Aleph-X builder's thread.

jarek said:
Taco, try to compare the ripple on the scope on power lines between to channels (before and after last capacitors). If both channels draw same currents than ripple on lines should be also the same. If not there is problem with psu (likely with capacitors).


Good idea, but I have no scoop :cannotbe:. But I can try measure the AC component on my rails with my DVM (or is this nog a good idea). But this exercise has to wait, until I find some time :).
 
I have recently experimented with my Aleph-x on a battery supply!
The results are much better than expected.It is very very close to the huge soundstage and liquidity of my 300bxls set amp but with very punchy and tight bass - better than the AC PS.
At the moment I can only get about 2-3hrs operation on the existing battery supply - but oh what glorious hours.My listening sessions don't last much longer anyway.I am tempted though to go get a pair of 75ah SLA's plus build a 2x8amp charger which could power up everything indefinitely for non-critical listening.At the moment it is 8x7ah sla's.
 
As I mentioned I have 8 SLA's of 7ah each in the battery supply +-12v.Each amp is biased about 4amps into 4 ohm 90db speakers.
I am experimenting with some caps near the mosfets.At the moment I am using 4caps of 680uf Black Gate N non-polar.I will try and go lower with some film caps.In my view you do need a little bit of capacitance even with batteries.Otherwise there might be a tendency to sound slightly bright .Of course just leaving all the original huge electro caps in negates some of the advantages of the battery sound.You can play around with various caps inthis case to suit your particular tastes/system .
 
Hi...

After some time, I will try to put my AX back in play...
So does any one have some idea what could be wrong when I measure 5V on output ( around 14 to 0 on one side and 9 to 0 on other )...

Whell as some of you mide remember, I have working AX ( or at least that what I think ) but could not set 50% AC current gain...I have replace R12 with 10K pot and I belive I over kill it with triming...I got white smoke come out of amp ( don't know from where...it look like from one output mosfet but when I check it, it was o.k. )...so I belive all output mosfets are fine...
Now I have replace 2x0.22 with 4x0.5 to easier get those 50%...

Now, when I turn amp I get 5V on output and 14-0 & 9-0 V, and I think it oscillates by the sound from speakers ?

any hint what could be wrong ?

:smash: :smash: :smash:

Thanks
 
Hi, I don't have amplifier in front of me at the moment, will have tomorow...so I really don't remember how I connect input ( maybe I place capacitor betwin 0 and - ... I think that's it )...
But previously it worked, I used to have cca 0.05V on output and betwin 0 and -/+...
 
Taco said:



Good idea, but I have no scoop :cannotbe:. But I can try measure the AC component on my rails with my DVM (or is this nog a good idea). But this exercise has to wait, until I find some time :).


Thanks to Duck-Twacy for using his scope, here are my findings:

Defective Channel:
=============
on both + and - rails I measure a ripple voltage of 2 V on the first cap (directly connected to the rectifiers). On the - rail it drops after the resistor to ~1 V. On the + rail it's still 2V !!!

Better working channel (Good is misplaced here :bawling: )
==========================================
directly after the rectifiers, I measured on the positive rail, 1V and after the resistor it is about 100 mV or so. The negative rail also had 2 V and this was after the resistor 500 mV.

Conclusion my powersupply caps are crap :bawling: and it is a miracle that things still work and that one channel is silent. Very good job Nelson!
 
Bad quality electrolytic caps

Taco,

As far as I remember you used in your AlephX those Nippon Chemicon "Stubby's", 68000uF/50Volt, as sold by a number of different (surplus?) dealers?

I think there was a kind of Dutch group-buy of those caps?
Considdering the price of those caps these must be NOS (New Old Stock) and would be of a manufacturing date some longer time ago (as you mentioned already yourself)
Do you know of any other bad experiences from the other group-buy members regarding the quality of those caps?

I'm asking because I was considdering buying a number of these caps myself but after reading your posts I don't know if I still should do this.
 
Re: Bad quality electrolytic caps

byteboy said:
Taco,

As far as I remember you used in your AlephX those Nippon Chemicon "Stubby's", 68000uF/50Volt, as sold by a number of different (surplus?) dealers?

I think there was a kind of Dutch group-buy of those caps?
Considdering the price of those caps these must be NOS (New Old Stock) and would be of a manufacturing date some longer time ago (as you mentioned already yourself)
Do you know of any other bad experiences from the other group-buy members regarding the quality of those caps?

I'm asking because I was considdering buying a number of these caps myself but after reading your posts I don't know if I still should do this.


Byteboy,

I used BC Electronics 47.000 uF 63V caps. I do not exactly know how old these caps are, but I think mine are older than 4 years.

I never took part of the Dutch stubby group-buy. You can use those stubbys but take care those caps stay cool, I think mine died because of the heat in the case (~50 C) and the age of the caps itself.

Kilowattski,

you can savely use the ZTX450 or a MPSA18 instead of BC550C. I am not sure of the BC550B, maybe the hfe is a bit low.
 
Irf540

Question for you guys....

I am going to build an aleph-x with ~30v rails. I intend to create a 'quiet wind tunnel' using 4 - 7" x 15" heatsinks. I have thoughts on perhaps electrically isolating each of the heat sinks from each other and the chassis. I am then proposing ~12 IRF540 for each of the 4 heat sinks without isolating the tab (no mica/silicone) for better thermal conductivity. I would like to bias it pretty hard - get the 'juice' flowin'

I have looked at the schematic and I see no problem electrically doing this. I am probably missing something. Pls lemme know if I am.

Also, does anyone have issues in not 'thermally coupling' the ccs and outputs? (eg. thier not being on the same heat sink)

(OH, why IRF540? Because I can get them cheap, and I am cheap!)

Thanks in advance
Marc
 
The thermal conductivity of a good silicone pad is comparable to the conductivity acheived with thermal grease, and better than a bare tab/heatsink junction. I see no reason to avoid them.

Is the extra 0.27K/W (Bergquist K-10) the killer issue in your thermal design?
 
Mr Pass and jwb,

Thanks for your reply. Yes, i did not mention that I would of course use grease, without question. What is your opinion/experience in mounting thansistors w/out insulators? Are the thermal benefits a real advantage or is it a waste of time and effort? It seems to me that getting the heat off of the chip itself as fast as possible would be of paramount importance. Basically..in a nutshell...is it a bad:hot: idea.

Thanks
 
mpmarino,

Instead of just hacking and conjecturing, you should sit down and work out the thermal problem. The devices you are using (IRF540) are much worse thermally than a proper power device like IRFP244. The 540 will get twice as hot for any given power level and heatsink, owing to its the higher R-theta of the (smaller) TO-220 package as compared to the (larger) TO-247. The question is: what is your per-device power dissipation and what is the device temperature for various devices and mounting methods? Does the desired device operate at a decent temperature or do you need a bigger one?

As for the circuit's performance, you should analyze whether or not it's ok to connect, electrically, the drains of all the devices. Is it OK to have multiple current paths to the drain? Does it cause any kind of unwanted loop? How will you keep your pet aardvark from licking the electrically-live heatsink?

You might want to go through the site and read Nelson Pass' advice about substituting devices for the IRF*44 series. There's a reason people use those. It's a really nice device.

Best,
jwb