Has anybody built an Aleph-x to this spec???

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

D3

Member
Joined 2003
Hi I would like to know if anybody is building or has built an Aleph-X to around this sort configuration..............

1000va 18-0-18 simple POWER SUPPLY 'C'
16 IRFP044N 8A BIAS.................................

I would like to discuss build issues etc ...............
:cool: :cool: :cool: :apathic: :apathic: :rolleyes: :rolleyes:
 
Well, if you are using only capacitors in the power supply (no resistors or inductors) then the rails will be just over +/-24V, to give a power output of about 120W into eight or four ohm loads, with its peak output somewhere around 5-6 ohms. This will also mean about 1A of bias per fet so you should be able to get away with rather large source resistors (of up to about one ohm without getting into power dissipation probelms, but around half that would probably be best) which will mean your FET's will not have to matched quite so accurately (which is nice).
 
Thanks for replying...

I am unsure what the rail voltage.........will be under load........
unloaded it sould be approx 18x1.4=25.2v.....................
I was thinking that loaded........after interpolating other threads...
I would be looking at 21 or 22volts..............given that I have quite a heavy bias................

This being the case I noticed thain the original Wiki page that the
IRFP044 was one mosfet suggestion - ideally not running over 20v.......however I belive this has been largley superceded with IRFP044N. From what I can gather.........the 'n' variation has a slightly lower voltage tolerance and lower capacitance.

Am I pushing the envelope too much voltage wise?????

At the end of the day this amp is for 'pleasurable' listening......
I don't wan't something unreliable......and likely to smoke!!!!!

....can the power supply be tweaked to bring me into the 'comfort' zone.....without losing performance................
or am I paranoid................and just go for it!!!!

Arghhh......

Kind Regards
Richard

(Thank you for your reply Mr Pass......nice to know highly regarded XA100 ..... is a similar spec)
 
Hi, I think you may have under estimated your rail voltage a bit there, as I have generaly found that trasformer manufacturers specify the load voltage for the secondaries, so that if you are drawing less than the rated power (which you will be at draw of around 300-400W from a 1000VA transformer). Therefore I would if anything say that your supply will be slightly higher than rated (with a typical regulation of the transformer being about 4% for that sort of size), around 26.5V (or 25.5V after the rectifier diodes).

This also means that if you do run the simple, capacitor only, filter on your PSU you will be dissipating around 400W, which even with the big heatsinks you have may be pussing the limits on safe opperating temperatures for the MOSFETS a little (but will probaby be ok, as the die temperature should stay below 100 degres C so long as your room temp isn't too high).

However, this does mean that you could include either resistors or inductors into your power supply filter to drop a couple of volts from the supply, giving a slightly lower rail voltage to your FET's (I think the 044's should be fine still on 25V though if you don't want to do this) and a slightly better smoothing on your power supply.

But, basicaly, I would say that as is the rails will be somwhere between 24 and 26 volts at your amp, and either way you should be fine running as is, as there will only be very marginal differences in power dissipation and the voltage shouldn't be too high for the IRFP044's either.

One question that I do have though, is if there is any reason that you only want to run capacitors to smooth your power supply? Is it just too keep things simple, or do you want to try nad get as much power as you can out from the design?
 
simplicity!

I am a firm believer in simplicity...............I sort of come from the school of american engine design..............why bother with 4 valves percylinder and complex cam shafts.............when a simple large capacity engine..........will do the job..................
nascar engines verses formula 1.................

My idea was to use a large transformer and 6 of the large 68,000 caps+ your bridge per monoblock................for a strong basic relatively low ripple supply................however I didn't want to fry my mosfets......



Kind Regards
Richard
 

Attachments

  • d3 amp.jpg
    d3 amp.jpg
    35.9 KB · Views: 1,595
something like this

My idea was to use a large transformer and 6 of the large 68,000 caps+ your bridge per monoblock................for a strong basic relatively low ripple supply................however I didn't want to fry my mosfets......

You mean something like this? the suply delivers about 25 volts
to feed 16x IRFP240 per channel.Im aiming at 6.7A of bias per channel.

grtz Jojo
 

Attachments

  • topview2.jpg
    topview2.jpg
    57.4 KB · Views: 1,400
Hey jojo, I was messing around with layout ideas for my amp on sunday and I came up with exactly the same arrangement as you have in your pic. I can imagine a few other people doing it this way as well now.

sorry D3 to clutter up your post with something not exactly relevant. :whazzat:

Thanks
Mark:)
 
My amp

Thanks everybody for posting..................

I have been drawing up my Aleph monoblocks on CAD....when the drawings are a little more polished I will post on here.............

However........I notice that jojo's amp is a very similar layout!

kram.....I might be able to get a reasonable deal on metal work......if you use a similar design to myself.....................
I will try and post it later today.................

Cheers everbody .......keep posting!!!!!
 
Ok Richard, as requested one power supply schematic for your amp. Now, I have assumed a few things when calculating the values for the components here:

First, the rails after rectification will be at 24.5V and you wold like them to be 20.5V (so we have to lose 4V in the supply some where)

Second, you will be drawing 8A bias curent (this is what you stated earlier and I assume you haven't changed your mind)

Third, you don't want your resistors in the supply getting too hot, so I have suitably over rated them for the job.

Ok, with all of that out of the way (and looking at the schematic below), you can see how the transformer will be wired to the +/-AC points on the rectifier PCB; with the two centre taps joined together and atached to your star grounding point. You will then need to run a line from this star ground point to the 0V section on the capacitor part of the power supply (you could use this point as your star ground, if you are going to use a large aluminum plate like a lot of people seem to do, that way you won't need as many cables in your amp).

Now, as you want to drop 4V accross the supply and you will have 8A flowing in the circuit, that will mean that we need a total of 0.5ohms in series with each supply rail. Now, I mentioned in the e-mail about tring to optimise these vaues and their placement, but for now, I will simply say that the total resistance should be split equaly between the two resistos on each rail (R1 and R3, or R2 and R4), so that you will have 0.25 ohms for each one. Also, with 2 volts accross each resistor and 8A flowing through them, each 0.25 ohm resistor will need to dissipate 16W.

Now, as 0.25ohms is not a standard value and 16W is quite a high rating, the best way to mke these resistors up, would be to parallel a number of resistors together to get the desired result. There are two main combinations which could be used for this, firstly, four 5W 1ohm resistors to get exactly 0.25 ohms and have a bit of spare dissipation. The second would be to use two 0.47ohm 10W resistors which will give a slightly lower 0.235 ohm resistance, but will still drop nearly 4V overall. The choice is up to you in terms of cost and space as to which combination you go for (you could even use 0.22 ohm 25W metal clad resistors if you were to mount them on a suitable heatsink surface), but all of the ones I have mentioned above should work fine in reducing your supply rails to close to what you want as well as helping to reduce the supply ripple you will get.

If you have any more questions though, feel free to ask and I will try and give some sort of useful answer.

andrew.

P.S., I will test the bridges in so much as they have the correct conections made and so forth, but I haven't got any real means of testing them at the 8A supply current you will be using them at, as I haven't got anywhere at all in putting my supply together yet to test it (the woes of only having one room to live in, and too much junk).
 

Attachments

  • d3-psu copy.gif
    d3-psu copy.gif
    16.8 KB · Views: 1,084
Thanks!!!

Thanks BP.........you really have helped I ideally want to get to arround 20v to get my IRFP044N back in a realistic operating zone.

Will this supply give good ripple results..........from the threads I have read it should work pretty well.

Cheers, cant wait to get the bridges.....by the way how much heat do you think will be emitted by the bridge???

Cheers
D3
:devilr: :devilr:
 
Hi, for the bridges, with your 8A draw, you should be looking at about another 16W in total from all four diodes, so you shouldn't get any problems with this.

As far as the ripple goes, I guess the best way would be to simulate the circuit in some software, however, I have never tried to do that before (I guess I should learn at some point) so you may need to find someone else to look at it for now, but I imagine it will be very low anyway.
 
SILPAD

Mr Rollins

Should there be more thermal mass than just 'silpad' .....would it be a good idea to use a thickish section of aluminium....to give thermal mass......or do the two transistors in question....need to be directly...........

I seem to remember reading that this was one possible solution to thermal instability....
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.