I added a new power supply kit to the website, using the new snubber power supply:
http://www.chipamp.com/supply.shtml
I am offering 2 kits for the power supply, a stand-alone kit, and one with the components needed for retrofitting to the previous LM3875/LM4780 boards:
I am also now selling mounting kit, consisting of 4 - 5/8" standoffs and 8 - 1/4" screws, to simplify the process of obtaining parts for the projects. These are now on the order page under the power supply kit section. They can be ordered seperately to go with any kit, and cost $2.75 for the set. (had several requests for this in the past, so I decided to offer this option).
I am aiming to get all of the current LM3886 orders shipped out by the end of the week, along with any new orders for the LM3886 kits/power supply kits.
Anyway, I am finishing the first LM3886 amp with the new boards now, and will post more pictures of it later, and have some friends come over to compare the amp against my other gainclones when it is finished.
Drop me an e-mail if you have any questions.
--
Brian
http://www.chipamp.com/supply.shtml
I am offering 2 kits for the power supply, a stand-alone kit, and one with the components needed for retrofitting to the previous LM3875/LM4780 boards:
An externally hosted image should be here but it was not working when we last tested it.
I am also now selling mounting kit, consisting of 4 - 5/8" standoffs and 8 - 1/4" screws, to simplify the process of obtaining parts for the projects. These are now on the order page under the power supply kit section. They can be ordered seperately to go with any kit, and cost $2.75 for the set. (had several requests for this in the past, so I decided to offer this option).
I am aiming to get all of the current LM3886 orders shipped out by the end of the week, along with any new orders for the LM3886 kits/power supply kits.
Anyway, I am finishing the first LM3886 amp with the new boards now, and will post more pictures of it later, and have some friends come over to compare the amp against my other gainclones when it is finished.
Drop me an e-mail if you have any questions.
--
Brian
homer09 said:
Yes, I spend the whole day finishing the assembly of the new LM3886 amp and comparing with my other 2 gainclone amps. Here is a picture of my first LM3886 unit with the new boards:
An externally hosted image should be here but it was not working when we last tested it.
The power supply isn't as big as it looks in the previous picture. More pictures here:
http://www.briangt.com/gallery/lm3886amp
I ran comparisons, and I thought that the LM3886 amp sounded quite good with the new supply. I compared 3 gainclones:
An externally hosted image should be here but it was not working when we last tested it.
(amps from left to right)
- LM3875 with premium resistors, BG N 1000uF power supply caps in Peter's integrated chassis, riken 22k series attenuator
- new LM3886 in my old prototype chassis, 10k ladder attenuator
- Peter's AMP-1 with premium resistors, BG STD 1000uF power supply caps, noble 20k attenuators
Speakers = Seas Thor TL kit: http://www.briangt.com/gallery/thor
The felt ring around the millenium tweeter seems to help the highs be less glaring, without any negative effects.
I liked the new LM3886 amp a whole lot, especially considering that it isn't broken in yet. I have about 6 hours of playing time on it now. Here are some comparisons of the three units:
LM3875 with BG N - best highs of all 3 amps, but the bass seemed loose, and not as refined as the other 2 amps. (this amp has almost a year of play on it as my main amp so this shoudn't be a break-in issue) This has been my favorite amp in the past, due to the great highs and sound stage.
LM3875 AMP-1 - bass better than LM3875 with BG N, but highs seem to crisp and analytical. I felt that the sound stage was too wide sometimes, without enough focus. I have compared this amp many times to the LM3875 amp above
LM3886 with snubber supply - great soundstage (just about the same as the first amp), best mids/bass of all 3 amps. The highs seemed a little more recessed than the other 2 amps, and slightly less crisp on cymbals, something that could improved through break-in.
As for the room I listened to the speakers in, it was far from optimal, but given the controlled setup (same speaker placement, same amp placement, as close as possible amp volume levels, nothing else running on the same house power circuit), there was a discernable difference between the amps. All 3 amps were dead silent with no hum, except for a brief 5 second humming noise that the transformers make when being powered on. As for turn-off, the LM3886 amp turned off fastest (assuming that the bleeder resistors control this combined with the mute circuitry). I am rebuilding another room in my house (with new wiring and wall treatments), so I should have a better listening area sometime in the future.
Overall, the LM3886 amps seemed to be the easiest to listen to, being slight more laid back, and the better mids made the vocals stand out more. With my Thor TL speakers The lower bass was also better, making the overall setup sound more full. I am going to build another version of this amp in the integrated chassis with a better volume control and do more comparisons. For now, I will use the new LM3886 as my main amp. Given that the parts cost of the LM3886 is 1/3 that of the LM3875 with BG N caps (the 4 - 1000uf caps alone cost more than the entire LM3886 setup), I definately feel that the LM3886 with the snubber is the best value. I will try the new supply with a basic LM3875 and see how it compares.
If anyone else in the Atlanta area wants to try these amps out and see what they can gather from the differences, drop me an e-mail. I find it easier to listen to 3 amps, and see what stands out more with each one, rather than doing a blind comparison with only 2 amps.
I almost have the new kits ready to start shipping out, so we should see more comparisons when people start receiving the kits/new power supply boards.
Josh (edjosh23) was here today as well, and is finishing up his own LM3886 snubberized amp tonight, so hopefully he will also post pictures/comparisons too.
--
Brian
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
http://www.cardomain.com/memberpage/621686/3
That is my web page with a few other pictures.
I didn't finish my amp but Brian and I did some major comparisons.
My review:
Amp1, had the widest and my favorite sound stage of all three amps. The other 3875 and the snubberized 3886 had similar sound stages, both very wide compared to most amps. The bass of the 3886 was the deepest, but I think the other 3875 had more predominant bass. The highs/ cymbols were (just like Brian said) not as crisp, not as clean and defined as they were on the other 2 amps, I definately agree that after a few hours of burn in the amp can sound better.
Hope everyone likes the kits,
Thanks,
Josh
I guess now it's time for my comparisons.
I was quite impressed initially by the snubber big caps supply and was using it for some time. It seemed to definitely add to the bottom end, making the sound more dynamic and full, more dimentional with better presence ( the vocals definitely stand out more) and the highs somewhat "cleaner".
Well, that was untill someone actually pointed out to me, that the sound was not natural. At that time I was using 100u BGN at the chip and Jensen 10,000/63 at the bridges. I was already posting about that here: http://www.diyaudio.com/forums/showthread.php?postid=574431#post574431
I did more tests and comparisons, and indeed, when using snubber (not necessarily with 10,000u caps but even with 1000u) the effect is very similar: you loose the "noise" between the notes, something that can be related to as acoustic space or resolution. The effect of the snubber is almost like filtering, sound becomes drier (but with great dynamics) with somewhat pleasing presentation, but again, not very natural and not as liquid. Someone compared it to a synthesized sound.
I'm not saying that snubber or big caps are bad, some people will definitely like it, and I liked it too. But presently the snubber is out of my PS and I'm using 100u BG N caps alone (with AC power supply). If it wasn't for a snubber, I would have never tried that, but I also got very pleasantly surprised. I've been using it for more than a week now, and presently it's the best PS arrangement I had in a GC. The bass may be a bit weak, but OTOH it's very articulate and natural, with much better detail than the fat bass one gets with bigger caps, mids and highs though are extremaly good. While all of the filter caps have certain defficiencies, those single 100u BG N on amps board seem to be almost perfect. I can't really find any faults (except for weaker bass, but that can be fixed with biamping at 100Hz crossover frequency). The 100u BGN directly after bridges is way better than 1000u BG N (much more clarity and resolution in mids and highs), and it is better than using same caps with the batteries. I prefer AC PS to the batteries now. I tried adding different 1000u caps at the bridges, but the effect was always similar: dulling of highs, loosing speed and loosing the feeling of being there. The effect of batteries can be compared to the caps: since they store energy, I'm coming to conclusion that the less stored energy the amp is dealing with, the fresher the sound becomes.
1000BG STD are preferred by people who want deeper bass and more crisp sound. By comparison, 1000 BG N are somewhat mellow, the highs are nice but they don't stand out that well, and some people actually prefer STD version in direct comparison. 100u BG N when used alone, have none of deficiencies of bigger N caps: the highs are very crisp, clean and detailed, but at the same time have none of the traces of harshness, brightness or being analytical.
With regards to snubber and big caps, it's always a trade off: you either go for dynamics and strong bass, or you choose naturality, silky smooth and detailed highs available with small caps only.
If you are smart, you can choose bi-amping and have both
I was quite impressed initially by the snubber big caps supply and was using it for some time. It seemed to definitely add to the bottom end, making the sound more dynamic and full, more dimentional with better presence ( the vocals definitely stand out more) and the highs somewhat "cleaner".
Well, that was untill someone actually pointed out to me, that the sound was not natural. At that time I was using 100u BGN at the chip and Jensen 10,000/63 at the bridges. I was already posting about that here: http://www.diyaudio.com/forums/showthread.php?postid=574431#post574431
I did more tests and comparisons, and indeed, when using snubber (not necessarily with 10,000u caps but even with 1000u) the effect is very similar: you loose the "noise" between the notes, something that can be related to as acoustic space or resolution. The effect of the snubber is almost like filtering, sound becomes drier (but with great dynamics) with somewhat pleasing presentation, but again, not very natural and not as liquid. Someone compared it to a synthesized sound.
I'm not saying that snubber or big caps are bad, some people will definitely like it, and I liked it too. But presently the snubber is out of my PS and I'm using 100u BG N caps alone (with AC power supply). If it wasn't for a snubber, I would have never tried that, but I also got very pleasantly surprised. I've been using it for more than a week now, and presently it's the best PS arrangement I had in a GC. The bass may be a bit weak, but OTOH it's very articulate and natural, with much better detail than the fat bass one gets with bigger caps, mids and highs though are extremaly good. While all of the filter caps have certain defficiencies, those single 100u BG N on amps board seem to be almost perfect. I can't really find any faults (except for weaker bass, but that can be fixed with biamping at 100Hz crossover frequency). The 100u BGN directly after bridges is way better than 1000u BG N (much more clarity and resolution in mids and highs), and it is better than using same caps with the batteries. I prefer AC PS to the batteries now. I tried adding different 1000u caps at the bridges, but the effect was always similar: dulling of highs, loosing speed and loosing the feeling of being there. The effect of batteries can be compared to the caps: since they store energy, I'm coming to conclusion that the less stored energy the amp is dealing with, the fresher the sound becomes.
1000BG STD are preferred by people who want deeper bass and more crisp sound. By comparison, 1000 BG N are somewhat mellow, the highs are nice but they don't stand out that well, and some people actually prefer STD version in direct comparison. 100u BG N when used alone, have none of deficiencies of bigger N caps: the highs are very crisp, clean and detailed, but at the same time have none of the traces of harshness, brightness or being analytical.
With regards to snubber and big caps, it's always a trade off: you either go for dynamics and strong bass, or you choose naturality, silky smooth and detailed highs available with small caps only.
If you are smart, you can choose bi-amping and have both
The silence is between the tracks. Between the notes, there is always something going on, like for instance the reverberations and decays of the original sounds, created by the recording space. I'm very well aware that this may be distortion, but to my ears it makes for much more real presentation.
On anechoic recordings there is indeed silence between the notes, but those recordings sound very artificial.
What I'm also noticing when using very small caps (100u in this case), is certain harmony and righteousness of the sound, that simply doesn't leave any question marks in your head if it is correct. You just know, that it can't be any other way (at least for a time being)
On anechoic recordings there is indeed silence between the notes, but those recordings sound very artificial.
What I'm also noticing when using very small caps (100u in this case), is certain harmony and righteousness of the sound, that simply doesn't leave any question marks in your head if it is correct. You just know, that it can't be any other way (at least for a time being)
Peter Daniel said:What I'm also noticing when using very small caps (100u in this case), is certain harmony and righteousness of the sound, that simply doesn't leave any question marks in your head if it is correct. You just know, that it can't be any other way (at least for a time being)
Peter Daniel said:What I'm also noticing when using very small caps (100u in this case), is certain harmony and righteousness of the sound, that simply doesn't leave any question marks in your head if it is correct. You just know, that it can't be any other way (at least for a time being)
Peter, your results are in the context of your system, mainly your speakers.
I once tried just 220uf caps and the amp didn't drive my Epos 11 speakers, the bass was not tight at all, it was more like a continuous tone that made everything vibrate, no definition (you may not notice this with your speakers).
As a result, the bad bass killed the midband and treble detail (as it always does).
I understand what you are saying about the detail, as with less demanding acoustic music the detail was very good.
But I felt like something was not really right, like pushed a treble tone control.
I have very good results in the mid/treble and good (with all the harmonics and vibration of the instruments, as I also like), tight, dynamic bass.
Btw, you can try two 4,700uf (or even three) caps in parallel (which I recommend) instead of single 10,000uf.
Yes, those results definitely apply to easy to drive speakers. I tried it at home using Raven/Triangle combo (Triangle used without crossover, full range), and in other system, with Omega speakers using Fostex drivers. I'm pretty much sure that it will not work with all setups, I'm quite surprised it works at all.
I'm not dismissing snuber and big caps, it definitely have certain appeal (well, I was impressed myself, if you remember). I'm only suggesting that it should not be considered as an obvious upgrade: it will greatly depend on personal taste, the system and listening priorities. It's just another way of enhancing your listening experience, but again, it may not be suitable for everyone. But I definitely encourage everybody to try it and decide for themselves.
However, I would also encourage anybody who have access to 100 BG N caps to try them alone in AC PS and report the findings. I don't think, myself, I will be going back to other configurations. But for more realistic bass (only on certain material), aditional low pass amp (for bi-amp setup) could be further enhancement.
PS: As you may notice in a picture, I also removed a series resistor a the input. As good as Caddock (or Vishay) was, I'm getting better resolution without it.
I'm not dismissing snuber and big caps, it definitely have certain appeal (well, I was impressed myself, if you remember). I'm only suggesting that it should not be considered as an obvious upgrade: it will greatly depend on personal taste, the system and listening priorities. It's just another way of enhancing your listening experience, but again, it may not be suitable for everyone. But I definitely encourage everybody to try it and decide for themselves.
However, I would also encourage anybody who have access to 100 BG N caps to try them alone in AC PS and report the findings. I don't think, myself, I will be going back to other configurations. But for more realistic bass (only on certain material), aditional low pass amp (for bi-amp setup) could be further enhancement.
PS: As you may notice in a picture, I also removed a series resistor a the input. As good as Caddock (or Vishay) was, I'm getting better resolution without it.
Attachments
I am liking your design!
It looks to me as this would be the cheapest amp with power supply kit having great reviews I've found recently!
Which version should I get, LM3886 or LM3875?
It looks like LM3875 is from another source and comes without PS, so I have to pay shipping twice. Is that so?
Is LM3886 Dual Mono kit more powerful than Stereo Kit due to double PS?
And finally, how many watts does this design deliver?
Thanks!
It looks to me as this would be the cheapest amp with power supply kit having great reviews I've found recently!
Which version should I get, LM3886 or LM3875?
It looks like LM3875 is from another source and comes without PS, so I have to pay shipping twice. Is that so?
Is LM3886 Dual Mono kit more powerful than Stereo Kit due to double PS?
And finally, how many watts does this design deliver?
Thanks!
I finally finished the amp! I hate high school, too much school work. Well this is the best sounding amp I've built so far, dc offset was -10mv and 11mv. There is no turn on or off thump, no hum. I got an emi filter from Brian that I think fixed that little problem. But I've now listened for about 4 minutes, and it sound quite good. I'll post later with another review.
Thanks,
Josh
Thanks,
Josh
I finished prepared the first batch of LM3886 kits tonight, and will be taking them to the post office tomorrow morning.
As for the packages, on the first 1/2 of them, I taped the 1N4004 diodes (used for rectification for the LED power supply) to the outside of the padded mailer (which will be put inside the USPS priority mailer), so don't discard the diodes when you unpack the kit. I forgot about this component until 1/2 way through the kitting process tonight. Everything else should be in order. It has taken me a cumulative average of 15 minutes per kit for putting together the new kits, which is considerably more than the LM3875 kits, but should decrease now that I have the method worked out.
I will get the remaining (majority was done tonight) orders put together and out by the end of week.
I will be unexpectively be out of town from this Saturday until Monday (family issues, flying home), and also next week from Wednesday until Saturday (last minute business trip to California), but I will try to keep up with e-mail during that time, and resume filling the orders when I am back in Atlanta.
The US orders that shipped out should have received a USPS confirmation mail.
--
Brian
As for the packages, on the first 1/2 of them, I taped the 1N4004 diodes (used for rectification for the LED power supply) to the outside of the padded mailer (which will be put inside the USPS priority mailer), so don't discard the diodes when you unpack the kit. I forgot about this component until 1/2 way through the kitting process tonight. Everything else should be in order. It has taken me a cumulative average of 15 minutes per kit for putting together the new kits, which is considerably more than the LM3875 kits, but should decrease now that I have the method worked out.
I will get the remaining (majority was done tonight) orders put together and out by the end of week.
I will be unexpectively be out of town from this Saturday until Monday (family issues, flying home), and also next week from Wednesday until Saturday (last minute business trip to California), but I will try to keep up with e-mail during that time, and resume filling the orders when I am back in Atlanta.
The US orders that shipped out should have received a USPS confirmation mail.
--
Brian
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.