| kittikun |
Hi All,
I'm just new in this forum and especially class D amp. After going through many threads, I think that it is a time to start a new experience.
I want to try some high power class D amp for my diy project but most of the thread I read is about technical discussion. There is very few information about the sound quality. I get confuse because I do not know exactly which amp should I start with my project. The problem is there are too many amps in different brand and varieties, range from Ucd(w & w/o AD8260), Lc audio and ICE.
I use Magnepan 1.6 in my system and it always hungry for power. Although I fed my monster with SS 500w/4 ohm amp, it is never enough. The impedance of magnepan is very low at 4 ohm and I love tube sound. But all my tubes cannot drive magnepan.
Can anyone help recommend me? Thanks you in advance
Kittikun |
|
|
| dmfraser |
Personally I am very happy with both ICE Power and the UcD amplifiers, both the 180 and the 400.
The ICE Power amps drive 4 ohms fine. They also have a version of the 500A, called the 500AC that is supposed to be 2 ohm stable. However, I have not had a chance to try the samples they sent me into a sub 4 ohm load.
The UcD180 operated fine into 4 ohms for me as well. However, again, I am unsure about sub 4 ohm loads.
However, if the amp chokes on the Maggies, note that the Maggies are almost purely resistive and do not go crazy as much as conventional speakers if you have to put in a bit of series resistance to keep the power amp stable. |
|
|
| kittikun |
Hi Dan,
Thanks for your recommendation.
The problem when I use SS amp with Maggie is the bass is very thin. When I use a PP 75 W tube amp(Mcintosh MC 75) with Maggie through the 4 ohm output, the bass is much better but it's still lack of the body.
If I were to put, let's say 4 ohm resistor in series to Maggie whether it has any adverse effect to both amp and Maggie. I think that the sound pressure will drop a bit for 3 db. Anyway, I will try your idea.
Can you please tell me where I can find the ICE Power? Thanks.
Kittikun |
|
|
| Jocko Homo |
Well, you can't. They are for OEM use only. I know of guys who have used commercially produced amps using ICEpower on current hog speakers. They seem to be pleased. I don't consider Maggies to be current hogs, as compared to something like Apogees.
You would probably be better off trying the UcD modules......easily obtainable, and probably more DIY-friendly in terms of implementation.
Jocko |
|
|
| dmfraser |
Yes Jocko you are correct. ICE Power modules are only available on an OEM basis. I'm a bit spoiled being the chief electronics engineer for one of those OEMs. I get evaluation samples of all the ICE products as soon as they are available. I can count about half a dozen within little more than arms reach.
But I have heard of some people getting their hands on some. However Jan Peter ot Hypex sent me samples of the UcD180 and UcD400 his company makes. I have to admit that the performance was very good and to tell you the truth, the iCE power people should worry about the Hypex products. As well, the UcD modules are quite a bit easier to use to replace Class AB amplifiers, specially in the power supply as the UcDs use a conventional bi-polar power supply while the ICE use a single supply but need auxiliary ±12V supplies to work.
Also, ICE modules are H bridge and the output floats at half the supply voltage, making them unsuitable for anything other than powered speakers.
Finally, the UcD modules are actually available to the small quantity buyer.
Therefore, I think that if a DIYer wants a good class D amplifier, or for than matter, any kind of amplifier, that they can actually buy, it is hard to do better than the value offered by the UcD modules. No, they are not quite at the level of things like Mark Levenson but considering the price difference, the value offered is spectacular. |
|
|
| davey1 |
Hi all,
Just found this site and thought I'd get your take on my position....
I formed a company with some of my former collegues last year and we specialise in class D / switch mode PSU / DSP and networking. We've combined these technologies and produced a range of power modules - pretty similar to B&O's IcePower but with higher power, onboard DSP with PC control and packaged in a slightly more complete format.
We've been pretty successful to date and supply a number of high end loudspeaker manufacturers. We've also had a number of enquiries from relatively small outfits / individuals but I haven't identified a sensible sales model to supply these two ends of the market.
So, that's my dilemma - I would like to be able to supply the DIY community but in such a way as it doesn't compromise the work we've done with our major customers - what do you think?
BTW, I've not mentioned the company name - I'm not trying to sell anything! |
|
|
| Steven |
| quote: | Originally posted by dmfraser
As well, the UcD modules are quite a bit easier to use to replace Class AB amplifiers, specially in the power supply as the UcDs use a conventional bi-polar power supply while the ICE use a single supply but need auxiliary ±12V supplies to work.
Also, ICE modules are H bridge and the output floats at half the supply voltage, making them unsuitable for anything other than powered speakers.
|
My company also uses the B&O ICE modules, both the bigger 500W version for 2 Ohm load and the smaller one for higher impedances. They drive transformers for a 70V/100V output for Public Address use.
For us this single supply is convenient as we also have to deliver full power from a charged 48V battery for emergency situations. We combine them with a SMPS that delivers normally 58V from the mains and also the small +/- 12V supplies. Battery kicks in via a diode if the mains disappears.
Steven |
|
|
| km |
hi.
"Can you please tell me where I can find the ICE Power? Thanks."
i can give you a hint where you can buy some ;)
email me if you need more info.
rgds - k madsen - www.cadaudio.dk |
|
|
| raintalk |
| quote: | Originally posted by davey1
Hi all,
...
So, that's my dilemma - I would like to be able to supply the DIY community but in such a way as it doesn't compromise the work we've done with our major customers - what do you think?
|
So it seems you don’t consider DYI as a major market. :-)
DIY is no threat to your major customers. Your major customers can buy in volume and at lower prices. Think of it in terms of growing the market. The DIY are not in the same market as your major customers, one’s not loosing sales to the other.
Other modules are becoming popular and "standard" in the DIY market so jump in soon ... |
|
|
| classd4sure |
Hi,
Raintalk:| quote: | | The DIY are not in the same market as your major customers, one’s not loosing sales to the other. |
Nice point.
davey1, hi, pretty slick "not trying to sell anything" lol ;)
So, the Xpod huh.
I'd agree with Raintalk, the two are different markets, and as such, have different needs.
It sounds like apart from the smps, it comes with a built in audio card.
The odd DIY're might like an option for a DSP, most won't want to pay the extra $ for it when all they are interested is the amp, or SMPS.
Would the amp itself be better without the added noise and power consumption of the DSP?
Looks like this is your competition:
http://pro-audio.powersoft.it/page....=195&id_page=39
Modular can be a good thing.
Xpod sounds alot like Ipod too... that could become a problem for you, I'd come up with a more original name for an original product, before they sue you over it.
Regards,
chris |
|
|
| sven-ake |
| quote: | Originally posted by kittikun
Hi All,
I'm just new in this forum and especially class D amp. After going through many threads, I think that it is a time to start a new experience.
I want to try some high power class D amp for my diy project but most of the thread I read is about technical discussion. There is very few information about the sound quality. I get confuse because I do not know exactly which amp should I start with my project. The problem is there are too many amps in different brand and varieties, range from Ucd(w & w/o AD8260), Lc audio and ICE.
I use Magnepan 1.6 in my system and it always hungry for power. Although I fed my monster with SS 500w/4 ohm amp, it is never enough. The impedance of magnepan is very low at 4 ohm and I love tube sound. But all my tubes cannot drive magnepan.
Can anyone help recommend me? Thanks you in advance
Kittikun |
Zappulse from LC Audio should be something to try you can build them very powerful if you like.I have mine driving martin Logan SL3 and it works really good and they are very easy to build.With the PSU i use they can deliver approx.700w in 4 ohms |
|
|
| peranders |
| quote: | Originally posted by davey1
BTW, I've not mentioned the company name - I'm not trying to sell anything! |
:cop:
It's OK for you to mention the company name and URL, this time at least :) I think pretty are interested to know.
:cop: |
|
|
| classd4sure |
Hi,
Yeh, I have to applaude anyone who wants to supply the DIY community in any way shape or form.
I'm sure I'm not the only one nosy enough to have done this but, just click on his homepage link and you'll find:
http://www.linea-research.co.uk/
Proceed to products and you'll find:
http://www.linea-research.co.uk/product_x_pod.htm
I've also found there is a program related to the Ipod with the same name as their amp module.
Dave didn't make this hard to find:) Should have started a new thread for it though, it's worthy of it, more people would have seen it too.
Regards,
Chris |
|
|
| kepa1 |
yes, X-Pod is also the name of speakers stands here in france; i'm not sure the name is so relevant...
pretty interesting product, though, especially the DSP part.
alain |
|
|
| classd4sure |
Hi,
Yeh that's a good point, they turn on a little slower than they turn off, but that's only as seen under the "typical" ratings, max ratings are all the same, so it probably wouldn't be good to rely on that, and that's not dead time anyway.
IVX posted a nice example of how to use the 2110, search for IVX and 2110.gif and you'll likely find an excellent example showing how to implement it.
I'd consider looking at the HIP series of the same drivers, from intersil. The timing on them is much better according to the data sheet.
Regards,
Chris |
|
|
| classd4sure |
| quote: | Originally posted by kepa1
yes, X-Pod is also the name of speakers stands here in france; i'm not sure the name is so relevant...
pretty interesting product, though, especially the DSP part.
alain |
It isn't relevant, until somebody takes them to court over it.. It's also related to audio as well as the Ipod.. it just wouldn't look good in court, they've already lost if it ever gets to that.
Nobody wants to see that happen. |
|
|
| peranders |
Of some peculiar reason I didn't notice his www-button.
I have checked the homepage but I didn't find much, nice photos but very little about the products, not a single datasheet in sight. |
|
|
| classd4sure |
| quote: | Originally posted by peranders
Of some peculiar reason I didn't notice his www-button.
I have checked the homepage but I didn't find much, nice photos but very little about the products, not a single datasheet in sight. |
Yeah, I noticed the same thing. I guess they're just emerging on the market though, so we'll see how well they do getting the act together.
I'm can't say with absolute certainty but I swear there's additional products there that I didn't see a few days ago, or maybe I just didn't notice at the time.
Anyway, bring on those data sheets :) |
|
|
| davey1 |
Hi all,
Wow! I only posted on Friday so thanks for taking the time to reply - I'm amazed so many have....
For those who did find our website, yes of course I'm trying to sell the stuff, that's my job! - I'm just trying not to force feed people, as I know it's annoying.
Also, I'm well aware of the lack of information on the site - I've tried to give an impression of where we have expertise rather than specifics on a particular product although I do intend to improve this in the near future.
I certainly do not consider the DIY market to be marginal - it's how I started and I have much respect for people doing it their own way. My problem is how to sell effectively to the two ends of the market. B&O and Powersoft have minimum order quantities that basically preclude anyone other than an OEM from getting involved. Some other manufacturers will sell to 'anyone' but this is not as straightforward as it may seem. I think the two key areas are pricing and support;
Despite what some people tell me, effective integration of these products takes time and support. This overhead has to be paid for - ignoring this requirement for application support is a folly in my book; I certainly don't want to sell to someone on the other side of the world and just forget about them if they need some after sales support.
Pricing is the second issue. In my experience pro-audio manufacturers work on a 50% margin i.e. make for $50 and sell for $100. If I sell a unit to BassBins Inc. for $100 they will integrated it into their product and sell it for $200 - $300. I would need to match this kind of pricing so I don't compromise the larger volume OEM sales. The problem is it makes these products seem pretty pricey.
So, I suppose that's it, I feel like I have to sell low volumes at high prices to allow for the support costs and protect our OEM sales - it just doesn't feel right.....you guy's pay more to protect the interests of the larger customers out there.
One final point - the modules' name.
X-Pod comes from the fact that it's designed to be a speaker powering module so 'Pod' seems about right. The X is for its' built in crossover. At the time the only 'X-Pod' I could find was a fishing rod stand :o). Unfortunately Pod's are big business now and I'm sure they'll be A, B & C-Pod's before long. It's not trademarked in the UK and I haven't had a writ from Mr Gates. (yet) but if we do get some pressure I'm sure we'll change. In the end it's the product that's really important, not the name eh?
Davey |
|
|
| classd4sure |
Hi,
Many good points.
How you want to work out the support issue is up to you of course. Good instructions, data sheets, faqs, tips, readily available will help (OEMs as well), and should be a big part of it. Forums like this are another. Maybe a design example or two could be an option as well.
This way it isn't like you're having to spend on a 1-800# and full time staff just to answer questions, and should help keep costs lower.
Remember the OEMs just bought your product, not the exclusive rights to it, if you think it will make a larger customer want to walk, then you've got a dilema. OEMs will walk on you anyway when a better/cheaper product comes along or suits their purposes for whatever the reason.
Volume based price breaks seems the fair way for everyone, is commonly done, and I don't think anyone honestly expects to pay the same amount the OEMs paid for it in mass quantities anyway, but all within reason. Don't tell us what OEMs are using it, or what they paid for it :)
Looking forward to seeing more of this.
Regards,
chris |
|
|
| phase_accurate |
| quote: | | I'm just trying not to force feed people, as I know it's annoying. |
That depends heavily on the taste of the food. :D
At the current exchange-rate for the USD this would be an interesting price for 2 channels + PSU + signal processing (can you implement transient-perfect crossovers as well ?).
| quote: | | It's not trademarked in the UK and I haven't had a writ from Mr Gates. |
I assume he would actually be pleased if you caused confusion with the other ".....-pod". ;)
Regards
Charles |
|
|
| davey1 |
Hi Charles
Not actually quoting prices for the unit, just an example to illustrate the issue.
What do you mean by 'phase perfect' - FIR filters?
Thanks,
Davey |
|
|
| davey1 |
Oops!
Meant, 'transient perfect' |
|
|
| phase_accurate |
Hi Davey
Yes, that could be one solution. But there would also be possibilities for transient perfect IIR filters.
A possible candidate is a proposal by John Kreskovsky (a sporadic poster on this forum). He subtracts a lowpass with constant group delay (i.e. Bessel) from a constant delay (i.e. a simple circular buffer) to obtain the highpass. The summed response is exactly representing the input in time- and frequency- domain.
The advantage over a FIR crossover is that it doesn't have pre-ringing. The disadvantage is its asymmetric slopes.
Regards
Charles |
|
|
| The Rocker |
| quote: | Originally posted by sven-ake
Zappulse from LC Audio should be something to try you can build them very powerful if you like.I have mine driving martin Logan SL3 and it works really good and they are very easy to build.With the PSU i use they can deliver approx.700w in 4 ohms |
Another newbie to this forum. I read the info on the LC Audio website and it seems reasonably easy for someone slightly hamfisted (myself) to build a working amplifier. Add a decent sized transformer.....
But the real question is, how good is the sound from the ZAPPulse? I don't have any unused >1KVA traffos on the shelf so every nut and bolt has to be purchased for this project. Therefore I would like to know what to expect. The last sentence should read: what you found when you built your amplifier. Sound quality is a very subjective thing, but how does it compare to commercially available amplifiers? In round figures I reckon on a spend of €750.00 so is it reasonable to compare a self built amplifer with a commercially built unit retailing at €1.5K? Thanks for taking the time to read this post, hope my query makes sense. |
|
|
| ericpeters |
Hi,
The zappulse 2.2 SE is a very good sounding module but I think this strongly depends on the speaker you want to drive with it. The experiences with speaker with complex impedances are not so good aparantly. Low impedance is no problem at all the modules can deliver lot's of current
I've had experience with Zap and Ice and the Zap sounds better than Ice IMHO, and the zap also sounds better than a Krell KSA 100 MK2. They outperformed the Rotel RB980BX by a very large margin.
I never heard the UCD modules. |
|
|
| tim pattinson |
Eric's answer is similar to my experiences- I have M-L Ascents [v. difficult load], and with the PSU on the upper end of the recommended voltage rail I had some early rolloff of LF. This was not the case with the UCD modules.
I saw mentioned in the forum recently that you reduce the output impedance of the Zaps by lowering their rail volts, so that's next to try. |
|
|
| GregD |
I've experimented a bit with 2.2SE Zappulse and have found it to sound very good with a beefy power supply. I have a pair of Watt/Puppy 6s and my normal amp is a Threshold S/500 II which I've been very happy with over the years. My comments below are in comparison to the Threshold.
I tried LC Audio's Predator power supply with two modules side by side, and found the channel separation to be poor, the bass had an initial impact but no body, and a rather closed in sound like the high frequencies were rolled off.
Then I tried paralleling some big capacitors (84,000uf) with the Predator power supply. This dramatically improved the bass and suprisingly enough helped the highs quite a bit as well. The channel separation was still bad and the highs still sounded rolled off, but not as bad as before.
So I decided to try a totally maxed out power supply with separate monoblocks. Each amp has a 1500VA transformer with 328,000uf of capacitors with bridge rectifiers and some bypass capacitors. With this combination, the amp was finally a contender; separation was excellent (as you would expect), the impact, dynamics, and control were outstanding and superior to the Threshold. The sound is very clean with excellent clarity and a great soundstage, but the highs still seem a bit rolled off; there's not the same airy spaciousness to the sound I get with the Threshold. Also, at medium to high volume levels the midrange gets a recessed sound, almost like it's at the end of a long tube, kind of a weird effect that is hard to describe.
I plan on trying a little RC snubber on the main caps with a 100uf cap attached directly to the modules like the chip amp folks have been discussing to see if this will make any difference to the sound. Also, I really want to try some of the UCD modules when they start making them again; hopefully they'll do the nice things the Zaps do without the odd things.
In summary, there's a number of things I like about the Zappulse amps better than the Threshold including having a generally clearer and less distorted sound, but the weird recessed sound effect and the less open sounding top end have me sticking with the Threshold. But they sound better than a couple of other amps I've tried in my system, an Adcom GFA-555 and a Cary Audio CAD572SE, and I think better than many other amps that I've tried in other systems over the years.
Just my observations and opinions. YMMV! :) |
|
|
| GregD |
| Here's a picture of the test amp configuration I put together. I had a couple of old computer cases laying around, so they were pressed into service for experimental purposes, some modification required. :smash: |
|
|
| serengetiplains |
Greg, thanks for your informative comment. Interesting how that particular amp came together for you using the mondo-supply approach. I wonder if a judicious use of film capacitors might help in the airiness department? But might also augment certain of the weird effects of which you speak?
Cheers. |
|
|
| GregD |
| quote: | Originally posted by serengetiplains
Greg, thanks for your informative comment. Interesting how that particular amp came together for you using the mondo-supply approach. I wonder if a judicious use of film capacitors might help in the airiness department? But might also augment certain of the weird effects of which you speak?
Cheers. |
I'm planning to try the snubber circuit that has been discussed on the chip amps forum in this thread http://www.diyaudio.com/forums/show...&threadid=48942
I'm hoping this will help the airiness, but will be surprised if it ameliorates the weird effects I noticed, but keeping my fingers crossed. It's a little frustrating because on some pieces of music, the amps sounded absolutely killer with a tubelike midrange, clean elegant highs, and a tight, deep, effortless bass that would put a Krell to shame. But then other times they would just sound kind of dead and a little like they were in a cave. They didn't sound BAD when this happened, just not GOOD, and often just on certain parts of a piece of music.
:cannotbe:
I had them in my system for a little over a week, and never felt a need to pull them out which is a good sign. When I went back to the Threshold I immediately noticed a more distorted and hashy sounding high end and relatively loose control of the bass, but despite that, it had a lifelike quality to it that the Zappulse couldn't match, and the sound is always consistent, at least after a 15 to 20 minute warm up period. The Zappulse based amp changed it's sound quite a bit for the better in the first hour or so of operation. After that, I didn't notice much difference.
A couple of other notes. The Zappulse had more hiss from the tweeters than the Threshold without music. I could hear the Zappulse from about 2 to 3 feet away, the Threshold only inside of 6 inches. I don't think it's too meaningful since both sound silent from my listening position. Also, something I had never checked before, the Threshold puts out a very low level 60HZ hum when I practically stuck my head in the midwoofer, no hum at all from the Zaps.:)
Here's a picture from the top of the amp. I like the way the star ground ended up looking like a star. :nod: |
|
|
| ericpeters |
This was originally replied to a question I got per email but I share it with anyone wo is interested:
I will first explain what steps I went trough, what the final result is , and tell you what I would use when I built it again.
At first I tried to build one stereo amplifier. That amplifier was built with the follwing power supply module:
BMM 90V
Also one 625 watt / 2*55VAC torroid transformer was used in this stereo amp
This did not sound good at al, lack of real bass and grainy mid and highs! I went listening to someone else his monoblocks and they were much better. So I decided to build something similar but improved the power supply according the zero distortion article.
The new transformers (2* 40VA / 1000VAC) arrived before the rest of the components so I tried one of this transformers in the stereo amp, and immediately the sound was very good.
I'm pretty sure you should keep the voltage around the 40 Volt even if the specifications of the Zap module alow more. I also believe: the bigger the transformer; the better the sound.
So I built my mono-blocks using the following components per channel:
*) 1000 VA / 2*40 VAC torroid (Even bigger is probably better but these are still standard size and relative cheap. (under 80 euro at www.Farnell.com )
*) 4* 33000uF BC components, 63V Capacitors. (so that is 132.000uF per channel, maybe less is possible but I wanted to be sure everything was perfect.)
I got them at Schuro for reasonable prices: BC 154 series
*) 8 Hi speed diodes ( Schuro ) You can built a single bridge with 4 diodes but the zero distortion document recommended double bridges to separate the power supply physically from the mains. I used this for the Zappulse module to prevent the ground loop related high frequency noise from occurring) If I built new ones I would probably use normal square rectifier bridges (but always double bridge). Aparantly there is an article from Nelson Pass somewhere that explains that these are better for Power amplifiers and they are a lot cheaper and makes it easier to built as well. You can get them everywhere.
*) All smaller capacitors as indicated in the zero distortion document, they are relative cheap.
*) 2 pcs of 25Watt 6k8 leak resistors over the power supply rails.
*) a softstart circuit ( I used this one ) I needed to change the timing because the standard values were too small for this size power supply.
I made 3 solid copper plates for the supply rails.
Signal cabling is all solid core silver in teflon tubing.
Connectors are WBT for Speaker and Chinch and Neutrik for XLR.
There a 5 things thing I could improve:
1) Even a bigger transformer
2) Shorter cables between zap module and power supply, layout needs to be different especially the heatsink location.
3) RF filters on the mains connection
4) The normal rectifier bridges to get rid of the chaotic cabling around the self made bridges.
5) Fully aluminum casing, It currently has a MDF bottom. As you see I put some aluminum tape on this, I did that when I had some noise problems with one amplifier. It turned out that one of the soldering joints was loose but I thought it had to do with the wooden part of the casing.
Please keep in mind that I have Apogee speakers that are very low impedance and pull a lot of current out of the amplifier. You might be satisfied with a lot smaller power supply. |
|
|
| ericpeters |
This is a side view of the same mono, as you can see construction wise it is what you could call "proper DIY"
I skipped on using any high quality bypass capacitors because the Zap 2.2SE module takes care of that with it's Blackgate's mounted directly on the module. |
|
|
| ericpeters |
[| quote: | Originally posted by GregD
I've experimented a bit with 2.2SE Zappulse and have found it to sound very good with a beefy power supply. .......... |
Greg, the modules on your pictures don't look like the 2.2SE modules. The SE version has 2 rubycon black gate capacitors on the module. I see only 2 blue elco's on your pictures.
Also the size of my transformers appears to be similar to yours (even bigger, they wouldn't fit in a pc mini tower width wise) but you claim yours is 1500VA and mine are 1000VA???
My modules are very consistent and no sign of harschness. I also cannot hear any hiss at all even if I put my ear on the speakers.
What's the voltage on the supply rails you got?
do you got mains filtering installed? If not the sound quality could vary depending on what else is happening on your mains net. |
|
|
| UrSv |
| quote: | Originally posted by ericpeters
[
Greg, the modules on your pictures don't look like the 2.2SE modules. The SE version has 2 rubycon black gate capacitors on the module. I see only 2 blue elco's on your pictures.
--snip-- |
The BG caps where replaced a while back. |
|
|
| The Rocker |
Guys, many many thanks for all the replies. Those are seriously large caps!!!!! My speakers are (oldish) Ruark Crusader 2 which are I think nowhere as severe a load as some speakers, I think I will dip my toes into the water by building a stereo amplifier with a traffo greater than 1 KVA and see how it sounds. Better than a Krell???, that is saying something!!
Can't wait to hear it speak. |
|
|
| GregD |
| quote: | Originally posted by ericpeters
[
Greg, the modules on your pictures don't look like the 2.2SE modules. The SE version has 2 rubycon black gate capacitors on the module. I see only 2 blue elco's on your pictures.
Also the size of my transformers appears to be similar to yours (even bigger, they wouldn't fit in a pc mini tower width wise) but you claim yours is 1500VA and mine are 1000VA???
My modules are very consistent and no sign of harschness. I also cannot hear any hiss at all even if I put my ear on the speakers.
What's the voltage on the supply rails you got?
do you got mains filtering installed? If not the sound quality could vary depending on what else is happening on your mains net. |
I ordered the SE modules and that's what it showed on the invoice, and my modules match the pictures on LC Audio's website http://www.lcaudio.com/index.php?page=35 I remember when I ordered them reading that they had recently changed the capacitors on the modules, so I assume that's where the difference in appearance lies.
The case is actually a medium tower, the transformers are 8" in diameter and about 4" tall with dual 45V secondaries, seems about right for 1500VA. I'm getting +/-60V on the rails with dual bridge rectifiers. I have a Schaffner mains filter integral with the IEC socket, and a soft start circuit of course. The big electrolytics are United Chemicon 100V 82,000uf each U32D series. I assume that the zero distortion PSU you're talking about is here http://www.zero-distortion.com/start.htm Did you try your amp without the small capacitors?
I also have heard no sign of harshness, just the strange recessed sound effect I mentioned earlier at medium and higher volume levels. I'm not surprised you don't hear the hiss with your speakers since the sound is spread over the whole surface of the ribbon. By the time you get far enough back from the speaker to be getting the sound from a significant portion of the ribbon, the level will be too low to hear; it's not a point source tweeter like my speakers.
Which Apogees do you have? I have a pair of Divas in storage; they wouldn't fit into my new place when I moved. :bawling: |
|
|
| GregD |
| I like your amp construction; it looks and sounds like excellent quality. One thing I was wondering about though was your leakdown resistors, "2 pcs of 25Watt 6k8 leak resistors over the power supply rails". With the 40V transformer, I assume that your output rails are about 55V. Watts = 55 * 55 / 6800 = .44 watts, so why 25 watt resistors? I have 3k6 2 watt leakdown resistors on mine, and with the big capacitors it takes a long time for the voltage to come down to safe levels; the amp will play at a medium to high volume level for over a minute after it's unplugged. :D |
|
|
| sven-ake |
| I was told by Henrik at LCaudio not to use more capacitors than max 30.000uf because they had experimented and found that the midrange suffered when they used bigger caps.As i see you DIY:ers testing bigger ones and with good results i begin to wonder a lot. |
|
|
| GregD |
| quote: | Originally posted by sven-ake
I was told by Henrik at LCaudio not to use more capacitors than max 30.000uf because they had experimented and found that the midrange suffered when they used bigger caps.As i see you DIY:ers testing bigger ones and with good results i begin to wonder a lot. |
When I have a little time, I'm going to try some smaller caps while leaving the rest of the power supply the same. I plan to try 20,000uf per rail, and 40,000uf per rail using these. |
|
|
| classd4sure |
Hi,
At what point do you suppose bigger caps and bigger transformers passes the point of diminishing returns?
I get the feeling you're there already. |
|
|
| serengetiplains |
| quote: | Originally posted by classd4sure
Hi,
At what point do you suppose bigger caps and bigger transformers passes the point of diminishing returns? |
Probably around 100VA. Returns are returns, unless you're Nortel. |
|
|
| classd4sure |
You meant 1000VA right? I agree.
Nortel, yeah.. or a politician...
Looking forward to seeing the results with the smaller caps anyway. |
|
|
| serengetiplains |
| No, I meant 100VA. The returns line probably begins curving well before 1000VA. |
|
|
| GregD |
| quote: | Originally posted by classd4sure
Hi,
At what point do you suppose bigger caps and bigger transformers passes the point of diminishing returns?
I get the feeling you're there already. |
Probably, but a little overkill is fun sometimes. :smash:
I think I can borrow a couple of 600VA 40+40V ouput transformers from a friend, so hopefully I'll be able to experiment with these as well. It would be nice if the finished amp ends up being smaller than the current configuration but I will not sacrifice any sound quality to do so. I kind of put my experimenting on hold over the holidays. I guess I should get going on this.:) |
|
|
| GregD |
| quote: | Originally posted by serengetiplains
No, I meant 100VA. The returns line probably begins curving well before 1000VA. |
Yep! 1000VA is probably well past the knee in the curve. Just trying to get that last 1% out of it. :) |
|
|
| classd4sure |
Yeh ok,
I woulnd't say you're there yet with a 1000W amp at 100VA though. Regulation factor, current... etc.
I guess what I meant was the extreem end of it, the -3db point of diminished returns, where more gets you nothing .
With a 1000W amp, what do you get with a bigger transformer than 1000VA? Tighter regulation? Probably not.. current? None that you can use..
With caps... 20 000 to 30 000 seems to be the agreed limit, without power factor correction. |
|
|
| serengetiplains |
| quote: | Originally posted by classd4sure
... where more gets you nothing. |
If we're talking diminishing returns, a return = something does not = nothing. Greg, I'm with you, I like that extra 1%. |
|
|
| classd4sure |
Ok,
I would argue that if at 1000VA you dont' have that last 1%, a 2000VA wont' help you either, but it won't hurt anything other than your pocket book and possibly your back.
Now caps are a different story.
Going bigger than required will hurt performance.
So I look forward to the results with the smaller caps. |
|
|
| blaaberg |
>Now caps are a different story.
>Going bigger than required will hurt performance.
>So I look forward to the results with the smaller caps.
Me too. I`ve plan to use 4 x 33.000yF /channel. It`s hard to understand that more yF, should sacrified sound quality.
Venlig hilsen Jan Jensen |
|
|
| tiki |
Hi,
is the difference of the before/after-sounds measurable? THD/IMD vs. frequency? Then try it! You do not need an AP2 for it. IMHO it is easier to believe, what you can see on the paper. And it is easier to establish congruent measuring setups too, a blind A/B-comparison is much more complicated.
Regards, Timo |
|
|
| classd4sure |
Hi,
I'm not saying it would degrade sound quality, but the more capacitance you add for any given transformer, the more your power factor degrades. Too much of a good thing? |
|
|
| tiki |
I agree fully.
Unfortunately I'm not very experienced in SMPS design.
Timo |
|
|
| GregD |
| quote: | Originally posted by classd4sure
Yeh ok,
I woulnd't say you're there yet with a 1000W amp at 100VA though. Regulation factor, current... etc.
I guess what I meant was the extreem end of it, the -3db point of diminished returns, where more gets you nothing .
With a 1000W amp, what do you get with a bigger transformer than 1000VA? Tighter regulation? Probably not.. current? None that you can use..
With caps... 20 000 to 30 000 seems to be the agreed limit, without power factor correction. |
I agree with you about the transformer, but the difference in cost for the extra 500VA was only about $40 per transformer, and I figured this way it should be more than enough for any other project I'm likely to do in the future, and I knew it wouldn't be the limiting factor in the amp.
For me, the -3db point is not good enough. I can hear the difference in two signals 1db to 1.5db apart, so I'm willing to do what I can to get the returns to the -1db point or less if possible.
The stock LC Audio Predator power supply has 20,000uf on it, and it just sounded plain bad to me. Adding another 20,000uf helped a bit, but when I paralleled two of the 82,000uf capacitors, it sounded a lot better, although still only ok, not good. But that's why I want to try the smaller capacitors with the home made PSU, to see if there's an optimum amount of capacitance. |
|
|
| GregD |
| quote: | Originally posted by tiki
Hi,
is the difference of the before/after-sounds measurable? THD/IMD vs. frequency? Then try it! You do not need an AP2 for it. IMHO it is easier to believe, what you can see on the paper. And it is easier to establish congruent measuring setups too, a blind A/B-comparison is much more complicated.
Regards, Timo |
Sorry, I don't have a scope, just a DMM.:sigh:
As for comparison testing, all I'm using or really care about is my ears, since I'm the one who's going to be listening to it. But I'm pretty fussy and I do want the reproduction to be as realistic as possible. I find a blind A/B comparison to be not very effective when evaluating components in my system. Assuming that the component is of a quality close to what I find acceptable, I usually like to live with it for a few days seeing how it sounds with a variety of music and movies, then switch back to my standard system whose pluses and minuses I'm very familiar with. If it's really close, I might switch back and forth a couple of times, but usually I know what I like after the first multi-day trial.
I know this is all subjective and we all have different preferences, but it's all I have to offer. |
|
|
| classd4sure |
Sounds great to me, not my intention to give you a hard time.
I like that you're experimenting and providing results! I just hope it doesn't become a standard that one requires a 2000000VA transformer to get a decent, 10W amp, a la class A, that'd be rediculous!
In your place I would have gone the same way... what's an extra 40$ for 500VA right? Seems crazy not to... I believe we suffer a like illness.
Such a transformer might help offset the size of the capacitors as well (with respect to the power factor). But then again...
"328,000uf of capacitors with bridge rectifiers and some bypass capacitors. With this combination, the amp was finally a contender; separation was excellent (as you would expect), the impact, dynamics, and control were outstanding and superior to the Threshold. The sound is very clean with excellent clarity and a great soundstage, but the highs still seem a bit rolled off; there's not the same airy spaciousness to the sound I get with the Threshold. Also, at medium to high volume levels the midrange gets a recessed sound, almost like it's at the end of a long tube, kind of a weird effect that is hard to describe."
WOW, that's no small amount. So you notice power dropping off at medium to high power levels.... with a tank of a supply like that??? Hmmmmm... Maybe this is the power factor thing creeping up on you. Once it starts to draw it's all reactive(current first .... voltage second... 90 degrees out of phase.. real power =zero...
That's worst case, but you see how it can be a problem. Maybe you can look forward to that going away when you use small caps (please let us know).
I also noticed you used the copper ground bar method as well as a plain star point.... which technique served you better?
Anyway, I really like what you did for a case :) Seems big enough to house everything easily, and you put the money into what really matters, well done. |
|
|
| GregD |
| quote: | Originally posted by classd4sure
WOW, that's no small amount. So you notice power dropping off at medium to high power levels.... with a tank of a supply like that??? Hmmmmm... Maybe this is the power factor thing creeping up on you. Once it starts to draw it's all reactive(current first .... voltage second... 90 degrees out of phase.. real power =zero...
That's worst case, but you see how it can be a problem. Maybe you can look forward to that going away when you use small caps (please let us know).
I also noticed you used the copper ground bar method as well as a plain star point.... which technique served you better?
Anyway, I really like what you did for a case :) Seems big enough to house everything easily, and you put the money into what really matters, well done. |
I'm just a hobbyist, a few EE courses a VERY long time ago, but I don't understand how power factor would be an issue for the DC supply. Can you explain or refer me to something where I can read up on it?
In any case I plan on trying different amounts of capacitance since I've read so many different opinions on diyaudio in both directions. I figure the best thing to do is to just try it, and see what I like, although I would actually prefer it if the best version is able to be made physically smaller. I just started out with the biggest parts first since I knew they would be the most difficult to fit in. Swapping them out for smaller parts later will be no problem.:)
I only used a plain star point. Ericpeters amp is the one with the copper bus bars. Personally, I think either approach should be fine as long as you connect everything to one grounding point on the bus bars. I did the wires primarily so that it would be easier to swap different capacitors in and out. For a final version, I think I'd probably go to bus bars for ease of assembly and looks. :cool:
The case has actually worked out great for putting the amp together: easy to modify, adequate space, no apparent performance compromise, and easy access for getting to and working on all of the parts. For a final version of the amp, I'd like something that looks nicer, but for experimentation, and for anyone who doesn't care too much about the looks, I'd highly recommend it. Besides, it's funny to have what looks like a couple of computers on the floor functioning as amplifiers.
The tower of power! :D |
|
|
| ericpeters |
Hi Greg,
I only have the small caps to bypas the diodes I don't have them on the main rails, never tried it without them. A friend of mine had the same model with lower transformer voltage and normal square bridge rectifiers without caps and this sounded also very good. (He is running Scintilla's in 4ohm, he now has a KSA100 and runs them in .5 ohm setting, he didn't want to risk bypassing the short circuit protection in the modules))
I've got the zap amps on Duetta Signatures, another interesting thing with the Zap monos is that the D-sigs appear to have added another octave on bottom end.
I've got my B&O Ice module stereo amp on a pair of Graz KLM ribbon restored Calipers in my home office. This works well for the purpose but is no contest for the zap's
You can sell me your Diva's, there is no better room than mine to put Diva's I can pick them up myself. I have to be in Southern California anyway next week. :cool: The calipers will fit everywhere. ALtough I'm not sure if the airline will accept diva's as hand luggage.
The zero distortion link is the one I meanth (Designing your own power supply). I followed their calculation regarding the leak-resistor sizing and estimated 10watt would be OK, but I found only these at my local supliier.
I've had discussions with LC-audio regarding the required size of Caps and transformer as well, but listening proved them wrong (or maybe lesser standards)
In my experience size of transformer (= low output impedance of the PS) is most important. And the quality of the of Capacitors.
My theory is that the high efficiency of the modules is the main reason for this: the current they draw from the ps will fluctuate almost linear with the music levels, this will result in an unstable spply voltage if your transformer has not got enough guts. (anyway it is just a theory) |
|
|
| GregD |
| quote: | Originally posted by classd4sure
WOW, that's no small amount. So you notice power dropping off at medium to high power levels.... with a tank of a supply like that??? |
One additional note, I wouldn't say that the power seemed to be dropping off, unless it was just at certain frequencies. The bass in all cases was very impressive, and some music played loud sounded great. It's kind of hard to describe what I was hearing, having never heard anything quite like it before. Also, the lack of airiness, or space, to the sound bothered me at least as much. |
|
|
| serengetiplains |
| Amplifiers can draw very high peak currents which can swamp a transformer's capacity, causing distortion. Any such distortion, of course, is augmented by DC on the AC line. Suffice it to say I see many benefits, including lower DCR, to using so-called over-spec'd transformers. |
|
|
| Luke |
| if capacitance is so important, maybe a capacitence multiplier or voltage regulator would work well. Any one tried this? |
|
|
| GregD |
| quote: | Originally posted by ericpeters
You can sell me your Diva's, there is no better room than mine to put Diva's I can pick them up myself. I have to be in Southern California anyway next week. :cool: The calipers will fit everywhere. ALtough I'm not sure if the airline will accept diva's as hand luggage.
The zero distortion link is the one I meanth (Designing your own power supply). I followed their calculation regarding the leak-resistor sizing and estimated 10watt would be OK, but I found only these at my local supliier.
I've had discussions with LC-audio regarding the required size of Caps and transformer as well, but listening proved them wrong (or maybe lesser standards)
In my experience size of transformer (= low output impedance of the PS) is most important. And the quality of the of Capacitors.
My theory is that the high efficiency of the modules is the main reason for this: the current they draw from the ps will fluctuate almost linear with the music levels, this will result in an unstable spply voltage if your transformer has not got enough guts. (anyway it is just a theory) |
Yes, I think Divas would have to be checked in. Of course the excess baggage charges at about 200 pounds each might be costly. :eek:
Even a 2 watt version of your leakdown resistors should be enough since they should only need to dissipate about .44 watts. Not that the bigger resistors will hurt anything, but you could save some space and maybe some money in the future. From the zero distortion site:| quote: | | You can work out the values yourself, remembering that current equals voltage divided by resistance, and power equals this current times the supply rail voltage. Make sure the power rating of the resistor is at least four times what you will need, because it will tend to heat up quite considerably in the period it takes to discharge the capacitors. I would prefer to see twice that; so, if you have a 2W resistor, work out its value so that current times voltage comes out as no more than 0.4-0.5 watts. |
I basically agree with your theory, but I plan to do the experimentation to prove or disprove it. At lease in my system. ;) |
|
|
| classd4sure |
| quote: |
I only used a plain star point. Ericpeters amp is the one with the copper bus bars. Personally, I think either approach should be fine as long as you connect everything to one grounding point on the bus bars. I did the wires primarily so that it would be easier to swap different capacitors in and out. For a final version, I think I'd probably go to bus bars for ease of assembly and looks. :cool:
[/B] |
OOps, sorry, good point though. Yeh the bus bars are a handsome approach, heard some bad about them, but you're probably right, apply it properly and all.
| quote: |
The case has actually worked out great for putting the amp together: easy to modify, adequate space, no apparent performance compromise, and easy access for getting to and working on all of the parts. For a final version of the amp, I'd like something that looks nicer, but for experimentation, and for anyone who doesn't care too much about the looks, I'd highly recommend it. Besides, it's funny to have what looks like a couple of computers on the floor functioning as amplifiers.
The tower of power! :D [/B] |
Yep, cheap too. Dime a dozen. I put off an amp due to cost and I was factoring in the case with that, it's a big chunk which really can wait. I didn't think of this though. I'm waiting on the day I see someone put an amp in one of those newer modded cases with the see through sides and the lights. It might look sharp and still alot cheaper than a ~500$ designer aluminum job.
| quote: | Originally posted by GregD
One additional note, I wouldn't say that the power seemed to be dropping off, unless it was just at certain frequencies. The bass in all cases was very impressive, and some music played loud sounded great. It's kind of hard to describe what I was hearing, having never heard anything quite like it before. Also, the lack of airiness, or space, to the sound bothered me at least as much. |
I can see that happening, XL is inverse of XC right(with frequency that is)? Maybe ~5Khz is the point at which one swamps the other and your power factor goes right out of wack. I hope someone with more of a clue will step in eventually and correct me..... I'm no expert and as you may have seen it is difficult finding this information for a non switching power supply, but I know it exists, Bruno uses power factor as the reason for not going over 20 000uF. His word is good enough for me, but I would like to know more about it as well. I'm sure there is more to it!
| quote: | Originally posted by serengetiplains
Amplifiers can draw very high peak currents which can swamp a transformer's capacity, causing distortion. Any such distortion, of course, is augmented by DC on the AC line. Suffice it to say I see many benefits, including lower DCR, to using so-called over-spec'd transformers. |
Personally I disagree with the OEM 70% of calculated transformer sizing. I would think 100% of calculated would be just fine, and statistically speaking you're already 30% over-rated. Anything more is wasted unless it gets you tighter regulation. |
|
|
| sven-ake |
| In my zappulse i use a 2*50 volt 1000VA transformer (and it`s huge physically 5 Kg) and a Predator power supply with 15000uf 80V Chemicon capacitors one for each channel.The thing is when i play rather loud but still so you can listen the modules shortcircuts starts blinking and shut down the amplifier.Unfortunally something happened with the right cannel and it`s still blinking very heavily when you start up the amplifier (sent in return to LC audio waiting for a new module) it also killed the woofer in in mine SL:3:s.As i see it the modules did`nt get enough power could it be the softstart,designed for max 1000VA causing this problem or do i need more capacitors.I don`t think the modules should clip so early with this amount of power supply.Anyone have any suggestion to this problem? I don`t want to kill more Woofers and modules. |
|
|
| Luke |
| Has anyone used regulated supply rails? |
|
|
| lawbadman |
Hi Ericpeters,
In post 58 (page 6) you mentioned that you have a B&O Ice module.
Have you done a comparison between the Ice and the Zap?
If you have which amp sounds better?
I am trying to decide whether I should build a zap 2.3se or buy the stero Ice amp from Acoustic Reality.
Thanks
Lawrence |
|
|
| serengetiplains |
| I just posted this question on the UCD400 thread. I'm wondering how these DIY class D amps compare to commercial offerings, including digital-only varieties like Tact? Anyone? |
|
|
| ericpeters |
| quote: | Originally posted by lawbadman
Hi Ericpeters,
In post 58 (page 6) you mentioned that you have a B&O Ice module.
Have you done a comparison between the Ice and the Zap?
If you have which amp sounds better?
I am trying to decide whether I should build a zap 2.3se or buy the stero Ice amp from Acoustic Reality.
Thanks
Lawrence |
Hard to say, because I am not able to do a fair comparison.
Im the current setup the Zap sounds a lot better than the Ice variant i got. But the Zap has a lot bigger power supply. Also the ICE I got costs me about 1/5th of the zap (including everything). On the other hand the first variant I had of the zap amplifier was a lot worse than the current Ice amp I got.
If you spend enough on good and big enough compenents the zap amp is quite easy to get to really good levels. The average beginning DIY-er will not manage to get there with the Ice modules. But a proffesional amp builder like Acoustic reality or the manufacturers of the H2O amp should be able to do that.
From a module point of view in my opinion the zap builders used better components and put more effort in to qualtiy where B&O is much more focussed on a compromise between manufacturing costs and quality. |
|
|
|
