AndrewT said:but it's the bad advice wrapped up in gobble de gook that is the problem.
Well, if that's complicated, there's certainly a problem. See photo below that illustrates the components. This does hinder the efficiency of the diodes by several volts in this application. However, it doesn't hinder any "useful" efficiency, in my opinion. There's just less voltage peaks.
PHOTO: A snubbed rectifier.
Attachments
i come in peace... hehe 😀
Hi rhythm,
i will suggest to avoid the bridged amp for your 3way..
not about the sound quality..
but its better to check what network inside the 3-way.
bridged amp has 'active' at both output, no speaker return, so i think all speaker, especialy the mid and the tweeter which has lower watt rated,will need crossovered at their both input, or they potentialy destroyed by the high power. this issue will mess u up.
my smaller bridged amp did destroy my tweeter's 10' fullrange.
but if you choose the BPA, i will agree with sangram for bi-amping or tri-amping( without passive crossover ), but again,i think that will be too much.. the single lm3886 should just enough.. even for your current 3-way.
for my example,my 100W bridged amp just strong enough to drive 2 x 10' kenwood subwoofer in series without any problem.
I maybe wrong, i'll like to thanks if other friends will shows me the related threads to read?
Hi rhythm,
i will suggest to avoid the bridged amp for your 3way..
not about the sound quality..
but its better to check what network inside the 3-way.
bridged amp has 'active' at both output, no speaker return, so i think all speaker, especialy the mid and the tweeter which has lower watt rated,will need crossovered at their both input, or they potentialy destroyed by the high power. this issue will mess u up.
my smaller bridged amp did destroy my tweeter's 10' fullrange.
but if you choose the BPA, i will agree with sangram for bi-amping or tri-amping( without passive crossover ), but again,i think that will be too much.. the single lm3886 should just enough.. even for your current 3-way.
for my example,my 100W bridged amp just strong enough to drive 2 x 10' kenwood subwoofer in series without any problem.
I maybe wrong, i'll like to thanks if other friends will shows me the related threads to read?
danielwritesbac said:
🙄
No, it doesn't. Repeating a wrong statement again and again will not make it right.
pacificblue said:
🙄
No, it doesn't. Repeating a wrong statement again and again will not make it right.
You must have tested with a steady tone. However, test again with music that has a "bass beat" and you'll see a difference in the amount of voltage fluctuations. This concept is commonly called "Diode Drop." And there are several things you can do about it.
Daniel; would you explain to me why are you pushing this thing with small caps? Have you some good results with that? And I mean by listening not measuring.
This is normal PSU voltage sag (fluctuation). Adding rectifier diode snubbers does not alter PSU sag.danielwritesbac said:
Sangram;
I would say that selection of power supply components affects the sound of the gainclone. All components affects the sound of the gainclone. I had different results with standard bridges, MUR860, U860 and SB560 diodes.
The catch is not in ''golden ears''; it's just about hearing the difference. If not...then you are blessed. You can build the amp with ''what gets in your hands''.
I would say that selection of power supply components affects the sound of the gainclone. All components affects the sound of the gainclone. I had different results with standard bridges, MUR860, U860 and SB560 diodes.
The catch is not in ''golden ears''; it's just about hearing the difference. If not...then you are blessed. You can build the amp with ''what gets in your hands''.
Igla said:
OH yes, by listening indeed! For audio variances, the power supply has the greatest effect.
The LM3886 presents a very light load to the power supply when it is doing only midrange and treble; however, during bass, the voltage may drop. This makes for less bass--in proportion. The effect is seen mostly in higher power applications, such as using inefficient speakers to reproduce dance music, for example. I've also had results with cooler heatsink temperatures. Its really saved me a lot of money. And, I do like good bass and an amplifier without the peakish "blare." The effect is minor, but, to me, its incredibly inexpensive for how much good it does me. 😉
I did, recently, also have some luck with a parallel amplifier, because each chip spots only half the load (8 ohm speakers, but the chips see 16 ohm). A lighter load has less power supply requirements. This effect is also minor. EDIT: The parallel amplifier sounds different (more laid back) even when used on the exact same power supply board and transfo as the (previously employed) single chip.
I think that a lot of minor effects retains fidelity better than one or more large effects. Its especially nice when you don't have to eq your solid state equipment. 😉
sangram said:
That is one of the most delightfully written descriptions of audio building I have read.
😉
Igla said:
I do know they're different, but the post was not meant to be read literally, it was tongue-in-cheek. Changes are usually small, but Daniels' post would lead one to believe (if you could believe) that drastic changes are caused by power supplies. In reality, there are effects - but not
Now I've built chipamps with capacitance from 1000 to 100000uF/channel, with individual diodes, Schottky rectifiers and composite 6 amp bridges (in a real pinch), and the real truth is somewhere in the middle, but neither extreme would give you the kind of results in the post above, unless you're trying to run a 80dB/w (I guess) speaker off a chipamp, which it was not meant to do. Gross frequency distortions are not the results of not having snubbers on the rectifier.
I apologise if you assumed I meant that there would be no difference.
No need to apologise sangram.
Daniel; I did try snubbers on rectifiers and I didn't like the result. But I have efficient and easy to drive full range loudspeakers (if this matter). By the way; the best results I get with MUR860 and SB560 diodes.
Ordinary rectifiers (KBPC and similar) gave me very different results - one was shouting with no bass; the other was bassy with moulded mids.
Daniel; I did try snubbers on rectifiers and I didn't like the result. But I have efficient and easy to drive full range loudspeakers (if this matter). By the way; the best results I get with MUR860 and SB560 diodes.
Ordinary rectifiers (KBPC and similar) gave me very different results - one was shouting with no bass; the other was bassy with moulded mids.
Igla said:
Thank you!
OH yes, that latter is close. It just needs a bit of a tuneup. Go with the bassy, sink the mids option if your amp needs to push 85db or lesser speakers, and then. . . reduce the size of and increase the quality of the pair of caps at the amplifier board in order to "clarify" (not boost) the midrange. Range = 150uF to 600uF. This way, most of the capacitance is at the (seperate) power supply board. For the last step, choose one capacitor for V+ to V- at the amplifier board (CarlosFM2005 footnotes) in the range of 2uF to 4.7uF. When its working in harmony, the heatsink temp will drop considerably. We're after a "tube taper" to the audio, which can be described as euphonic and clear. An amplifier with this nature will retain fidelity even when used to its power output limits.
IMO, don't snub "soft switch," MUR860, MUR756, MR, etc. . .
PLAN B: 😉 Many of the customary designs, including the chipamp.com kit, can work appropriately (for hi-power application) when used in parallel. There's less need for extensive fine tuning (the audio) on a parallel amplifier because they're not going to strain so much when shoving inefficient speakers. Caveat: You'll need to spend a bit on the heatsink. But, that's certainly the easier way to go. 😉
I stopped playing with snubbers because I didn't get good results. I have tryed a lot of things with all of my gainclones (LM1875, LM3875, TDA2050, TDA2030 and OPA549) and the best results I get with:
PS: tranny + single or double bridge of MUR860 or SB560
Clone: 1000uF on PS pins, no input cap and no cap on negative input to ground.
PS: tranny + single or double bridge of MUR860 or SB560
Clone: 1000uF on PS pins, no input cap and no cap on negative input to ground.
Igla said:
Did you try larger ceramic caps? Other types have a high internal resistance, frequently resulting in an unintentional RC network.
Did you try larger ceramic caps? Other types have a high internal resistance, frequently resulting in an unintentional RC network.
I have tried ceramic caps (can't remember exact value), then MKT and MKS caps (Wima and Epcos) - small values in the range of 0.01uF to 0.33uF. The result was too bright top end; somehow shouting character I couldn't stand and listen. The same thing happened not just with gainclones but also with my Pass Zen amp. Have also tried snubbers on my TDA1543 DAC power supply and didn't like the sound. I had standard KBPC rectifiers on ZEN, U860 on DAC and MUR860 and SB560 on gainclones.
I have tried ceramic caps (can't remember exact value), then MKT and MKS caps (Wima and Epcos) - small values in the range of 0.01uF to 0.33uF. The result was too bright top end; somehow shouting character I couldn't stand and listen. The same thing happened not just with gainclones but also with my Pass Zen amp. Have also tried snubbers on my TDA1543 DAC power supply and didn't like the sound. I had standard KBPC rectifiers on ZEN, U860 on DAC and MUR860 and SB560 on gainclones.
Igla said:
1
Cool! That finding is a "signpost." When adding 1 cap per diode as a filter and you get a shout, the cap value is too large. Try 10x smaller capacitor value or 100x smaller capacitor value. The "shouting" indicates that you "hit" the audio band. . . so "dodge" it instead. 😉
2
Other place that its easy to tell by listening is the 2 big electrolytics on the amplifier board. . . If you get boomy/blurry, then those are too big. 😉 There are some seriously inappropriate designs that cover up a shout with a boom/blur; instead, I prefer this next bit (#3). . .
3
Have you tried the Rail2Rail cap to stop the shout? In the footnotes of the CarlosFM2005 power supply is a mention of a small (2uF to 4.7uF range and 250v) cap that goes from V+ to V- at the amplifier board itself. This is otherwise undocumented. It sinks the mids, and I think it also increases clarity. This, depends on a careful selection, though. A poly of some sort can work well here, or at least make the selection easier. Personally, I don't use snubber circuit at the output of the power supply board.
4
Not enough bass? Okay, then a couple of larger ecaps (average value 3300uF) can work at the Rectifier board. Example: One audiosector customer reported moving the supplied 1500uF's to the rectifier board and then using (adding) specialty 1000uF (blackgates) at the amplifier board. Well, that structure is quite similar to cap, pole, cap, as seen in the CarlosFM2004 supply. I think that particular customer had mistakenly purchased an undersized transformer. An alternative (much more simple) fix could have been in purchasing the recommended model or larger (VA) transformer, which would have preserved the "gainclone style" intact. Edit: You can do it either way.
No.
At the risk of hijacking this thread back to the original topic-
No, the BPA 300 cannot get you Plinius quality. I have owned a lot of high end and even "luxory" amps, and you will not get the full frequency response and base kick that you will get with Plinius. Not going to happen with these little chips. Cannot even come close to the quality you get with Krell Clone.
Exception- if you have very efficient speakers you may get close.
That said, the BPA and other chip amp variations are well worth trying out for tweeter amps and for fun. They do sound very good. You just have to give it a try. Try out the snubbers and everything else, see if you like it. Maybe you will surprise yourself.
At the risk of hijacking this thread back to the original topic-
No, the BPA 300 cannot get you Plinius quality. I have owned a lot of high end and even "luxory" amps, and you will not get the full frequency response and base kick that you will get with Plinius. Not going to happen with these little chips. Cannot even come close to the quality you get with Krell Clone.
Exception- if you have very efficient speakers you may get close.
That said, the BPA and other chip amp variations are well worth trying out for tweeter amps and for fun. They do sound very good. You just have to give it a try. Try out the snubbers and everything else, see if you like it. Maybe you will surprise yourself.
First of all..I have tried a lot of things with my gainclones - bypass caps, snubbers, regulated supply (LM338 and Mosfet), big PS caps, small PS caps, few fast and shotky diodes, different transformers, metal film and carbon resistors, Alps blue...etc,etc. Not to mention cooper blocks and Al heatsinks and
Every change by itself brought me different sound (sometimes better, sometimes worse).
But after every change after a few days listning I allways came to the conclusion that stripped gainclone with 1000uF on power pins with no input cap and with cooper block for heatsink and wooden base (BYOB style) and with one bridge of SB560 shottky diodes sounds best to me.
The second thing is that I like TDA2050 more then LM1875 and LM3875 in the same configuration - it remainds me more to SE tubes.
I guess I have somehow different ears than you guys.
Every change by itself brought me different sound (sometimes better, sometimes worse).
But after every change after a few days listning I allways came to the conclusion that stripped gainclone with 1000uF on power pins with no input cap and with cooper block for heatsink and wooden base (BYOB style) and with one bridge of SB560 shottky diodes sounds best to me.
The second thing is that I like TDA2050 more then LM1875 and LM3875 in the same configuration - it remainds me more to SE tubes.
I guess I have somehow different ears than you guys.
Re: No.
There it is! And now the point! 😀
Yes, LM3886 BPA300 is capable of a speedy and fierce low bass kick, along with excellent dynamics. The quality difference will be at the power supply and at the buffer (see Decibel Dungeon, 3 transistor buffer). You can match the performance of a high end amp. This is not less difficult than building the high end amp. 😉
EDIT: But this difference is why I rambled on SO much about the power circuit. Therein is endless variety, including the one you're looking for.
lgreen said:
There it is! And now the point! 😀
Yes, LM3886 BPA300 is capable of a speedy and fierce low bass kick, along with excellent dynamics. The quality difference will be at the power supply and at the buffer (see Decibel Dungeon, 3 transistor buffer). You can match the performance of a high end amp. This is not less difficult than building the high end amp. 😉
EDIT: But this difference is why I rambled on SO much about the power circuit. Therein is endless variety, including the one you're looking for.
- Status
- Not open for further replies.