[dewardh]

Welcome to AudioCircle

Well thanks . . . I'd visited before, but it was the experience of others with the IRS2092 based amps that got me to register.  They seem to be the best thing going at this time for the Class D DIYer, and the 2092  boards from classDaudio are a great deal with only a couple minor, and easily corrected, issues.  As you know, they sound great . . .

[Nick77]

        only a couple minor, and easily corrected, issues                                                       

Care to enlighten us. 

[dewardh]

Care to enlighten us. 

I don't know how "enlightening" it is, as it's all pretty obvious, but a few issues, in no particular order:

- the SMD input coupling cap is less than ideal (the manufacturer boasts that it's "as good as solid tantalum").

- the output filter is optimized for 4 Ohm loads, and high frequency response varies with speaker load impedance (this is documented in the reference design pdf).

- the output inductor is undocumented . . . it may or may not be a problem, but see the distortion graph in the reference docs . . .

- the power supply could use some snubbing/decoupling, especially if two amp boards will be used on the same supply.

- the gain/input impedance issues already discussed.

Nothing really serious . . . all easy to deal with . . .

[srb]

I don't know how "enlightening" it is, as it's all pretty obvious, but a few issues, in no particular order:...................Nothing really serious . . . all easy to deal with . . .

I have always wanted a "dual mono" amp, using 1 enclosure and 1 power cord.
 
I had thought about using 2 of the 250W at 4 ohm boards, bridged, with 2 transformers and 2 HD power supplies.
 
Both pairs of my current speakers are 4 ohm.  Would this configuration accomplish anything meaningful or address any of the power supply or output limitations?  Or might a different configuration make more sense?
 
Steve
 
 

[gitarretyp]

I agree 100% with dewardh. I've optimized the low pass filter on my amp for 8 ohm loads, and it does improve the top end a little. As a caveat, optimizing the filter for 8 ohm loads will lead to a fair bit of roll-off in the top octave for 4 ohm (8 ohm mono) and lower loads.

[dewardh]

Both pairs of my current speakers are 4 ohm.

None of the IR designed boards are speced for use bridged into 4 Ohm loads.  Look at the table on page 4 of the Reference Design pdf for suitable configurations.

[jtwrace]

Look at the table on page 4 of the Reference Design pdf for suitable configurations.

How about providing links?

[srb]

None of the IR designed boards are speced for use bridged into 4 Ohm loads. Look at the table on page 4 of the Reference Design pdf for suitable configurations.

I was going by what Tom had written in an email reply:  "The sonic quality is the same in either stereo (half bridge) or full bridge mode. Only the 250W X 2 into 8 ohms is not recommended for 4 ohm in full bridge mode. Our other amps will run 4 ohm with no problem, either stereo or bridge."
 
Steve

[Geardaddy]

To use who are currently using these units, how is the bass performance? Tom told me the damping factor was around 120 which seemed surprisingly low....

[dewardh]

How about providing links?

I did, in #628 above.  Here it is again:

http://www.irf.com/technical-info/refdesigns/iraudamp7d.pdf

[pjchappy]

I was going by what Tom had written in an email reply:  "The sonic quality is the same in either stereo (half bridge) or full bridge mode. Only the 250W X 2 into 8 ohms is not recommended for 4 ohm in full bridge mode. Our other amps will run 4 ohm with no problem, either stereo or bridge."
 
Steve

How old was that e-mail from Tom?  In a recent e-mail to me he said all higher powered amps will have a new heatsink, allowing 4 ohm operation.


Paul

[dewardh]

he said all higher powered amps will have a new heatsink, allowing 4 ohm operation.

For bridge mode into 4 Ohms they would probably need changes to the output filter as well, to avoid high frequency rolloff, and possible inductor core saturation.  The output filter is load sensitive.  The output FETs have some current limits (which also effect turnoff timing).  Because of the wide variation in loudspeaker loading there is no single turnkey solution for Class D amplifiers . . . there really is a "DIY" component in getting it right.

[srb]

How old was that e-mail from Tom?  In a recent e-mail to me he said all higher powered amps will have a new heatsink, allowing 4 ohm operation.

2/26/2010
 
He prefaced the paragraph I quoted above with "Yes, the heatsink for the 250W X 2 into 8 ohm is standard now"
 
And then went on to say "Only the 250W X 2 into 8 ohms is not recommended for 4 ohm in full bridge mode. Our other amps will run 4 ohm with no problem, either stereo or bridge"

Steve

[atroder]

dewardh - At a given price point such as this, one can expect certain design and build compromises. Have you had a chance to compare these modules against other well known class D designs such as the B&O ICE modules? Although I don't fully grasp all of the technical details you present, I certainly appreciate your airing of the information.

[dewardh]

At a given price point such as this, one can expect certain design and build compromises.

Oh, absolutely.  What’s astonishing about the classDaudio boards, considering their price, is how few compromises there are overall.  There’s nothing particularly obscure about the *reason* for the compromises, either . . . “audiophile grade” input coupling caps would cost $10-20 each, and be the largest parts on the board as well, and folks would (will) argue all day about whether they’re worth it.

Have you had a chance to compare these modules against other well known class D designs such as the B&O ICE modules?

I’ve been watching Class D evolve into a useable technology just like everyone else . . . studying, and listening to when possible, the various offerings as they came to market (and watching the manufacturers struggle with some of the same problems discussed here).  I almost bit on the Hypex modules when they became available, but the price was too much, and there were aspects of the design that didn’t satisfy.  I haven’t seen any DIY friendly ICE modules, although they are now used in a number of commercial amps, and TriPath has its own problems, and not enough power.  The overall *sound* of almost all the Class D devices is good, though.  I attribute a lot of that to the absence of crossover distortion . . .

The IRS2092 driver design, and the classDaudio copies of the reference board, were the first that really said to me “right stuff at the right price”.  They sound as good as the 3886 chip amps, without the power limitation.  They do have issues and limitations of their own . . . but all amplifiers do, and so far I haven’t found any “show stopper”.  At the moment I think they (along with the 3886) are the best deal going in DIY audio electronics.

[AmpDesigner333]

This has really gotten me thinking about the importance of impedance matching.

Some info on impedance matching:
http://www.audiocircle.com/index.php?topic=78263.msg743813#msg743813

[roymail]

They sound as good as the 3886 chip amps, without the power limitation.  At the moment I think they (along with the 3886) are the best deal going in DIY audio electronics.

I'm glad you brought that up because I was wondering how the two would compare.  If the ClassD amps can do what the 3886 amps do with more power and authority, then I finally understand what all the *hoopla* is about.

When I got my 3886 recently, it just "unveiled" the sound so I could hear "through the music."  Amazing  spacial clarity, resolution, depth...    Of course system matching and efficient speakers help.

[Nick77]

   Oh, absolutely.  What’s astonishing about the classDaudio boards, considering their price, is how few compromises there are overall.  There’s nothing particularly obscure about the *reason* for the compromises, either . . . “audiophile grade” input coupling caps would cost $10-20 each, and be the largest parts on the board as well, and folks would (will) argue all day about whether they’re worth it.

                                                         

How would one go about this mod, are these the same caps a few others have swapped out? The ones located behind the speaker posts? What caps would be reccomended, Sonicaps for instance? Is this one of the upgrades with the new board revision Tom is working on? What mods have you done to the board and results achieved? Thanks for all the insight! 

[krikor]

I haven’t seen any DIY friendly ICE modules

Check out Marchand Electronics... they are selling the 1000w ICEPower 1000ASP modules for $490 each (SMPS power supply is integral, but you need two modules for stereo).

http://marchandelec.com/amplifiers.html

A bit pricey for my tastes (though not compared to finished versions being sold) which is why I've been lurking on this thread contemplating the ClassDAudio kits for some time now.  I may have to give them a try, but want to make sure my Reference Line Preeminence Passive Pre will work with them first.

[dewardh]

are these the same caps a few others have swapped out? The ones located behind the speaker posts? What caps would be reccomended . . . What mods have you done

I have only seen comments about changing the output filter caps (the large blue cubes by the output terminals), although others may have changed the inputs.  The input coupling caps are the orange surface mount devices (3-4 mm on a side) by the input terminals.  They are physically tiny by comparison to the output caps (but have somewhere around 100 times the rated capacitance).  Replacing them on the board would be near impossible because of size diferences and lack of holes for physical attachment . . . it would be done between whatever RCA input connector is used and a (modified) input on the amp board.  I'm not religious about caps, and would (will) use any good quality metalized film (which may already be present in your crossover (as with mine) or pre-amp output).  The easiest or best mod depends on your specific enclosure and signal source . . . in some circumstances (decoupling and proper grounding at your source's output) it may not be needed at all (but there *must* be decoupling *somewhere* between the source and the amp input).  I'll give detailed reports on the mods I've done after they've been well tested and demonstrated safe.  If you bypass the input coupling cap, for example, you CANNOT just plug in any source and expect it to work . . . you have to know the source's output circuit, and that it is compatable with the (modified) amp input.