I want to build a pair of x1000s.

Status
Not open for further replies.
Jump to Latest
#21
Clarkcr

C please be serious,

If you build an Aleph 1.2 you don't save money by using 2 mosfets less per side.

If you want to save money don't bother building the 1.2's.
Just wait till you hear the cost of heatsinks and as NP put it "heatsinking and ps capacitance is never to much".

By using fewer mosfets you don't increase bass.

A high quality amp like the Aleph never "increases" or "decreases" bass - it just reproduces music the way it was recorded.

Mosfets benefit a lot by high bias current and since 10 or even 8 IRFP240's per side can do the task, all you have to do is to see that they don't fry.

The original 1.2 was biased at about 4 Amps. To build the 1.2 and bias it, say, at 6 Amps, you need heatsinks at least the size of the ones used at the XA200 or equivalent. Take a look at the XA200 and you will understand.

Add to all this the fact that PassLabs uses TO3 cased IRF244's and not TO247 plastic cased IRFP240's as output devices and you will understand the need for laaaaaaaaaarge heatsinks.

That is, if you want your amps to last long without any reliability problems.

Good luck with your project, whichever amp it is you finally decide to build.

Nick
 
#22
I know....I didn't understand your post...sorry. I didn't understand why someone would willingly put fewer MOSFETs in their Aleph 1.2 build effort. I thought you were saying put fewer in than the original design, push them hard and listen to the great bass respoinse. But, I take it you were just saying they sound great period. I agree.

Can anyone send me the schmatics for the Aleph 1.2? I can't seem to find them on the PASDIY site.


C
 
#23
Calling the MOSFETs "FETs" was just sloppy of me. Though I could be nitpicking and say that both junction field effect transistor and metal oxide semiconductor field effect transistors are indeed F ield E ffect T ransistors.

The balancing act Petter has described may be an issue. I am interested in trying the design, but haven't done so yet.

I am also not too well equipped and only partly experienced in discrete circuit building. I put together a Mini Aleph by recalculating and rearranging the parts and using veroboard, with different type MOSFET, changind the input stage to Aleph30, different rails etc., but I don't consider this a major achievement. The Aleph just seems to be a forgiving design ;-)

This being said - a fair understanding of the principles at work and the general issues, of each part, capability of reading and understanding data sheets, as well as some equipment, and dedication, and some literature, can go a long way. Just go slow, safe, breadboard a lot, ask questions to the forum, etc. Just common sense.
 
#25
The reason to use fewer output fets is to allow you to experiment without risking blowing up massive amounts of parts. I should have first built my Alephs with only 2 output fets and bias to the same .4A per fet and test drive with 8* 8 or 64 ohm load. I could then tweak the front end until I liked it (with scope and maybe headphones). If I had a problem in the front end, It would only melt 2 fets and source resistors instead of up to 16. It wouldn't put out enough current to drive speakers with any real volume. It would current starve and distort beyond 2 watts peak, but you could test offset and stability etc. Then add the rest of the fets. Trust me, You don't want to have to rematch and replace output fets often when they cost $2.00 each and need to be matched with 50% cull rate and in sets of 8 That's $32.00 each time you touch one without a static wrist strap.


Heat sinks for my 2 channels cost something like $400 BTW.
 
#26
Is it the bias voltage that I want to be careful of? When you say fry them, I assume you're not talking about a physical mistake like a cold solder or mis-soldered connection, or short or some other goof on my part right? Are you talking about getting the resistance values wrong or the rail voltage or something like that?

Thanks for the advice. It's all stuff I want to hear before I commit to being the ginnie pig. I keep getting pulled back into building the X for some reason. I love it when people try to talk me out of something!!!

In regards to the X amp, what is it exactly that we lack in cracking the mystery? The bias values? Petter has the resistance values...the capacitor values aren't that critical nor the size of the transformer right?

C
 
#28
Beyond that though, what don't I know about the X amps that would lead to me frying components and melting metal?

I'm beginning to wonder if building the X 1000 is as simple as matching the FETs, setting the correct bias voltage, knowing what rail votage to use and being vewy vewy careful with the static energy. Well.....besides how amplifiers work and what rail votage actually is.....

Again, I ask all of you....what else do I need to know???? Nelson? Where will I stumble? Can Petter's version of a SuSY circuit be scaled to X 1000 size? Nod if I can do it, shake your head and my middle name is Aleph 1.2.

C
 
#29
Sometimes you just have to bite the bullet and buy some
extra parts and find out. It's not reasonable to want to
build what would commercially be an exotic and expensive
amplifier and then worry excessively about a couple of bucks.

(Mr Krab clutching a $20 icon)
 
#33
Mr. Krab

It's been a long time I think since HH's last post.
I wonder, would he post something like that as Mr. Krab?


(Sorry Nelson but I don't have a pic of Tracy)

Nick
 

Attachments

  • skylos.jpg
    skylos.jpg
    62.7 KB · Views: 387
#34
As is well known by now, I'm the kind of guy who prefers to throw a few parts together rather than fret over whether this or that will bite me. It's surprisingly rare that I actually burn up parts. I'm not saying that it never happens, but it doesn't happen anywhere near as often as you might suppose.
Armchair theorizing is nice for a while (actually, I generally ponder things while driving to and from work), but will only take you so far. Eventually you have to actually try things or face fear-induced paralysis.
Go slowly, use your noggin at each step, and don't hook a new design into your system until you've flogged it to death on the bench.

Grey
 
#35
One has to wonder what would you really need a 1000 watt amplifier for when you can biamp on a 1/4 scale or less for a better result.

Given the effort and resources involved, engineeing an active speaker makes more sense.....to me an audition of the Rushmore is proof of that. (courtesy of Passlabs)

Ian:nod:
 
#36
Happy Father's Day everyone.

Okay then, a few questions.

Why does Petter's X100 circuit contain P-Positive type MOSFETS as well as N-Negative type MOSFETS? Which are better for the amplification of teh audio signal?

Am I to understand that when the gate voltage is increased (the bias right?) more elctrons flow from the Source to the Drain, there-by the signal is amplified?

In run-of-the-mill amps these gate are actually closed to barely open most of the time, therfore they run cooler? In Class A they are wide open all the time? So in Push/Pull circuitry, the signal is being pushed and pulled by this constant manipulation of the gate voltage?

A while ago someone meantioned using zeners to protect MOSFETs from static electriciy. What is different about the zener diode than a regular diode? Does it allow electrons to pass if it is of a certain current? Kind of like a dam that opens when the water rises to a certain preasure? Where are these zener diodes placed in the circuit? I don't see them in Petter's X100 diagram.

I read last night about the X Series amps PCB boards. Double copper, double sided and even the hole were lined with copper. Surely this is a longevity selling point? Showroom selling point? Does this exotic (I don't know where I could get this done at) PCB add anything sonically to the amps quality?

Thanks for everyone's help. I know this sounds remedial, but I'm getting there!!


Chris
 
#39
Okay, I know, I re-read the service manual for the Aleph and saw that there are "N"elson type MOSFETs used as well as "P"*** types. I'm learning more and more about this and that is cool.

Why not all N type or all P type?

C
 
Status
Not open for further replies.
Top Bottom