DIY digital PWM amp project

It's hard to say. I don't think I was as blown out of the water with my pre-amp when I got it as with the power amp, it may be though that Arcam was strangling it's capabilities a little. However, I've just looked back on my review of the NOH and I said....errm..... "Certainly the biggest since I got my DAC64" :shame:

On balance, I'd have to say that they were probably an equal improvment but the combination of the two was greater than the sum of the parts so the power, being the final link in the chain gave the impression of a bigger improvement. I'm now fully balanced from source to power amp aswell which possibly has helped.

Michael.

 

Hi Michael,

I'd love to hear more of your thoughts on your new amp now that you've spent more time with it.

I'm stilling thinking of a 1000VA and 4 modules, but probably in a few months time realistically.

regards,
Rob.

 

Not much more to report other than that I'm loving it

Couple of things though:

- mate of mine here who, in contrast to myself, actually knows a thing or two about electronics has suggested I stick a small (eg 470uF) electrolytic cap in parallel with the power LED - he says LEDs are notoriously noisy and the cap will shunt that noise off. Might try that this w/e.

- The amp has XLR and RCA inputs in parallel and there's no switch. I wondered what would happen if I had them connected at the same time: RCA to my AV receiver (front pre-outs) and XLR to my pre. Well, it worked pretty well for the RCA inputs (sounded fine) but then trying to listen to the stereo system via the XLR inputs (not at the same time of course) something was horribly wrong. It sounded like hearing music through a telephone . So, Plan A it is - use one of the RCA inputs on the pre for the AV amp

- I got my Rothwell XLR attenuators the other day. They are pretty big - about 3-4 inches long. They came in one of those velvet inside jewelrey boxes....I should think so too at the price . Anyway, hooked them up and it's much better. No sonic difference, just that I can use more than 3 or 4 stops on the volume control now.

1000VA for 4 modules should be fine. However, on a thread I started on DIY Audio.com a guy just swapped out a 500VA xformer for a 1000VA one (for 2 channels) and said the improvement was scarcely believable!

Michael.

 

Superb. I'm glad you are enjoying it.

I read your diyaudio thread too. He also went down from 2x50V to 2x45V (230W in to 8 ohms), as lcaudio recommend 2 x 50V AC 600VA (286W in to 8 ohms).

I think using a 500VA when pushing the voltage to 69V after rectification (72V max for the 2.1SE) sounds like it was asking too much.

Although this does contrast with "Our V4P power Supply board is perfect for multichannel amplifiers, as it has connection for up to 8 channels. For normal use one 500 VA transformer is enough to drive 5 channels of 200 Watts RMS (total 1000 Watts Audio Power) because there is no thermal loss in the amplifiers, and even if the power at clipping level is 1000 Watts total, the average power cointent of a signal integrated over a period of ½
hour (thermal time constant of the transformer), is only less than 50%. "

I have noticed the alternative "Predator Power Supply for 2ch. ZAPpulse 2.2" power supply, but can't find any specs.

At 2x53.70 euros, plus the Digital Safe Softstart unit max. 10 Amp at 40.24=147.64 euros. I suspect it may just be a pair of caps and rectifiers like here: http://www.lcaudio.dk/com/zpn9.jpg though, excluding the DC filter of the V4P.

This would allow a pair of 1000VA transformers to be used in a dual-stereo 4-ch amp.

Like yourself, I do like the look and simplicity of the V4P though.

Running 2x42V (200W) x4 looks like it should work off a single 1000VA.

 

By all means try it but I'm afraid its a bit of a misconception. LEDS are forward-biased junctions, therefore exceptionally low noise - in fact they make good voltage references for this reason.

Looks like a great project, and one I'd really like to hear; *very* tempted myself, but with 93db/W speakers I don't need that kind of power

 

Yes - I think that too. The improvement is IMO more likely to have been due to the reduction in voltage than the increase in VA.

Michael.

 

Interesting - perhaps I won't bother then. I'll put that to my mate and see what he says :torkmada:

Michael.

 

Well try it anyway - yu never know. But often when this trick works, it's because the rest of the circuit at the other end of the wires really needed the extra decoupling...

All the best,
F.

 

Interesting project. Glad you're getting good results Michael. I'll be considering these in a few years time. I'm thinking along the lines of replacing my amp, speakers and sub with a DIY full-range active system (with 12 inch woofers at least :torkmada: )

Got my eye on some ATC drive units but also considering the Tannoy dual-concentrics. We'll see. That's a few years away yet though.

 

Michael - could you take some really close up pics of the modules (pref with the output inductors lifted out of the way, such that the PWM chipset can be identified?

 

New job got you doing a bit of industrial espionage eh Issac?

When I get back to Lisbon (end of next week) I'll get out the macro lens and give it a go but I think you won't learn anything. LC Audio aren't using an off the shelf PWM chipset - it's their own proprietary circuit of (mostly) discreet components rather than ICs.

There are two very small ICs on each board which have been covered in some black gunk to obscure their identity - I've no idea what they do but they've presumably been obscured for a reason.

Underneath the output inductor is merely the 2 output MOSFETs and the voltage reg for the circuit.

Michael.

 

From the pics I've seen there are three national semiconductor ICs on the board. If they're all dual op-amps then they're probably just doing a discrete circuit. The obscured ICs are probably the gate drivers. I'm just trying to gauge where the performance of the LCaudio modules is likely to be at, and possibly identify things you could do to improve them.

One improvement that does spring to mind is to use a high-accuracy clock generator (eg trichord) and appropriate filter to generate a better triangle wave for the PWM comparitor, and use the same clock to drive both channels synchronously. Another would be to change the op-amps for something else (which is likely to just give you a change in sound, not necessarily better or worse).

 

That's an option with the modules. They can be driven by an external clock instead of oscillating at their natural frequency. LC Audio recommend doing this if you have "several" modules in a single box allthough they don't specify how many "several" is.

They can be clocked at anything from 500kHz to 1Mhz. Their natural oscillation freq. is about 490kHz.

Naturally, they suggest that their own LClock XO3 is a great candidate for the job

Here's probably the best image on the LC Audio site of their latest 2.2SE (the ones I'm using) module without the inductor coil:



..on the bottom left and on the right side you can see the two ICs that have been obscured. Underneath each board are 4 light blue (polypropylene?) caps - they're the only components not mounted on the top surface.

Michael.

 

Yes, that's the pic I saw. Interesting... The chip on the left hand side that is obscured is almost certainly the gate driver, probably something like an Ixys IXD series. Not sure about the one on the output side of the board though.

It's a very simple setup there (simplicity is good of course), and about £20 production cost, although the custom inductors may add some to that. It would be interesting to see how a DIY job on veroboard would compare (obviously not as good, although it could be reasonably close and a stonking bargain).

Do you happen to know if the output is bridge-tied load or single ended? I'm assuming single-ended because of the single output pair, in which case you could use two modules bridged per channel, using the +ve and -ve connections from your balanced input to drive each module individually.

 

You could always ask Mr C about that one, I believe its actually a 'borrow' of one of his earlier designs

 

That's going a little beyond my knowledge of electronics but I know the modules can be bridged because they diagram how to do it in the "cookbook" PDF doc that accompanies the modules. You can get really silly power outputs when bridged

If you're looking for the cookbook on the site it seems to have gone. I think they're waiting 'til they have an English translation of the 2.2SE version (I have the old 2.1SE version of the doc at home which I can e-mail you next week if you're interested - it's not significantly different from the new one).

The Danish version of the 2.2SE manual is here

Michael.

 

Hi Michaelab what about a review of your power amp + pre together

GAZZ

 

I love them, they sound great. How's that?

Seriously though, I'm no good at doing reviews. Suffice it to say that I'd have to spend considerably more on an integrated or pre/power combo to get even close to the performance I'm getting

Michael.

 

That great Michaelab, i hope they give you years of pleasure.