The Truth About Digital (Class D) Amplifiers

[mtrycrafts]

Well, I can't argue with your position as it relates to your choices. .

I just happen to remember those product more so than some others

 

[silversurfer]

Is there a white paper that I can read about the the TACT EQ system?

 

[mtrycrafts]

There are two sides to the stories. PS Audio belongs to those few companies where the CEO himself takes the time to discuss on do-it-yourself electronics forums like diyAudio.com.

I don't know him or have a real opinion about his amps or other products. They may be great amps, no bs or voodoo, I am sure, but he is not immune from it, as demonstrated by that one product. So, to some, that voodoo product unfortunately is a detractor

One of their truely unique products is the HUMBUSTER, a device that can eliminate the mechanical humming sound of toroidal transformers. In the DIY forum he openly describes how their unit works, and confirms that you can probably build the device yourself for a fraction of what they're charging for the unit ($300 MSRP).

Transformers do hum, the nature of the beats. Not sure about the toroidal design, so, this would seem to be a legitimate product, unlike that detractor.


Also, the PS Audio power cables have a unique feature: The ground pin is removable (it unscrews). This is a fantastic feature to have if you experience ground loops. Those that frequently swap around gear must love this feature.

Yes, ground loop is an issue, unfortunately. But, is that a safe product, lifting the ground like that? Does it meet UL stamp of approval?

Now, again, I don't believe in the Noise Harvester (I pretty much ignored it), but PS Audio does have products like the Humbuster that does work.

However, how is one to know, if uninformed, or gain trust in a company when they have both on the market, a good product and a voodoo of a product???
That is the issue, not that they also make good products. Just like a sighted comparison. At times, you do get the same results as from a DBT, but how is one to know when?

 

[gene]

Are there any other peculiarities with the Sunfires that I should know of? It's hard to find any info about them. Strangely enough, my Parasound dealer didn't seem to regard them very highly, although he also sells Sunfire. He tells me that the New Classic (entry level) Parasound is better than the Sunfire amps... Maybe I will get burnt by the Sunfire and better stay away?

Parasound makes solid amps.

 

[Peter Nielsen]

However, how is one to know, if uninformed, or gain trust in a company when they have both on the market, a good product and a voodoo of a product???

How come some people trust NordOst and gladly pay them $10,000 for a pair of Valhalla speaker cables?

Considering that, how much trust do you really need to buy an inexpensive product like the $75 PS Audio Noise Harvester?

'nuff said.

Peter

 

[AdrianMills]

I am not familiar with those products, so how much is that roll off? This link shows how audible some roll off is compared to bandwidth:

http://www.provide.net/~djcarlst/abx_crit.htm

A broad bandwidth roll off is more audible than a fraction of an octave.

Thanks but I think Clark's explanation and graph are about as clear as mud.

The roll off of the 1000ASP module that B&O publish is just under 1dB at 20kHz starting at approx 2kHz and it gets to about -0.1dB at 5kHz. That's with an "open load" - with ruler flat 4ohm loads the freq response is also flat. So the freq response is load dependent; I've no idea if this approx 1dB drop is worse case either.

 

[mtrycrafts]

Thanks but I think Clark's explanation and graph are about as clear as mud.

The roll off of the 1000ASP module that B&O publish is just under 1dB at 20kHz starting at approx 2kHz and it gets to about -0.1dB at 5kHz. That's with an "open load" - with ruler flat 4ohm loads the freq response is also flat. So the freq response is load dependent; I've no idea if this approx 1dB drop is worse case either.

His paper shows what is audibly detected with pink noise if the level is different by the amount in the left column and over the different octaves. 1/3 octave is the hardest to detect, need a bigger volume difference; 3 octaves of difference is the easiest to detect, down to .2 db spl, except at the ends of the spectrum.

But, that small amount you are talking, .1 dB at 5kHz, and 1 dB at 20kHz, no way. Many cannot even hear 20kHz, the JND at 16kHz is 3 dB on average and is 1dB spl at 10kHz.

 

[AdrianMills]

His paper shows what is audibly detected with pink noise if the level is different by the amount in the left column and over the different octaves. 1/3 octave is the hardest to detect, need a bigger volume difference; 3 octaves of difference is the easiest to detect, down to .2 db spl, except at the ends of the spectrum.

But, that small amount you are talking, .1 dB at 5kHz, and 1 dB at 20kHz, no way. Many cannot even hear 20kHz, the JND at 16kHz is 3 dB on average and is 1dB spl at 10kHz.

So why is there such a big kafuffle about the freq of these ICEPower modules being so load dependant if the variations aren't even audible? Or am I missing something here?

 

[ar-t]

This raises the question of the ICEpower and its power reserves. It's so small that there is no room for a lot of capacitors...

And you are basing this supposition on what??????? Because some manufacturers stuff an ASP module into a tiny little box because they think it looks cool? If so, you are barely scratching the surface of what is available.

So why is there such a big kafuffle about the freq of these ICEPower modules being so load dependant if the variations aren't even audible? Or am I missing something here?

Yes, you are.

A.) It exists.
B.) It is audible.
C.) Some listeners object to it. Others are not concerned. As a manufacturer, I have to be concerned.
(Yes, we still use them. And will most likely continue to do so.)

BTW..........it is "20 kHz", not 2 kHz. If it was that low, it never would have made it off of the factory floor.

As for power rating/requirements:

My personal system uses Maggies. I have run them to levels that I can not endure with a 60 W amp. This was with the amp we produced for years, and d/c'ed when we decided to go Class D. Our least powerful Class D offering is capable of supplying more power than our previous offerings. So, nope..............I have not run into a situation where any of them runs out of steam.

 

[AdrianMills]

A.) It exists.
B.) It is audible.
C.) Some listeners object to it. Others are not concerned. As a manufacturer, I have to be concerned.
(Yes, we still use them. And will most likely continue to do so.)

BTW..........it is "20 kHz", not 2 kHz. If it was that low, it never would have made it off of the factory floor.

OK, it's not 2kHz... but it’s close and I was wrong about the -0.1dB at 5kHz; according to the graphs in the B&O 1000ASP datasheet the roll off gets down to about -0.2dB at 5kHz and from there it's a steady drop down to -0.6 dB at 15kHz. That's with an "open load" and I'm guessing that B&O would probably have picked the more flattering results to publish.

If the module was originally intended for sub applications then that roll off would be no big deal really.

As for power rating/requirements:

My personal system uses Maggies. I have run them to levels that I can not endure with a 60 W amp. This was with the amp we produced for years, and d/c'ed when we decided to go Class D. Our least powerful Class D offering is capable of supplying more power than our previous offerings. So, nope..............I have not run into a situation where any of them runs out of steam.

More watts in a box is... compelling even if it's not that rational at times.

 

[Sleestack]

If the module was originally intended for sub applications then that roll off would be no big deal really.


More watts in a box is... compelling even if it's not that rational at times.

IMO, it really isn't a big deal in any application, especially if you do any sort of EQ'ing or correction. Of course, some may feel differently. Not saying it matters, but I take it you haven't even heard the amps?

More watts is compelling, and at some times can be completely rational. When you use an advanced room correction system, the massive amounts of power come in hand. 500W may be excessive, but not by much.

 

[Peter Nielsen]

IMO, it really isn't a big deal in any application, especially if you do any sort of EQ'ing or correction. Of course, some may feel differently. Not saying it matters, but I take it you haven't even heard the amps?

More watts is compelling, and at some times can be completely rational. When you use an advanced room correction system, the massive amounts of power come in hand. 500W may be excessive, but not by much.

Also remember that doubling the power only increases the sound level by 3dB.

Thus, a 96dB/1W/1M speaker with a 30W amp will play just as loud as a 87dB/1W/1M speaker with a 240W amp...

So the best thing you can do is to get an efficient speaker!

(93->96db requires 60W, 90->96dB requires 120W, 87->96dB requires 240W)

Peter

 

[mtrycrafts]

So why is there such a big kafuffle about the freq of these ICEPower modules being so load dependant if the variations aren't even audible? Or am I missing something here?

If it really is load sensitive, that would mean it has a high output impedance. I have no idea if it does or doesn't. If it doesn't but is similar to SS amps output impedance, then it cannot be load sensitive. If it has a high impedance, then it will follow the speaker's FR response and most likely be audible, but that would depend on the speakers FR as most would be.

 

[mtrycrafts]

How come some people trust NordOst and gladly pay them $10,000 for a pair of Valhalla speaker cables?
Peter

Human gullibility, that is how. Same applies to PSaudio. Some will buy that product because a reputable amp maker is selling it.

 

[Peter Nielsen]

Human gullibility, that is how. Same applies to PSaudio. Some will buy that product because a reputable amp maker is selling it.

And considering earlier posts here, and using your logic, I now condemn Theta Digital to be crap They claim that a digital signal must be volume controlled in the analog domain...

Ok, so PS Audio is crap. Nordost is crap. Theta Digital is crap... Parasound is crap using the same logic. The CTC blowtorch must really suck. Krell is probably crap too, and so is Classé.... Oh and Bryston is DEFINITELY crap, since they're basing their 10B crossover on similar assumptions.

Is TacT good? What GOOD brands do we have left?

Or, let's just turn this the other way around: What brands DO YOU trust? Let's see if we can't find any disputable issues there

Oh my,... I'd love to see you go car shopping! Try to find a car dealer you can truely trust

Peter

 

[AdrianMills]

IMO, it really isn't a big deal in any application, especially if you do any sort of EQ'ing or correction. Of course, some may feel differently. Not saying it matters, but I take it you haven't even heard the amps?

No I haven't - SilverSurfer got me thinking about getting some of those nice looking Midgard monoblocks though. I'm considering mailing them and asking if I can go for a listen at their workshop which is only about 3 hours drive from where I live... but there's a HiFi show close by in early November and I'll probably just wait to see them there. One thing I considered was to ask them if they would put together an amp built from the UCD700 for me - the UCD amps are supposed to be more linear aren't they? Although I haven't seen any hard data on that.

Do you think Audessey would by able to correct for this sort of problem?

More watts is compelling, and at some times can be completely rational. When you use an advanced room correction system, the massive amounts of power come in hand. 500W may be excessive, but not by much.

In all honesty, in my case I don't think wanting all those watts is very rational - I'd probably be in pain long before I used that much power.

 

[silversurfer]

No I haven't - SilverSurfer got me thinking about getting some of those nice looking Midgard monoblocks though.

Hey! Don't blame anything on me!

This has been a great discussion.

The UcD amps, I have read, are suppose to be more linear, but I have not seen any data. Also, I have been told that they have very little headroom, but again, I have seen no data.

 

[Peter Nielsen]

Damping factor

I contacted TacT and asked them about the damping factor of the BOZ amps. It's approximately 45 (!) Isn't this figure pretty bad? The ICEpower 1000ASP has a damping factor in the thousands, right?

Peter

 

[Sleestack]

I contacted TacT and asked them about the damping factor of the BOZ amps. It's approximately 45 (!) Isn't this figure pretty bad? The ICEpower 1000ASP has a damping factor in the thousands, right?

Peter

I'm no expert, but I don't think low v. high damping factors necessarily translates to bad v. good. Pass Labs flagship amps have very low damping factors, while their cheaper amps have high damping factors. Not saying that means anything either, but the few discussion I have read on the matter indicate that other factors are relevant.

 

[Peter Nielsen]

I'm no expert, but I don't think low v. high damping factors necessarily translates to bad v. good. Pass Labs flagship amps have very low damping factors, while their cheaper amps have high damping factors. Not saying that means anything either, but the few discussion I have read on the matter indicate that other factors are relevant.

Yeah, after reading up some more on this, I came to the same conclusion. It appears that as long the damping factor is 50 or more, it really makes no difference. TacT's 45 should be just fine. (It appears the high damping factor in ICEpower et al. is more of a sales pitch really).

Peter