Hi,
thanks Andrew,
the 2n5401 i found are ok .
not to easy to find all parts ! and can't find this 4.7 POLY/Film in the appropriate dimensions and don't want to play without it. (all that i find is bigger).
jf
thanks Andrew,
the 2n5401 i found are ok .
not to easy to find all parts ! and can't find this 4.7 POLY/Film in the appropriate dimensions and don't want to play without it. (all that i find is bigger).
jf
Hi jef9200,
I used both 3u3 and 1u0 with good results. Most speakers and rooms will not support the lower registers of bass. I don't think you will actually miss anything.
Remember that you can tack another cap on the foil side.
-Chris
I used both 3u3 and 1u0 with good results. Most speakers and rooms will not support the lower registers of bass. I don't think you will actually miss anything.
Remember that you can tack another cap on the foil side.
-Chris
I disagree.anatech said:
I used small speakers for years (over a decade) and did not realise what low bass they were capable of until I added a new pre-amp that had a wideband ability.
That then let me experiment further with power amp bandwidth and I was staggered at the improvement in sound that comes from extending the low bass response to a full decade below the normally accepted 20Hz limit.
I can hear the difference between a 90mS high pass filter and a 20mS high pass filter on the front end of the power amp. This almost defies the explanation. The speakers are Acoustic Energy AE1 with a small 5inch bass/mid that claim a response to about 50 or 60Hz.
Andrew,
I agree to that a low cut off frequency is preferable! Even if I use much bigger bass reflex speakers (-3 dB at about 26 Hz) there is absolute no need to use a high-pass filter to avoid cone flutter. If a record player is used then use a sub-sonic filter in the RIAA amplifier. In PA-systems subsonic filters are used to save amp power, but for home use they are not needed IMHO. Using DC-coupled amps make a big difference when listening to good recordings. The "short" bass reproduction I have achieved now partly due to extensive room treatment is awesome.🙂 And this bass reproduction I thought was nearly impossible to obtain affects the perception of the mids and high frequencies as well making it hard to stop listening....😉
I agree to that a low cut off frequency is preferable! Even if I use much bigger bass reflex speakers (-3 dB at about 26 Hz) there is absolute no need to use a high-pass filter to avoid cone flutter. If a record player is used then use a sub-sonic filter in the RIAA amplifier. In PA-systems subsonic filters are used to save amp power, but for home use they are not needed IMHO. Using DC-coupled amps make a big difference when listening to good recordings. The "short" bass reproduction I have achieved now partly due to extensive room treatment is awesome.🙂 And this bass reproduction I thought was nearly impossible to obtain affects the perception of the mids and high frequencies as well making it hard to stop listening....😉
May I add that this is my experiences and that drivers with a very low fs (resonant frequency) in a big closed box ( a total Q for the system around 0.5) or even drivers with a low fs in an a bass reflex cabinet may behave differently compared to my system. My point is however that the need for subsonic filters are often taken for granted, and this may be detrimental to the final sound quality.
Hi MBK,
May I know why does the low hfe input transistors causes turn-on transient?
Hi MikeB,
I'll use a smaller voltage transformer then. Thanks for your advice. I have completed component placement of two boards. Will test it today. 🙂
Thanks.
May I know why does the low hfe input transistors causes turn-on transient?
Hi MikeB,
I'll use a smaller voltage transformer then. Thanks for your advice. I have completed component placement of two boards. Will test it today. 🙂
Thanks.
Cabling output transistors apart
Hi !
I already posted this request, possibly in the wrong thread.
What about wiring apart the 2SA1943/2SC5200 output power transistors ?
Purpose : to use existing chassis and heatsinks where direct mounting is impossible.
Caveats, potential problems (hum, oscillation) ?
Has somebody already tested that ?
I plan to deport the devices about 10cm (4") away with 14AWG wires.
TIA for your feedback, positive, I hope
Hi !
I already posted this request, possibly in the wrong thread.
What about wiring apart the 2SA1943/2SC5200 output power transistors ?
Purpose : to use existing chassis and heatsinks where direct mounting is impossible.
Caveats, potential problems (hum, oscillation) ?
Has somebody already tested that ?
I plan to deport the devices about 10cm (4") away with 14AWG wires.
TIA for your feedback, positive, I hope
While I am not completely qualified to reply to this, I have been told that it has to do with less control of the output stages by the front end via feedback, during power up - it takes longer to control the DC level.
Hi Ipanema,
Firstly, they have higher base current. This is dropped across the resistors to the feedback and also ground at the input. This will give you a higher static DC offset.
Secondly, because there is higher base current, it takes longer for associated caps to charge up. It takes longer for the amplifier to settle into it's steady state conditions.
In any event, they must be matched for beta. This is very important as this is where the distortion (non-linearities) are canceled out. A mismatched pair will both increase DC offset and distortion. The mismatch can be used to reduce the DC offset at the expense of much increased distortion. Not recommended.
-Chris
Low hFE transistors in the diff pair cause two things at one time.
Firstly, they have higher base current. This is dropped across the resistors to the feedback and also ground at the input. This will give you a higher static DC offset.
Secondly, because there is higher base current, it takes longer for associated caps to charge up. It takes longer for the amplifier to settle into it's steady state conditions.
In any event, they must be matched for beta. This is very important as this is where the distortion (non-linearities) are canceled out. A mismatched pair will both increase DC offset and distortion. The mismatch can be used to reduce the DC offset at the expense of much increased distortion. Not recommended.
-Chris
Hi MBK and Anatech,
Thanks for the explanation. 🙂
I've fired up one of my board yesterday. Unfortunately, R10 heats up and smoking. Don't know why it is so. I'm using MPSA18 for input stage pair and MPSA42/92 for other small signal transistor. Since I don't have any 100pF at hand, I just use 33pF at C2 and C7 for testing. Will this cause problem? Will test again today at lower power supply +/-16V instead of +/-34V. Any other point that I need to check for this problem?
Thanks.
Thanks for the explanation. 🙂
Which caps do you mean?
I've fired up one of my board yesterday. Unfortunately, R10 heats up and smoking. Don't know why it is so. I'm using MPSA18 for input stage pair and MPSA42/92 for other small signal transistor. Since I don't have any 100pF at hand, I just use 33pF at C2 and C7 for testing. Will this cause problem? Will test again today at lower power supply +/-16V instead of +/-34V. Any other point that I need to check for this problem?
Thanks.
Hi MJL21193,
I just want to find the problem of R10 overheat. Do you have any suggestion?
Thanks.
I just want to find the problem of R10 overheat. Do you have any suggestion?
Thanks.
hi Ipanema,
Be careful with C7 as it is Cdom. This cap can affect the stability of an amplifier and too low a value can cause oscillations.
regards
Be careful with C7 as it is Cdom. This cap can affect the stability of an amplifier and too low a value can cause oscillations.
regards
Hi Ipanema,
Greg is bang on there. Actually, all your caps are important to be the correct value (all the pf values).
The caps that need to charge are the input coupling cap and the feedback capacitor. They have to charge up to whatever the bases are sitting at once the amp has settled down. If both bases were sitting at 300 mV, you would have no DC offset. It's the getting there that is the hard part. The output will swing around as the (-) base of the diff pair tries to equal the (+) base.
Of course this is in a perfect world. There are some other reasons why a balanced diff pair may lead to an offset (or more correctly, conditions that force a diff pair to be unbalanced to maintain a low DC offset).
Hi Andrew,
However, with large speakers you are able to get below the cut off frequency of the room. Then the bass rolls off rapidly. Your smaller speakers may be getting into frequency doubling where it appears to have deeper bass. There are other things that may be going on as well. No matter what, the laws of physics must be obeyed.
My tests with the SymAsym showed no discernible roll off with music. As always, YMMV depending on all kinds of things. Never mind that there is very little really going on down there unless you are into sound effects or organ music.
-Chris
Greg is bang on there. Actually, all your caps are important to be the correct value (all the pf values).
The caps that need to charge are the input coupling cap and the feedback capacitor. They have to charge up to whatever the bases are sitting at once the amp has settled down. If both bases were sitting at 300 mV, you would have no DC offset. It's the getting there that is the hard part. The output will swing around as the (-) base of the diff pair tries to equal the (+) base.
Of course this is in a perfect world. There are some other reasons why a balanced diff pair may lead to an offset (or more correctly, conditions that force a diff pair to be unbalanced to maintain a low DC offset).
Hi Andrew,
That's fair.
However, with large speakers you are able to get below the cut off frequency of the room. Then the bass rolls off rapidly. Your smaller speakers may be getting into frequency doubling where it appears to have deeper bass. There are other things that may be going on as well. No matter what, the laws of physics must be obeyed.
My tests with the SymAsym showed no discernible roll off with music. As always, YMMV depending on all kinds of things. Never mind that there is very little really going on down there unless you are into sound effects or organ music.
-Chris
I'm too lazy (or busy) to search for this but has anyone done any real world testing on this amp? If you fine gentlemen can point the way to the results, I'd be much obliged🙂
I agree with Chris. The bass perception issue is quite complex. The Q of the rolloff for instance does change bass perception, and here either a smaller OR a larger input cap may give you a stronger bass impression... (because it will mess with the speaker's Q). But at the discussed cutoffs, all well below 10 Hz, I don't see the coupling cap as a major factor in bass magnitude.
Other factors that are much more important include the room. Example: I used to run my speakers down to 20 Hz (EQ'd dipoles, so the cutoff and Q is at my will depending on EQ). Then I noticed that my room has a strong resonance at 50 Hz, and others at multiples: 25, 75 and 100 Hz. At 40 Hz there is a bottomless dip. Then I EQ'd the 50 Hz peak out. But after experimentation in the end I realized that I could run the system at its natural 69 Hz rolloff, the 50 Hz resonance would then "EQ" and lift up the response somewhat below, and below 40 Hz my room is hopeless (the 25 Hz peak won't change your music impressions much). So, I realized things often work out in a different way, and the input coupling cap in all likelihood has nothing to do with it.
I am using 2.2 uF input caps on my Symasyms now with no discernible ill effects.
Other factors that are much more important include the room. Example: I used to run my speakers down to 20 Hz (EQ'd dipoles, so the cutoff and Q is at my will depending on EQ). Then I noticed that my room has a strong resonance at 50 Hz, and others at multiples: 25, 75 and 100 Hz. At 40 Hz there is a bottomless dip. Then I EQ'd the 50 Hz peak out. But after experimentation in the end I realized that I could run the system at its natural 69 Hz rolloff, the 50 Hz resonance would then "EQ" and lift up the response somewhat below, and below 40 Hz my room is hopeless (the 25 Hz peak won't change your music impressions much). So, I realized things often work out in a different way, and the input coupling cap in all likelihood has nothing to do with it.
I am using 2.2 uF input caps on my Symasyms now with no discernible ill effects.
Hi John,
I built a couple 5.3's up. They outperformed my expectations and so are suffering "performance creep". 😀
These are really nice little amplifiers. The do sound great. I didn't bother running many tests on them as yet. Once I begin experimenting I shall be doing that. I just got a bunch of AAK boards to compare with once I build them up. Then I can build different versions.
One day they may be installed where they were originally intended. 🙂
Is there anything you'd like to know? I may have an answer. Understand my units live on a temporary heat sink and are run from an old Marantz 1180DC supply turned down with a variac.
-Chris
I built a couple 5.3's up. They outperformed my expectations and so are suffering "performance creep". 😀
These are really nice little amplifiers. The do sound great. I didn't bother running many tests on them as yet. Once I begin experimenting I shall be doing that. I just got a bunch of AAK boards to compare with once I build them up. Then I can build different versions.
One day they may be installed where they were originally intended. 🙂
Is there anything you'd like to know? I may have an answer. Understand my units live on a temporary heat sink and are run from an old Marantz 1180DC supply turned down with a variac.
-Chris
anatech said:
Hi Chris,
Over here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=108860&perpage=25&pagenumber=1
I'm playing around. Tonight, I re-built my SymAsym model and I'm doing some measurements (simulated). I had done some test before, but didn't do any square waves at lower frequencies. The results were strange.
I'd just like to see the square wave output of this amp at 1KHz on a real oscilloscope.
Hi John,
The square wave was pretty clean. I think there were square wave tests on Mike's site, or Pavel's site. I'm pretty sure they are there.
I will hook it up again at some point and I'll run that test. It is something I ran in passing to give me an idea of lurking nasties. They were not there.
Keep in mind that I used MJL0302A and MJL0281A for outputs. They allowed me a much lower bias current. Everything else was per MikeB's BOM.
What you should have done was to build this amp and simulate it. You would then see correlation between life and sim. Also keep in mind that PCB layout can make a large difference. Mike's original thread bears witness to this.
-Chris
The square wave was pretty clean. I think there were square wave tests on Mike's site, or Pavel's site. I'm pretty sure they are there.
I will hook it up again at some point and I'll run that test. It is something I ran in passing to give me an idea of lurking nasties. They were not there.
Keep in mind that I used MJL0302A and MJL0281A for outputs. They allowed me a much lower bias current. Everything else was per MikeB's BOM.
What you should have done was to build this amp and simulate it. You would then see correlation between life and sim. Also keep in mind that PCB layout can make a large difference. Mike's original thread bears witness to this.
-Chris