| yoke |
Hi I have plan to build some active loudspeakers and would like to use Aleph X for bass , Aleph 30 for mids and mini for twiters...
Now I have build case for it, which is based on Fischer SK 41 heatsinks. In spreadsheat it states that c/w at lenght of 150mm is 0.28...muhahahahhhhaaa...
In my case the temperature at heatsinks is 72*C...
on picture is one cahnel of Aleph X with 8 mosfets per chanel, 18V and 4A bias ( 1A per mosfet )
so is this 72* too hoot for mosfets ???
how much can it take ???
I know that it's recomended not above 60-65 , but :-(
Thanks |
|
|
| yoke |
| and one picture of case ( inside would be Aleph X,30, and mini ) |
|
|
| GRollins |
Might be time to consider forced air cooling.
Grey |
|
|
| AndrewT |
Hi,
18V @ 4A is only 72W.
The temp rise for a 0.28C/W sink should be about 20Cdegree.
Measuring 72degC indicates there is something far wrong.
What is ambient?
Could there be oscillation?
Is the airflow blocked, maybe at the bottom? |
|
|
| moamps |
Hi yokash:
If each transistor draws 1A and the voltage is 18V then the total dissipation for 8 transistors is 8x18x1=144W. If the heatsink has a coefficient of 0.28 deg C/W, it will heat up at 40 degrees above the ambient temperature (in theory). In practice, things are much worse. From my experience with Pass amps, your chassis might accomodate maximum 1 channel of AX or Aleph 5.
Regards,
Milan |
|
|
| Blues |
| You can reduce your bias to about 0.75A for each transistor or about 3A total. This will theoretically give you about 30C above ambient. Total dissipation for a channel of your X will be about 110W. |
|
|
| AndrewT |
post1 pic shows four devices on each of the two sinks.
But are they really 6inches tall? |
|
|
| yoke |
Hi all...
I have problem with Internet at home, seams like my router have drop dead ...
so...I will not access to web all the time...( writing from office )
The heat sinks are fischer SK41 at length of 150mm ~ 6 inch...ambient at around 20*C...
AndrewT _ I don't think that there is any thing oscillating, it's just the bead spreadsheet from fischer...I believe...
here is the tread I ask about those heat sinks
http://www.diyaudio.com/forums/show...&threadid=76428
O.K. the butom of heat sink do lay on the floor but , could that relay affect that much ? ( I mean all amplifiers that I see a not far from surface, less then 1 inch )
How tall should I rise it from the floor ?
I have place the top sheat of case just to see what I can expect if I add more surface...the case would be all aluminum, as you can see from pictures...
so now with this one, temperature drops to 64*C...which is not that bad :D ...I hope :angel: |
|
|
| vuki |
| Friend and I made AX with SK110 (150mm, ca. 0.35C/W) and 4 IRFs on each heatsink. There are approx. 90W dissipated on each SK110 and the temperature is ca. 60C with room temerature at 22-24C. |
|
|
| yoke |
Gray ... could you please tell me, if you know, what resistor value should I use to set ac current gain to 60%...
It says that I should place trim pot in place of R12/R34 ( schematic from wiki tread revision 1.0 )...but last time I played with this I manage to burn one of my output mosfets :-(...
well, at lest should I go below this 1.2 or higher ?
since this will be driving bass driver I think that I can sacrifice sound quality regarding power, at 60% I will get more power to control bass ... am I right ? |
|
|
| yoke |
| quote: | Originally posted by vuki
Friend and I made AX with SK110 (150mm, ca. 0.35C/W) and 4 IRFs on each heatsink. There are approx. 90W dissipated on each SK110 and the temperature is ca. 60C with room temerature at 22-24C. |
hmmm...will try to rise heatsinks from the floor, maybe that's the problem... |
|
|
| yoke |
| quote: | Originally posted by moamps
Hi yokash:
|
Hi moamps...will have to bug you regarding mox crossover...still waithing for bass driver, so I'am not in the hury...but soon thay will come
one picture of speakers..work in progres |
|
|
| Zen Mod |
| quote: | Originally posted by yoke
Gray ... could you please tell me, if you know, what resistor value should I use to set ac current gain to 60%...
It says that I should place trim pot in place of R12/R34 ( schematic from wiki tread revision 1.0 )...but last time I played with this I manage to burn one of my output mosfets :-(...
well, at lest should I go below this 1.2 or higher ?
since this will be driving bass driver I think that I can sacrifice sound quality regarding power, at 60% I will get more power to control bass ... am I right ? |
even if AC gain setting procedure is pretty simple-I think that's easier to bug ( ;) ) Moamps or Vuki via phone to make that clear for you ;
besides-why I have impression that lower AC gain value (less than 50%) can be more desirable :clown: ?
anyway- it's safer to replace R12/34 with resistor and trimpot in series ( instead ,say,1k2 - 680E resistor + 470 multiturn).....in that case you can't burn anything........
to Vuki- I never say this-but I really like your A3 bastardz :devilr: :clown: |
|
|
| moamps |
| quote: | Originally posted by yoke
Hi moamps...will have to bug you regarding mox crossover... |
No problem.
I wouldn't want to discourage you but it seems to me that you should think very carefully before taking any further steps in cooling.
In the Fisher .pdf file you are referring to, the temperature graph for SK41 is obviously incorrect. If you look at the heatsink SK144 further down the same page, you'll see it has worse characteristics despite it being considerably bigger in size than SK41!! Something's definitely wrong with the data.
Besides, when heatsinks are placed next to one another, as they are in your case, you can't do your calculations based on the length of each heatsink separately and its power dissipation. In that case, you need to use the coefficient for the total length and figure in total dissipated power. Your situation is even more complicated because by placing the heatsinks next to one another, you have reduced the effective radiation surface of the sides which are closest together.
The metal chassis is of no great help here in addressing the cooling issues because the coefficient of the flat surfaces (horizontal ones in particular) is relatively poor. It will only take longer for the amplifier to reach its operating temperature. Also, the total temperature of the chassis will be even higher due to the components within the amp that tend to heat up the system, like rectifier diodes, transformers, resistors and such.
As I said earlier, I don't think your chassis can manage more than 1 AX channel without some sort of forced cooling.
BTW, the chassis is b-e-a-u-t-i-f-u-l!! You may wish to consider cutting some venting slits or holes at the top and the bottom.
The speaker box looks like a ton of hard work. Amazing.
Regards,
Milan |
|
|
| moamps |
| quote: | Originally posted by Zen Mod
to Vuki- I never say this-but I really like your A3 bastardz :devilr: :clown: |
What? Where? |
|
|
| AndrewT |
Hi,
I see some saying that the Fischer data for this sink is possibly wrong.
I see the 50mm @ 0.65C/W.
For 150mm long Rth s-a = 0.65/[sqrt(150/50)] =0.65/1.7=0.375C/W not 0.28 as drawn. But this assumes that 0.65 is correct.
The correction of delta T is about 1.28 for 80 to 35 Cdeg.
That increases the Rth s-a from 0.375 to 0.48C/W.
Sit it on the floor and you will lose a bit more maybe 10% and the slight loss due to radiation another 2%. giving a guessed Rth s-a =0.55C/W , almost exactly double the starting figure. If the fin depth were very low, then sitting on or near the floor would be a small error. For a deep section that is half the height I would expect the floor correction to be quite large and the gap would at a guess be approaching half the section depth (say 30 to 40mm) to minimise air flow constriction.
The manufacturer has probably got the back surface exposed to free air flow as well including the devices bolted onto it.
It is important to carry the calculation and installation corrections all the way to conclusion, else you end up breaking something.
Electronics is either a.) copy a successful recipe exactly, or b.) get involved in a lot of arithmetic.
Of all the engineers, the Electronics Engineer is among the most mathematically competent. |
|
|
| yoke |
well don't worry about discouraging me :-)
I’m aware of problems you mention, about placing one sink next to another, but the problem was in bad spreadsheet...
Even If I didn't completely believe it, I thought that I still can get let's say 0.5 c/w...what would be 35*C above ambient for 70W...
but...
will see what I can do...
For now I raise heatsinks from the floor...around 4-5cm...temperature is 60* now...
I was thinking that placing one heatsink by another will make something like an a efect of chimney...when it heats up it will suck the fresh air...off course I don't have that much high for real effect but :-)
for now 60*C sound fine...will se tomorrow about buying some PC ventilator...even I don't like the noise it make...will drive it at small speed...
Here is one more picture of mid/high cabinet for loudspeakers |
|
|
| yoke |
| Thanks AndrewT for explaining calculation...0.48 seams realistic for temperature that I'm getting... |
|
|
| djucki |
Hi Yoke,
for heatsinks try at www.mali-sp.si
Price is a half of Fisher Electronic's.
Miroslav |
|
|
| yoke |
Hi djucki...
Cases are already built...as you can see...
no way I'm gona build'em again..maybe I would, if I could get rid of those :-)))))
they cost me 270 Euro in material...
I paid one SK41 around 15-16Euro ( best price I could get ) and aluminum 100 Euro + 35% for custom/taxes...
aluminum sheets are from MALI :-)... |
|
|
| Zen Mod |
| quote: | Originally posted by moamps
What? Where? |
Sorry Moamps......I meant on YOU and your Seven of Nine ;
mea culpa
I hope not maxima ;) |
|
|
| moamps |
It's OK. Seven of Nine (i.e. Aleph3 clone) sounds so good that I've almost forgotten about my valves.
Regards,
Milan |
|
|
| yoke |
Is it normal that bias drops as temperature rise ?
...in my case it drops from 4.5A cold to 3A hot ... ( well I have al. case ( on picture above ) which is not over jet, so it have no ventilation holes ) ... inside gets really hot ... when I remove top of case, the bias start to rise ... |
|
|
| Zen Mod |
| quote: | Originally posted by yoke
Is it normal that bias drops as temperature rise ?
...in my case it drops from 4.5A cold to 3A hot ... ( well I have al. case ( on picture above ) which is not over jet, so it have no ventilation holes ) ... inside gets really hot ... when I remove top of case, the bias start to rise ... |
it's normal |
|
|
| moamps |
That could mean that there is a big difference between the temperature of the output transistors when the amplifier is cold and when it is up and running. It might be a good idea to investigate things a bit further to see what's going on in that department. For instance, you may wish to measure the temperature of the output transistors. If it's too high, I'm afraid you're in for more trouble in the near future, most probably due to the issue of undersized heatsinks that we discussed earlier.
Regards,
Milan |
|
|
| yoke |
I believe that it has more to do with front end mosfets IRF9610...If the case is open there is more air around them so the temperature is lower. When I place the top, the temperature in case start to rise and so the temperature of front end mosfets...
I will place one small heatsink in be twin IRF9610...
If case is open...the bias stay at the sime level and after couple hours...( from 0.5V drop to 0.47V on source resistors which are 0.47Ohm or something like that )
Temperature on heatsinks,on closed case, are 64-65*C...( for testing I place 2 chanel of Aleph X - 18V 4A in one case ) |
|
|
| Nelson Pass |
| quote: | Originally posted by yoke
Is it normal that bias drops as temperature rise ?
...in my case it drops from 4.5A cold to 3A hot ... ( well I have al. case ( on picture above ) which is not over jet, so it have no ventilation holes ) ... inside gets really hot ... when I remove top of case, the bias start to rise ... |
The Aleph X has positive current sources for the output stage
which are regulated by NPN bipolar device junction voltages.
This voltage declines with temperature just like a diode, so
you expect maybe a 20% drop, depending on temperatures,
and of course with the cover off, the bipolars will stay cooler. |
|
|
| yoke |
Dear Mr. Pass thanks for explanation ... :D ...
I'm not an EE so I learn every day some thing new about those little monsters :D
I think that my bias is in be twin those 20% ... will check next time I power it up... |
|
|
| yoke |
One more question ;)
Here is the thing, I drive my Aleph X with 18V and bias at 4A, and have use 8 x IRFP240 per channel (instead of 4 IRFP044 as on original schematic)...
Now I have small room and ''small'' loudspeakers and some how I have impression that it's missing a bass...on some recording there is enough and on some I fell like it's missing...
So I’m in doubt is it because I use IRFP240 on this or is it just because of my current setup...?
The rest is very nice sounding, crystal clear hights and mids and only thing that misses me are just a little more low bass...
and I have use 2k2 instead 4k7 for R46/47...
Is there any thing I can do to get more bass (except playing with loudspeakers placement), optimize circuit for IRFP240 and 18V/4A???
I can't rise voltage, or I can, but will have to lower bias...the heatsinks could only take something like 70W ( 65*C for 4mosfets x cca18W )
Or is it that a need to rearrange loudspeaker placement?
I'am asking this bacause I have plan to use AlephX for bdriving bass in 2.5 active loudspeakers that are in proces of building...
Thanks
:smash: |
|
|
| Nelson Pass |
The Aleph X's are generally known to have better bottom
than the original Alephs, and I would expect twice as many
240's to outperform 044's. Have you had an opportunity to
measure bass performance in terms of power, damping,
frequency response and distortion?
:cool: |
|
|
| lineup |
| quote: | Originally posted by yoke
Now I have small room
and ''small'' loudspeakers
and some how I have impression that it's missing a bass
...on some recording there is enough and on some I fell like it's missing...
------------------
Or is it that a need to rearrange loudspeaker placement?
I'm asking this because I have plan to use AlephX for driving bass in 2.5 active loudspeakers that are in process of building...
. |
Most 'small' speakers in a small room
roll off at freq response
several times before
most any amplifier - even a 'small' power amplifier.
If anybody have a lack of bass
they should check up
IN THIS VERY ORDER
1. speakers bass performance, low and high cut-off
2. line level CD - preamp - power input R+C filtering effects
- that is: some resistor/potentiometer in combination with output/input caps
- can make frequency roll off, both in high and low end of audio band
- be upset and not what intended
3. see if the power amplifier can perform well within full audio frequency range
and a bit beyond this
-- if you have any problem, in point 1. and point 2.
anything you try in your frequency linear power amp
is not going to help
-- unless you want to have BASS BOOST filters in your power amp
lineup :) |
|
|
| yoke |
| quote: | Originally posted by Nelson Pass
I would expect twice as many
240's to outperform 044's. Have you had an opportunity to
measure bass performance in terms of power, damping,
frequency response and distortion?
:cool: |
I do remember those words so that's why I use 2XIRFP240 in place of 044 . (First I thought to build Aleph 4 or 5 so I have those 240 )
Unfortinaly I only have one fluke multimeter 87III...so before today I didn’t measure anything beside voltage, ac current gain and dc offset.
Now I have played test cd with sinus tones and this is the results that I get:
sinus tone Hz L (mV) R(mV )
12.5------76.8-----76.3
16------78.4------77.9
20------79.5------79.2
25------80.2------79.9
31.5------80.7------80.4
40------81.1------80.7
50------81.3------81.0
63------78.3-85 ------ 78-83 ???
80------81.6------81.3
100------81.6------81.5
125------78.3-85-----78-83 ???
160------81.8------81.5
200------81.9------81.6
250------81.1------ 80.4
315------81.6-------81.8
400------81.9------ 81.6
500------81.3------ 81.2
630------81.9------ 81.6
800------82.0------ 81.6
1000-----82.5------81.9
1250-----81.7------81.5
1600-----81.2------81.5
2000-----81.1------81.2
2500-----81.3------81.1
3150-----81.1------81.3
4000-----80.8------81.0
5000------80.8------81.4
6300------ 80.8------81.7
8000 ------ 81.4------82.5
10000------81.9------83.2
12500------82.2------84.2
16000------82.8------85.4
20000------83.3------86.8
How does it seams...the power suplay is not fineshed...I have two chanels on one toroid, will change that, I belive thats where the hum/brum is coming ( those readings on 63Hz and 125 ) I think ???
The preamp is S&B TX102 and I have 4.7uF on input and from ground to - ....
On the picture is my curent setup...distance be twin side wall and closet ( right on picture ) is something like 2.2m and from loudspeakers back wall to front is 4m...
Those are the loudspeakers http://www.audioepilog.hr/old.asp under chocolate to peyote ( 150W 87db 43Hz-20kHz 420mmx215mmx412mm 14.5kg ) - drivers are peerless HDS165 and scan-speak ( not shure which model - I forgot ) |
|
|
| yoke |
| lineup, thanks for replay...I tought to ask about the amp before I turn all room up side down :-) |
|
|
| vuki |
Did you measure your amp with speaker load or resistor load? If with speaker than measured voltage variation could be result of speaker impedance. Anyway - as measured your amplifier doesn't have problem with bass.
The speakers you have are a bit bass shy and bass is not TVC's strongest feature either. Try it with potentiometer volume control. |
|
|
| yoke |
agggrrgggghhhhh
I just disconected speakers and mesure ac without load...agggrrggghhh...
:headbash:
will mesure again...
so those mesurment I take, doesn't work ? there was no load exept multimeter...
as I mention before I'am not EE so please :D |
|
|
| moamps |
| TVC = transformer volume control |
|
|
| yoke |
| quote: | Originally posted by moamps
TVC = transformer volume control |
thanks
:) |
|
|
| yoke |
One more question...
those mesurment I take are not god ? ( since no load at the output )
So, I can take new mesurment with real load ( loadspeakers ) or I can place some resistors ? I have 8 x 75 Ohm 5W...in paralel 8x75 would give 9.37Ohm ... would that do ? |
|
|
| moamps |
| quote: | Originally posted by yoke
So, I can take new mesurment with real load ( loadspeakers ) or I can place some resistors ? I have 8 x 75 Ohm 5W...in paralel 8x75 would give 9.37Ohm ... would that do ? |
That may do. I'd suggest you measure the frequency response of your amp at the output voltage of about 5V.
Regards,
Milan |
|
|
| lineup |
| quote: | Originally posted by yoke
lineup, thanks for replay...I tought to ask about the amp before I turn all room up side down :-) |
Looks like your amp have like -1dB at 40-50 hertz
Quite normal, to me
Remember that for a loudspeaker that we buy
and is told to have frequency response 40-20.000
this value means -6dB data, mostly
and -6dB at 40 hertz, IS NOT A SMALL speaker
it is most often a floor stander with height 1 meter (3.3ft)
small speakers, monitor style, may even have -6dB at 70-80 hertz, at best!
and -12 to-15dB at around 40-50 point
Compare this to your amp with -1dB at 40 hertz
============================================
Many stepped volume attenuator controls ha got -3dB steps
because -2dB -3dB is around where we can hear any clear difference
My guess is your loudspeakers are way above -2dB -3dB value
while your amplifier obviously,
as your simple CD-tone Resistor load test shows this,
is good and normal in frequency response
it is such data we may get and expect
from a Nelson Pass construction
he doesn't make too bad amplifiers, you know
============================================
One thing that, in my opinion, is vital for low frequency play,
good bass in power amps
is POWER SUPPLY
and the single most important component in a power supply is TRANSFORMER
SMALL transformers will sag, drop in voltage a power delivery
at a bit higher power output
and especially at lower frequencies, say below 100-120 hertz
and certainly around and below 50-60 hertz
number 2 in important component in POWER SUPPLY
is Big Filter CAP = Electrolytic
the size and location of these
so especially if your transformer is not biggest type
you may be able to compensate this a little bit with having MORE CAPACITANCE
Nelson Pass has told us, that he has found
some rather big CAP or CAPs rather CLOSE to output power transistors,
is what he prefer.
That is at the PCB where power MOSFET are attached
But to ensure a lot of power available
at frequencies of 30-200 hertz
we should:
1. make sure our transformer is 3-4 times more VA rated than output power
2. see to we have enough much energy stored in supply caps
to deliver currents to MOSFET output
==============================================
The above was a bit about amplifier and BASS.
But as I told you,
your amplifier does NOT show anything bad, not normal, at those BASS tones.
so, as I have told before, You should make sure:
Point 1. speakers have good response of BASS reproduction level
Point 2. seems okay, as you used CD-player tones for your amp test
and so the whole line level chain in your audio system (anything before power amp input)
was tested along with your amplifier
Regards
lineup |
|
|
| Nelson Pass |
-1 dB at 40 Hz is not good enough IMHO. I suggest you try
to get a couple more octaves out of your bottom end, probably by
quadrupling the value of coupling capacitors.
:cool: |
|
|
| yoke |
thanks everyone for replay...unfortunate vikend is over...so I have to get back to ''normal'' life :-)))))...my girlfriend doesn't like all that wires on the floor :-) can't see reason why :-) so I will have to wait for another vikend...
in mean time, could some tell me what are the coupling caps in this amplifier :D
thanks |
|
|
| vuki |
| quote: | Originally posted by Nelson Pass
-1 dB at 40 Hz is not good enough IMHO. I suggest you try
to get a couple more octaves out of your bottom end, probably by
quadrupling the value of coupling capacitors.
:cool: |
I guess yoke measured it via his TVC, so this is where that -1dB came from and I doubt he can do much about it even with larger coupling cap.
Yoke: coupling cap would be that 4u7 at your TVC. |
|
|
| yoke |
I just look at the part list for AlephX and it only mention C12/13 as decoupling cap ( 47uF ), is that it?
...it's paralleled with power supply capacitors..
or does it mean that I need more capacitors in power supply ?
For now the power supply is not finished, I have 3x12000uF per rail...will have 3x22000uF ( 6x22000 per channel ) |
|
|
| yoke |
Thanks Vuki :D
and yes, I mesure trough TVC |
|
|
| yoke |
just quick report...
I remove 4.7uF from input of TVC and the bass is back :D
right now the aleph30 is playing :)
Thanks everyone for the answers and help...especially to the ''one and only''
back on slamming
:smash: |
|
|
|
