, I have both types of transistors so I guess I'll keep an eye open and try to make some decisions when my copper clad boards get here.
Thanks, Terry
Terry,
You're not really relying on the protection circuit unless you run too few output pairs. So it is a tradeoff - do you want to make your own boards for the double barrel version or use Jens' board loaded up, or Professor Leach's board and add extra output devices? (you may want to make a board to handle the source and current sensing resistors.) I'm going for the third option because my heat sinks and chassis won't accomodate Jens' board.
Bob
You're not really relying on the protection circuit unless you run too few output pairs. So it is a tradeoff - do you want to make your own boards for the double barrel version or use Jens' board loaded up, or Professor Leach's board and add extra output devices? (you may want to make a board to handle the source and current sensing resistors.) I'm going for the third option because my heat sinks and chassis won't accomodate Jens' board.
Bob
- Status
- Not open for further replies.