That was done on another very good amp with Sziklai outputs - LJM's MX50SE. But LJM kit has dedicated heatsink specially produced for this.
been finishing my first Alchemist Forseti to have repaired .... tough issues the thermals was a complete nightmare in this amplifier but i managed my way around it
As discussed before the APD 15A is almost identical with the P3A while features a world wide unique output stage formed by double an sziklai pair .It features 4 outputs and 4 drivers per ch .
To keep stable uses miller caps over power transistors in the astronomical size of 680 pf which manage in a way the stability issues while kill most of the air and sonics of the circuit .
In total and since the APD 15A is a well designed integrated amp the result is very good with a very nice kick, very clean , very good drive ability, obviously more power but the P3A will outperform it in every aspect except power ...
Kind regards
Sakis
As discussed before the APD 15A is almost identical with the P3A while features a world wide unique output stage formed by double an sziklai pair .It features 4 outputs and 4 drivers per ch .
To keep stable uses miller caps over power transistors in the astronomical size of 680 pf which manage in a way the stability issues while kill most of the air and sonics of the circuit .
In total and since the APD 15A is a well designed integrated amp the result is very good with a very nice kick, very clean , very good drive ability, obviously more power but the P3A will outperform it in every aspect except power ...
Kind regards
Sakis
Quoting my self from time to time made me think and come up with a hypothetical conclusion in steps please .....
---APD 15 senses the idle of the amplifier in the wrong place as we now know .
--- It senses over the outputs and not over the drivers , Plus that drivers all 4 of them share no heatsink at all ( may be not needed) but no heatsink at all means also no common behavior to the 4 drivers since these are behaving on their own when it comes to temperature
---Then APD 15 features huge miller caps of the size 680pf to keep stable in the output these may prevent oscillation but will do nothing to thermal stability .
Result :
Is there a chance that the designer tried to cure stability issues with the wrong way ? it is common practice that miller caps put a brake in oscillation issues but will only make the circuit worst if the source of the problem is thermal stability .
Why something is telling me that this amplifier may also work without the sonic killers miler caps on the outputs or have them replaced with something a lot lot smaller that will work as an oscillation stopper without killing the sonics ?
The only thing you have to do is redesign the thermals in the drivers and stability will come to visit you ....
Think please ...
Last edited:
Member
Joined 2009
Paid Member
In linear power supplies, it's common to use a ferrite bead or small choke in the collector lead of CFP regulator (output) transistors. Because it's only DC, that works well and a small bead has little affect on low impedance loaded audio frequencies anyway.
I know audiophiles will gasp and have fits when you mention ferrite but I'm only illustrating a principle that there are more and often subtle ways to limit instability. Regulators also use low Ft and Hfe semis which are less troublesome and the same would apply to audio by using lower Ft output stages, where it is possible. For example, instead of using LAPTs on everything, use something like Motorola MJL21193-6 and moderate Ft drivers instead of VHF transistors. Experiment and a good 'scope is the key.
I know a lot of people believe that extreme bandwidth gives wonderful audio extension but I personally think that is coincidental rather than a necessary design feature when we are considering frequencies only in the order of 10kHz. Sure, clobbering bandwidth with 680pF turns your output devices into slugs but a Zobel could have been used with better bandpass characteristics. Perhaps a bit more than a basic, kludge technique would have given better results in the APD 15.
I know audiophiles will gasp and have fits when you mention ferrite but I'm only illustrating a principle that there are more and often subtle ways to limit instability. Regulators also use low Ft and Hfe semis which are less troublesome and the same would apply to audio by using lower Ft output stages, where it is possible. For example, instead of using LAPTs on everything, use something like Motorola MJL21193-6 and moderate Ft drivers instead of VHF transistors. Experiment and a good 'scope is the key.
I know a lot of people believe that extreme bandwidth gives wonderful audio extension but I personally think that is coincidental rather than a necessary design feature when we are considering frequencies only in the order of 10kHz. Sure, clobbering bandwidth with 680pF turns your output devices into slugs but a Zobel could have been used with better bandpass characteristics. Perhaps a bit more than a basic, kludge technique would have given better results in the APD 15.
Gareth you cannot be sure since these results cannot be simulated ... you will only face the problem in real life conditions .
As we speak i am actually trying to locate the original designer to find out ( if possible ) the way he planed this and if miller caps was originally there or they had to add them afterwords to keep the APD 15 stable .
Given the chance if anyone knows the whereabouts of Alchemist HiFi Products | Designers | Glen Gayle i will be more than happy to know
Tim De Paravichini name is also involved there which i also try to email to see how far i can go from there
I will eventually construct to solve that but in between id like to listen first
Thanks for the input
I will have to partially argue with you IAN
I too think that amplifier for audio use will be just fine with 20-20 0db if possible but to be on the safe side lets make that 20-40 To my opinion though 100 khz or even 200 that has absolute NO use in the audio area will present an amplifier that is solid rock ...Now if that also helps the listeners just let it be .
The rest i see as commercial competition about numbers that will eventually help no one ...
It remains an issue that simple sziklai circuits with one pair of outputs have superior performance and produce a sonic signature that listeners like very much The real life opinions we get from costumers is that the content of harmonics that often have a tubish signature is what they go for .
Now if we manage to safely remove the sonic killer miller caps from the APD15 output style we may have a P3A with more power and plenty of kick ...
Thats the target !!!
Adding a pair in the P3A is not the solution to my understanding it will make more power but the amp will sound worst
Lets see ...
Thank you all for your time
This is very joyful for me !
Kind regards
Sakis
According to your link details, Gayle designed the basic range and De Paravachini "engineered" the signature series versions. I take it that is a polite way of saying he tweaked them and matched or used better componenents. At near 1% THD, there is obviously a lot of room for fixing a few problems without necessarily losing much "sound "quality"
The standard Forseti is pretty cheap and nasty with some slow old parts for an up-market product too - TIP35/6 output transistors, TIP31/32 drivers and even the VAS transistor is TIP32!
That should bring tears to few eyes and leave heads shaking.
1500x935.gif
The standard Forseti is pretty cheap and nasty with some slow old parts for an up-market product too - TIP35/6 output transistors, TIP31/32 drivers and even the VAS transistor is TIP32!
That should bring tears to few eyes and leave heads shaking.
1500x935.gif
Last edited:
Correct Ian but for some weird reason i have never seen an APD 15 with this type of outputs at least as seen in the schematic Always they had faster things on them
Let us not forget that this is not the Alchemist original web page ... Its just a site created by Joel Goodman for or about the Alchemist ...
I already emailed Tim was easier to reach, I wonder if he will bother to listen or he had a real part to play in the forseti issue
Let us not forget that this is not the Alchemist original web page ... Its just a site created by Joel Goodman for or about the Alchemist ...
I already emailed Tim was easier to reach, I wonder if he will bother to listen or he had a real part to play in the forseti issue
That's what I suspected, without having seen more than one. I think there will be a variety of parts grades across the different models because there are several versions of each basic model, as I read the review comments and try to follow the mixed sequence of release dates.
Member
Joined 2009
Paid Member
I don't have your listening experiences but from what I have read this seems to be a widely held observation. It lead me to use such an approach with my "TGM8" amplifier where only a single CFP pair in the output is used - but yet is safe to use with speakers that have big impedance dips because of supporting FETs that stay out of the way until needed; they are biassed in Class C. It works well.
@ IAN ...
Your comment about power supply, ferrite rings ,and so on brought memories of the past and times that GSM didnt exist
Then at the time VHF -UHF professional transmitters was the solution of high power 25W allowed in Greece and often Superphone 505 or bigger was also a solution
Reminds me of the power supplies we used at that times, often also for commercial applications 24/12 volt converters in a bus or big truck where you had two options :
LM723 driving NPN transistors ( some times even 741 ) or LM 317 or L200 with external PNP transistor option A is ""EFP "" while option B is "CFP" in terms of reliability especially when driving RF amps the "CFP" was far more stable and capable of plenty of current !!
Kind regards
Sakis
Your comment about power supply, ferrite rings ,and so on brought memories of the past and times that GSM didnt exist
Then at the time VHF -UHF professional transmitters was the solution of high power 25W allowed in Greece and often Superphone 505 or bigger was also a solution
Reminds me of the power supplies we used at that times, often also for commercial applications 24/12 volt converters in a bus or big truck where you had two options :
LM723 driving NPN transistors ( some times even 741 ) or LM 317 or L200 with external PNP transistor option A is ""EFP "" while option B is "CFP" in terms of reliability especially when driving RF amps the "CFP" was far more stable and capable of plenty of current !!
Kind regards
Sakis
Member
Joined 2009
Paid Member
I don't think that's the best approach, having drivers and output devices on the same heatsink because then you have the output devices unnecessarily heating up the drivers. You want the bias transistor in thermal contact with one (or both) of the drivers, but not in thermal contact with the output devices.
I think we have been over this point about the Vbe multiplier and driver heatsinks many times, also with yourself, in this thread: http://www.diyaudio.com/forums/solid-state/201446-rod-elliot-p3a-layout-critics-30.html
If you want to make up a clamp so that you can use a T092 transistor rather than just bolting a TO126, that's fine - the NPN transistor type is not important, the BD139 is used because its easier to mount and conducts heat better.
If you want to just push a small transistor close to the driver heatsink as Rod Elliott showed in his pic, that's ok too - note that this will be slower to stabilise and respond to changing power levels. If you read the thread and look at the pics in #152 here, you'll see that Sakis did something in between:http://www.diyaudio.com/forums/solid-state/164756-p3a-comparsion-table-long-16.html
Last edited:
Hello All,
I've just finished reading through this thread and found it a fascinating read. I'm not brilliant at electronics but I have built stuff before and got it working so I'm going to jump in with both feet and order 2 PCB's from Rod Elliott s web site.
Hello Mr Sakis, Id just like to ask, - In all your years experience have you ever come across the Nad 3020A amplifier? I am using this at the moment as a power amplifier and was wondering how the P3A will compare to it.
Kind Regards to all
Roy
I've just finished reading through this thread and found it a fascinating read. I'm not brilliant at electronics but I have built stuff before and got it working so I'm going to jump in with both feet and order 2 PCB's from Rod Elliott s web site.
Hello Mr Sakis, Id just like to ask, - In all your years experience have you ever come across the Nad 3020A amplifier? I am using this at the moment as a power amplifier and was wondering how the P3A will compare to it.
Kind Regards to all
Roy
Well i have been repairing plenty of those so i know exactly how these work .
We need to be fare here let us not forget that the P3A will be a linear amplifier while the 3020 A is an integrated amplifier ... so the test is not fare for the NAD
The truth is that if you combine the P3A with any decent preamp behind it, to my understanding you are going to throw the NAD out of the window with in the first 5 minutes .
The P3A made as supposed to be or as said in the ESP pages will have about 5 times the speed and bandwidth of NAD you are going to listen what low end means and what crisp means in the high .Expect "air " in your system that you never heard before .Totally dark and things and details are going to jump in to your speakers out of nowhere !!! Speed and bandwidth make it also super analytic .
As said let us not fool with that the NAD is a consumer integrated amplifier and it is supposed to be guarded in many ways to avoid troubles with costumers and cost increase .
It is expected to be less clean than the P3A or any other similar amp because its integrated and that will include more noise and many other issues originating from teh complexity of an integrated amplifier .
Also not to forget that level presets and gain of pre are different from today's standards so more gain more pain and noise and so on .
The NAD if working properly after all these years ( expect it to play but to my understanding not properly ) is a machine that includes a few goodies that others don't have and made by NAD as an exercise to prove to the rest of the world that a good amplifier CAN be made with cheap parts if designed properly .
Let us also not forget that the NAD has been refined to meet English style of music. detail that made NAD belong to this school of amplifiers at the time ..
As a closing argument and if one forgets that NAD is integrated and see it as a linear amplifier only then yes the P3A is 5 class better than the NAD .
Other than the above comment you cannot compare the 3020A and the P3A due to the nature of the two machines being different .
I f you go to a complete step and design a proper P3A amplifier and connect behind any decent preamp to be able to have a more complete comparison the NAD will be just a memory within a few minutes
Kind regards
Sakis
We need to be fare here let us not forget that the P3A will be a linear amplifier while the 3020 A is an integrated amplifier ... so the test is not fare for the NAD
The truth is that if you combine the P3A with any decent preamp behind it, to my understanding you are going to throw the NAD out of the window with in the first 5 minutes .
The P3A made as supposed to be or as said in the ESP pages will have about 5 times the speed and bandwidth of NAD you are going to listen what low end means and what crisp means in the high .Expect "air " in your system that you never heard before .Totally dark and things and details are going to jump in to your speakers out of nowhere !!! Speed and bandwidth make it also super analytic .
As said let us not fool with that the NAD is a consumer integrated amplifier and it is supposed to be guarded in many ways to avoid troubles with costumers and cost increase .
It is expected to be less clean than the P3A or any other similar amp because its integrated and that will include more noise and many other issues originating from teh complexity of an integrated amplifier .
Also not to forget that level presets and gain of pre are different from today's standards so more gain more pain and noise and so on .
The NAD if working properly after all these years ( expect it to play but to my understanding not properly ) is a machine that includes a few goodies that others don't have and made by NAD as an exercise to prove to the rest of the world that a good amplifier CAN be made with cheap parts if designed properly .
Let us also not forget that the NAD has been refined to meet English style of music. detail that made NAD belong to this school of amplifiers at the time ..
As a closing argument and if one forgets that NAD is integrated and see it as a linear amplifier only then yes the P3A is 5 class better than the NAD .
Other than the above comment you cannot compare the 3020A and the P3A due to the nature of the two machines being different .
I f you go to a complete step and design a proper P3A amplifier and connect behind any decent preamp to be able to have a more complete comparison the NAD will be just a memory within a few minutes
Kind regards
Sakis