Building a DIY amp with NVA amplifier boards

[karatestu]

Starting to plan this music centre built in to a sideboard like enclosure. I will be using better quality wiring between psu and amp board - that orange stuff was found in a 50 metre reel in an old caravan It is 2.5mm stranded 3 core but is donkeys years old and probably starting to oxidise. I soldered some male spade connectors on to the nva amp boards for ease of connection.

When adding local decoupling (big smoothing cap) at the amp board I can actually solder one of the legs to the amp board. The other leg obviously must go to ground but not on the board itself. Don't want to mix power ground into signal ground. In a NVA system with multiple power amps the 0V all connect up in the preamp at the sockets along the back panel. I thought about copying that but what I have now seems to work. I have soldered the 0V wire from each psu to the 0V bus bar I have linking up each amplifier board. The two 0V wires from the mid bass psu join each other half way along the bus bar between the amp boards they are powering. Same for the tweeter amp psu's. It works and there is absolutely no hum.

Doc wasn't a fan of star earthing so I have tried to avoid that which is hard when you've been doing it that way for so long. Doc would probably have joined all the various psu 0V wires in the psu and then sent one 0V wire from the remote psu to the head section. Choices choices. I remember in an interview Doc said screw up the earth at your peril.

[karatestu]

I have got the latest version of my amplifier wired up and tested. Now have all six NVA boards in use and six remote PSU's. wiring still all over the place and the two 0V wires from the two added PSU's are joined on the bus bar half way between the two boards they power. Still absolutely zero hum so I must be doing ok but still wonder if the six psu grounds should be joined together in the remote psu and one taken to the head unit and joined to the bus bar between the back panel sockets and the first amp board in the chain



The psu's aren't arranged tidily but you get the idea, nothing is attached to the chipboard. I am now using all EI frame transformers with twin centre tapped secondary windings. Three transformers and one winding per amplifier board. These transformers are smaller in voltage and VA than my big Canterbury Windings transformers but you wouldn't know it.



I prefer the presentation with EI transformers it seems. Never tried C cores or R cocers but I can't be bothered. Less hash let through from the mains is bringing out low detail I could not quite resolve before. I don't play loud and the speakers are a very easy load as there are NO components between any of the amps and the drivers, just wire The four mid bass amps see the 7.5 ohm impedance of the mid bass driver only and the tweeter amps see the 20 ohm impedance of the four tweeters wired in series and nothing else.

Well, the concept works so time to try adding some psu caps at the load. Then I need an input selector switch and a volume control to sit between all the power amp boards. I can then move on to the phono amp - the head unit is going in the integrated and the psu's are going down at the bottom next to the power amp psu,s. Eventually my CDP psu's will go down there as well and possibly the Turntable psu as well.

[karatestu]

Some observations.

When you change the impedance seen by the amp from 16 ohms to 4 ohms by changing two mid bass wired in series to wired in parallel the volume of the mid bass goes up which wrecks the balance of the speaker (unless you have two tweeters which you can do similar with).

When taking two mid bass wired in series and giving them an amp each instead of sharing the same one the two amps see half the impedance that the sole amplifier did. But the overall volume doesn't go up. So what's the point in doing it ? Well there us an increase in SQ but I am not sure why. Maybe the 16 ohm impedance effects the drivers qes like when you put a resistor in front of a mid bass to increase qts.

When checking the dc offset of the six amps I noticed the tweeter amps had a slightly higher amount. Nothing major but interesting none the less. Must be to do with the high pass filter I have put on the input.

Edit: I removed some bollox that I wrote

[karatestu]

After adding local smoothing caps at the amplifier boards my next job is to remove the 10uf coupling caps in my cdp. I then need to put some Wima 3.3uf film coupling caps at the input of the four mid bass power amp boards. The tweeter amps already have 560pf polystyrene coupling caps which together with the 47K resistor to ground already on the nva amp board make up the line level high pass filter for the tweeters. They also do the job of blocking DC ofc which is essential when I have removed the coupling caps from the source.

The coupling caps that I need to put on the input of the four mid bass amp boards are for blocking DC only. A 3.3uf capacitor together with the 47k resistor to ground will form a high pass at 1Hz. This will not roll off the audible bass as even 40hz is over five octaves above 1hz and first order filters roll off at 6dB per octave.

Doing all this will mean that I won't have two high pass filters in the signal chain before the tweeters, only one. In effect that will amount to the removal of one filter (less evil) as everything else is the same (the coupling caps in the mid bass circuit have just being moved from source to amplifier.

The only fly in the ointment MAY be that any DC going through a volume control MIGHT make it noisy in operation. But I won't know that for sure until I get a volume control. Suck it and see as Doc would say.

He advised me that doing a tweeter high pass at line level was no better than doing it at speaker level. I don't know what his motive was there - maybe he thought I was talking about opamp based active line level filters or the fact that the tweeters have no protection anymore. Anyway, I will never know. What I do know is that the music is better doing the high pass filter passively at line level and all it required was adding a cap at the amplifier input. Because of the much larger resistor to ground (47k compared to the speaker's 8 ohms impedance) I get to use a much smaller capacitor. In this range there are polystyrene caps available which are said to be the ultimate for passing signal

It's a win win for the music, a total no brainer