diyAudio Forums Archive - Has anybody experience with the ICEpower ASP modules?

diyAudio Forums Archive > Top > Amplifiers > Class D

Has anybody experience with the ICEpower ASP modules? - Click HERE for Original Thread

ghemink

Hello all,

Has anybody experience with the ICEpower ASP modules?

Jeff Rowland uses them in 10k$ amps, so they can't be that bad.

I have checked the data sheets of these amps. They do some things less nice than the UcD amps. Such as a freq. response that is more dependent on the load and distortion that increases with frequency. I also don't like the fact that they have half the power supply voltage at the speaker outputs.

I have never heard them, anybody any comments?

Best regards

Gertjan

unclemonty

yes, I have heard them and own some, but I have not used them or tested them in a DIY way, yet.

I recently bought a PS Audio GCC500 amp, to replace a Krell KRC HR / FPB600, powering Martin Logan Ascent i speakers. This amp uses two 1000ASP modules. It's a very impressive sounding amp, not necessarily worse or better than the Krells it replaced. Seems to have limitlesss power & authority, and a load of detail too.

Now, I would love to get hold of a further two 1000ASP modules, to bi-amp the bass units of my Ascent i speakers. I've only found one place that will sell them to me, but they are really expensive, 530 Euros each I think. It's almost better for me to buy two eAR 1001 monoblocks, with their nice cases, to acccomplish this.

Does anyone know where or how to buy the ASP modules? I know you can get the A versions easily enough but I want matching ASPs!

Thanks,
unclemonty

ghemink

quote:

Originally posted by unclemonty
yes, I have heard them and own some, but I have not used them or tested them in a DIY way, yet.

I recently bought a PS Audio GCC500 amp, to replace a Krell KRC HR / FPB600, powering Martin Logan Ascent i speakers. This amp uses two 1000ASP modules. It's a very impressive sounding amp, not necessarily worse or better than the Krells it replaced. Seems to have limitlesss power & authority, and a load of detail too.

Now, I would love to get hold of a further two 1000ASP modules, to bi-amp the bass units of my Ascent i speakers. I've only found one place that will sell them to me, but they are really expensive, 530 Euros each I think. It's almost better for me to buy two eAR 1001 monoblocks, with their nice cases, to acccomplish this.

Does anyone know where or how to buy the ASP modules? I know you can get the A versions easily enough but I want matching ASPs!

Thanks,
unclemonty

Hi Unclemonty,

Thanks for the info, yes the EAR cases look nice. Maybe we get some more replies of people that have heard both UcD and these ICE modules. Access to UcD is easy and they are not expensive, they are very nice for DIY. Just would like to know how these ICE amps compare with them.

Best regards

Gertjan

Septimus

Just happened across this thread.

I work for Bang & Olufsen. I'm into DIY audio as a hobby despite working for B&O. We invented the ICEpower technology and license it out to other companies. ICE stands for Intelligent, compact, efficient FYI.

If you ever want to hear these amps, stop by a store sometime, bring a favorite CD of yours and ask to listen to the $17,000/pr BeoLab5's. These have digital input, analog devices DAC's, room correction for the bass freq. with built in mic and active XO's. They have 4 ICE amps in them, 250W on the 1" tweet, 250W on the 3" dome midrange, 1000W on a 6.5" midbass, and another 1000W on a down-firing 15" sub. Even the power supply uses the ICE technology in some way i'm told!

They're quite interesting, and they SLAM with an immediecy i've not heard from any other speaker, you should give them a listen

Yves Smolders

The sweetest thing about ICEpower modules is the ASP series, if you can get your hands on these puppies it's plug&play, power supply in incorporated in the modules.

It would be so nice to have a small efficient switching power supply to feed UcD modules - sure the modules are small, but the toroids needed are *huge* !

phase_accurate

quote:

If you ever want to hear these amps, stop by a store sometime, bring a favorite CD of yours and ask to listen to the $17,000/pr BeoLab5's.

Thanks for the short commercial, but where can we listen to the amp technology in a DECENT system ?? ;)

What is the minimum amount of modules that have to be ordered that B&O wants to make a deal at all ?

Regards

Charles

unclemonty

You can buy them here: http://cgi.ebay.co.uk/ws/eBayISAPI....ssPageName=WDVW

but it's way too expensive for something that costs <$100 OEM.

Perhaps Karsten at http://www.cadaudio.dk/ might be able to obtain them as he seems to be able to source the A versions of the modules?

Karsten?

unclemonty

ghemink

quote:

Originally posted by unclemonty
You can buy them here: http://cgi.ebay.co.uk/ws/eBayISAPI....ssPageName=WDVW

but it's way too expensive for something that costs <$100 OEM.

Perhaps Karsten at http://www.cadaudio.dk/ might be able to obtain them as he seems to be able to source the A versions of the modules?

Karsten?

unclemonty

Hi Unclemonty,

I had a closer look at the cadaudio link. I'm not really considering buying ICEpower modules at this stage, just interested in whether somebody knows how they perform (the new ASP series), especially in comparison with UcD.

I find the switched power supplies on the CADaudio site however very interesting. It seems that they only sell them in combination with an amp. I was thinking of using a switched power supply for my UcD modules. The switched mode supply would be regulated and the power factor could possibly be better than with a linear supply. A linear supply with large caps will draw very large current peaks during a very short time (at the peak of the sine wave, 100 times per second). This maybe better (less) with a switched supply. And since the amp is already switched, why not using a switched supply?

Maybe a good idea for Hypex to develop switched supplies as well?

Does anybody know of other vendors for switched power supplies for audio applications?

Best regards

Gertjan

Septimus

Well, believe me, they are more than decent, just didn't know if anybody knew where these amps came from and he had asked where he could here some. It wasn't a commercial, if I thought they were the end-all be-all I wouldn't be doing the DIY thing as a hobby.

Septimus

Oh, and I have no idea what amount you would need to purchase, but its more on the order of like, buying the rights to manufacture yourself, B&O wouldn't sell any modules themselves.

phase_accurate

quote:

B&O wouldn't sell any modules themselves

But it seems that some manufacturers do indeed use ICE-power modules manufactured by B&O themselves (like the aforementioned amps and also P.A. companies like EAW).

regards

Charles

Jan-Peter

quote:

Oh, and I have no idea what amount you would need to purchase, but its more on the order of like, buying the rights to manufacture yourself, B&O wouldn't sell any modules themselves.

I don't think you can get the rights to produce the ICEpowers by yourself. As far as I know ICEpower has a plant in Malaysia for producing the ICEpowers........

Regards,

Jan-Peter

Yves Smolders

I'd love to find a nice switch-power supply also. The UcD's are very small and compact, and our company would like to test them as amplifier for a high-end(ish) home theater PC. Only thing that wouldn't fit in a PC case is the toroid/caps. Switchmode would be the solution, if we only could find a decent DIY kit or completed supply we'd be fine!

I didn't have the chance to test the modules yet, as I haven't received them. Apparently Hypex has big problems with GLS, they were due to bring us the modules on monday.

GLS was just here with a package from another company, still no Hypex delivery...

Yves

Jan-Peter

quote:

Apparently Hypex has big problems with GLS, they were due to bring us the modules on monday.

We do not have problems with GLS, the problem is GLS has a new driver in your location in Belgium.............

Don't worry the best things comes slowly :D

We are looking for a SMPS solution, see above ;)

Jan-Peter

Yves Smolders

Jan-Peter,

This wasn't meant as an "attack" towards you or Hypex. I'm just tired of transport companies.

We deliver PC's to end-users and know how these things go...

So no hard feelings towards you or Hypex, but I can't vent my anger towards GLS.

Poor Niels, I've sent him quite some mails lately... :D

I hope we all can figure out what went wrong

I'll shut up now, this is not the discussion on this forum :angel:

classd4sure

"Custom Design

ICEpower offers to design according to your specifications and requirement for high volume projects. Our team of designers has extensive experience with design from specifications all the way to 100% approved solution ready for manufacturing.

Our design experience ranges from for pro audio & consumer electronics to automotive applications. ICEpower offers to manufacture custom designs as a turnkey solution or alternatively on a production license."

"Optional production license
Optional production license is offered for high volume projects to reap the cost benefits of optimized logistics. For production license projects ICEpower delivers a prototype ready for EMC & safety approval."

I read on another site which isn't affiliated to B&O that a min. of 1000 would be required for them to produce them for you.

The problem with Icepower aside from it being unobtanium, and too expensive, is its load dependance.

As I understand it, a SMPS would still require beefy caps for energy reserve, but would save on the transformer size.

It's only less expensive from a mass production view point. If you're not into making supplies, you have little to save by producing a SMPS, while the extra design time/costs would only be passed on to the user. They are far less reliable than your average power supply, and would be troublesome to use properly, much more so than a standard PSU.

Personally, I'd prefer Hypex's time was spent further improving their UCD modules, and releasing new ones, than amping up the cost of them by implementing a sub optimal supply.

I never want to see the thread titled "How to mod your UCD module for true power and quality by removing the SMPS"

Regards

Yves Smolders

Too expensive, well...

Acoustic Reality

Then click "Order Info":

quote:

eAR� 202, 2x200 Watt RMS - US$499 (ex. VAT), EURO 399 (with VAT), 2999 DKR (with VAT)

This is actually 2x100W into 8 omhs, based on ICEpower ASP modules, fully assembled.

So if someone wants a hold of the modules, this might be a way to get them... it will be ASP's of course.

Then I saw the Hypex modules... so I never ordered those eAR's!

Samsung has also announced a 7 channel receiver with all whistles & bells present, with 7*200W/8 ohms based on ICEpower, THX Ultra2 certified (THX might have lost its good name, but the Ultra2 specs are quite difficult to live up to)

Yves

unclemonty

Yves - this one you mention is based on the cheaper ASC module. Probably not equivalent with the 'full' ASP series.

unclemonty

Jaka Racman

Hi,

I am a bit curious about real power capability of those modules without additional heatsinking. In SMPS business I am accustomed that 1000W power supply gives 1000W continously up to the maximum rated ambient temperature. Now 1000W power supply that gives out 40W at maximum ambient temperature is in my book a 40W power supply and not a 1000W one. See attached part of 1000ASP module data sheet.

I know that FTC has now lowered preconditioning from 1/3 rated power to 1/8 rated power and I think that rated power is tested for 5minutes, but i would like to have your opinion what is a realistic continious power requirement for audio SMPS.

Best regards,

Jaka Racman

Yves Smolders

Somebody should set up a nice "open standard" for all these things.

A public domain THX certification of sorts, maybe a kind of "benchmark" for amps and other audio equipment, purely based on things that can be measured.

All this power nonsense is driving me nuts. Does anyone even know what 1W is these says? Or 50? Or 100?

Good thing the PMPO craze is over. If they would be measuring car perfomance like they measure audio performance, we would have amazing specs :bigeyes:

"Yes sir-eee, 14.000 revs peak engine power output, just before she blew up"

But even "RMS" power can't be trusted these days...

:rolleyes:

classd4sure

Hi Jaka,

Wow, thank you.

If it is a thousand watt amp, then it should obviously be expected to sustain 1000W continuously.

Everytime the thousand watt capability of these modules is referenced, everyplace I've seen, it _does_ say "continuous" beside it, but they always leave out the "for 15 seconds, after it's taken out of the freezer" aspect.

These figures seem shockingly low to me, and I was expecting something like this.

40W continuous at 50 deg. C ( a reasonable in-case temp) is a far cry from 1000W, not even close, I wonder what their 200W amp is? Can you show us that same part of the data sheet for that one as well?

It does say it can be increased with heatsinking or forced air, but I think the only way to actually get close to 1000W out of it for any length of time would be liquid nitrogen.

Who wants to use fans and big honking heat sinks just to get a reliable ~100W of power anyway, may as well go with class B right, hell, may as well go with a real supply. One has to note the figure of it's max capability with all the additional cooling isn't given either.

Such embellishment forces me to ask how far they stretched the truth on all their other specs as well.

For instance, what's with the FTC rating?? Bandwidth limited to 3Khz, by protection that..... isn't normally in the amp? So they squeeze a little more power out? Isn't FTC test supposed to be "across the rated bandwidth (so it should be up to 20Khz that's what they said it is), at the specified ratings". I guess they only specificed all the ratings up to 3Khz? Something stinks.

Now, be serious, do you really want your UCD modules rated this way? The cost of a real SMPS that can "do" the job would be insane.

Regards

Jaka Racman

Hi Chris,

datasheets are on their web page. Here is 1000ASP, here is 200ASC and here is 250ASP .

It is interesting that 250W module has almost the same continious ratings as 1000W one.

But the most shocking digram is from the 200ASC datasheet. It shows how do they dimension Zobel network. That is also the reason why 1000ASP has FTC rating only up to 3kHz. See attached picture.

Best regards,

Jaka Racman

classd4sure

Hi,

...and it all comes crashing down. Thanks for those links Jaka, I hadn't seen those before. I'm really taken aback by those.
I'm surprised they haven't yet evolved beyond the obvious problems they have, too busy marketing it perhaps.

At great restraint that's all I'm going to say about it, but it looks as though if anyone truly wants an amp capable of 1000W RMS it's a ZAPpulse for you, for now.

Best regards,
Chris

ghemink

If I read all this, it seems that people are not so enthusiastic about switched power supplies.

Why is that? The reason why I consider them is as follows:

1. better power factor, a linear supply will draw high peak current, especially when big caps in the power supply are used. I live in a 100V country, so those big current peaks can be an issue and limiting for the performance
2. regulated output voltage, so even if my 100V that I get from the outlet is not that stable or hard, I will still have a pretty stable supply voltage for the power amps
3. I have to feed many amps since I use them in an active system, this means many power supply caps that need to be charged, and therefore, high current peaks as mentioned in 1 as well, when a linear supply is used

Actually, I`m also considering to make a regulated supply (conventional) using the THEL regulator modules. However, those modules may not be so easy to get since I`m located in Japan, can`t order with credit card and so on. What I would do with those modules is to have relatively small caps in front of the regulator and a 20V higher input voltage than needed at the output. The voltage before the regulator will have quite a lot of ripple because the caps would be relatively small, however, that ripple will be regulated away. So after the regulator, I can have bigger caps and a nice stable voltage that is almost load and AC outlet independent. That way, I can again reduce the peak currents that are drawn from my wall outlet.

Any comments on the above ideas? Any way to get some SMPS for experimental purposes? An easy way to get those THEL regulators? When can we have an Hypex switched power supply? These could be as simple as a bridged UcD module, if the only objective is to have a regulated supply (will have no power factor correction) or do I see these things to simple.

Best regards

Gertjan

phase_accurate

It IS possible to make decent SMPS but this is definitely not an easy task.
I once heard an amp (that by itself was class AB) that is fed bay a beefy SMPS and this thing had steam at the low end while sounding very effortless and clean. It was a Chord BTW.

Regards

Charles

wiligruen

Anybody have an idea what Jeff Rowland is doing with the ICEpower 1000ASP to achieve such specs?

Model 501 Specs

Output Power
Continuous RMS

8 ohms 500 Watts
4 ohms 1000 Watts

see

http://www.jeffrowland.com/Model%20501%20Page.htm

FuriousD

quote:

Originally posted by phase_accurate
It IS possible to make decent SMPS but this is definitely not an easy task.
I once heard an amp (that by itself was class AB) that is fed bay a beefy SMPS and this thing had steam at the low end while sounding very effortless and clean. It was a Chord BTW.

Regards

Charles

Ok, I'll allow many things to pass, but not this. Good Bass, from a Chord? You've got to be joking. They have no bass and an ear bleeding top end.

Take it from someone with more experience with Chord than anyone else araound ...They use a very noisy unregulated SMPS that has over 4uF of X capacitance across the mains to snub it's own switching noise. The AB amplifier itself is not biased properly. in fact, the bias changes with PSU fluctuations (Ever heard of a VBE multiplier... They haven't!) causing amplifitude modulation distortion. The output MOSFET's are biased at a ridiculous 50mA for 32 devices. (Now there is no way that is class AB) and the units are usually shipped with nearly 2v of HF oscillation. Above 100kHz the amplifier distortion is so bad that it folds down at a near 1:1 ratio, so good luck with an RF nearby!

Charles, if you heard good bass on a Chord, then please, don't ever trust your ears again!

unclemonty

"...and it all comes crashing down"?

You've written off the ASP technology before even testing it? That's funny!

jazzy

We are using icepower modules( 250A, 500A and 1000A) in our production, almost one year with " zero defect"!
rgds
jazzy

classd4sure

quote:

Originally posted by unclemonty
"...and it all comes crashing down"?

You've written off the ASP technology before even testing it? That's funny!

I'm pleased that you're amused.

I'm amused that you want to bi-amp these "1000W RMS" "load independant" modules, should get you a clean 250W for at least 20 seconds, if you open the window, for a subwoofer anyway. Now _that_ is amusing.

Further to my amusement, statements like "Seems to have limitlesss power & authority, and a load of detail too." Just read the data sheet and tell me how limitless the power really is. I know, "seems" was the key word right? So it has a bit of dynamic headroom, so what, that term went out of style right along with PMPO did it not? In reality, it can't even come close to half of what their rating claims, let alone across the audio band. So yeah, I'll write that off and never look back. I think you're going after a hugely expensive and sorely dissapointing 300W at best.

Gertjan,

No problems with a SMPS here, at least not if they actually deliver as claimed, like they should, it doesn't save on caps apperently, just on size and weight of transformer. I just wouldn't want to see it come on the module, possibly limiting performance (as below), certainly limiting peoples choices, and causing the price to skyrocket, inversly to reliability, which will plummet. That's only my opinion.

Maybe as an additional module it would be alright, but it might be hard to use properly that way. I know Bruno got funding for working on a real SMPS for audio, his opinion was that non existed that can really do the job, would be most interesting to see what his results were one way or the other.

I support your DIY supply effort.

I'll have to read that SMPS thread, there's alot of ideas in there.
Lots of them with respect to PFC as well. Analogspiceman even provided a few circuits with the UCD in mind.

Regards
Chris

unclemonty

Jeez - this is a tetchy forum! I was only expressing an opinion in reply to the original poster, and comparing sound quality (not measured performance) to Class AB stuff I owned before. So, I did feel partly qualified to speak, as an actual owner of IcePower units, if not a tester. Never mind...

Hows about this then - there's a guy who sometimes posts here, DM Fraser, who actually uses the IcePower units for his companys' products. Surely he'd be a good person to ask about the measured performance?

My entire reason for wanting this kind of technology in my house was to reclaim real estate that large Krell amps had been taking up, to lower my electricity bills, and to keep the house cooler in summer. Oh, I wanted good sound quality too! All of that has been achieved & I'm happy.

I'd still like to bi amp with another set of 1000ASP despite the previous posters' derision for that idea. If anyone tests the 1000ASP then that's great I think. Ultimately if they didn't meet the spec I would probably still go ahead anyway, if the price were right.

From a customers perspective it seems good, successfully implemented technology - I'm happy with the sound, as a former (and current) Krell owner. I guess if some of you are in the industry then the measured performance ultimately is whats important. Whether it's better/worse than UcD/Intersil et al, who knows? If I could buy two boards at a good price I'd be more than happy to let someone here measure & test them as long as I got them back undamaged afterwards.

Regards

unclemonty

classd4sure

Hello,

Perhaps I was too harsh? You're perfectly qualified to speak, I never implied you weren't.

It's good to know they sound good, I'm sure they do, and not having heard them, will not dare to say otherwise.

The reason I'll personally turn my back on them is because, according to their own data sheets, they perform far from advertised.

Really, so you have an integrated power supply, let's say with a huge heatsink, that's fan cooled, you can actually squeeze 1000W continuous out of it. You just negated all its benefits(small size, lightweight, efficiency) what's the point?

Their problems with the zobel network... I think, seriously require some attention? Surely Karsten Madsen is capable of addressing these issues, or comming up with an equally good sounding design that doesn't require one, he certainly is no fool.

In the past he's done extensive research, I'd hate to think it ends at this, but I don't see the new versions being improved upon. Why? Maybe because people are buying it as it is, that's just my theory.

I think it's great you gave class d a try, but this isn't the only option out there, and if you compare it to other technologies, it seems to come up short, so with that in mind, I don't really have to hear it to think it sucks, once again, not at all refering to the sound quality of it, just their (false) marketing BS, which I have no tolerance or respect for.

Regards,
Chris

TNT

quote:

Originally posted by classd4sure
....

....., I don't really have to hear it to think it sucks, once again, not at all refering to the sound quality of it, just their (false) marketing BS, which I have no tolerance or respect for.

Regards,
Chris

This is, hidden, but anyway, marketing !!!! Be ware !

/j

ghemink

quote:

Originally posted by classd4sure

I'm pleased that you're amused.

I'm amused that you want to bi-amp these "1000W RMS" "load independant" modules, should get you a clean 250W for at least 20 seconds, if you open the window, for a subwoofer anyway. Now _that_ is amusing.

Further to my amusement, statements like "Seems to have limitlesss power & authority, and a load of detail too." Just read the data sheet and tell me how limitless the power really is. I know, "seems" was the key word right? So it has a bit of dynamic headroom, so what, that term went out of style right along with PMPO did it not? In reality, it can't even come close to half of what their rating claims, let alone across the audio band. So yeah, I'll write that off and never look back. I think you're going after a hugely expensive and sorely dissapointing 300W at best.

Gertjan,

No problems with a SMPS here, at least not if they actually deliver as claimed, like they should, it doesn't save on caps apperently, just on size and weight of transformer. I just wouldn't want to see it come on the module, possibly limiting performance (as below), certainly limiting peoples choices, and causing the price to skyrocket, inversly to reliability, which will plummet. That's only my opinion.

Maybe as an additional module it would be alright, but it might be hard to use properly that way. I know Bruno got funding for working on a real SMPS for audio, his opinion was that non existed that can really do the job, would be most interesting to see what his results were one way or the other.

I support your DIY supply effort.

I'll have to read that SMPS thread, there's alot of ideas in there.
Lots of them with respect to PFC as well. Analogspiceman even provided a few circuits with the UCD in mind.

Regards
Chris

Hi Chris,

Thanks for this reply. The power factor correction is the main thing I was after with a switched power supply, and of course the regulation. I don't mind using big transformers and big caps, but I know that the 100V we have here in Japan is not that solid. I see my lights go down a bit when I turn on my Accuphase E407, and I'm quite sure that thing has some sort of inrush limiter. I was actually thnking about buying an active power conditioner (like Accuphase has) that can supply the huge peak currents that occur at the top of the rectified sinewave. This will help me as well but will be far more expensive than a SMPS.

ANyway, maybe I'm too much of a perfectionist. As I use my UcD180 modules right now, they sound very good, so probably I should not worry too much.

Best regards

Gertjan

Jocko Homo

I don't know why you guys are obsessed with a module that is OEM only. Doesn't matter if it works or not.............it is not for DIY.

We have member here who does make modules for DIY'ers. Stick with his, and stop kvetching.

Jocko

Pierre

I have just seen the graphs from B&O showing the max. power vs. time at high (10,15,20KHz). I assume that's due to power dissipation in the Zobel network resistor, isn't it?

Someone pointed to the way they design the zobel network, please, could you tell me where to see that info?

Best regards,
Pierre

Jaka Racman

Hi,

There is no info how they design Zobel. I would not care for that either, given the results. IMHO in properly designed classD amp there is no need for Zobel.

Only info is on page 6 of the ASC200 data sheet:

quote:

5) The power bandwidth is limited due to the Zobel-network. (See further details on page 11-12)

.

Best regards,

Jaka Racman

dmfraser

The ICE Power 1000ASP uses a full H bridge on the
primary. The regulator is UCC3808 from TI and others. The IC drives a pair of complementary current boosters which then drive a dual FET for the lower pair and a 3 winding switching transformer as a level translator for the upper pair of FETs.

The drain leads of the upper FETs feed through two windings of a 3 winding transformer on their way to the 120V supply. This is a current monitor for protection purposes.

On the output side, the winding for the main 120V supply has a tap that is connected through an inductor to the 80V output terminal. The 80V output has a 680�Fd 100V cap on it. The 120V supply which is also available on an external connector has three 2200�Fd 160V caps in parallel.

The main switcher also has a winding for the �12V outputs and a low voltage winding to supply power to operate the UCC3808.

There is also an output winding with a center tap that appears to not be used. Possibly the transformer is meant for additional models.

The switching transformer core in the 1000ASP is made by Kasche core size ETD44 which is rated for some 800W+ continuous. The power supply is meant to drive the internal 1000W power amp plus a n external 1000A to full power, with music but not sinewaves.

dmfraser

I operated a sample 1000ASP on the bench delivering 350W average of pink noise into a 4 ohm load for over one hour with no additional heatsink and the metal case stayed below 55�C.

However, higher levels would make the power supply voltage go down to act as a thermal compensation. Much nicer than just shutting down.

However, with 1214W of sine wave, partly into clipping, after about 35 seconds, the output level would drop to about 600W by the protection circuitry. Remember this is with no additional heatsink.

The 200ASC, I was able to operate with sine wave at 100W into 4 ohms continuously. Turning it up, with sine wave, the module would shut down at about 214W.

However, with pink noise at 100Waverage and a 6 db crest factor, the 200ASC would run all day without shut down, until I was tired of tying up one of my Audio Precisions with it. It would also run all day with music, driven hard where the output limiter was set to a peak limit of 210W maximum. That is, the limiter would allow small amounts of clipping. Driven so the limit light in our external optical limiter was on essentially continuously.

We're quite satisfied and will be ordering production quantities of the 200ASC shortly.

dmfraser

I also just got a sample 500W per channel 2 channel module from Powersoft with integrated power supply with PFC to test. It also has a socket for a DSP to be plugged into it. I'll be looking at this shortly and from the looks of it, I expect good performance from it. I'll start a new thread for the Powersoft one when I have tested it.

classd4sure

Thanks for the additional info on ICEpower, and I'm really looking forward to reading about your experience with the Powersoft module.. too bad it's another unobtainable one to the DIY'r, will be nice hearing about it anyway.

dmfraser

If anyone will be in Orlando from March 9 to 12, I will be at the NSCA show, in the Renkus-Heinz demo room. Also at the show will the people from Powersoft, at the Martin Audio booth. Maybe a DIYer from Florida can try to find if they will release product to a dealer for the DIY market.

ghemink

quote:

Originally posted by dmfraser
I operated a sample 1000ASP on the bench delivering 350W average of pink noise into a 4 ohm load for over one hour with no additional heatsink and the metal case stayed below 55�C.

However, higher levels would make the power supply voltage go down to act as a thermal compensation. Much nicer than just shutting down.

However, with 1214W of sine wave, partly into clipping, after about 35 seconds, the output level would drop to about 600W by the protection circuitry. Remember this is with no additional heatsink.

The 200ASC, I was able to operate with sine wave at 100W into 4 ohms continuously. Turning it up, with sine wave, the module would shut down at about 214W.

However, with pink noise at 100Waverage and a 6 db crest factor, the 200ASC would run all day without shut down, until I was tired of tying up one of my Audio Precisions with it. It would also run all day with music, driven hard where the output limiter was set to a peak limit of 210W maximum. That is, the limiter would allow small amounts of clipping. Driven so the limit light in our external optical limiter was on essentially continuously.

We're quite satisfied and will be ordering production quantities of the 200ASC shortly.

Hello DMfraser,

I was waiting for you to reply :-)

Any comments on how they sound in comparison with UcD?

Thanks and best regards

Gertjan

Jaka Racman

Hi Dan,

thanks to your report, B&O has regained some of the respect back. It is nice to see that they underrate their products in the data sheet by a significant number. However I am interested if you had any problems with Zobel networks. I expect that it would be quite usual that some prior stage of amplifcation (not necessarily ICE module) would burst into high frequency oscillation when some inadvertly puts finger into the input.

Best regards,

Jaka Racman

dmfraser

I have a pair each of UcD180 and UcD400 which Jan Peter kindly sent me for evaluation. I find both the UcD and ICE power units to have an excellent sound quality. However, they do NOT sound the same. Both excellent but distinct.

The UcD amps sound closer to an excellent class AB amplifier while the ICE sound slightly "tubey".

I think this is partly because the UcD is a half bridge design and does an excellent job simulating an excellent class AB transistor design. Basically a near perfect reproduction of the audio in as clean and essentially unbiased sense. Where the amplifier imparts very little of its own sound in the signal. Nominally the goal we are striving for. Not perfect but as close as the musical truth all but the most esoteric high end amplifiers of say 1990 could give. And with a lot more efficiency and in a much smaller package. And available to the DIY community at a reasonable price.

On the other hand, the ICE Power is a full bridge design and seems to come closer to simulating the sound of a very, clean tube amplifier because the bridged output stage does a great job in cancelling out harmonics like a tube amp does. As well, their soft limiting circuit makes the sound, when clipped, more like the sound of a tube amp in clipping. Imagine the sound of a tube amp would give if you had a perfect output transformer with zero loss, no hysteresis and near perfect frequency response. Again very good and the sound that many people are looking for. However, for the small quantity user, essentially unobtainable.

And for those who might ask, I do work for an OEM with access to all these unobtainable OEM products, the agreements with the module makers do not allow resales other than in manufactured products. Service spares are on an exchange basis only. SO please, don't ask to buy these from me. I only run the engineering department. I don't make policy.

As for the Zobel networks, in the ICE 500A and 1000A, in early experimenting, I did blow the cap off the TO-220 4 ohm resistor in the zobel network of a couple of these before I learned to NOT to sustained high frequency testing. These amplifiers are for audio only, not for sine waves. The problem is not the amplifier but on the physical size. To get the module under 100mm square, this means the zobel resistor must be smaller. They sized it for music. To sustain a 15KHz full power sine wave, would require a far larger part, one that is not necessary for audio work. This is called designng the product to fit the application. Do not fault the product for being able to blow up the resistor in the zobel network by driving it at full power with a sine wave over 10KHz. We don't listen to sine waves. We listen to a complex signal woth a varied spectral content, with less energy in the high end than the low, called "music".

And yes, any product will make noise if you put your finger on the input. Some may even oscillate. That's the price you pay for having good transient response. My recommendation is to not have the inputs floating in the air. When nothing is connected anyway, leave it turned off. Turn it on only when its all connected up.

ghemink

quote:

Originally posted by dmfraser
I have a pair each of UcD180 and UcD400 which Jan Peter kindly sent me for evaluation. I find both the UcD and ICE power units to have an excellent sound quality. However, they do NOT sound the same. Both excellent but distinct.

The UcD amps sound closer to an excellent class AB amplifier while the ICE sound slightly "tubey".

I think this is partly because the UcD is a half bridge design and does an excellent job simulating an excellent class AB transistor design. Basically a near perfect reproduction of the audio in as clean and essentially unbiased sense. Where the amplifier imparts very little of its own sound in the signal. Nominally the goal we are striving for. Not perfect but as close as the musical truth all but the most esoteric high end amplifiers of say 1990 could give. And with a lot more efficiency and in a much smaller package. And available to the DIY community at a reasonable price.

On the other hand, the ICE Power is a full bridge design and seems to come closer to simulating the sound of a very, clean tube amplifier because the bridged output stage does a great job in cancelling out harmonics like a tube amp does. As well, their soft limiting circuit makes the sound, when clipped, more like the sound of a tube amp in clipping. Imagine the sound of a tube amp would give if you had a perfect output transformer with zero loss, no hysteresis and near perfect frequency response. Again very good and the sound that many people are looking for. However, for the small quantity user, essentially unobtainable.

And for those who might ask, I do work for an OEM with access to all these unobtainable OEM products, the agreements with the module makers do not allow resales other than in manufactured products. Service spares are on an exchange basis only. SO please, don't ask to buy these from me. I only run the engineering department. I don't make policy.

As for the Zobel networks, in the ICE 500A and 1000A, in early experimenting, I did blow the cap off the TO-220 4 ohm resistor in the zobel network of a couple of these before I learned to NOT to sustained high frequency testing. These amplifiers are for audio only, not for sine waves. The problem is not the amplifier but on the physical size. To get the module under 100mm square, this means the zobel resistor must be smaller. They sized it for music. To sustain a 15KHz full power sine wave, would require a far larger part, one that is not necessary for audio work. This is called designng the product to fit the application. Do not fault the product for being able to blow up the resistor in the zobel network by driving it at full power with a sine wave over 10KHz. We don't listen to sine waves. We listen to a complex signal woth a varied spectral content, with less energy in the high end than the low, called "music".

And yes, any product will make noise if you put your finger on the input. Some may even oscillate. That's the price you pay for having good transient response. My recommendation is to not have the inputs floating in the air. When nothing is connected anyway, leave it turned off. Turn it on only when its all connected up.

Hi Dan,

Thanks for your comments, very usefull. I would not ask you to supply me with ICE modules. One way to get them is from Acoustic realitily, of course then they are encased and of course more expensive than UcD modules. Maybe one of the differences between UcD and ICE is caused by the higher output resistance of ICE at higher frequencies, giving a smoother sound? I agree with your comment on listening to music and not to sine waves. However, I like it that the UcD does not need the zobel, it seems that class D amps with a burned out zobel can be pretty dangerous.

Basically I was interested in how these ASP amps sound with that SMPS in comparison with ICE with a conventional supply. I would like to experiment with UcD and SMPS but I can't find anything suitable and I don't know if SMPS would degrade the sound. However, if it works well on ICE, a well designed SMPS should work well on UcD as well.

Thanks for your feedback and best regards

Gertjan

dmfraser

The 1000ASP had a tighter bottom end with somewhaty more authority at higher power levels than the 1000A with a linear power supply. I suspect this is because the power supply on the 1000ASP is fully regulated while the linear power supply tended to droop under load.

HF performance could not be compared as the 1000A is limited to 3 KHz while the 1000ASO is a fukk range device.

However, neither had quite the overall punch of an Renkus-Heinz P3500 class AB amplifier on some overprocessed modern music. I suspect this is caused by the somewhat higher output impedance of the ICE amplifier compared to the P3500 caused by the output inductors. The network impedance is low on the ICE but it has an effect. However, unless you are playing "boom car" music I feel the amount of punch on the ICE is within the needs of any sound played on an actual instrument. It is just there is a difference on notes produced electronically.

The UcD has decent bottom end as well. However, since I was not able to test it with an SMPS and the UcD180's power was about 1/8 that of the P3500, comparisons are not really meaningful. For its power, and the unregulated power supply used, the UcD180 had good punch and again, great for any sort of natural or "real" music.

ghemink

quote:

Originally posted by dmfraser
The 1000ASP had a tighter bottom end with somewhaty more authority at higher power levels than the 1000A with a linear power supply. I suspect this is because the power supply on the 1000ASP is fully regulated while the linear power supply tended to droop under load.

HF performance could not be compared as the 1000A is limited to 3 KHz while the 1000ASO is a fukk range device.

However, neither had quite the overall punch of an Renkus-Heinz P3500 class AB amplifier on some overprocessed modern music. I suspect this is caused by the somewhat higher output impedance of the ICE amplifier compared to the P3500 caused by the output inductors. The network impedance is low on the ICE but it has an effect. However, unless you are playing "boom car" music I feel the amount of punch on the ICE is within the needs of any sound played on an actual instrument. It is just there is a difference on notes produced electronically.

The UcD has decent bottom end as well. However, since I was not able to test it with an SMPS and the UcD180's power was about 1/8 that of the P3500, comparisons are not really meaningful. For its power, and the unregulated power supply used, the UcD180 had good punch and again, great for any sort of natural or "real" music.

Hi Dan,

Thanks again for the very useful comments.

Kind of have given up on using a SMPS.

I now plan to use two UcD400 bridged for the woofers. Not only to get more power. The main reason is that in a bridged situation, the power will be drawn from both rails at the same time. As a result, one rail needs to supply only half the current in comparison with a single module unbridged delivering the same power. The current consumption on each rail would be more continuous since each rails needs to deliver power during both the positive and negative part of the sinewave. There would also be no power supply pumping and the + and - rails always see the same load, so + and - power voltage would always be the same. Also the transformer may like it to always see a symmetrical load. This may also make it easier to use a capacitance multiplier like stabilizer.

Thanks again and best regards

Gertjan

Alfetta87

quote:

Had the chance to compare directly an ICE against an UcD (180, older version), and the verdict is clear: go for the UcD's.
The first test with the older UcD180 (driven with a single tranny) gave also better results in imaging (localisation of voices and instruments than the ICE.

I ordered some UcD 180 and 400 with the AD8620 OpAmps (fast delivery despite GLS...), hooked up already the UcD 400 and sold the ICE (ASP500)...
Tighter bottom in the ICE: to me the ICEpowers lack some bass, as well as some warmth in the lower midrange. The UcD 400 has it all, also a sweet HF.
The only thing that I could say against the UcD is that the ICE sounded faster. Also, both are similarly (and unbelievably) transparent in the mids.

Beside this, price is very reasonable, especially if you compare to finished products with similar performance, where you'd spend probably a couple Grands.

Indeed, the happiness would be perfect if some switching power supplies would be available: amateurs, if you know some...

Dan

IVX

quote:

Originally posted by ghemink

Maybe one of the differences between UcD and ICE is caused by the higher output resistance of ICE at higher frequencies, giving a smoother sound?
Gertjan

I think so too. My buddys, after listening the UcD180, said, that this is an inertialless amp, and even too much (actually, i don't think so).. For the "tube smoothing" simulation, as easy way, we can try to add resistor or a coil to the output. Although, U-source it's a still standard requirement for audio amplifiers, but we also have a choice between true fidelity and more pleasure.

ghemink

quote:

Originally posted by Alfetta87

Had the chance to compare directly an ICE against an UcD (180, older version), and the verdict is clear: go for the UcD's.
The first test with the older UcD180 (driven with a single tranny) gave also better results in imaging (localisation of voices and instruments than the ICE.

I ordered some UcD 180 and 400 with the AD8620 OpAmps (fast delivery despite GLS...), hooked up already the UcD 400 and sold the ICE (ASP500)...
Tighter bottom in the ICE: to me the ICEpowers lack some bass, as well as some warmth in the lower midrange. The UcD 400 has it all, also a sweet HF.
The only thing that I could say against the UcD is that the ICE sounded faster. Also, both are similarly (and unbelievably) transparent in the mids.

Beside this, price is very reasonable, especially if you compare to finished products with similar performance, where you'd spend probably a couple Grands.

Indeed, the happiness would be perfect if some switching power supplies would be available: amateurs, if you know some...

Dan

Hi Dan,

Good info. UcD is a real bargain. In Japan you have to pay about 10000$ to get an integrated Jeff Rowland amp that uses ICE modules. While we can probably make something that sounds better for a few hundred $. Besides that it is more fun, DIY saves you a lot of money too.

Best regards

Gertjan

Pan2

quote:

Originally posted by ghemink

Hi Dan,

Thanks again for the very useful comments.

Kind of have given up on using a SMPS.

I now plan to use two UcD400 bridged for the woofers. Not only to get more power. The main reason is that in a bridged situation, the power will be drawn from both rails at the same time. As a result, one rail needs to supply only half the current in comparison with a single module unbridged delivering the same power. The current consumption on each rail would be more continuous since each rails needs to deliver power during both the positive and negative part of the sinewave. There would also be no power supply pumping and the + and - rails always see the same load, so + and - power voltage would always be the same. Also the transformer may like it to always see a symmetrical load. This may also make it easier to use a capacitance multiplier like stabilizer.

Thanks again and best regards

Gertjan

Gertjan,

For a given current into the speaker the current via both halves will be the same as if one half would supply it. This is since the + and - halves of the supply are in series with eachother AND the load.

/Peter

Yves Smolders

quote:

For the "tube smoothing" simulation, as easy way, we can try to add resistor or a coil to the output. Although, U-source it's a still standard requirement for audio amplifiers, but we also have a choice between true fidelity and more pleasure

I don't feel there is any need to "simulate" anything with UcD's. I've tried the 400AD's with some SACD (Diana Krall Besame Mucho) and it was the sweetest sounding treble I've ever heard - not tiresome, just perfect. The "S"-sounds in Diana's voice were never harsh.

Strangely enough, with a good source signal, my B&W N804's, which some people look at as having an ear-piercing treble, are tamed perfectly. Mind you, if the source provides this extreme treble, they still perform - Some DVD's I've tested were very extreme on the treble, even more so than on my older class AB - but with good source, it sounds like it should. Midrange is super transparent, but never boring, bass response is now what I would expect from N804 - the bass is like I heard on Class� amps. (803 still has better bass, but the cabinets are much larger)

I really believe the Hypex UcD's are something truly special, being very very neutral.

A good set of speakers, UcD's driving them, is all you really need. If a recording sounds "harsh" or not "sweet", it's not your amp, or your cables, or your cd-player, or whatever; it's the recording that is "boring" (see comments that Bruno made) - I've tested my UcD's in a truly experimental setup (nothing was "perfect", not the cables, not the interconnects... the interconnects where 2 cables connected together in "free air"...) still sounded better than my old receiver with good cabling (mind you, nothing esoteric, just good quality interconencts, never exceeding 25�/channel, speaker cable is 2.5mm2 copper, no fancy stuff)

If you want to "sweeten up " these "bad recordings", maybe you could use that resistor or coil, but make sure you can switch it out of your system, or you'll be losing out on a truly amazing experience at a perfectly good price.

As a matter of fact, when my amp is ready, i'm going to take it to my dealer to "audition" speakers, I want to see the question marks on his face :bigeyes: :xeye:

GregD

quote:

Originally posted by dmfraser
The UcD amps sound closer to an excellent class AB amplifier while the ICE sound slightly "tubey".

I think this is partly because the UcD is a half bridge design and does an excellent job simulating an excellent class AB transistor design. Basically a near perfect reproduction of the audio in as clean and essentially unbiased sense. Where the amplifier imparts very little of its own sound in the signal. Nominally the goal we are striving for. Not perfect but as close as the musical truth all but the most esoteric high end amplifiers of say 1990 could give. And with a lot more efficiency and in a much smaller package. And available to the DIY community at a reasonable price.

On the other hand, the ICE Power is a full bridge design and seems to come closer to simulating the sound of a very, clean tube amplifier because the bridged output stage does a great job in cancelling out harmonics like a tube amp does. As well, their soft limiting circuit makes the sound, when clipped, more like the sound of a tube amp in clipping. Imagine the sound of a tube amp would give if you had a perfect output transformer with zero loss, no hysteresis and near perfect frequency response. Again very good and the sound that many people are looking for. However, for the small quantity user, essentially unobtainable.

Very interesting info. Do you think that bridging a pair of the UCD modules would accomplish some of the same cancelling out of harmonics that you note the ICE Power does. I'm tempted to buy another pair of the UCD modules, but I'd rather not spend the money unless the difference will be noticeable.

:)

classd4sure

Hi,

From all I've read on the topic, bridged UCD sounds no different than the half bridge, so if that's your only goal and you don't need or want the extra power, I wouldn't do it, the half bridge already senses the load differentially.

Yves Smolders, regarding the tube like effect, some people just like it dirty.

Regards,
Chris

bandini1

>Does anybody know of other vendors for switched power supplies for audio applications?

http://www.a-and-t-labs.com/K6_Sw_Amp/index.htm

ghemink

quote:

Originally posted by bandini1
>Does anybody know of other vendors for switched power supplies for audio applications?

http://www.a-and-t-labs.com/K6_Sw_Amp/index.htm

Thanks for the link, interesting supply, don'tknow of another supply like that for DIY.

Thanks

Gertjan

ghemink

quote:

Hi Dan,

I plan to order the SMPS from

http://www.a-and-t-labs.com/K6_Sw_Amp/index.htm

This thing is pretty flexible, one can set the output voltage with a switch, or of course you could modify a few resistors to get the output voltage you want. It should also be able to deliver 1000W continuous.

I have also ordered the active power supplies on

http://www.wnaudio.com/

Plan to use those after the SMPS (if needed) to reduce ripple voltage for the amps that feed the midrange and tweeter. Probably I'll feed the amps (plan to bridge two UcD400) for the woofers directly from the SMPS.

I'll let you know how experiments proceed, need to get the stuff at home first and then build it (these are kits).

Best regards

Gertjan

classd4sure

quote:

Plan to use those after the SMPS (if needed) to reduce ripple voltage for the amps that feed the midrange and tweeter.

Hi ghemink,

That sounds like a really neat idea to try.

Keep posted.

Regards,
Chris

ghemink

quote:

Originally posted by classd4sure

Hi ghemink,

That sounds like a really neat idea to try.

Keep posted.

Regards,
Chris

Hi Chris,

Thanks for the support.

The point is that this thing is then getting expensive, so many people may see it as overkill. However, I plan to feed 3 amps (or 4 if you count the bridged one for the woofers as two) with one SMPS and one active supply, don't need a tranny and don't need very big powersupply caps and power supply voltage should be pretty stable. That would be tough with a normal tranny feeding those 4 amps. Some people use one transformer per amp, so I guess compared with such a solution, this one maybe even cheaper.

Hope I have these toys in home soon.

Gertjan

DukeNude

quote:

Hello,
you do NOT need 1000ASP!!! look at the B&O-Page, there you can see, that you only need a 1000A because you can use power supply of you ASPs!

dmfraser

The A and T SMPS looks good for a split supply SMPS for a class AB amplifier. Maybe even for several UcD400s.

However, the active filtersays it uses RC filtering, the R being the power transformer secondary and the resistance of the wiring.

This is like a car dealer saying to buy his car because it has wheels with skinny tires.

A high transformer internal impedance and lossy wiring is another way of saying the wire used in the transformer is too thin and they use skinny power supply wiring. What it really means there will be high source impedance, which will result in impaired impact in the low end. The caps will not be able to recover very well from strong bass notes.

With modern designs exceeding 80 db power supply rejection ratio, careful grounding is more important than high source impedance in power supply design.

I would recommend some series inductance except that the A and T power supply already includes this.

I suggest the A and T power supply be tried but the WN Audio active power supply NOT be combined with it. The WN Audio power supply should be confined to mid and high amplifiers, not full range units, at least not if you expect super tight bass.

As for the amp with 2 1000ASPs in it, the reason the people did it is that the 1000A is limited to 3 KHz. If you want 1000W per channel full range, you need 2 1000ASPs. I have been told, as of last October by the ICE Power people, that there were no plans at that time to extend the 1000A to full range as they have to many other interesting projects to work on and not enough talented people.

DukeNude

what about 1000ASP and 500A?

ghemink

quote:

Originally posted by dmfraser
The A and T SMPS looks good for a split supply SMPS for a class AB amplifier. Maybe even for several UcD400s.

However, the active filtersays it uses RC filtering, the R being the power transformer secondary and the resistance of the wiring.

This is like a car dealer saying to buy his car because it has wheels with skinny tires.

A high transformer internal impedance and lossy wiring is another way of saying the wire used in the transformer is too thin and they use skinny power supply wiring. What it really means there will be high source impedance, which will result in impaired impact in the low end. The caps will not be able to recover very well from strong bass notes.

With modern designs exceeding 80 db power supply rejection ratio, careful grounding is more important than high source impedance in power supply design.

I would recommend some series inductance except that the A and T power supply already includes this.

I suggest the A and T power supply be tried but the WN Audio active power supply NOT be combined with it. The WN Audio power supply should be confined to mid and high amplifiers, not full range units, at least not if you expect super tight bass.

As for the amp with 2 1000ASPs in it, the reason the people did it is that the 1000A is limited to 3 KHz. If you want 1000W per channel full range, you need 2 1000ASPs. I have been told, as of last October by the ICE Power people, that there were no plans at that time to extend the 1000A to full range as they have to many other interesting projects to work on and not enough talented people.

Hi Dan,

That active power supply is a capacitance multiplier (although the web-site does not clearly say that), note that the web-site has a couple of power supplies.

The active one should have a very low output impedance as it is basically an emitter follower with the base voltage being an RC filtered version of the power supply voltage at the input (and a bit of voltage drop of course to create a bit of room for the emitter follower to do its work). This should give a very stable output voltage and very good noise suppression.

These active supplies work well when a relatively constant current is drawn, such as for class A amps. However, with Class D, where current consumption can vary over a wider range, they may get into problems (too fast voltage drop at the input). However, if that circuit is being fed by a regulated power supply, it should be able to work very well. That's why I want to combine the two.

Best regards

Gertjan

dmfraser

The 1000ASP with one or two 500A modules for a bi/tri-amp system is fine. There is plenty of power for this in the setup.

The WM Audio power supply write up implied to me that the filter multiplier effect for the high power portion was an RC filter. The head capacitance multiplication is for the lower voltage supply. It doesn't really matter. If what they are doing is a true capacitance multiplier, that is active, all a capacitance multiplier is, is a regulated power supply that may have a relaxed regulation spec. When you are using an SMPS, that runs at high frequencies anyway, the normal capaciatnce would be 10 uFd. By using caps in the thousands, you have super capacitance already. As well, any ripple will be above the range of hearing.

As for capacitance multipliers, I had one on my very first amp, in 1968 because the crappy single ended designs of the day had poor supply rejection, a problem we had solved by 1975 in the amp design. Hum and noise are not your problem. And as far as an electronic curcuit giving you more oomph in the bass, the best way to do this is to either use bigger caps or increase the power frequency, something you are doing by a factor of 1000 by getting the SMPS. Remember that if you increase the frequency, theoretically, a given amount of capacitance will be 1000 times as effective. In practice, maybe 50 times but more than a capacitor multiplier will do for you.

However, you can't hurt anything besides your wallet by doing both.

ghemink

quote:

Originally posted by dmfraser
The 1000ASP with one or two 500A modules for a bi/tri-amp system is fine. There is plenty of power for this in the setup.

The WM Audio power supply write up implied to me that the filter multiplier effect for the high power portion was an RC filter. The head capacitance multiplication is for the lower voltage supply. It doesn't really matter. If what they are doing is a true capacitance multiplier, that is active, all a capacitance multiplier is, is a regulated power supply that may have a relaxed regulation spec. When you are using an SMPS, that runs at high frequencies anyway, the normal capaciatnce would be 10 uFd. By using caps in the thousands, you have super capacitance already. As well, any ripple will be above the range of hearing.

As for capacitance multipliers, I had one on my very first amp, in 1968 because the crappy single ended designs of the day had poor supply rejection, a problem we had solved by 1975 in the amp design. Hum and noise are not your problem. And as far as an electronic curcuit giving you more oomph in the bass, the best way to do this is to either use bigger caps or increase the power frequency, something you are doing by a factor of 1000 by getting the SMPS. Remember that if you increase the frequency, theoretically, a given amount of capacitance will be 1000 times as effective. In practice, maybe 50 times but more than a capacitor multiplier will do for you.

However, you can't hurt anything besides your wallet by doing both.

Hi Dan,

Thanks for the additional comments. The reason why I'm thinking of using the capacitance multiplier is to remove the 75kHz switching frequency ripple of that SMPS since that ripple will depend on the power I draw and on the quality of the caps at the output of the SMPS. Besides that, I don't know how well the SMPS is regulated, so I wanted to improve the power supply decoupling between the amps for the woofers (directly fed from SMPS) and the amps for mid and high that would be fed from the active supply.

Of course, this stuff with the active suplies after the SMPS may very well be overkill, just want to try it. I plan to do measurements (using a scope) on that SMPS and active supply, first without amps. What would be a good load? I was thinking of just connecting a few 100W 100V lightbulbs between the + and - rail of the supply to get a load of a few 100 Watt. This is cheap and 100V lightbulbs are easy to get here, besides that, you have immediate visual feedback :-). Of course can only simulate a steady state load that way but that is good enough for a start. Later when connecting the amps, I may load the amps with lightbulbs as well, has anyone tried this before. Much cheaper than huge dummy load resistors and if the frequency is not too high, it may work pretty well.

Best regards

Gertjan

classd4sure

Hi, this discussion took a fun turn.

How about some small series resistance with the lights to limit their inrush, I don't know what their impedance would be when cold, but it will certainly draw a surge.

No doubt it is overkill, but if you want to try it, I'm eager to hear of your results :)

Of course you know, all the added complexity significantly increases the odds of something going wrong with it, I hope you get to test it to your satisfaction without anything like that happening.

Regards.

ghemink

quote:

Originally posted by classd4sure
Hi, this discussion took a fun turn.

How about some small series resistance with the lights to limit their inrush, I don't know what their impedance would be when cold, but it will certainly draw a surge.

No doubt it is overkill, but if you want to try it, I'm eager to hear of your results :)

Of course you know, all the added complexity significantly increases the odds of something going wrong with it, I hope you get to test it to your satisfaction without anything like that happening.

Regards.

I'll keep you updated, will of course first only use the SMPS.

The Inrush current of the bulb would be interesting :-) This is nice actually, can check the transient response of the regulation loop of the SMPS that way :-)

By the way, has anybody ever investigated the impedance of a light bulb as a function of frequency? I guess it has a bit of series inductance. Yes, this really becomes a fun project.

Gertjan

marconist

quote:

Originally posted by ghemink

I'll keep you updated, will of course first only use the SMPS.

Gertjan

New on this forum, but very interested in Class-D and SMPS. I experimented with Class-D in the 70's (small HF transistors, 8 Watts output, not a nice sound ...... :(

I wonder if an SMPS can handle the low load of a Class-D amplifier during the silence periods and does not switch off. If an artificial load is neccessary in normal musical operation it would be really a pity.

Anyhow, I will follow this thread with high interest!

Success,
Art

ghemink

quote:

Originally posted by marconist

New on this forum, but very interested in Class-D and SMPS. I experimented with Class-D in the 70's (small HF transistors, 8 Watts output, not a nice sound ...... :(

I wonder if an SMPS can handle the low load of a Class-D amplifier during the silence periods and does not switch off. If an artificial load is neccessary in normal musical operation it would be really a pity.

Anyhow, I will follow this thread with high interest!

Success,
Art

Still waiting for the kit to arrive

Gertjan

beppe61

quote:

Originally posted by unclemonty
You can buy them here: http://cgi.ebay.co.uk/ws/eBayISAPI....ssPageName=WDVW
...
but it's way too expensive for something that costs <$100 OEM.
...
unclemonty

Dear Sir,

please excuse me if I got back this extremely interesting information.
I understand that also Rotel has jumped on the ICE wagon.
If the price you mention is real they have an amazing mark-up on their latest amps.
Are you referring at the ASP version with the embedded switching power supply?
This is unbelievable.
A ASP module with just some cable anc connectors is a ready to use high quality power amp !
I think that in the next future we will see an huge growth of hi-fi and pro amps based on these modules, if the price stay this low.

Thank you very much again for the information.
Kind regards,

bg

fredos

Nice forum, lot of interesting and funny information! First of all, I want to make a comment about zobel circuit. A zobel should be always over the audio bandwidht, in that way, it only serve for stability and do not dissipate lot of power or blow in continuous hight frequency measurement. Please everybody learn this!!!

Second. Power measurement. A good way I think to rate amplifier is to rate the VA, not the watts. And best is maximum continuous (30 minute for our power amp) at 0.5% distortion. After this, we design heat sink to handle 50% of maximum power for 8 hours. If you whant to rate your amplifier at maximum power just before it blow, it's your right, but because of this, a lot of good product with real spec can be sold.

Least, about switching power supply. I think their is no rule here. I have try some competitor amplifier, and some did not work very well with our power supply, especialy the ICE module. Our switching power supply can provide over 6.5Kw of peak power, with 4 Kw continuous capacity, and always blow with ICE modula at near 600W continuous. But with our power amp, a 4 Kw continuouscan be measured continuous. We never understand what happen with the ICE module, protection in the power supply never trigh. Lot of analog amplifier sound better too with a switching power supply, but lot too sound very bad. I think that is no rule for this, just experimentation.

My comment now about the ICE with a 1.5Kw transformer. It sound very good in the mid and hight frequency, that's impressive. Maybe better than our D-Amp amplifier. But the lack of power and bass sustain make it unemployable for PA.

I dont agree with a lot of statement in this tread, lot of people will not agree with me too, but that's the purpose of this forum!

Fredos

www.d-amp.com

fredos

Sorry I miss a letter...I want to say that lot of good product CAN'T be sold because of this...

Fred

classd4sure

Hi Fredos,

Do you have any hypothesis as to why it might sound thin in the bass, something with the topology(feedback) perhaps?

You'd heard a UCD I presume...thin bass has been observed by many with that too, any correlation there in your view?

About measuring amps in VA instead of watts. As far as I know, watts=VA, volts times amps. What you rate your amps for "reactive power" would seem to imply VAR, volts amps reactive, but that's not true power delivered to the load given the angle by which the volts and current are out of phase with one another, producing no real power.

Please do correct me if I'm wrong.

Regards,
Chris

classd4sure

Fredos,

Sorry I posted too fast and confused a few things, like watts doesn't =VA in the sense which you obviously used it, which was that of "apparent power" and not "reactive power" as I'd implied.

Obviously I confused myself there.

Still, I maintain my point, although slight reformatted, VA would seem to artificially inflate the rating, since it doesn't take phase into account. True power (watts) is what the load is actually using, I'd think it preferred to continue rating amps for true power, anything less would seem to be kin to such things as PMPO, hate to say it though.

Regards,
Chris

fredos

VA is the real power delivred to the load, including inductive, reactive and resistive power. VA is more when you measure power dissipated in a load, but less when you measure from a source! So that's why people won't believe the power of our power amp! When you put a QSC of 1500Watts again a D-Amp of 1500Watts (VA), the D-Amp is always louder (a lot!)!

About the tinny bass sound of competitor, I cannot say why, but I suspect more the PSU than the amplifier itself. Or it's just maybe the damping factor of the amplifier...If feedback cannot compensate for the resistive component of the output filter (and maybe power supply), the bass will by tinny or with no control..That's the reason why all the power stage of my amplifier have only X12 (X24 for my HVI serie) gain. I give the voltage gain before the class d amplifier with a class a voltage amplifier...So in this way, I have a real control over the output.

Fredos

www.d-amp.com

iwok

Hi. I managed to place my hands on a pair of used 500asp modules thanks to trading posts on this forum. I have however a small problem with them. During transportation one of the capacitors was crashed. This is the cap connected straight to the transformer. It is Vishay`s BC Components RVI136 series: 680uF 100V capacitor - practically unavailable online.
Because I can`t get BCC cap I am experimenting with other brands` caps. I used Panasonic FC series cap but the sound lost its tubey character and became too bright. Caps were replaced after few days with Nichicon PF (M) series. This time there is sufficient amount of lower frequencies but the bass is soft and uniformed. I bought some ELNA RJH series on ebay but they are still in hands of transportation company.
What do you think guys, which capacitors can be closest to RVI136 Vishay`s series?
If somebody has this capacitor in his/her drawer I would really appreciate contact. Regards.

iwok

I`ve just installed ELNA RJH series capacitors (old brown-sleeve version) and there is an immediate improvement. These are much better than panasonic FC and Nichicon PM (F).

domcheng

iwok,

If somebody has this capacitor in his/her drawer I would really appreciate contact. Regards.

If you send me the spec, I maybe able to get it for you here in China.

iwok

Hi domcheng,

Thanks for your help. I already ordered it at Vishay. I had to book 100 pieces, need only 2 so if somebody needs some of Vishay`s 680uf 100v rvi136 series caps - please contact me.

iwok

I used Panasonic FC series caps in my icepowers but the sound lost its tubey character and became too bright. Caps were replaced after few days with Nichicon PF (M) series. This time there was sufficient amount of lower frequencies but the bass was soft and uniformed. I bought some ELNA RJH series on ebay - these were closest to Vishay`s RVI136 but not such `juicy`and full like the original. I finally got the BCC caps and they are just perfect in the implementation of B&O Icepower.

domcheng

iwok,

Really? I didn't know the E-cap brand names will matter the sound characteristics. In fact, I am building an PSU + Digital amplifier myself using Zetex solutions. I may want to try out different E-caps to see its differences.

Which BCC series are you using, can you send me an spec sheet or URL?

iwok

domcheng,

B&O used RVI136 series. Specs are here:
http://www.vishay.com/docs/28321/136rvi.pdf

buzz1939

Maybe this is mentioned above but just a reminder, be sure you're clear on whether the modules have any protection against shorts, dc - a little discussion by 2 engineers here (Jack Hidley, formerly of NHT) and here

"I