My Class D PWM amp project.

[michaelab]

Just finished my first major DIY project, it's a Class D switching amp using the ZAP Pulse 2.2SE modules from LC Audio. OK, so I didn't really do anything more than screw and solder the various bits together but hey, it's my first project

Here are a few pics detailing some of the project (there's a parts and cost breakdown at the end of the post):

The main parts, together with the main part of the case:

An externally hosted image should be here but it was not working when we last tested it.

The rear panel of the case with all the necessary holes drilled and sawed into it. This was by far the toughest part of the project!

An externally hosted image should be here but it was not working when we last tested it.

The rear panel now populated with sockets etc:

An externally hosted image should be here but it was not working when we last tested it.

The PSU, transformer and input sockets and front panel power LED all wired up. Just needing to solder the connections to the actual modules themselves:

An externally hosted image should be here but it was not working when we last tested it.

Everything wired up and powered up for the first time - yay! :

An externally hosted image should be here but it was not working when we last tested it.

Closeup of one of the modules in action. If you're very observant you'll see I wired up the -ve and GND wires from the input the wrong way around (I switched them). I noticed that and fixed it before trying it out :

An externally hosted image should be here but it was not working when we last tested it.

Another view of the completed amp, including front panel with obligatory blue LED :

An externally hosted image should be here but it was not working when we last tested it.

Now sitting in my rack, and sounding very nice indeed:

An externally hosted image should be here but it was not working when we last tested it.

Parts and cost breakdown:
http://www.luacheia.com/misc_images/amp_project/AmplifierCosts.htm

The cost could have been significantly lower if I'd made my own PSU instead of going for the LC Audio one but for my first project I wanted something as "plug and play" as possible. I also ordered WAY too much cable from LC Audio. I actually used about 1/10th of what I ordered. The case from www.par-metal.com is very nice and surprisingly cheap.

It sounds very, VERY good. Far better than the Arcam A22 integrated I was using (in power amp mode) previously. Particularly noticable is the bass control, but then you'd kind of expect that with 230W per channel (into 8ohms, 460W into my 4ohm Dynaudios!). Probably needs a bit of running in but I'm already very impressed and pleased.

Any comments appreciated.

Michael.

 

[pburke]

sweet looking setup.

I just noticed the price of the 100,000uF power supply you have was slashed in half at DIYCable.com because of the new 2-channel setup they are selling.

Since i want to build a 4-channel unit for bi-amping, I may just get one of these, since at $150 it actually is a pretty good deal, including soft start and having 100,000uF on the board. And since I would prefer to build 2-channels first and compare to my existing amps, I'll probably end up building two 2-channel setups like yours (or I'll have one 2-channel for use with future subwoofers if it fails the comparison test with my Odyssey mono extreme edition amps).

Which chassis did you get from par-metals? The RFI shielding version, or just plain aluminum? I'm still thinking of just building something on top of a fat aluminum plate with lots of cool acrylic and copper to enclose the modules - Peter Daniel gainclone style in a way.

Do some more listening and then let us know how it breaks in

Peter

 

[peranders]

Nice job... but if I were you I should try to keep all cables with as small loop areas as possible, meaning twisted speaker wires or at least keep them closer together, the same thing with the input wires. Since this is a switched amp, this must be even more important than for a regular one. That's all, nice job.

 

[chris ma]

Isn't that a CHORD below the TEAC there? How does it sound?

Chris

 

[michaelab]

Thanks guys .

Another reason I got the V4P PSU was because I might add a 3rd channel later to power the centre channel of my AV setup but tbh the thought of dismantling it all to drill the necessary holes doesn't fill me with warm thoughts

I got this case from Par-Metal. 16x16x4 inches, anodized front plate, grey painted rest. It's just standard aluminium.

peranders - you're not the first person to suggest I should twist the input wires. I'll definitely give that a try but there's no evidence of any RFI problems so far. As for the speaker wires, I'll try twisting them too but I wanted to avoid having the speaker wires running close to the input wires if possible.

Michael.

 

[michaelab]

Isn't that a CHORD below the TEAC there? How does it sound?

Yep - Chord DAC64. I just love the sound of it. It's the only digital source I've heard that really brings music to life. It's fast, detailed and just incredibly musical. I wouldn't change it for anything

The Teac is a T1 transport which has been modified: it has the Clock 4 clock upgrade from Trichord Research with it's own PSU which is also from Trichord Research - it's one of their new "Never Connected" PSUs. Other than that, the captive power lead has been replaced with an IEC socket and some of the internal power cabling has been changed. Also, the case has had Dynamat damping material liberally applied.

The pre-amp is a Bent Audio NOH transformer based passive pre.

Michael.

 

[michaelab]

Been listening to it for a while now and here are some more thoughts on the sound:

- incredibly fast and detailed, and I think this adds to the sense of transparency.
- extremely well controlled bass and excellent extention too. Absolutely no "bloat" or overhang.
- really fluid mid-range, I've only heard the same kind of fluidity in Class A and valve amps
- pin sharp and detailed treble but without any harshness or edginess.

I had the case open again to change the resistor infront of the front panel LED for a more meaty one as the existing one was pretty much burning up . At the same time I took the chance to twist the input wires and also to earth the casing (and hence have a hard rather than "floating" ground). None of this had any noticable effect on the sound though

Michael.

 

[pburke]

you're making me yealous - I'm still saving for the parts

Sounds a lot like the reviews i have heard of other zappulse based amps (www.solar-hifi.com - see audioasylum amp forum for a review a few months ago)

did you ever check into their active filter? one thing I am missing on the web site is info on what kind of power supply I would need to run that crossover.

Peter

 

[ericpeters]

Hi,

How is the temperature around the zappulse modules, I notice you don't use any specific cooling....just the bottom plate?

I also see you use both XLR and Cinch input parallel wired , do you hear any difference when you hook them both up at the same time?

are the speaker terminals WBT? I don't see them in your cost overview


I've build an amp using the same modules but a higher voltage and lower spec'd power supply:

500 VA / 2*50VAC transformer and this power supply module:

http://www.bmm-shop.nl/Product.asp?Product_ID=1170

Probably because of the power supply specs the results are a bit dissapointing, I don't think the suplly has the current capacity as mentioned on this website.
I compared the amp with 2 rotel RB980BX and with one Krell KSA 80B and in both cases but of specifically in comparision with the krell the Zap had sharp treble and lacking bass pressure. Also at louder levels the zap started clipping sooner than the Rotel.

I will rebuild it soon as 2 monoblocks with probably lower voltage and overdimensioned (1000VA / 132mF ) power supply. But I'm a bit more cautious as I don't want to make the same mistake but also don't want to spend more money than nessecary.

 

[phase_accurate]

I think that is a mistake that is often made: Implementing weak PSUs for switching amplifiers.

Some PA guys I know complained about bad bass reproduction of switching amps, though I experienced the contrary. I therefore think that also manufacturers sometimes take too weak PSUs for switching amps in order to save weight and money. Even a generously dimensioned PSU for a class-d amp would still give weight and cost savings, one should just not be too scrimpy.
Regards

Charles

 

[Jean Benoist]

BMM electronics power supply 2014

Hi Eric,

Implementing the PSU on every kind of power amplifier is one of the most underestimated things with DIY audio and professional builders. The BMM electronics 2014 power supply is designed for use with the 4008 2 x 200W analogue power amplifier and is capable of delivering up to 14 Amp continuous current and 30 Amp peak current so this won’t be the problem. I think that the O88042 500VA 2 x 55V 4.55Amp transformer you use causes the problem in your application. In my opinion you should have used at least 800VA and best a 1000VA transformer.

Regards,

Jean Benoist
BMM electronics

 

[michaelab]

did you ever check into their active filter? one thing I am missing on the web site is info on what kind of power supply I would need to run that crossover.

I didn't really look into it but I'm aware they make an active filter. See http://www.lcaudio.dk/com/xover.htm for more info.

How is the temperature around the zappulse modules, I notice you don't use any specific cooling....just the bottom plate?

They get slightly warm but no more than that. From the ZAPPulse "cookbook" it says:

Cooling requirements depends on the applied supply voltage, as most of the hear is generated in idle mode, with no load. If you want to be sure to have appropriate cooling, simply use a A5 size aluminium sheet per module, but less will do. A 200W stereo amplifier can be cooled just using the bottom of a thin steel sheet 19" rack enclosure. Even playing at full power for extended periods will not thermally overload the ZAPPulse amplifier. So the obvious cooling solution would be to place the modules on the enclosure floor or side walls. With smaller PSUs of +/- 35-40V which gives 50-80W into 8ohms no heat sinking is called for at all. Not all power is dissipated in the heat sink, some 3-5W are dissipated in the power resistors on the board and 1-2W in the filter choke all depending on signal load. A thermal protection circuit will shut the module down at around 100C.

I also see you use both XLR and Cinch input parallel wired , do you hear any difference when you hook them both up at the same time?

I haven't tried that yet. I did it mainly for flexibility and also because I was thinking of using the RCA terminals for my AV setup (front channels) and XLR for stereo but I think I'll leave the AV going through the pre for better isolation (prevent ground loops caused by the AV gear being hooked up to CATV).

are the speaker terminals WBT? I don't see them in your cost overview

Yes, they are WBT-0765 terminals and I forgot to include them because I got them through a friend. They cost me 71 Euros for the set of 4, including the small torx screws for securing the wires (I did this rather than soldering them).

I've build an amp using the same modules but a higher voltage and lower spec'd power supply:

500 VA / 2*50VAC transformer and this power supply module:

http://www.bmm-shop.nl/Product.asp?Product_ID=1170

Probably because of the power supply specs the results are a bit dissapointing, I don't think the suplly has the current capacity as mentioned on this website.
I compared the amp with 2 rotel RB980BX and with one Krell KSA 80B and in both cases but of specifically in comparision with the krell the Zap had sharp treble and lacking bass pressure. Also at louder levels the zap started clipping sooner than the Rotel.

Hmm - 2*50VAC is right at the limit of what the ZAPPulse modules (even the "SE" version) can handle. 2*50VAC, rectified would give you +/-69VDC, giving 286Wpc into 8 ohms . The limit for the modules is +/-72VDC giving 290Wpc into 8ohms. It's possible that close to the limit they don't work so well.

Michael.

 

[medum]

........

did you ever check into their active filter? one thing I am missing on the web site is info on what kind of power supply I would need to run that crossover.

Peter

I'm the happy owner of the LC-Audio active X-over.

The active X-over filter is purchased with a complete 230V AC PS.
The trafo should deliver appr. 15-0-15 V and the rest is inboard.

Be careful and check the DC on the board. There was a failure in the board-design in my deliverance. Hope Lars have corrected it (I have told him!)

Sounds really OK , but I'm using a battery powered supply. Sounds even better!

 

[ericpeters]

Jean,

you're absolutely right about the transfromer should be 1000VA. But I also think that for my Ribbons (they go down to 2 ohm and have low sensitivity) probably 14A continious is not enough either.

I will start with a bigger tranformer and check it with the current ps first in stereo mode. If this is still too small I always can make 2 mono's

 

[pburke]

I'm the happy owner of the LC-Audio active X-over.

The active X-over filter is purchased with a complete 230V AC PS.
The trafo should deliver appr. 15-0-15 V and the rest is inboard.

Be careful and check the DC on the board. There was a failure in the board-design in my deliverance. Hope Lars have corrected it (I have told him!)

Sounds really OK , but I'm using a battery powered supply. Sounds even better!

well, a 230V AC PS is useless for me unless it runs on the plug I have my clothes dryer hooked up to Maybe for export they can leave that off. their reseller here in the US doesn't carry the crossover.

And just like you, I'd rather run it with a battery supply, just like my DAC. How do you get to the 15V? Are you using two 12V gel cells in series, regulated down to 15V? Probably needs quite the heat sink to do that.

Peter

 

[medum]

well, a 230V AC PS is useless for me unless it runs on the plug I have my clothes dryer hooked up to Maybe for export they can leave that off. their reseller here in the US doesn't carry the crossover.

And just like you, I'd rather run it with a battery supply, just like my DAC. How do you get to the 15V? Are you using two 12V gel cells in series, regulated down to 15V? Probably needs quite the heat sink to do that.

Peter

If you could get the board you just need a trafo 115 V prim. and 15-0-15 V secondary

I use 4 x 12 V lead-acid batteries in series and regulates the 2x24 V down to 2x15V using lm317/lm337 and not really that much heat sink, but I will add more heat sink in the future - se the thread I started

earlier thread on battery power

 

[ericpeters]

I just replaced the 55V / 500VA transformer with a 42V / 1000VA transformer and the difference is amazing.

Even at very low volume there is a substantial difference in bass, treble and also soundstage.

Adding some bass because of a bigger transformer is understandable, but I can hardly beleive that this also improves the treble and soundstage.

I still get the amp to clip (altough at very loud level) so I think a further enlargment of the entire power supply will not do any harm.

 

[phase_accurate]

Even though self-oscillating class-d amps have better PSSR than their carrier-based ones they still profit from a generously dimensioned PSU.
Keep in mind that the gain of the output stage is determined by the ratio of supply-voltage/carrier-signal (the latter measured at the input of the comparator).
With a more generously dimensioned PSU you have less "muddying-up" of the midrange by the LF signals.

A more generously built amp does generally play much more effortless than its weak counterpart.

Regards

Charles

 

[xman737]

Could anybody tell me what different between ZAPpulse 2.2SE and ZAPpulse 2.2? Power rating? Sound Quality?
Thanks!

 

[michaelab]

The SE version uses Black Gate caps. According to LC Audio this makes it less susceptible to the PSU caps you use and alltogether improves the sound quality, particularly in the mid and high frequencies. The SE version can also handle a higher PSU voltage and therefore you can run it at a higher output power than the normal version.

LC Audio recommend using the normal version only for subwoofer applications and the SE for full range audiophile amplification.

All this info is on their website btw

Michael.