Building an amp

[frd1996]

Famous Nelson Pass' bulb amp :) Oh, I'll be definitely watching your progress Fran if you decide to share it :) Best of luck :D

Frd.

[Fran]

Thanks lads.


OK, so the first bit then is the power supply. FWIW, what you actually are hearing through your speakers is the mosfets (or whatever) modulating the power supply - so in nearly every case the power supply is a big factor. Now if you cast your eyes back to teh schematic on the previous page, the PS shown there is very simple. A tranformer with 50V secondaries (so its going from 240V mains to 50V or thereabouts), a simple bridge, 2 big caps (he shows 15 or 20kuF) and a 1 ohm resistor. Now Nelson goes on to state that this power supply is fairly rudimetnary and can be improved on, but just to get you going, it will work AOK.

So very luckily, I have the transformer from my salvage amp (thanks jaybee) plus the box to put it in plus heatsinks I'm going to need later. So all I really need is the rectifier bridges. Now this transformer I have has 2 secondary windings, so I will have 4 wires coming out of it, a pair of 50V. Also the transformer is rated at 650VA - normaly split half and half, so 325VA per secondary which very neatly is pretty much what Nelson recommends. If you were buying such a transformer new, expect to shop around a bit and pay 50-80 or so.

I also have some bridges salvaged, and since I'm a believer in dual mono, what I will do is set up one winding per channel with everything after that separate for each channel. I'm short on big caps, but I have some I can at least try this out with and then get some if it sounds anyway OK. That way little money goes out until I know its worth it. This is actually an important point, because very few people have built this amp, and there are certainly no commercial reviews of it - it is experimental after all. With the F5 amp, it was reviewed commercially so I had a good idea that it should be good. All I'm going on here is the word of a few members on DIYaudio, Nelsons own coments and the comments of a few people who heard it at the burning amp festival in SF.

So I go and physically wire up the power supply. Connect the transformer to L and N of your mains IEC adaptor (with fuse and switch) and connect the earth to the chassis. No point in being dead for an amp! Turn on and verify the secondary voltage is 50V AC - remember to swtich the DMM from DC to AC!!! So assuming that is OK, and the transformer is OK and at the right voltage, next turn off again and wire in the caps and resistors as shown in the schematic. This combination is called a CRC filter - cap, resistor,cap and the values of each smooth the ripple after the rectifier to pure DC (or at least purer DC). See the following pic for what the votlage looks like before and after:



We want that ripple to be a small as possible, but for now, the CRC filter will suffice. So I have 2 15000uF caps and 2 6800uF caps of a suitable voltage rating to try out, so thats what I'll use. Not quite what is specc'd but will do for a tryout anyway. So what I ahve is 50VAC>>bridge>>65VDC with lots of ripple>>15000uF>>big 10R resistor>>6800uF and at the end of all that I have about 65VDC with no load. Right what Nelson suggests...


Fran

[Fran]

mmm, stikes me I should haev mentioned safety earlier. Look, you are going to be dealing with mains supply, and DC voltages >50V in this amp. You need to have a healthy respect and a desire to live - I'm not going to go into the safety aspects here, but please, be careful with this kind of stuff. We don't want to see anyone (or smell pork either)

[Adrian]

Great Stuff Fran, You're on a roll, keep up the excellent work.

Looking forward to your next update on this little baby.

Adrian

[Fran]

So the next bit is to wire up the actual "audio" bit of the circuit. So if you remember, I decided to go for the version that uses IXTH6N50D2 mosfets, because the voltage I had to hand suited it, but also, I didn't need an input capacitor. So heres the actual audio circuit:

ScannedImage-2.jpg

(79.51 KiB) Downloaded 329 times

So thats the next step. The circuit is so simple I decided to just wire it up using a little bit of perf board to support the components. See that output capacitor - that blocks the DC voltage from your speaker - so its absolutely necessary to stop your speaker getting ~25VDC straight from the power supply. A lot of people (me included) find it hard to get their head exactly around whats going on in a circuit like this. The way I figure it is like this: There's pretty much limitless supply of current (oomph) from this power supply. So your input signal from your preamp "modulates" how much voltage the mosfet puts out. You really have to think of the mosfet like a tap - with the music signal being the thing that turns on and off the tap. And that mosfet can do that quickly enough so that the same music that goes in comes out the far side, but bigger. How faithfully it does this, can be termed the distortion. But also how quickly it does it matters + a whole host of other things too.

So going back a step..... The lightbulb - this acts like a constant source of current to supply the mosfet, in this case you are getting about 1.2A of current at ~25V or so. So there is a constant 25V DC sitting there, and that mosfet is constantly getting fed 1.2A. So what happens with that mosfet being fed that current, but just sitting there? Well it gets hot.... dissipates the energy as heat. So we need to cool that fet, and we do that by mounting it on a big ole heatsink. As I mentioned before I had these.... but heatsinks can be dear, and I have seen lads use al sorts of things from bits of engines to alloy wheels to CPU cooler fans to do the job. So if you are tempted, don't let that stop you! So in real life, as the mosfet passes some signal and makes the speaker do some work it will actually run a little cooler (eh, I think). Now, if you look up the datasheet for these things, they will tell you what temperature the fet is happy at, and luckily enough they are pretty hardy, usually surviving 150degC Ok - so if ours gets to say half that, we should be OK for long life. So what's the point of this long ramble.... well you need to sort out the heatsink!! Cannot be ignored. Look in the magazines - all the class A amps have big fat heatsinks down the side.

The job of the output capacitor (C3 in the drawing) is to stop all of the DC voltage and only allow the AC (ie like a sine wave) through. That's what a capacitor in series does. The size of the capacitance (along with the speaker) determines how low a frequency can get through... too low a value, say 100uF, and you might only get 100Hz, below that no signal. So we want a fairly large capacitor here. Nelson says 1000uF will get you to about 30Hz or so..... I'm using 6800uF so I should be going down to the low 20Hz or so. So this is one of the areas that will determine how this amp sounds and really you need good quality here. I'm using some Elna silmic audio grade caps, and I've done another little trick with them - I've added another smaller value higher quality capacitor in parallel with them - a thing they call "bypassing". So realistically, the big 6800uF capacitor must be an electrolytic type:



Normally these are not great for sound.... and often can affect high frequencies in particular and make them more harsh. So I've added some film caps as I mentioned above:



which should help sweeten the treble without making it too splashy. There's a bit of trial and error with this, so be prepared to try a few different types. The way I made up that littel circuit, I'll be able to swap in and out the filmn cap easily enough.

So more in another night or 2.....

Fran

[Ivor]

I must see if Google can translate this thread into English for me.

[JAW]

Wow! That amp's about as close to "The straight bit of wire wire with gain" definition of an amp as you'll get.

[Fran]

Ivor: where do you get them!!


JAW: yes, thats exactly what is so apprealing about this amp, very few components. In fact it is pretty much exactly the solid state equivalent of a SET valve amp, but even simpler. My guess is that some experimentation will be warranted as each component change is likely to affect the sound quite a bit. Of course there are some penalties - theres a relatively high distortion value, (still at least as good as a really well made 300B amp) and only a few watts per channel.

BTW, sorry my explanation was a bit garbled (well ok, more than a bit!) above, but internet died on me half way through so I was contending with flickering DSL lights etc. Still, hopefully its relatively amusing technobabble.....


Next up: actual construction - with pictures!!


Fran

[Ivor]

In fairness Fran I'm following it alright (just about!) but only by filtering up the more technical bits. Great thread, I might even be tempted at some stage.

[fergus]

I must see if Google can translate this thread into English for me.

LOL!!!....Nice touch Ivor!