Hi Mike,
I finally got around to measuring THD this morning. I gotta say Symasym is an incredible design, I've built many kits, and never seen any measure this well. Of course how it sounds is what really matters. I'll get back to you soon on that. I measured THD at 50,200,1Khz, and 5khz into 8 and 4 ohm loads at 25watt/50 watts with the fundamental at -10db. The voltage output across each load measured 14.5Vrms. My source input is a little noisey so I included it as reference. Here's the numbers.
@50Hz 8ohms/25watts, 4/50watts
Input 2HD=-100db,3HD=96,4HD=114,5HD=101,6HD=111
Ouput/8 2HD=-109db,3HD=96,4HD=112,5HD=101,6HD=110
Ouput/4 2HD=-107db,3HD=96,4HD=114,5HD=103,6HD=109
@200Hz
Input 2HD=-109db,3HD=101,4HD=106,5HD=104,6HD=114
Ouput/8 2HD=-110db,3HD=100,4HD=103,5HD=101,6HD=114
Ouput/4 2HD=-103db,3HD= 99,4HD=105,5HD=107,6HD=112
@1Khz
Input 2HD=-108db,3HD=101,4HD=111,5HD=104,6HD=117
Ouput/8 2HD= -99db,3HD=101,4HD=111,5HD=104,6HD=117
Ouput/4 2HD=-94db,3HD=103,4HD=103,5HD=105,6HD=106
@5Khz
Input 2HD= -97db, 3HD=118
Ouput/8 2HD= -90db, 3HD= noise floor
Ouput/4 2HD= -83db, 3HD= noise floor
I finally got around to measuring THD this morning. I gotta say Symasym is an incredible design, I've built many kits, and never seen any measure this well. Of course how it sounds is what really matters. I'll get back to you soon on that. I measured THD at 50,200,1Khz, and 5khz into 8 and 4 ohm loads at 25watt/50 watts with the fundamental at -10db. The voltage output across each load measured 14.5Vrms. My source input is a little noisey so I included it as reference. Here's the numbers.
@50Hz 8ohms/25watts, 4/50watts
Input 2HD=-100db,3HD=96,4HD=114,5HD=101,6HD=111
Ouput/8 2HD=-109db,3HD=96,4HD=112,5HD=101,6HD=110
Ouput/4 2HD=-107db,3HD=96,4HD=114,5HD=103,6HD=109
@200Hz
Input 2HD=-109db,3HD=101,4HD=106,5HD=104,6HD=114
Ouput/8 2HD=-110db,3HD=100,4HD=103,5HD=101,6HD=114
Ouput/4 2HD=-103db,3HD= 99,4HD=105,5HD=107,6HD=112
@1Khz
Input 2HD=-108db,3HD=101,4HD=111,5HD=104,6HD=117
Ouput/8 2HD= -99db,3HD=101,4HD=111,5HD=104,6HD=117
Ouput/4 2HD=-94db,3HD=103,4HD=103,5HD=105,6HD=106
@5Khz
Input 2HD= -97db, 3HD=118
Ouput/8 2HD= -90db, 3HD= noise floor
Ouput/4 2HD= -83db, 3HD= noise floor
Hi Mike,
I found the cancelling to be interesting too. It made me check my setup a couple times to make sure everything was Ok. The cancelling on the 2nd harmonic happens only at 50Hz, and 200Hz. The odd harmonics appear to be flat throughout. I'll try increasing power to 50watts into 8ohms, and 100 watts into 4 to see what happens. It might be that running Symasym with 50v rails improves THD.
I measured the souncard, a USB Audigy 24-bit Live External, by connecting the line in and signal out together and setting my TrueAudio signal generator to 0.315Vrms/-10dB and measuring THD. I then connected the signal generator to the amp input, and using a voltage divider to drop the amp output from 14.5Vrms to 0.315Vrms I measured the amps output THD. I used SpectaPlus Real-Time Spectrum Analyzer version 5.0 to measure THD.
Al
I found the cancelling to be interesting too. It made me check my setup a couple times to make sure everything was Ok. The cancelling on the 2nd harmonic happens only at 50Hz, and 200Hz. The odd harmonics appear to be flat throughout. I'll try increasing power to 50watts into 8ohms, and 100 watts into 4 to see what happens. It might be that running Symasym with 50v rails improves THD.
I measured the souncard, a USB Audigy 24-bit Live External, by connecting the line in and signal out together and setting my TrueAudio signal generator to 0.315Vrms/-10dB and measuring THD. I then connected the signal generator to the amp input, and using a voltage divider to drop the amp output from 14.5Vrms to 0.315Vrms I measured the amps output THD. I used SpectaPlus Real-Time Spectrum Analyzer version 5.0 to measure THD.
Al
50 volt rails ? Not if you are using mpsa18, it has max rating of 45v...
IIRC, you use 2n5551, no problem then...
As Pavel might confirm, measuring is an "art". I spend some time configuring and testing, and finally got my soundcard distortionfree.
The signal generator included in spectralab is ... (bad c-word) ... useless.
Mike
IIRC, you use 2n5551, no problem then...
As Pavel might confirm, measuring is an "art". I spend some time configuring and testing, and finally got my soundcard distortionfree.
The signal generator included in spectralab is ... (bad c-word) ... useless.
Mike
Attachments
I cranked up the power, still looks impressive. Just measured at 1Khz this time. Here's the measurements.
At 8ohms / 50watts, and 4ohms / 100 watts, 20Vrms
Input: 2HD=102, 3HD=107,4HD=107,5HD=96,6HD=112
8/out: 2HD= 95, 3HD=107,4HD=106,5HD=96,6HD=107
4/out: 2HD= 90, 3HD=100,4HD=100,5HD=95,6HD=102
At 4ohms / 132watts, 23Vrms
Input: 2HD=99,3HD=103,4HD=107,5HD=108,6HD=114
4/out: 2HD=89,3HD= 93,4HD= 94,5HD= 95,6HD=105
At 8ohms / 100watts, 28.2Vrms
Input: 2HD=94,3HD=103,4HD=105,5HD=101,6HD=114
8/out: 2HD=90,3HD= 97,4HD=100,5HD= 98,6HD=107
I also noticed Symasym is dead quiet. With the inputs grounded the noise level is flat at -120dB. Not even 60Hz or harmonics of it anywhere. I hooked it to my speakers and with my ears touching the dust caps, I couldn't hear a hum or a hiss, amazing!!
At 8ohms / 50watts, and 4ohms / 100 watts, 20Vrms
Input: 2HD=102, 3HD=107,4HD=107,5HD=96,6HD=112
8/out: 2HD= 95, 3HD=107,4HD=106,5HD=96,6HD=107
4/out: 2HD= 90, 3HD=100,4HD=100,5HD=95,6HD=102
At 4ohms / 132watts, 23Vrms
Input: 2HD=99,3HD=103,4HD=107,5HD=108,6HD=114
4/out: 2HD=89,3HD= 93,4HD= 94,5HD= 95,6HD=105
At 8ohms / 100watts, 28.2Vrms
Input: 2HD=94,3HD=103,4HD=105,5HD=101,6HD=114
8/out: 2HD=90,3HD= 97,4HD=100,5HD= 98,6HD=107
I also noticed Symasym is dead quiet. With the inputs grounded the noise level is flat at -120dB. Not even 60Hz or harmonics of it anywhere. I hooked it to my speakers and with my ears touching the dust caps, I couldn't hear a hum or a hiss, amazing!!
AAK, nice measurings ! Were these tests already done with 50v supply ? (obviously with the 40v peak for 100wrms into 8ohm)
Its good visible that at 132w into 4ohms symasym runs out of power, giving a rise in 3rd harmonic. The 50w into 8 ohms seem to be done effortless, showing no 3rd harmonic yet.
Yes, the silence of symasym is one of its "outstanding" features...
Amazing regarding the amount of openloop gain, or because of it ?
Have you already listened to music through symasym ?
About my soundcard setup...
The first important thing is to reduce levels, my soundcard has an option to reduce outputlevels from "consumer" to "-10dBV".
The next thing is signalsource, more complex than one would believe. If this is not carefully programmed you get jitter/crossoverdistortion because of rounding errors. "Goldwave" does a good job here, using its expression evaluator you can generate any testsignal needed. I used ~-1db level here (90%). I generated and played this signal with 24bit resolution.
Samplerate, i tried all samplerates, sadly my soundcard degrades above 48khz. Playback and recording must be done on identical samplerates, or you will get resampling. I also have the ability to lock the soundcard to a specific internal samplerate, in this case of course again 48khz to avoid any resampling. (resampling = bad)
All volumesliders need to be on default, not touching the signal.
Surprisingly, most distortions seen on soundcards seem to be generated by lousy digital signal processing, not analog circuitry...
My soundcard is a PCI - M-Audio Delta 24b96.
Mike
Its good visible that at 132w into 4ohms symasym runs out of power, giving a rise in 3rd harmonic. The 50w into 8 ohms seem to be done effortless, showing no 3rd harmonic yet.
Yes, the silence of symasym is one of its "outstanding" features...
Amazing regarding the amount of openloop gain, or because of it ?
Have you already listened to music through symasym ?
About my soundcard setup...
The first important thing is to reduce levels, my soundcard has an option to reduce outputlevels from "consumer" to "-10dBV".
The next thing is signalsource, more complex than one would believe. If this is not carefully programmed you get jitter/crossoverdistortion because of rounding errors. "Goldwave" does a good job here, using its expression evaluator you can generate any testsignal needed. I used ~-1db level here (90%). I generated and played this signal with 24bit resolution.
Samplerate, i tried all samplerates, sadly my soundcard degrades above 48khz. Playback and recording must be done on identical samplerates, or you will get resampling. I also have the ability to lock the soundcard to a specific internal samplerate, in this case of course again 48khz to avoid any resampling. (resampling = bad)
All volumesliders need to be on default, not touching the signal.
Surprisingly, most distortions seen on soundcards seem to be generated by lousy digital signal processing, not analog circuitry...
My soundcard is a PCI - M-Audio Delta 24b96.
Mike
Hi Mike,
Thanks for the info on the soundcard setup.
I'm actually playing symasym right now. First impressions, incredible! They sound beautiful. I'll break them in over the weekend, and get back to you after I do some serious listening. Of course I think It's safe to say that the amp I used to do all those measurements probably needs a well deserved break.
Thanks for all your help!
Al
Thanks for the info on the soundcard setup.
I'm actually playing symasym right now. First impressions, incredible! They sound beautiful. I'll break them in over the weekend, and get back to you after I do some serious listening. Of course I think It's safe to say that the amp I used to do all those measurements probably needs a well deserved break.
Thanks for all your help!Al
Hi AAK,
Would you be so kind as to summerise your amplifier with a parts list, schematic and PCB layout?
I think this design qualifies for this thread. Sounds like you did a very good job on it as well. My limited testing agrees with yours. A really nice little amplifier. I am convinced that 50W is enough for most applications.
-Chris
Would you be so kind as to summerise your amplifier with a parts list, schematic and PCB layout?
I think this design qualifies for this thread. Sounds like you did a very good job on it as well. My limited testing agrees with yours. A really nice little amplifier. I am convinced that 50W is enough for most applications.
-Chris
Hi Chris,
I'm working on it. I'll provide a schematic with the actual part numbers, and BOM. Also gerbers for top, bottom, and silk for those interested in building there own. The PCB program that I used is called Ivex, they went out business about three years ago. The program is very easy use, much easier than eagle, but not as powerful. I'd be glad to email the .exe plus provide any help necessary for those who would like to modify the board.
But, before I provide all this info on this forum I'd feel better emailing it to Mike so he can go over it to make sure everything is Ok for others to build.
As you can see from the pic, the board is very compact (5"x3") requiring the output and vas transistor to be mounted to the heat sink under the board. It can be a little tricky. I'll also provide some simple assembly instructions.
Best regards,
Al
I'm working on it. I'll provide a schematic with the actual part numbers, and BOM. Also gerbers for top, bottom, and silk for those interested in building there own. The PCB program that I used is called Ivex, they went out business about three years ago. The program is very easy use, much easier than eagle, but not as powerful. I'd be glad to email the .exe plus provide any help necessary for those who would like to modify the board.
But, before I provide all this info on this forum I'd feel better emailing it to Mike so he can go over it to make sure everything is Ok for others to build.
As you can see from the pic, the board is very compact (5"x3") requiring the output and vas transistor to be mounted to the heat sink under the board. It can be a little tricky. I'll also provide some simple assembly instructions.
Best regards,
Al
Why, is it because AAK used a different layout or used 2N input parts? Does this not defeat the very purpose of separating the 'sequel' thread from the 'original Symasym' thread?
IMO such information would prove more useful in the original thread. Moreover, the last many posts have been focussing on the original Symasym and I wonder as to what happened to the idea of the cascoded sequel.
Hi AAK,
By all means, have Mike review your work first.
Hi Samuel,
AAK extended the basis Symasym in an easy to reach step that may satisfy some other builders. It is definitely the same basic amplifier and build upon the original design. Placing the information allows this step to be examined by the original designer (Mike) for comment and possible inclusion of it's elements in future extensions. Pavel may also have some comment as he worked with Mike on this design.
Also, as Pavel himself suggested earlier .........
-Chris
By all means, have Mike review your work first.
Hi Samuel,
AAK extended the basis Symasym in an easy to reach step that may satisfy some other builders. It is definitely the same basic amplifier and build upon the original design. Placing the information allows this step to be examined by the original designer (Mike) for comment and possible inclusion of it's elements in future extensions. Pavel may also have some comment as he worked with Mike on this design.
The name of this thread is "Symasym - the sequel", this seems to be an extension of the original design.
Also, as Pavel himself suggested earlier .........
It would appear that AAK has done exactly that. Although, it appears that AAK may have skipped a step and simply constructed the suggestion.
-Chris
Thanks Chris, its fine with me as long as other forum members and visitors benefit. As I recall the very first postings of destroyerX in the Symasym thread showed that he had used MikeB's initial layout whereas John Mateus had used his own different layout (like AAK now) and several others' variation ideas might be buried in the original thread. Hence, I thought it would be more logical for this flavour of SymAsym to be included there. A pointer to a newbie or a search by anyone would then lead to the original thread when one seeks various iterations and layouts that have already been tried out by others; it does get tedious when the thread is rather long but atleast it remains more logical.
Anyway let be.
Anyway let be.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.