Welcome! This thread is for all those who bought the PCB for Mike Bittner's Symasym 5.3, this PCB model designed and tested by "AAK", and organized / produced by "Ryssen". AAK's design implements several incremental changes, most notably very linear Toshiba output transistors (and drivers), a change also rcommended by Pavel Macura (PMA), and 50V rails instead of 36V. All this is done in a compact PCB design of 5x3", intended to mount on top of a 4x5" heatsink.
It is for all discussion related to the actual building process: BOM, components, modifications, testing, functionality, and eventually, one might hope, sound quality
Related threads and posts:
PCB group buy , with all PCB related complementary info in the first post of the thread. This round of the GB is closed, but another round may be forthcoming if interest keeps coming in strong.
AAK's design
Mike Bittner's original design
Pavel Macura's review and technical recommendations
It is for all discussion related to the actual building process: BOM, components, modifications, testing, functionality, and eventually, one might hope, sound quality
Related threads and posts:
PCB group buy , with all PCB related complementary info in the first post of the thread. This round of the GB is closed, but another round may be forthcoming if interest keeps coming in strong.
AAK's design
Mike Bittner's original design
Pavel Macura's review and technical recommendations
I asked Ender Hoffman (roender) about changing the 2 input transistors to j-fets,and that shouldn´t be a problem.
Schema j-fet
Might be a problem if you use 2x50v DC though,as it says max 40v in the data sheet:Datasheet 2SK170 I don´t know how much voltage is over the j-fet .
Anyone know of a substitute?
Schema j-fet
Might be a problem if you use 2x50v DC though,as it says max 40v in the data sheet:Datasheet 2SK170 I don´t know how much voltage is over the j-fet .
Anyone know of a substitute?
ryssen:
i don't know any conventional jfets that could be used "bareback" with 50v rails. however, there are some power jfets that several folks have been happily experimenting with, stimulated by a nelson pass project. sorry, i don't remember the drain-source voltage limitation of this device. also don't know how suitable it would be as an input device (have to check device capacitances, transconductance, etc.)
but even dropping the rails lower (for example) to 40v, for the 2sk170, reputable designers on this board strongly recommend cascoding to keep the drain-source voltage in the 10-15 volt range. the issues with not doing so would show in a search, i think.
mlloyd1
i don't know any conventional jfets that could be used "bareback" with 50v rails. however, there are some power jfets that several folks have been happily experimenting with, stimulated by a nelson pass project. sorry, i don't remember the drain-source voltage limitation of this device. also don't know how suitable it would be as an input device (have to check device capacitances, transconductance, etc.)
but even dropping the rails lower (for example) to 40v, for the 2sk170, reputable designers on this board strongly recommend cascoding to keep the drain-source voltage in the 10-15 volt range. the issues with not doing so would show in a search, i think.
mlloyd1
They are http://www.lovoltech.com/pdf/LD1014D.pdf I have a few of them. They only take 24v and input C is 1147pf compared to 2SK170 30pf.You mean to cascode with the LD1014?
The thing is,if I start to experiment with this they will probably blow up before I have switched the power on..
Never done such a thing.
And I´m thinking of using 2x50v DC,maybee a little lower if I find such a transformer.
The thing is,if I start to experiment with this they will probably blow up before I have switched the power on..
Never done such a thing.
And I´m thinking of using 2x50v DC,maybee a little lower if I find such a transformer.
Here is sommething that looks like what I´m searching for:http://www.diyaudio.com/forums/attachment.php?s=&postid=1145866&stamp=1172607377
It is from Symasym Roender style
Maye it works,will se how much it differs from the orig schema.
Edit:Quite much...
It is from Symasym Roender style
Maye it works,will se how much it differs from the orig schema.
Edit:Quite much...
Some questions about the BOM
Due to the higher rails, and some design choices, some resistor values differ from the original Symasym. In some cases the reason was clearly given by AAK, in some I am not so sure whether the changes are a must, or if there is some flexibility. Here is a summary comparing component values other than the transistors, between Symasym 5.3 original ("5.3") and "AAK" BOM:
- base stoppers: R1, R3 AAK: 10R (5.3:1.2R); R8, R9 AAK: 100R (5.3: 22R). In the sequel thread, Mike Bittner said he used 22R for "speed issues", and AAK justified these values because they are like in the AKSA and other reputed amps. Is there a definitive answer to this question?
- thermal transistor T1 (BD139): R22 AAK 100R, 5.3: 470R; R23 AAK: 1k, 5.3: 2k; R24 AAK: 316R, 5.3: 499R. I would like to know whether this change has anything to do with the higher rails, or whether it was simply more convenient. I will be using pretty much AAK's BOM but with 36V rails only, so I'd like to know which combo is better for this case, the 5.3 values or the AAK values.
- R26 AAK: 150R necessary for thermal stability, 5.3: 33R
- output zobel: C18: AAK not in BOM but exists on PCB, 5.3: 47nF; R4: AAK 10R, 5.3: 4R7. I suppose the point is to maintain the same corner frequency, which would put the AAK C18 at 22nF.
- output inductor: AAK none, 5.3: 10R // a dozen windings of 0.6mm dia magnet wire around it. Optional?
- ground link R2 AAK: cannot find R2 on BOM and PCB, 5.3: 10R.
Is IGND and AGND tied or separate on the AAK PCB?
- emitter resistors for the input diff pair if reduced gain desired by increasing R30 to 1k: AAK: not on PCB, 5.3: R16, R19.
Additional comments that may be useful:
Pavel Macura, using the same output transistors as AAK, recommends C14 at 3.3pF and setting Iq at 150 mA for minimum distortion. He also used 220pF for C3, C4 (AAK and 5.3: 330 pF) and 0R15 for R27, R28 (AAK, 5.3: 0R2).
Due to the higher rails, and some design choices, some resistor values differ from the original Symasym. In some cases the reason was clearly given by AAK, in some I am not so sure whether the changes are a must, or if there is some flexibility. Here is a summary comparing component values other than the transistors, between Symasym 5.3 original ("5.3") and "AAK" BOM:
- base stoppers: R1, R3 AAK: 10R (5.3:1.2R); R8, R9 AAK: 100R (5.3: 22R). In the sequel thread, Mike Bittner said he used 22R for "speed issues", and AAK justified these values because they are like in the AKSA and other reputed amps. Is there a definitive answer to this question?
- thermal transistor T1 (BD139): R22 AAK 100R, 5.3: 470R; R23 AAK: 1k, 5.3: 2k; R24 AAK: 316R, 5.3: 499R. I would like to know whether this change has anything to do with the higher rails, or whether it was simply more convenient. I will be using pretty much AAK's BOM but with 36V rails only, so I'd like to know which combo is better for this case, the 5.3 values or the AAK values.
- R26 AAK: 150R necessary for thermal stability, 5.3: 33R
- output zobel: C18: AAK not in BOM but exists on PCB, 5.3: 47nF; R4: AAK 10R, 5.3: 4R7. I suppose the point is to maintain the same corner frequency, which would put the AAK C18 at 22nF.
- output inductor: AAK none, 5.3: 10R // a dozen windings of 0.6mm dia magnet wire around it. Optional?
- ground link R2 AAK: cannot find R2 on BOM and PCB, 5.3: 10R.
Is IGND and AGND tied or separate on the AAK PCB?
- emitter resistors for the input diff pair if reduced gain desired by increasing R30 to 1k: AAK: not on PCB, 5.3: R16, R19.
Additional comments that may be useful:
Pavel Macura, using the same output transistors as AAK, recommends C14 at 3.3pF and setting Iq at 150 mA for minimum distortion. He also used 220pF for C3, C4 (AAK and 5.3: 330 pF) and 0R15 for R27, R28 (AAK, 5.3: 0R2).
Hi Ryssen,
There are several good threads on cascodes with supporting literature data showing that a bipolar-bipolar cascode has the lowest noise and best performance, with the JFET differential and bipolar cascode the next best performer. The MAT02-MAT03 dual low noise BJT are often used in hi-end equipment. Many favor the JFET-Bipolar - it just sound better and solves some box-to-box connect problems. The bipolar-bipolar would be better if an active crossover is right in front of the amp....maybe on the same PCB with PASS JFET opamps.
For JFET-JFET cascodes, the self biasing using the Vt difference between 170 differential and 246 cascode transistors as Borbley champions is very difficult to construct because of JFET VT matching....be prepared to purchase and hand match dozens of JFETs.
Anyway...that's what I've learned.
There are several good threads on cascodes with supporting literature data showing that a bipolar-bipolar cascode has the lowest noise and best performance, with the JFET differential and bipolar cascode the next best performer. The MAT02-MAT03 dual low noise BJT are often used in hi-end equipment. Many favor the JFET-Bipolar - it just sound better and solves some box-to-box connect problems. The bipolar-bipolar would be better if an active crossover is right in front of the amp....maybe on the same PCB with PASS JFET opamps.
For JFET-JFET cascodes, the self biasing using the Vt difference between 170 differential and 246 cascode transistors as Borbley champions is very difficult to construct because of JFET VT matching....be prepared to purchase and hand match dozens of JFETs.
Anyway...that's what I've learned.
I will Search later and try to learn on cascading,if no one have an finished solution.Will look in the original Explendid amplifier thread,but its a Looooong thread.
The output transistors 2sc5200-OQ/2sa1943-OQ Are there any bennfits sonical to use them compared to MJL21193/94 or MJL1302/MJL3281?I have lots of them.
The output transistors 2sc5200-OQ/2sa1943-OQ Are there any bennfits sonical to use them compared to MJL21193/94 or MJL1302/MJL3281?I have lots of them.
Hi MBK,
Thanks for pointing out that C18 is not listed in the BOM. It should be 0.1uf, same as C5,C6. The components that are critical on my board are R26 (150 ohms) to maintain thermal stability, R22,R23,R24 to correctly set the bias, and C14 (10pf or less for selected U3,U4) to keep the amplifier from oscillating.
It's getting late. I'll get back to answering your other questions tomorrow.
Al
Thanks for pointing out that C18 is not listed in the BOM. It should be 0.1uf, same as C5,C6. The components that are critical on my board are R26 (150 ohms) to maintain thermal stability, R22,R23,R24 to correctly set the bias, and C14 (10pf or less for selected U3,U4) to keep the amplifier from oscillating.
It's getting late. I'll get back to answering your other questions tomorrow.
Al
Hi Al,
thanks, I appreciate. This is only my second discrete amp, and that is in building, not designing. My usual DIY efforts are all in speakers and op amp based signal circuits / chip amps... So while I vaguely know what the parts are for, functionally, I am not always sure of what change will affect what parameter.
thanks, I appreciate. This is only my second discrete amp, and that is in building, not designing. My usual DIY efforts are all in speakers and op amp based signal circuits / chip amps... So while I vaguely know what the parts are for, functionally, I am not always sure of what change will affect what parameter.
Hello !
I'm VERY interested in a higher power version of Symasym as I have 8Ohms speakers.
I built Symasym, and I'm amazed of how good it sounds, despite I still searching for a higher power version, I can't leave the good sound of Symasym.
-> So, HOW does it sound? Does it sound as good as symasym 5.3 ?
I built mine with Styroflex and Mica capacitors, Wima 3.3uF for the input, and On Semi MJL3201 and MJL1302 transistos.
-> How much power with a 8Ohms load?
-> What is the recommended supply rails for a VERY safe operation with the highest power possible at 8 ohms?
-> How much costs the PCB ?
Thank you very very much, I'm anxious for the feedback !
edit: I'd like to keep the output devices, can I use MJL1302 / 3201 ? I can use a pair if is necessary.
Best Regards
I'm VERY interested in a higher power version of Symasym as I have 8Ohms speakers.
I built Symasym, and I'm amazed of how good it sounds, despite I still searching for a higher power version, I can't leave the good sound of Symasym.
-> So, HOW does it sound? Does it sound as good as symasym 5.3 ?
I built mine with Styroflex and Mica capacitors, Wima 3.3uF for the input, and On Semi MJL3201 and MJL1302 transistos.
-> How much power with a 8Ohms load?
-> What is the recommended supply rails for a VERY safe operation with the highest power possible at 8 ohms?
-> How much costs the PCB ?
Thank you very very much, I'm anxious for the feedback !
edit: I'd like to keep the output devices, can I use MJL1302 / 3201 ? I can use a pair if is necessary.
Best Regards
The beginning...
Been playing with some stuff I had laying around..settled on a 2x24v trafo,good for heat if raising bias..
At least a start as I have this habbit of doing many projekts (to many)at the same time.
An externally hosted image should be here but it was not working when we last tested it.
Been playing with some stuff I had laying around..settled on a 2x24v trafo,good for heat if raising bias..
At least a start as I have this habbit of doing many projekts (to many)at the same time.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.