jethari said:
Hi Hari,
Yes you can eMail me the file. I will eMail you via diyAudio with an eMail address to send the file to.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 00:56
Official Quasi Thread Researcher
Re: Re: Performance report
Keypunch,
I do my simulations with Electronics Workbench 5.0. This runs only on Windows98. I know this is obsolete and I need to upgrade to a better simulation package that runs on Win2K and above. I am saying this because I haven't tried importing .ewb schamatics to post them as pictures. Hence, I won't be able to post a schematic, as per your request.
However, it is as simple as this: the supply for the front end is tapped off from the main supply board. First there are the IN400x diodes and in series are 10R 2W resistors, followed by the caps in shunt. The diodes do not allow reverse flow of current and hence, isolate the front end supply from varying too much in sympathy with the main supply.
Caps of different values help to bring down the total inductance. Having different charge/discharge periods could be a problem when the current draw is high, but the current draw of the front end of the amp is low and hence, this arrangement works well here.
R12/R15 need to be 68R and not 680R. I corrected the typo in a subsequent post.
Protecting the front end of the amp with rail fuses is superfluous. I think a properly rated fuse in this position will only blow after some damage has occured in the components, particularly transistors. Moreover, the whole amp is biased from back-to-front by the LTP and cutting off supply to this section through blown fuses will result in one of the rail voltages appearing at the output. Even with DC protection, this is an unwarranted situation. Currents flowing in the front end of an amp are low compared to the output stage and most often damages are minimal. I have not seen the front end being protected separately in any amp design.
I can't say whether my experiences are related to the use of KSE parts, rather than MJE; but compared to the experiences of other builders as reported here, it might seem like it.
I think that there was some current starvation in the front end with higher rail voltages. I have encountered a similar problem with a 100 watt Lateral Mosfet design which appeared in EW&WW. In that case I used a regulated supply for the front end, reduced the supply for the front end compared to the output stage. The amp had fantastic resolution but lacked a bit in low end punch. Fortunately in that situation I had installed a powered sub, so the phenomenon of lacking-bass, came as a blessing.
Keypunch, my suggestion is that you go ahead an build your amp. It looks like you have done enough research on this thread. I don't mean to sound negative, but something better is likely to come along, either as an offshoot of this thread or otherwise and this kind of occurences, leave many audiophiles with numerous incomplete projects. I might have over a dozen such circuits, all half done and many more that have been thoroughly tested but not housed in a chassis.
keypunch said:
Keypunch,
I do my simulations with Electronics Workbench 5.0. This runs only on Windows98. I know this is obsolete and I need to upgrade to a better simulation package that runs on Win2K and above. I am saying this because I haven't tried importing .ewb schamatics to post them as pictures. Hence, I won't be able to post a schematic, as per your request.
However, it is as simple as this: the supply for the front end is tapped off from the main supply board. First there are the IN400x diodes and in series are 10R 2W resistors, followed by the caps in shunt. The diodes do not allow reverse flow of current and hence, isolate the front end supply from varying too much in sympathy with the main supply.
Caps of different values help to bring down the total inductance. Having different charge/discharge periods could be a problem when the current draw is high, but the current draw of the front end of the amp is low and hence, this arrangement works well here.
R12/R15 need to be 68R and not 680R. I corrected the typo in a subsequent post.
Protecting the front end of the amp with rail fuses is superfluous. I think a properly rated fuse in this position will only blow after some damage has occured in the components, particularly transistors. Moreover, the whole amp is biased from back-to-front by the LTP and cutting off supply to this section through blown fuses will result in one of the rail voltages appearing at the output. Even with DC protection, this is an unwarranted situation. Currents flowing in the front end of an amp are low compared to the output stage and most often damages are minimal. I have not seen the front end being protected separately in any amp design.
I can't say whether my experiences are related to the use of KSE parts, rather than MJE; but compared to the experiences of other builders as reported here, it might seem like it.
I think that there was some current starvation in the front end with higher rail voltages. I have encountered a similar problem with a 100 watt Lateral Mosfet design which appeared in EW&WW. In that case I used a regulated supply for the front end, reduced the supply for the front end compared to the output stage. The amp had fantastic resolution but lacked a bit in low end punch. Fortunately in that situation I had installed a powered sub, so the phenomenon of lacking-bass, came as a blessing.
Keypunch, my suggestion is that you go ahead an build your amp. It looks like you have done enough research on this thread. I don't mean to sound negative, but something better is likely to come along, either as an offshoot of this thread or otherwise and this kind of occurences, leave many audiophiles with numerous incomplete projects. I might have over a dozen such circuits, all half done and many more that have been thoroughly tested but not housed in a chassis.
Re: Performance report
Samuel,
Thanks for your informative reply and taking the time to reply.
I was aware of the typo correction of 680R that should of been 68R. I simply chose just your one posting as it was the main source of my questions.
I was not implying a need for a simulation when I asked for the driver/front end PSU configuration you used. I just wanted to be certain I understood how the amp module elements for the driver/front end supply were arranged as I am not an EE. I asked for the schematic segment so I was sure I had your approach correct rather than assume. I believe I understand with your verbal schematic reply where the resistor and diode are in series with the PSU rail supply follows with the caps in parallel to the rail to ground.
The main reason I asked about the different value electrolytics was I have sometimes read some who have the opinion of using a mix of different value smoothing electrolytic caps in the PSU. With my limited knowledge of electronics this seemed to me like a poor idea for the reason you indicated that it might pose a problem if not for the low current draw of the driver/front end stages. I knew the reason was for paralleling to reduce inductance/ESR. I simply was not sure if you had some specific rational in design vs what parts you had available to effect a lower ESR/inducatance.
Thanks for your insight to using the KSE parts and your past experiences with current starvation that can happen.
I am not offended with your suggestion to go ahead and build. I would like very much to actually do so, but I am still having some difficulty locating certain parts to allow me to go forward or at reasonable cost.
I have not located any suitable heatsinks yet, used or new. I would lend to used to control costs as I am building for an active crossover based configuration. I do have a backup plan to make my own heatsinks if necessary. Doing so will add time in not only building, but going to a local metal supplier to buy the raw materials.
PSU filter capacitors are still a hard to find at reasonable cost. I picked up some excellent toroids a couple years back in surplus at an excellent price which is great. The ones I would use are really big and heavy physically at almost 1000VA capacity. Sadly this puts the PSU filter capacitors rating in the 85V-100V. Had I had transformers of about 35-0-35 I could of used 63V filter capacitors where were available in excellent quanity and price locally.
Then there is the case. I have seen others build cases from scratch and suspect I could as well. That would add alot of extra time which I have no personal issue with, but I simply have not had that kind of time available. The case has to handle the 1000VA toroids that are about 20 pounds and 7 inches in diameter and about 4 inches high.
The most important reason is aside from a number of software projects that were a must for me to attend to for my own system and very time consuming, there have been a number of life realted crisis over the past couple years. These to some extent are still ongoing that prevent me from knowing I can start building and be able to finish a 4 or 6 module unit from one of the 1000VA toroids or try to build a small power version using a 20V-0V-20V toroid I have, still to source last type of parts aside and learning how to make a PCB.
That said I am not sure if you have noticed, but this is the thread I have spent most of my time with. I have spent time when time allows in some other threads, but mostly for my education or generic design and learning. I very much intend to build the quasi amp. I am not waiting for something better. I am sure quasi's amp will meet my needs.
At some point I might consider low power Class A for the tweeters, but that has always been the case even before I found a n-channel AB class amplifier I would be happy with. I am very happy with the quasi design and results many have had. I really like the idea if I wish I can build the same basic design using the BiPolar output design variant quasi designed and now there is a hint of a Class A design that again has similar front end. That is great so if I wish or need to build a BiPolar or Class A I can do so with all the same basic design for the front and and/or predriver which enables one to have as close as one can have to same for all amps in an active crossover configuration, yet choose the class or output device one wishes based on needs or preference. At some point I will want to also build a pre-amp, but that is not a priority as much as building some amps are. Building new speakers are the next priority after the amplifiers.
Initially I had lots to learn before I embarked on building. I still have lots to learn, but if the last few parts I need can be sorted out I would be happy and comfortable I know what PSU design I will use, likely what balanced input design I will use and what choices I have for a VU meter to start building. I am comfortable editing the PCB to a accomodate some changes I wanted even though it is via a graphics editing program and not a eCAD. I just have to merge the changes onto one PCB.
As you know even when building you need more time than expected as kinks arise that were not expected. That is ok for me, but the last 18 months have been short to no time available. I have actually spent alot of time looking for parts. I think I finially found a relay that is suitable for the DC protection after searching for almost 2 years, not constant of course, but looking as I cannot order from the US due to odd policies imposed on one to do so. the MJE340/350 parts were a challenge to find as well in the quanity I needed, but a kind gent was able to supply me the KSE versions. So basically the parts sourcing process is one of the key reasons for me not starting a build yet. I am patient, and I suspect have far more than many people do even outside the diy audio types.
In addition I have taken on and still have some research to do in the thread to answer some questions some posters have asked. I have a nack, due in large part to my patience, to go through the thread and create postings of links collections of the key items of interest to posters that do not have time to read the whole thread, let alone organize the information. I am behind on that count as well due to the life crisis the last 18 months and my limited time still. Tis is one skill I have and is one way I can give back for all those that have had to endure my many questions and sometimes long postings 😉 Think this is one of those. Sorry Samuel.
Again thanks for taking the time to reply to my inquiry Samuel and thanks for your informative experience and results in your build of the quasi NMOS design.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 12:52
Official Quasi Thread Researcher
Samuel,
Thanks for your informative reply and taking the time to reply.
I was aware of the typo correction of 680R that should of been 68R. I simply chose just your one posting as it was the main source of my questions.
I was not implying a need for a simulation when I asked for the driver/front end PSU configuration you used. I just wanted to be certain I understood how the amp module elements for the driver/front end supply were arranged as I am not an EE. I asked for the schematic segment so I was sure I had your approach correct rather than assume. I believe I understand with your verbal schematic reply where the resistor and diode are in series with the PSU rail supply follows with the caps in parallel to the rail to ground.
The main reason I asked about the different value electrolytics was I have sometimes read some who have the opinion of using a mix of different value smoothing electrolytic caps in the PSU. With my limited knowledge of electronics this seemed to me like a poor idea for the reason you indicated that it might pose a problem if not for the low current draw of the driver/front end stages. I knew the reason was for paralleling to reduce inductance/ESR. I simply was not sure if you had some specific rational in design vs what parts you had available to effect a lower ESR/inducatance.
Thanks for your insight to using the KSE parts and your past experiences with current starvation that can happen.
I am not offended with your suggestion to go ahead and build. I would like very much to actually do so, but I am still having some difficulty locating certain parts to allow me to go forward or at reasonable cost.
I have not located any suitable heatsinks yet, used or new. I would lend to used to control costs as I am building for an active crossover based configuration. I do have a backup plan to make my own heatsinks if necessary. Doing so will add time in not only building, but going to a local metal supplier to buy the raw materials.
PSU filter capacitors are still a hard to find at reasonable cost. I picked up some excellent toroids a couple years back in surplus at an excellent price which is great. The ones I would use are really big and heavy physically at almost 1000VA capacity. Sadly this puts the PSU filter capacitors rating in the 85V-100V. Had I had transformers of about 35-0-35 I could of used 63V filter capacitors where were available in excellent quanity and price locally.
Then there is the case. I have seen others build cases from scratch and suspect I could as well. That would add alot of extra time which I have no personal issue with, but I simply have not had that kind of time available. The case has to handle the 1000VA toroids that are about 20 pounds and 7 inches in diameter and about 4 inches high.
The most important reason is aside from a number of software projects that were a must for me to attend to for my own system and very time consuming, there have been a number of life realted crisis over the past couple years. These to some extent are still ongoing that prevent me from knowing I can start building and be able to finish a 4 or 6 module unit from one of the 1000VA toroids or try to build a small power version using a 20V-0V-20V toroid I have, still to source last type of parts aside and learning how to make a PCB.
That said I am not sure if you have noticed, but this is the thread I have spent most of my time with. I have spent time when time allows in some other threads, but mostly for my education or generic design and learning. I very much intend to build the quasi amp. I am not waiting for something better. I am sure quasi's amp will meet my needs.
At some point I might consider low power Class A for the tweeters, but that has always been the case even before I found a n-channel AB class amplifier I would be happy with. I am very happy with the quasi design and results many have had. I really like the idea if I wish I can build the same basic design using the BiPolar output design variant quasi designed and now there is a hint of a Class A design that again has similar front end. That is great so if I wish or need to build a BiPolar or Class A I can do so with all the same basic design for the front and and/or predriver which enables one to have as close as one can have to same for all amps in an active crossover configuration, yet choose the class or output device one wishes based on needs or preference. At some point I will want to also build a pre-amp, but that is not a priority as much as building some amps are. Building new speakers are the next priority after the amplifiers.
Initially I had lots to learn before I embarked on building. I still have lots to learn, but if the last few parts I need can be sorted out I would be happy and comfortable I know what PSU design I will use, likely what balanced input design I will use and what choices I have for a VU meter to start building. I am comfortable editing the PCB to a accomodate some changes I wanted even though it is via a graphics editing program and not a eCAD. I just have to merge the changes onto one PCB.
As you know even when building you need more time than expected as kinks arise that were not expected. That is ok for me, but the last 18 months have been short to no time available. I have actually spent alot of time looking for parts. I think I finially found a relay that is suitable for the DC protection after searching for almost 2 years, not constant of course, but looking as I cannot order from the US due to odd policies imposed on one to do so. the MJE340/350 parts were a challenge to find as well in the quanity I needed, but a kind gent was able to supply me the KSE versions. So basically the parts sourcing process is one of the key reasons for me not starting a build yet. I am patient, and I suspect have far more than many people do even outside the diy audio types.
In addition I have taken on and still have some research to do in the thread to answer some questions some posters have asked. I have a nack, due in large part to my patience, to go through the thread and create postings of links collections of the key items of interest to posters that do not have time to read the whole thread, let alone organize the information. I am behind on that count as well due to the life crisis the last 18 months and my limited time still. Tis is one skill I have and is one way I can give back for all those that have had to endure my many questions and sometimes long postings 😉 Think this is one of those. Sorry Samuel.
Again thanks for taking the time to reply to my inquiry Samuel and thanks for your informative experience and results in your build of the quasi NMOS design.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 12:52
Official Quasi Thread Researcher
Hari,
You should of found I had eMailed back your PCB layout in PDF form with a file size of about 75Kb. I made some comments about the PCB or possible considerations in that eMail. Before the demands of the work week, let alone domestic chores I need to do today, there are other comments I wanted to make I forgot to do so in the eMail of early this morning before I forget them.
Have you allowed for T6 and T7 to be heatsinked by providing enough space and clearance about the other parts to do so? I think T6 and T7 need to be heatsinked even though they do not need to share the main heatsink like T8, T9 and T10 need to with the output devices.
The trace from L1 to the the output tab should to be at least as wide as the traces you have from the output devices. The same curent load will be on that trace from L1 to the output tab.
The traces for rails and about parts of the traces to the output devices are much narrower than the rest of these traces in the R24 and R27 areas that feed R29 - R32.
VR1 appears to have a setscrew on the side to adjust the resistance, but it looks like R6 may block acess to the VR1 adjustment setscrew.
As a general comment any reason you like the VR parts to have side facing setscrew adjustment rather than a top side adjustment setscrew? It would seem to me that it would be much easier to set or adjust if these were top facing setscrew types. Then again you, like quasi, may have an existing stash of parts you like to use.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 13:16
Official Quasi Thread Researcher
You should of found I had eMailed back your PCB layout in PDF form with a file size of about 75Kb. I made some comments about the PCB or possible considerations in that eMail. Before the demands of the work week, let alone domestic chores I need to do today, there are other comments I wanted to make I forgot to do so in the eMail of early this morning before I forget them.
Have you allowed for T6 and T7 to be heatsinked by providing enough space and clearance about the other parts to do so? I think T6 and T7 need to be heatsinked even though they do not need to share the main heatsink like T8, T9 and T10 need to with the output devices.
The trace from L1 to the the output tab should to be at least as wide as the traces you have from the output devices. The same curent load will be on that trace from L1 to the output tab.
The traces for rails and about parts of the traces to the output devices are much narrower than the rest of these traces in the R24 and R27 areas that feed R29 - R32.
VR1 appears to have a setscrew on the side to adjust the resistance, but it looks like R6 may block acess to the VR1 adjustment setscrew.
As a general comment any reason you like the VR parts to have side facing setscrew adjustment rather than a top side adjustment setscrew? It would seem to me that it would be much easier to set or adjust if these were top facing setscrew types. Then again you, like quasi, may have an existing stash of parts you like to use.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 13:16
Official Quasi Thread Researcher
Hi All,
I happened to find a Windows based program, PDFCreator that may be helpful to those wishing to create PDF documents for posting. It also has a "Compression" tab that may make it easier. I say may, as I do not have a Windows System here to look and try it out. I am a Linux user and only install or bring up a Windows system when I must. Usually once or less a year. If any of you want to check out PDFCreator and determine how useful this program is perhaps you can let us know what you think of it. It is an Open Source program which means it is free for commerical and non-commerical use.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 17:06
Official Quasi Thread Researcher
I happened to find a Windows based program, PDFCreator that may be helpful to those wishing to create PDF documents for posting. It also has a "Compression" tab that may make it easier. I say may, as I do not have a Windows System here to look and try it out. I am a Linux user and only install or bring up a Windows system when I must. Usually once or less a year. If any of you want to check out PDFCreator and determine how useful this program is perhaps you can let us know what you think of it. It is an Open Source program which means it is free for commerical and non-commerical use.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 17:06
Official Quasi Thread Researcher
hi keypunch,
After a recommendation in the DX Amplifier thread, I downloaded and use PDFCreator. It works very well although I haven't done specific tests on how well it compresses.
BTW: I don't care what operatering system you use, you don't need to make apologies for using Linux. 😀
regards
After a recommendation in the DX Amplifier thread, I downloaded and use PDFCreator. It works very well although I haven't done specific tests on how well it compresses.
BTW: I don't care what operatering system you use, you don't need to make apologies for using Linux. 😀
regards
Hi Greg,
Great. I only noted I use Linux and had no Windows machine to test the PDFCreator program. Great you liked it. I had a sense it would be great as long as it was not buggy or had insufficient heaksinking 😉
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 18:03
Official Quasi Thread Researcher
Perhaps Odd Time diyAudio Reseacher 😉
Great. I only noted I use Linux and had no Windows machine to test the PDFCreator program. Great you liked it. I had a sense it would be great as long as it was not buggy or had insufficient heaksinking 😉
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 18:03
Official Quasi Thread Researcher
Perhaps Odd Time diyAudio Reseacher 😉
A Question of Gain
Hi,
I like to know if formula for amplifier gain is R17/R18 or 1/(R18/(R18+R17)) for the NFB design used in the quasi NMOS350/500/200 as is is also used on many other amplifier designs? The answers for these two equations are close, but not the same - 33 and 34 respectively. I have seen the R17/R18 type equation much more, but by chance I have now seen the 1/(R18/(R18+R17)) equation a couple of times.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 18:56
Official Quasi Thread Researcher
Hi,
I like to know if formula for amplifier gain is R17/R18 or 1/(R18/(R18+R17)) for the NFB design used in the quasi NMOS350/500/200 as is is also used on many other amplifier designs? The answers for these two equations are close, but not the same - 33 and 34 respectively. I have seen the R17/R18 type equation much more, but by chance I have now seen the 1/(R18/(R18+R17)) equation a couple of times.
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 18:56
Official Quasi Thread Researcher
Hi John,
You are doing a marvellous service to fellow DIYers and to Quasi!!
Try (R18/(R18+R17). That's the right one!!
Cheers,
Hugh
You are doing a marvellous service to fellow DIYers and to Quasi!!
Try (R18/(R18+R17). That's the right one!!
Cheers,
Hugh
AKSA said:
Hi Hugh,
Thanks. There are many excellent people here who are doing a marvellous job as well. I know there are many on this forum who think highly of your knowledge and products.
Did you make a typo and omit the "1/" in front of the (R18/(R18+R17)? I have made so many typos over the postings I have made ... ergo my self described forum signature 😉
Regards,
John L. Males
Willowdale, Ontario
Canada
08 July 2007 19:42
Official Quasi Thread Researcher
08 July 2007 19:46/19:53 Opps made a handful of typos with this short posting. jlm 😉
Thanks John,
Dang!! I did indeed!! My apologies!!
Correct relation is (R17 + R18)/R18, with R18 the lower of the two feedback resistor, that is, the shunt element. R17 is the series element which comes off the output.
So, in easyspeak, add the series and shunt resistor values, say 27K and 1K to give 28K, then divide by the lesser, giving a gain of 28.
This applies to non-inverting input amps. Inverting input amps have only the greater divided by the lesser, BUT this assumes zero source impedance driving them, which is near enough for CD/DVD players, but not quite true of tube preamps.
Cheers,
Hugh
Dang!! I did indeed!! My apologies!!
Correct relation is (R17 + R18)/R18, with R18 the lower of the two feedback resistor, that is, the shunt element. R17 is the series element which comes off the output.
So, in easyspeak, add the series and shunt resistor values, say 27K and 1K to give 28K, then divide by the lesser, giving a gain of 28.
This applies to non-inverting input amps. Inverting input amps have only the greater divided by the lesser, BUT this assumes zero source impedance driving them, which is near enough for CD/DVD players, but not quite true of tube preamps.
Cheers,
Hugh
keypunch said:
John
Thanks for the compressed file. As I had emailed to you, I had already identified two flaws in the pcb vis a vis the schematic and have corrected them.
There is enough space for heatsinks for T6 and T7. I'll be using stand alone heat sinks for both of them.
Good point about the trace from output of L1. Guess I was just getting a bit lazy by that time
I've taken care of that now.The feed traces to R29-R32 need not be that heavy IMHO as the trace lengths are pretty short. Very heavy traces are required over long lengths to offset possible current loss. All the same, I've increased the trace width marginally.
Dont bother about the shape of VR1, that's the only component footprint I have. I can very easily use the one with the screw on the top as the pad spacings are the same.
I'm still working towards creating smaller pdf files. If all else fails, I'll email the ps file to you 😀
Thanks
Hari
i have a question!
could i add a shortcircuit to this amplifier(nmos350)?
i found it on the web:
http://users.swing.be/edwinpaij/short_circuit_protection.htm
now i am wondering if i should add this circuit to the nmos350 the easy way(nmos350 and shortcircuit will be connected by wires) or it would be better to add it in the original pcb of nmos350?
thanks
could i add a shortcircuit to this amplifier(nmos350)?
i found it on the web:
http://users.swing.be/edwinpaij/short_circuit_protection.htm
now i am wondering if i should add this circuit to the nmos350 the easy way(nmos350 and shortcircuit will be connected by wires) or it would be better to add it in the original pcb of nmos350?
thanks
Hi Smoking,
no. That circuit applies to complementary pairs.
It works by shorting the input to the drivers to the common output line.
Quasi's circuit is Nchannel and requires the top half of that protection scheme in his top half output stage and a repeat of the top half protection applied to the lower half of the output stage but this time shorting the input to Vee.
This would need to be breadboarded (or simulated) to prove it still works. The ladder of three resistors will no longer exist, an alternative circuit will be needed here.
It also omitts the V part of IV detection with the result that it is not particularly effective.
Probably not worth the effort.
Other schemes could do it much better. But keep in mind, nearly all protection schemes are designed for complementary. Look out for common emitter output stage protection for ideas on how to implement the lower half.
Now the big question!!!!
Does a FET stage need IV detection?
I say yes, most others say Zener protection is sufficient when used in combination with close rated supply rail fuses.
Any IV protection must not trigger on valid audio signals. Otherwise it becomes audible and that is no good at all.
no. That circuit applies to complementary pairs.
It works by shorting the input to the drivers to the common output line.
Quasi's circuit is Nchannel and requires the top half of that protection scheme in his top half output stage and a repeat of the top half protection applied to the lower half of the output stage but this time shorting the input to Vee.
This would need to be breadboarded (or simulated) to prove it still works. The ladder of three resistors will no longer exist, an alternative circuit will be needed here.
It also omitts the V part of IV detection with the result that it is not particularly effective.
Probably not worth the effort.
Other schemes could do it much better. But keep in mind, nearly all protection schemes are designed for complementary. Look out for common emitter output stage protection for ideas on how to implement the lower half.
Now the big question!!!!
Does a FET stage need IV detection?
I say yes, most others say Zener protection is sufficient when used in combination with close rated supply rail fuses.
Any IV protection must not trigger on valid audio signals. Otherwise it becomes audible and that is no good at all.
Hi Andrew,
That was a very good post Andrew.
-Chris
Good question. A normal detection circuit should work fine across the source - output. The detect detect line can then be modified to clamp the lower transistors gate as well. It should have a time constant or latch.
That may be enough. I personally don't know.
A golden rule if I ever saw one!
That was a very good post Andrew.
-Chris
I don't know if a zenner circuit is enough for protecting the output stage but for shure it won't hurt.
Does anyone know how to choose the zenner value?
Does anyone know how to choose the zenner value?
Hi,
decide your safe output current for each device in the output stage.
Look up the datasheet and select the voltage from the Vds vs Id graph that matches your maximum current.
Find a Zener that is just above that voltage.
The problem with this method is that the Vds vs Id is not straight horizontal line. As Vds rises so does Id leading to a compromise; higher dissipation than the device can take versus adequate pass current without hitting the limiter. This becomes very significant into reactive loads where High Vds can co-exist with High Id.
The result is audible limiting, or inadequate protection. Take your choice.
decide your safe output current for each device in the output stage.
Look up the datasheet and select the voltage from the Vds vs Id graph that matches your maximum current.
Find a Zener that is just above that voltage.
The problem with this method is that the Vds vs Id is not straight horizontal line. As Vds rises so does Id leading to a compromise; higher dissipation than the device can take versus adequate pass current without hitting the limiter. This becomes very significant into reactive loads where High Vds can co-exist with High Id.
The result is audible limiting, or inadequate protection. Take your choice.
cd-i said:
Zeners cannot be used in this schematic. In case of the Nmos350 / 500 a zener on the emmiter of T9 will destroy itself or T9 or both if it ever turned on. The situation is similar for the zener connected to the collector of T10.
There is a description for short cct protection over a few posts commencing here http://www.diyaudio.com/forums/showthread.php?postid=1173157#post1173157 but I have not yet built or tested it. I think it's worthy of development as it could be a universal cct that can be used elsewhere.
Cheers
Q
Hello
There is also this Audio Amp Output Power Limiter from
Rod Elliott (ESP)
http://sound.westhost.com/project53.htm
Gaetan
There is also this Audio Amp Output Power Limiter from
Rod Elliott (ESP)
http://sound.westhost.com/project53.htm
Gaetan
One very good (all in one) amp protection:
http://www.diyaudio.com/forums/showthread.php?postid=649777#post649777
Regards,
Miodrag Pejin
http://www.diyaudio.com/forums/showthread.php?postid=649777#post649777
Regards,
Miodrag Pejin