| CheffDeGaar |
WAX, FAX, MAX, SAX, TAX, I miss the words !
Just let me introduce my anticipated Xmas presents.
For anybody here, it's just another pair of Aleph-X,
but for me two marvelous monoblocks of pure class A,
that just arrived in time to warm the freezy winter nights :).
As I said, I just miss the words... Don't know how to call them :
WAX - Wonderful Aleph-X ?
FAX - Fabulous Aleph-X ?
MAX - Marvelous Aleph-X ?
SAX - Superlative Aleph-X ?
TAX - Tremendous Aleph-X ?
Just don't know. Any suggestion ?
Anyway, a HUGE thanks to Nelson and Grey :drink: :drink: :up: :up: :up: !!!
May all your Aleph-Xmases be white !!!
Thanks for sharing such a great design !
I began with a pair of Zen V2, 30W, so I was tempted to say I knew
something great, especially about the "heavily biased mosfet
sound"... Was desperately wrong !! I couldn't believe my ears. The
new twins make me rediscover every record I own. Never experienced
such a multilevel listening. Everything is new to me, perfectly in
place and pace, nothing added, nothing removed, everything's
unveiled. Tigh and perfectly controlled bass, unreached medium
clarity, tremendous highs. Silences are silences, every recording
room characteristic is present, natural reverberations and
instruments extinctions are just perfect and natural. And the
imaging... Sorry, my english's too poor to describe what I feel.
Just WOOOOOOOOOOWWWWW !!!!!
And No, no Blue LED... Sorry, too cold color for such hot beasts.
And thanks again to Santa Nelson :hohoho: and Grey Claus :santa3: !
May be you would like to see a picture, so here it is :
Oooops, sorry, it's just the prototype, with messy wires everywhere...
So just let me introduce WAX (Left) and MAX (Right) |
|
|
| CheffDeGaar |
Rear view ?
Nice hefty babies : almost perfect cubes, 25x25x25cm
Just heatsinks with aluminium addons :) Ready for biwiring.
Unbalanced and balanced inputs, with a switch for selection
between them. At the bottom, mains switch, standby/full power
pushbutton (the small red point :) ), plus a 3.5mm stereo jack
input for external remote control.
WAX (Right) and MAX (Left) |
|
|
| CheffDeGaar |
Well, from the practical pont of view now, some characteristics
for the interested ones :
Differential DC offset around 10mV for both, and absolute DC
offset 'round 100mV.
8xIRPF044N per channel (4 per half-side), biased @ one small Amp,
950mA more exactly. I matched the mosfets for this current,
and even matched them on transconductance. Lot of time here,
since I had 60 of them...
Supply : a 2x18V / 250VA toroid, discrete double bridge of MUR1560
from OnSemi, followed by 2 x 22000uF/35V, followed by 2 x 2.2mH air
coils, followed by 6 x 22000uF/35V caps (THSA series from Panasonic).
Each electrolytic is bypassed by a 100nF Polyester SMD 2824 from Wima,
mounted on the copperside of the PCB. All of this giving a rough
+/- 21.9V rail voltage after the coils.
For the circuit itself : Nothing fancy here. I used the "TL431 current
source" ala Netlist ;) for the front end Mosfets, the BJT being
a MJE15030 (just 'cause I have a bunch of them). The "critical" resistances
in the front end are Welwyn RC55 0.1% with low tempco, Ian's magic resistors
;) being 3K65. Feedback resistors are 150K, and I didn't find on the scope
I had to use a small cap in the feedback path, although I had planned room
for it on the PCB.Power resistors from output to ground are 100R/7W, quite classical.
The power resistors gave me food for thought. I was somewhat reluctant
to use 5% or more power resistors with all the time I spent matching all
the mosfets. Dunno here if I was thinking right... So I began to look for
precision and low tempco power resistors. I found some, but the price
wasn't compatible with me. May be a kind of allergy... So I tried to
overcome this by paralleling low power low tempco 1% metal film resistors
(MRS25 from Philips, 0.6W @ 70°C). As I planned to bias the mosfets around
1 ampere, the theoretical dissipated power was only 0.5W. Thus I used 4x2R
small resistors as a 0.5R power resistor. Large safety margin, but it gives
a total of 128 resistors for the two channels :) More time to solder them,
but much much more cheaper than the 32 required power resistors !
I couldn't resist adding a "Cheff touch" to the design :) ... I added a
standby mode for the beast. Two relays add a 47K resistor in // with the
R11/V1 (and R33/V3) in Grey's original schematic. At standby, the total
resistance here is decreased, thus lowering the bias. In my configuration,
the mosfets idle at 250mA, keeping the monoblocks gently warm, and it lowers
the time to reach the cruise level when powered on.
All this stuff is controlled by some glue logic I designed. There's a mains
switch (and a soft start circuit for the power section), plus a very small
pushbutton that toggles the standby/full power mode. When actionned (from
standby to On), it immediately switches on the standby relays (removing the
additional resistance in the circuit), and 5 second later, the speaker relay
(God, I forgot to say there's a speaker relay...) is activated, just to let
the circuit the time to eliminate the transient behaviour when changing the
bias current. When switching from On to standby, the speakers relay is
immediately switched off, and 100ms later, the bias current is decreased.
In my case, the ridiculously small pushbutton is useless, since the
Monoblocks are triggered from and to standby by the preamp though a
5V pulse.
I first tested a "spaghetti" version, with wires everywhere, and it worked
fine (no listening...). Then I designed a PCB, trying to remove almost all
the wires, and trying to maintain the marvelous symmetry of the design. As
I previously built ZEN V2 monoblocks, idling at 4 amps, I knew how hot it
can be... So the PCB is designed with many many holes in it (almost 200
if memory serves), to allow a natural convection INSIDE the beast. I tried
to have all the components rated for high temperatures (105°C elecrolytic
caps for example) with a low tempco. The "equivalent power resistors" are
standing 3 or 4 mm above the pcb, not touching it, and a concentrated
"hole zone" is drilled underneath them. Although not quite necessary,
heathsinks have been used for the front end mosfets and current source,
just to help them to reach the global thermal equilibrium faster,
and to have a higher thermal inertia.
Here is a photo of the main PCB... |
|
|
| CheffDeGaar |
And from the copper side, with the small SMD bypass caps under the big
'lytics... |
|
|
| CheffDeGaar |
Another view of the main board, with the thermistance for ground to
chassis connection (near the bal/unbal switch). |
|
|
| Duck-Twacy |
Very, very neat.
And a merry Aleph-Xmas to you too! |
|
|
| CheffDeGaar |
Building the perfect beast (copyright Don Henley)
The base plate, 6mm thick, with holes for the air flow inside.
The "little" PCB here includes the soft start circuit, plus
the standby and relay logic. Heatsinks with lots of holes too,
but they were previoulsy used in my ZENs... |
|
|
| CheffDeGaar |
| Almost completed... |
|
|
| CheffDeGaar |
And the speaker relay arrangement... Not enough room on the main PCB
without destroying the symmetry... |
|
|
| grataku |
OH LA LA! Cheff!!
That's probably the best realization of an amp ever. It rivals the real thing. It's what I always wanted to do but never been able to! Congrats! |
|
|
| CheffDeGaar |
Gentlemen, thanks for the kind words :)
Nothing special here, I just took the time... No special tools, just a press drill and a belt sander. And a lot of "elbow oil" as we say here...
I just hope that it can give people the willing and desire of making their own A-X. A thousand times worth the time spent ! |
|
|
| wuffwaff |
Congrats Cheff,
very nice work!:up: :up:
I hope mine will look as nice as when they are ready.
Nice to see that someone else has already tried the trick using small resistors in parallel for the sources.
Do you have some more measurements / values? For example
sec. voltage of the transformer, dc after bridges (wich diodes do you use?), dc after chokes, ripple before and after chokes, value of chokes. Amount of caps before and after chokes etc.
Wich ac current gain do you use and what maximum power do they deliver?
william |
|
|
| tortello |
Very nice, Cheff!
The PCB recall old Tek stuff or Audio Research style: beautiful!
Marcello
Post Scriptum: but please, wrap spaghetti without helping with the spoon!! :rolleyes: |
|
|
| Andypairo |
Really beautiful, may I know how are made the aluminium corners?
Cheers
Andrea |
|
|
| Dennis Hui |
Cheff,
Your amps are really beautiful!
I'm wondering how you cut the holes on the top plates.
They give the amps a funky look.
I didn't notice any power devices mounted on the front heatsink.
Does it get warm at all?
I'm also curious if you needed to do anything special to get a
good thermal interface between the metal pieces with the
mosfets mounted and the heatsinks.
Merry X-mas,
Dennis |
|
|
| MikeW |
| WOW! THOSE ARE REALLY NICE.:D :D |
|
|
| CheffDeGaar |
Thanks all !
| quote: | Originally posted by wuffwaff
Nice to see that someone else has already tried the trick using small resistors in parallel for the sources.
Do you have some more measurements / values? For example
sec. voltage of the transformer, dc after bridges (wich diodes do you use?), dc after chokes, ripple before and after chokes, value of chokes. Amount of caps before and after chokes etc.
Wich ac current gain do you use and what maximum power do they deliver?
william |
Yes, the standby trick is great ! Works like a charm.
About the components, did you read post #3 ;) ? DC after the bridge is about 24.5V (loaded), and the DC resistance of the coil is about 0R7. For the measurements, I have not measured the ripple. I used the same arrangement in my previous ZEN, and there was no audible hum. And I still hear nothing ! Dead quiet.
The AC sensing resistors for the current sources (R12/R34 on Grey's schematic) are 931R. It was obtained by simulations I made, to get an AC current gain of 50%. I had hard times measuring the experimental value, since I don't have a true AC voltmeter. I did it on the scope, and get a rough 52%, modulo the errors. But it doesn't worry me a lot :D
And finally, concerning the output power, I actually didn't measure it. It plays as loud as I want, and I don't wanna go deaf :). Both simulations and your spreadsheet gave a rough 55W/8R in theory.
| quote: | Originally posted by Dennis Hui
Cheff,
I'm wondering how you cut the holes on the top plates.
They give the amps a funky look.
|
Just with this kind of tool (38mm in diameter)
| quote: |
I didn't notice any power devices mounted on the front heatsink.
Does it get warm at all?
I'm also curious if you needed to do anything special to get a
good thermal interface between the metal pieces with the
mosfets mounted and the heatsinks.
Merry X-mas,
Dennis |
First question, yes, they do get hot ! Only the bridge rectifiers are mounted on this heatsink, but there's a 1" aluminium square rod with multiple screws joining the heatsinks, and it's hidden by the triangular aluminum corners. This rod allows a heat transfer between the lateral heatsinks and the front one.
Second question, second answer. I used a graphite thermal film between the mounting plates and the heatsinks. Characteristics are here. It's a bit messy to work with (dirty and fragile), but it helps to fill the gaps between the two metal surfaces.
| quote: |
may I know how are made the aluminium corners? |
That was the trickiest part for me. As I said, there's a square rod at each corner. Its length is the total height of the heatsinks, minus 2x6mm, the thickness of the two triangular "captions" at the top and bottom. These rectangle triangles are drilled in their thickness (perpendicular to the rectangle sides) to be bolted on the two heatsinks they join, with M2 screws. Two other holes are drilled (perpendicular to the hypotenuse :confused: ), and M3-threaded, to welcome the front screws. A 1/8" aluminium plate with bevelled edges (very very time consuming to do only with a belt sander....) hides the square rod.... Am I clear enough ? Should do a drawing, but I'm too bad at this game :(
Thanks again, and merry X-mas :) |
|
|
| moe29 |
simply amazing! what excellent work :nod:
very cool design
A+
(what could you possibly build after that?) |
|
|
| Peter Daniel |
| quote: | Originally posted by moe29
(what could you possibly build after that?) |
A GainClone? ;)
But seriously, very well implemented design. Just looking at the pictures, I'm developing desire to build another Aleph X (and I will definitely do).
I've been using that hole saw tool to cut 1" openings in 1/4" panels with very good results, just use some lubricant. |
|
|
| jam |
Cheff,
Amazing work!
Peter Daniel take note....................;)
Regards,
Jam |
|
|
| Peter Daniel |
| quote: | Originally posted by jam
Peter Daniel take note....................;)
|
Since you asked....
I would be tempted to make all sides symetrical, this meaning that rear corners would look the same as front. It would require a more complicated back panel, but it would further improve good looks already ;) |
|
|
| Taco |
| Very neat amplifier, compared to mine (I still have to take some pictures) it's the difference between wires and no wires :D. |
|
|
| jam |
Peter,
You have a point, I probably would have used a heatsink for the back panel and used Mr.Pass's idea of milling out a few fins to mount the connectors, as he did in the aleph series of amps. Very fine work none the less.
I am still amazed at the quality of work that some of the members produce. Almost makes you want to go out and buy a CNC milling machine.
Regards,
Jam |
|
|
| Da5id4Vz |
| Time for a Bridgeport Group buy? |
|
|
| CheffDeGaar |
Et in terra pax :)
| quote: | Originally posted by moe29
simply amazing! what excellent work
(what could you possibly build after that?) |
Thanks ! And looking towards an Ono... Time is on my side :)
| quote: | Originally posted by Peter Daniel
But seriously, very well implemented design. Just looking at the pictures, I'm developing desire to build another Aleph X (and I will definitely do).
|
Thanks a lot, Peter ! But I'm not challenging you :D . Me little grasshopper. Still have to learn from the masters.
| quote: | | I've been using that hole saw tool to cut 1" openings in 1/4" panels with very good results, just use some lubricant. |
Yes, a lot!!! Hard times for cleaning the plates after.
| quote: | Originally posted by jam
Cheff,
Amazing work!
Peter Daniel take note....................;)
Jam |
Thanks Jam. May I have a note (or more) too :D ?
| quote: | Originally posted by Peter Daniel
I would be tempted to make all sides symetrical, this meaning that rear corners would look the same as front. It would require a more complicated back panel, but it would further improve good looks already ;) |
And here we go ! Once again, no, Peter, it's not a challenge :).
More seriously, I had the same idea, but it required toooooo much additionnal work, and, although the journey might be more interesting than the destination, I couldn't wait more to listen to music. And from where I stand, I can't see the rear of the amps ;)
| quote: | Originally posted by Taco
Very neat amplifier, compared to mine (I still have to take some pictures) it's the difference between wires and no wires :D. |
Thanks, and yes, WAX stands for Wondeful AX, but also for Wireless AX :D |
|
|
| grataku |
Cheff,
it definitely seems you like your drill press alot ;)
here is the question: why not mount the output devices directly on the heatsinks instead on making them go through another transition?
Oh yeah, how how do the little source resistors get? I am using 5w to be on the safe side and they get toasty. |
|
|
| Peter Daniel |
| Actually, if copper plates would be used here, this might be an andvantage (better heat transfer from devices). |
|
|
| Bricolo |
| quote: | Originally posted by Peter Daniel
Actually, if copper plates would be used here, this might be an andvantage (better heat transfer from devices). |
Peter, I used to work with copper, and it's even harder to use for visible (and touchable) parts.
The oxidation is much more visible, and fingerprints are awfull.
Do you use something like anoline, bur for copper? |
|
|
| Duck-Twacy |
| quote: | Originally posted by Taco
Very neat amplifier, compared to mine (I still have to take some pictures) it's the difference between wires and no wires :D. |
Taco, from what I remember your AlephXs looked quite neat too, maybe a little more "ordinary" (less fancy), but certainly not bad at all (functional is the word that comes to my mind).
Next time I bring a digicam ;) |
|
|
| Taco |
| quote: | Originally posted by Duck-Twacy
Taco, from what I remember your AlephXs looked quite neat too, maybe a little more "ordinary" (less fancy), but certainly not bad at all (functional is the word that comes to my mind).
Next time I bring a digicam ;) |
Thx duck, you're welcome and also listen the next time :D. |
|
|
| moe29 |
in looking closer, i noticed that the Inductors were mounted to the
PCB. I seem to remember reading somewhere that Mr. Pass
mentioned there was a problem doing it that way with the AX amps?
I've read too many threads to remember where i read that,
does anyone else remember reading that? or am i mistaken? |
|
|
| Nelson Pass |
In one of the Revs we mounted the inductors on the main
PCB but found that it made too much mechanical noise, with
the PCB acting as a sounding board. We had to relocate them
to the floor of the amp. |
|
|
| CheffDeGaar |
| quote: | Originally posted by grataku
Cheff,
it definitely seems you like your drill press alot ;)
|
Yep ;). But it's the only one I got...
Sure if I had a CNC mill, the drill press would feel a little bit neglected...
| quote: |
here is the question: why not mount the output devices directly on the heatsinks instead on making them go through another transition?
Oh yeah, how how do the little source resistors get? I am using 5w to be on the safe side and they get toasty. |
Well, I was first reluctant in doing so. Assuming I wanted no wires between the PCB and the Mosfets, I still found two reason for finally adding such a thermal bridge. First, while making my first Zens, the power mosfets were soldered on the PCB and directly bolted on the heatsinks. And it was a real pain to have the threaded holes in the heatsinks to be aligned with the mosfets... There was always one of them that was misaligned :(.
The second and more "serious" reason is purely mechanical. The PCB including the supply caps and the coils is quite heavy. And moreover, I even decreased its overall mechanical resistance drilling so many holes in it. The aluminium bar holding the power devices helps to rigidify (Is it english ?) the pcb with its three screws, and doesn't make the mosfets to hold all the weight of the populated PCB.
Source resistors now. Yes, I was afraid of the temperature they could reach. They have to dissipiate 0.5W @ idle. Four //ed 0.6W resistors give a 2.4W equivalent. So I should be safe... I made a little test feeding the 4 //ed with the current they will have to withstand, and, in free air, they were not very hot, I could hold a finger on them for a long time. I even investigated this with simulations on the whole circuit. They have shown that under the worst conditions (when reaching the full amp power with a continuous sine signal), the maximum (if memory serves) RMS power was near the Watt, i.e. half the maximum power. In the real life, I can hold my finger on them with no problems. Dunno why yours can toast bread..:) What power do they dissipate at idle ?
| quote: | Originally posted by moe29
in looking closer, i noticed that the Inductors were mounted to the
PCB. I seem to remember reading somewhere that Mr. Pass
mentioned there was a problem doing it that way with the AX amps?
I've read too many threads to remember where i read that,
does anyone else remember reading that? or am i mistaken? |
Haven't seen this post... So many on the AX... In fact, I was a little worried by the proximity of the coils from the current source mosfets and their steel bolts, but I don't know what happens in the real life... |
|
|
| CheffDeGaar |
| quote: | Originally posted by Nelson Pass
In one of the Revs we mounted the inductors on the main
PCB but found that it made too much mechanical noise, with
the PCB acting as a sounding board. We had to relocate them
to the floor of the amp. |
If you look at the picture on post #5, you will see a little grey thing at the base of the coil on the left. It's a little piece of plastic material, a little bit floppy and soft I salvaged long ago from a typewriter, and that I had in my junkbox. Dont' know what this material is, but it is fireproof (hard to melt with a soldering iron, impossible to burn with a lighter), and it I used it for two reasons : first, to avoid to damage the enamel on the coil wire on the PCB edges when strongly pulling the fixing ribbon, and second to damp the possible coil vibrations, since I had this little problem in my old Zens.
Seems to be efficient (no coil noise), but is it due to this material ? |
|
|
| moe29 |
Cheff,
it's evident that your amp was well thought out :nod:
good idea inserting that material under your inductors. |
|
|
| akb1212 |
| quote: | Originally posted by Da5id4Vz
Time for a Bridgeport Group buy? |
Oh, I already got one.... :D
I'm right in the midle of converting it to CNC now...... :cool:
If you want to know where I got the inspiration take a look at the frontpage of the second part of the A-75 article.
The thing now is that as the possibilities increases, so do the demand on design skill......
My plan now is to make my own heatsinks. Custome designed heatsinks :cool:
:smash: :smash: |
|
|
| Da5id4Vz |
Audio Accessories had a picture of a bench top mill in one of there old catalogs. It wasn’t CNC but did have digital readouts. It seems a much better way to drill the 192 holes for a patch field than a drill press.
http://www.patchbays.com/
I’ve been hooked ever since and often schemed for one at work but never succeeded. I think I finally have room at the house to do something modest. But there are a few other things needing to be done first. |
|
|
| Idefixes |
| quote: | Originally posted by CheffDeGaar
8xIRPF044N per channel (4 per half-side), biased @ one small Amp,
950mA more exactly. I matched the mosfets for this current,
and even matched them on transconductance. Lot of time here,
since I had 60 of them...
|
Juste one little question to clear about your lovely little blocs : Is the bias current you mention over there, per fet or total per bloc.
Thanks for your answer.
Marc |
|
|
| Mark A. Gulbrandsen |
Really nice professional job!! With all those vent holes in the pcb's it dfinately has a made by Audio Research look to it. Were you perhaps their pcb designer?
Mark |
|
|
| CheffDeGaar |
| quote: | Originally posted by Idefixes
Juste one little question to clear about your lovely little blocs : Is the bias current you mention over there, per fet or total per bloc.
Thanks for your answer.
Marc |
Hi Marc,
It's the bias per fer, so about 4 Amps per block.
| quote: | Originally posted by Mark A. Gulbrandsen
Really nice professional job!! With all those vent holes in the pcb's it dfinately has a made by Audio Research look to it. Were you perhaps their pcb designer?
Mark |
Thanks! And, believe it or not, I wasn't even aware of Audio Research's PCBs... I knew the make, but haven't paid attention to their pcbs... |
|
|
| Netlist |
| quote: | Originally posted by CheffDeGaar
I knew the make, but haven't paid attention to their pcbs... |
Maybe they could have a look at yours.
Very very nice amp.
/Hugo :) |
|
|
| CheffDeGaar |
| quote: | Originally posted by Netlist
Maybe they could have a look at yours.
Very very nice amp.
/Hugo :) |
Thanks Hugo. But you're one of those who started the fire ;) I whish I have all your instruments to measure something on my twins, but I'd rather listen to them.
By the way, I had a lot of requests in private e-mails (c'mon pals, don't be afraid to ask here...) for the schematic of the amps. Don't understand why, since it's quite classical now... But here it is. ASAX. Another schematic of an Aleph-X...
Zipped PDF, too large image for the forum. |
|
|
| CheffDeGaar |
Many requests too for the standby circuit... So here it is.
Just a few logic circuits put together, et voila :)
A triggering pulse is either generated by a debounced pushbutton, or by a pulse
coming from the preamp. These two activation sources are "summed" by a discrete
OR gate formed by D1, D2 and R2.
The working mode is slightly different in a stanby->on or on->standby sequence.
From standby to on, the pulse triggers a non-retriggerable monostable (U2a), with a pseudo-period of a rough
50ms (RA x CA). After this time, the first D ff (U3a) is toggled, and its output switches the standby relays
on, an the "On" Led as well. It also triggers the second monostable, with a pseudo period of 5s (U2b & RB x CB),
and after this time, the speakers relay D ff is toggled on. The relays and Led drivers are contained in an array
of darlingtons (U4, ULN2003), which includes free-wheel diodes for the relays.
The "shutdown" pulse acts a bit differently : it immediately shuts off the speakers relay
(through U3b's reset), and disconnects the standby relays 50ms after.
When the mains switch is activated, a simple reset circuit (RR, CR, and U1b) ensures a right
state of the circuit.
The "On LED" arragement allows to have a faint light when in standby mode (with the 22K resistor), and a full
light when on, through the 560R resistor. Tune these values to your likes, and according to the Led you're using.
The soft start circuit for the power transformer is not shown, but you you can find it here on the
forum. 120VAC/60Hz users should use a 150nF X2 cap for C1...
Just a raw estimation, so be careful and recalculate the values !!!! Letal voltages here.
Feel free to ask if anything is not clear. |
|
|
|
