"I've done two Leach amps, both for other people that requested specifically that amp, and in a first one output devices were MJL3281/MJL1302, that also worked marvelously. I developed PCB for stereo version using that case (TO247), and I can post it too if there is any interest. "
Can you post it please?
Can you post it please?
Thanks!
Is the part list same as this (expect output transistors):
http://users.ece.gatech.edu/~mleach/lowtim/part3.html
I think I will put some extra output devices so I have to change a little bit of your pcb.
Is the part list same as this (expect output transistors):
http://users.ece.gatech.edu/~mleach/lowtim/part3.html
I think I will put some extra output devices so I have to change a little bit of your pcb.
Macgyver, it seems you skipped one of my posts. For Q1 up to Q11 I used 2N5551/2N5401. To me, they are better choice than MPSA and I am using them a lot.
There are some other changes because Prof. Leach and I do not agree on some issues.
I think that Leach's value of output coil is too small, and at the same time, value of R49 is too high. Value of around 4 to 5uH is much more appropriate and it should be shunted with no more than 2 to 3 ohms (2,2 ohms in my case). Those values are applicable to most amplifiers for they are generally not related to configuration but to the load impedance and amount of ringing with fast rising edges. In a waste majority of cases of such measurements, rising time of the leading edge of the square wave used, is not stated, but it is the most important factor for accuracy and relevance of such measurements. With the rising time of 1uS for example, you will find no or minimal ringing with good number of amps. Try them with 20nS rising square wave and you will see completely different picture. Higher values of output inductor will reduce ringing and reduction of the value of shunting resistor will improve things even further. Normally, there are limits and for an 8 Ohm load it will be typically a maximum of about 6uH, any larger than this will cause unacceptable attenuation of high frequencies. A 6uH inductor will introduce a low frequency loss (assuming 0.03 Ohm DC resistance of the coil) of 0.03dB and will be about 0.2dB down at 20kHz. These losses are insignificant, and will not be audible. In contrast, ringing (or in extreme cases, oscillation) of the output devices will be audible (even at very low levels) as increased distortion.
Of course, that's my point of view and I am not advocating that anyone should accept my suggestions.
There are some other changes because Prof. Leach and I do not agree on some issues.
I think that Leach's value of output coil is too small, and at the same time, value of R49 is too high. Value of around 4 to 5uH is much more appropriate and it should be shunted with no more than 2 to 3 ohms (2,2 ohms in my case). Those values are applicable to most amplifiers for they are generally not related to configuration but to the load impedance and amount of ringing with fast rising edges. In a waste majority of cases of such measurements, rising time of the leading edge of the square wave used, is not stated, but it is the most important factor for accuracy and relevance of such measurements. With the rising time of 1uS for example, you will find no or minimal ringing with good number of amps. Try them with 20nS rising square wave and you will see completely different picture. Higher values of output inductor will reduce ringing and reduction of the value of shunting resistor will improve things even further. Normally, there are limits and for an 8 Ohm load it will be typically a maximum of about 6uH, any larger than this will cause unacceptable attenuation of high frequencies. A 6uH inductor will introduce a low frequency loss (assuming 0.03 Ohm DC resistance of the coil) of 0.03dB and will be about 0.2dB down at 20kHz. These losses are insignificant, and will not be audible. In contrast, ringing (or in extreme cases, oscillation) of the output devices will be audible (even at very low levels) as increased distortion.
Of course, that's my point of view and I am not advocating that anyone should accept my suggestions.
You are right, and I am aware of that clever detail from the first moment I saw the design in "Audio" magazine somewhere around 1976. My point is: why relying only on that when you can raise safety margin against ringing and oscillations with such simple modification which ,on the other hand, won't negatively affect or degrade other characteristics. There are piles of modern loudspeakers which will prove to be a very "nasty" loads to most amps (including even this one), and additional measures toward better stability won't harm.
As a designer that is frequently in a position to do hand made amps on custom orders and mostly without knowing speakers that will be used during the life time of my amp, I always incline to the solutions that will provide better safety margin without degradation in other areas.
As a designer that is frequently in a position to do hand made amps on custom orders and mostly without knowing speakers that will be used during the life time of my amp, I always incline to the solutions that will provide better safety margin without degradation in other areas.
Hi Boraomega,
re Thiel network on output.
I have seen a few designs with a pi network rather than the two L pad versions proposed by Thiel.
Would the pi version with RC to ground from both ends of the inductor help with the rf rejection senario?
What values should be used? How about doubling R and halving C for each CR?
re Thiel network on output.
I have seen a few designs with a pi network rather than the two L pad versions proposed by Thiel.
Would the pi version with RC to ground from both ends of the inductor help with the rf rejection senario?
What values should be used? How about doubling R and halving C for each CR?
boraomega said:
This feature was not in his original amp from 1976. It was added in a later revision. It was published in an article in May of 1988, in IEEE Transactions on Consumer Electronics. This feedback technique bypasses the pole in the output stage caused by the load capacitance for improved stability. It has been said the output inductor may not even be needed.
AndrewT:
>Hi Boraomega,
>re Thiel network on output.
>I have seen a few designs with a pi network rather than the >two L pad versions proposed by Thiel.
>Would the pi version with RC to ground from both ends of the >inductor help with the rf rejection senario?
.What values should be used? How about doubling R and halving >C for each CR?
Frankly, I don't have any "firm" and final opinion about that addition (another Zobel after the coil), mostly due to the fact that I have never devoted any significant time to test it "from all angles". Some first impressions were that it could (eventually) be useful if you assume that the capacitance of speaker cable will be of the order of several tens of nano-farads, which would really be a grotty affaire. Most of the currently available speaker cables will have somewhere between a few hundreds pF up to the maximum of, say, 1,5nF for reasonable room length wiring, hence, I don't see the point. I've tried it only once in practice, and I haven't noticed any influence in either direction (positive or negative). It could be that some further investigations will reveal much more, but as I said, first impressions haven't justified any significant time and effort investment from me. It is possible that I am equally wrong about that........
>Hi Boraomega,
>re Thiel network on output.
>I have seen a few designs with a pi network rather than the >two L pad versions proposed by Thiel.
>Would the pi version with RC to ground from both ends of the >inductor help with the rf rejection senario?
.What values should be used? How about doubling R and halving >C for each CR?
Frankly, I don't have any "firm" and final opinion about that addition (another Zobel after the coil), mostly due to the fact that I have never devoted any significant time to test it "from all angles". Some first impressions were that it could (eventually) be useful if you assume that the capacitance of speaker cable will be of the order of several tens of nano-farads, which would really be a grotty affaire. Most of the currently available speaker cables will have somewhere between a few hundreds pF up to the maximum of, say, 1,5nF for reasonable room length wiring, hence, I don't see the point. I've tried it only once in practice, and I haven't noticed any influence in either direction (positive or negative). It could be that some further investigations will reveal much more, but as I said, first impressions haven't justified any significant time and effort investment from me. It is possible that I am equally wrong about that........
pooge wrote:
This feature was not in his original amp from 1976. It was added in a later revision. It was published in an article in May of 1988, in IEEE Transactions on Consumer Electronics. This feedback technique bypasses the pole in the output stage caused by the load capacitance for improved stability. It has been said the output inductor may not even be needed.
--- Thank you for info! I never followed this design chronologically (I haven't follow it at all by all means!) apart from seeing it in "Audio" in our University Library way back in '76 and again more than 20 years later when customer brought schematic with request to build that amp for him. Without having original schematic to compare, I presumed that this feature was there from the very beginning which is obviously not the case ( If you don't have a photographic memory, don't rely on it!).
Anyway, I don't think that exact moment or revision in which that mod appeared for the first time, is of a crucial relevance for our current theme.
macgyver wrote:
What are 100ohm resistors in near of Q18-21 devices (TO-247 layout)? Where will the another head "leads"? I don't find them in the leach schematics or part list.
---No, you won't find them for they are my additions for connection of my design of universal protection system. They are actually sampling points, and you can freely neglect them unless you are interested to make that protection system too. That protection system is not designed specifically for Leach amp, but is pretty much universal and applicable to just about all modern amps with complementary pairs in output stage and split supply. Some of incorporated additional features are: protection against DC on speaker terminals, delay for speaker switching, thermal protection and protection against incorrect (too low) load impedance.
This feature was not in his original amp from 1976. It was added in a later revision. It was published in an article in May of 1988, in IEEE Transactions on Consumer Electronics. This feedback technique bypasses the pole in the output stage caused by the load capacitance for improved stability. It has been said the output inductor may not even be needed.
--- Thank you for info! I never followed this design chronologically (I haven't follow it at all by all means!) apart from seeing it in "Audio" in our University Library way back in '76 and again more than 20 years later when customer brought schematic with request to build that amp for him. Without having original schematic to compare, I presumed that this feature was there from the very beginning which is obviously not the case ( If you don't have a photographic memory, don't rely on it!).
Anyway, I don't think that exact moment or revision in which that mod appeared for the first time, is of a crucial relevance for our current theme.
macgyver wrote:
What are 100ohm resistors in near of Q18-21 devices (TO-247 layout)? Where will the another head "leads"? I don't find them in the leach schematics or part list.
---No, you won't find them for they are my additions for connection of my design of universal protection system. They are actually sampling points, and you can freely neglect them unless you are interested to make that protection system too. That protection system is not designed specifically for Leach amp, but is pretty much universal and applicable to just about all modern amps with complementary pairs in output stage and split supply. Some of incorporated additional features are: protection against DC on speaker terminals, delay for speaker switching, thermal protection and protection against incorrect (too low) load impedance.
"No, you won't find them for they are my additions for connection of my design of universal protection system. They are actually sampling points, and you can freely neglect them unless you are interested to make that protection system too. That protection system is not designed specifically for Leach amp, but is pretty much universal and applicable to just about all modern amps with complementary pairs in output stage and split supply. Some of incorporated additional features are: protection against DC on speaker terminals, delay for speaker switching, thermal protection and protection against incorrect (too low) load impedance."
Could you send me schematic to the protection circuit?
b8_beat@hotmail.com
Could you send me schematic to the protection circuit?
b8_beat@hotmail.com
Hy
You can faind that schematic and PCB, on Dr. Borivoje web page, and other useful staff related to DIY audio........ I made his Studio amplifier and I am very satisfied with that project....,
http://www.bas.abshost.net/mojiprojekti1.html
Pomoæni ureðaji / OMNI protektor
Vinko
(Radili ste i leach pojaèalo!, imam isprintan cili projekt za njega, koji su vaši dojmovi o njemu kao originalu te posli vaše prepravke, mislim na signalne te izlazne tranzistore koji su ipak nešto moderniji, )
You can faind that schematic and PCB, on Dr. Borivoje web page, and other useful staff related to DIY audio........ I made his Studio amplifier and I am very satisfied with that project....,
http://www.bas.abshost.net/mojiprojekti1.html
Pomoæni ureðaji / OMNI protektor
Vinko
(Radili ste i leach pojaèalo!, imam isprintan cili projekt za njega, koji su vaši dojmovi o njemu kao originalu te posli vaše prepravke, mislim na signalne te izlazne tranzistore koji su ipak nešto moderniji, )
boraomega ,
I would like to have your protection ckt too , or why not post it on the site for everyone ,
Rajeev
systems_rajeev@rediffmail.com
pinco
How can I read on that website , its not english , also it needs virtual jawa machine in the pc which I do not have in mine .
I would like to have your protection ckt too , or why not post it on the site for everyone ,
Rajeev
systems_rajeev@rediffmail.com
pinco
How can I read on that website , its not english , also it needs virtual jawa machine in the pc which I do not have in mine .
You must have java v.m. , then you have menu, click "pomoæni ureðaji" and "OMNI protector"
http://www.bas.abshost.net/mojiprojekti1.html
Vinko
http://www.bas.abshost.net/mojiprojekti1.html
Vinko
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.