The man builds a class A amplifier, and he wants to add an oven to it?
Oi!
I'm not sure that more heat, per se, is the answer. Keep two things in mind: That once you get the critters in a chassis, they won't be subjected to vagrant breezes, and that as long as the DC offset is in the absolute realm (as opposed to relative), it won't matter as long as it doesn't get too excessive.
I said somewhere, perhaps in this thread, that I had considered using a copper or aluminum block to increase thermal mass. Never got around to it, largely because I'm not set up to work metal. The boards I've seen here aren't laid out to do that, so it may be a moot point.
Realistically, back-to-back mounting is a pretty good solution. Alternatively, you can go for flat-against-the-PCB in order to insulate the device and reduce air flow around it.
Yes, dual devices are a good solution, at least in theory, but then you get into other tradeoffs. That's not to say that it can't be done.
I never said I was done fiddling...
(Ahem...)
Grey
Oi!
I'm not sure that more heat, per se, is the answer. Keep two things in mind: That once you get the critters in a chassis, they won't be subjected to vagrant breezes, and that as long as the DC offset is in the absolute realm (as opposed to relative), it won't matter as long as it doesn't get too excessive.
I said somewhere, perhaps in this thread, that I had considered using a copper or aluminum block to increase thermal mass. Never got around to it, largely because I'm not set up to work metal. The boards I've seen here aren't laid out to do that, so it may be a moot point.
Realistically, back-to-back mounting is a pretty good solution. Alternatively, you can go for flat-against-the-PCB in order to insulate the device and reduce air flow around it.
Yes, dual devices are a good solution, at least in theory, but then you get into other tradeoffs. That's not to say that it can't be done.
I never said I was done fiddling...
(Ahem...)
Grey
Ya'll I don't kow what's the deal with all the "drifts".
Get a contractor-e to re-insulate your house or stop blowing on your differential pair.
And what is with this idea of the "thermal flywheel"?
No matter what you do you can't get away with the fact that ambient temp. change will change your offset point. You could conduct a study over the course of a year and measure the average annual temperature in your house and set the DC based on that so that in the winter the DC will be a little positive and in the summer a little negative (or viceversa) and perfect in the spring an fall.
If you are obsessed you could put a small piece of neoprene sleeve around the differential pair, the mosfets doesn't get hot at all anyways.
Get a contractor-e to re-insulate your house or stop blowing on your differential pair.
And what is with this idea of the "thermal flywheel"?
No matter what you do you can't get away with the fact that ambient temp. change will change your offset point. You could conduct a study over the course of a year and measure the average annual temperature in your house and set the DC based on that so that in the winter the DC will be a little positive and in the summer a little negative (or viceversa) and perfect in the spring an fall.
If you are obsessed you could put a small piece of neoprene sleeve around the differential pair, the mosfets doesn't get hot at all anyways.
Better still, use the DC offset to measure the ambient temperature in your room and put a small digital display on the front of the amp to display the ambient temp, heatsink temp, day of the week, and Miss March's measurements.
On second thought, the first three might prove too distracting. Just pare it down to that last set of numbers.
Grey
EDIT: For those who are seriously wanting to get the DC offset down to zero, you could always go the servo route. I worked a up a discrete circuit while I was building the prototype, but never posted it. Someone else posted a similar circuit later on, as I recall. It'll be in the original thread somewhere. If you can't find it, I'll try to post one. I believe I ran across a hard copy of that schematic recently, so apparently it survived the death of my C drive back in December. Or you can roll your own. Ain't all that difficult.
On second thought, the first three might prove too distracting. Just pare it down to that last set of numbers.
Grey
EDIT: For those who are seriously wanting to get the DC offset down to zero, you could always go the servo route. I worked a up a discrete circuit while I was building the prototype, but never posted it. Someone else posted a similar circuit later on, as I recall. It'll be in the original thread somewhere. If you can't find it, I'll try to post one. I believe I ran across a hard copy of that schematic recently, so apparently it survived the death of my C drive back in December. Or you can roll your own. Ain't all that difficult.
Rollin' your own..........
"Or you can roll your own. Ain't all that difficult."
It never is when you seldom post a schematic is it? I can only say I am reminded of the fish story. The number of designs that "got away" are truly alarming. Do you realize the amount of knowledge the forum is loosing due to these missing designs. Can you archive some of these please? "Ain't all that difficult " is little nonspecific for useful design advice........... Is a servo for the common mode or differential DC offset that is desired?
"Or you can roll your own. Ain't all that difficult."
It never is when you seldom post a schematic is it? I can only say I am reminded of the fish story. The number of designs that "got away" are truly alarming.
Do you realize the amount of knowledge the forum is loosing due to these missing designs. Can you archive some of these please? "Ain't all that difficult " is little nonspecific for useful design advice........... Is a servo for the common mode or differential DC offset that is desired? 
Mark A. Gulbrandsen said:
You don't have to go as far as an oven, but putting the whole front-end of the amp in an aluminum enclosure might work. I don't have any of these problems with my headphone amp, but it uses a dual FET for the input and the whole works is in a very poorly vented box, so the temperatures are very stable. As opposed to my 40W Aleph-X rig, whose DC params are all over the map.
Re: Rollin' your own..........
Damned if you do or Damned if you don't
Where is Jam with his sidekick cartoon?
Why should Grey post stuff here? Just to get crap from you and your compadre about parroting NP or stealing his intellectual property and endangering his livelyhood?
Awhile back I encouraged Grey not to post anymore circuits but to just send them to me and a few other friend on his list he agreed and I am happy with the results so far. I can vouch that to be the truth.
Damned if you do or Damned if you don't
Fred Dieckmann said:"Or you can roll your own. Ain't all that difficult."
It never is when you seldom post a schematic is it? I can only say I am reminded of the fish story. The number of designs that "got away" are truly alarming.
Where is Jam with his sidekick cartoon?
Why should Grey post stuff here? Just to get crap from you and your compadre about parroting NP or stealing his intellectual property and endangering his livelyhood?
Awhile back I encouraged Grey not to post anymore circuits but to just send them to me and a few other friend on his list he agreed and I am happy with the results so far. I can vouch that to be the truth.
I encouraged Grey NOT to post anymore circuits?
"Why should Grey post stuff here? "
Why not.......... everybody else does. I think circuit details like servos and circuit subsections are great things to post and are often useful in other projects. I have seen several good circuits and dozens of hints on the forum that I have thought were worth saving. Approaches to "generic" circuits like current sources, power supplies, cascodes, and servos stand little chance of IP violation and demonstrate the posters attention to details and perhaps a novel way of doing something different from the standard cookie cutter circuits.
What I find annoying is the "no sweat ...... I've done that.......no big deal" claims that trivialize the design process and seem to serve no function other that self aggrandizing. If a circuit is that straight forward, what is the risk of posting it to help someone out? One or two of these episodes is fine....... people get pressed for time or might want to point to a resource that covers the subject in more depth than is practical here. But after at least half a dozen times, it is starting to sound like the same old story.
"Why should Grey post stuff here? "
Why not.......... everybody else does. I think circuit details like servos and circuit subsections are great things to post and are often useful in other projects. I have seen several good circuits and dozens of hints on the forum that I have thought were worth saving. Approaches to "generic" circuits like current sources, power supplies, cascodes, and servos stand little chance of IP violation and demonstrate the posters attention to details and perhaps a novel way of doing something different from the standard cookie cutter circuits.
What I find annoying is the "no sweat ...... I've done that.......no big deal" claims that trivialize the design process and seem to serve no function other that self aggrandizing. If a circuit is that straight forward, what is the risk of posting it to help someone out? One or two of these episodes is fine....... people get pressed for time or might want to point to a resource that covers the subject in more depth than is practical here. But after at least half a dozen times, it is starting to sound like the same old story.
*warning* this is my first big diy project, so please excuse me for my (probably) noob questions!
My A-X is running!!!
rails are @ 15 volts...
using grey's orginal plans/bom
Good and bad news, who wants what first?
Good: 21 mv dc offset from +out to -out...
11mv dc offset from +out and ground, and -out and ground
bad...
it soundsl ike a quieter headphone amp (the volume)..
but i may just be feeding it a bad signal?
it doesnt seem that any of the fets are turningon... its been running here for an hour and i have felt *no* heat... but i sense voltage at the fets...
the drain on the input gain fets is the -.015 you see below, the source is 3.9some volts...and the gate is -.001 v
Why arent they turning on?... any suggestions?
Output Fet Reading:
Ground to Gate on Q10 .670 v
Ground to Drain on Q10 11.74 v
Ground to Source on Q10 -.015
Same as above but on Q1...
.669 v
11.74 v
-.015
i'm using line level input [will this work if i ground the SHD to ground?]... or do i have to use mic level? if so... do you think that my problem is this?
ps... q1 and q 10 are matched
q2 and q11 are matched
q5 and q7 are matched q6a is matched to them too
is there a very simple circuit to change to mic level instead of line level?
/me wants to listen to his amp!
Thanks very much!
triple check to make sure all your active components are installed
correctly.
ie: the MPSA18's the 9610's the Zeners (D1) and the FETS.
do you have the righ jumpers installed?
the amp will run fine without an input source on the test bed.
if the FETS are turning on, you'll get heat immeadiately!
worry about DC Offset after the amp's working.
good luck,
m.
correctly.
ie: the MPSA18's the 9610's the Zeners (D1) and the FETS.
do you have the righ jumpers installed?
the amp will run fine without an input source on the test bed.
if the FETS are turning on, you'll get heat immeadiately!
worry about DC Offset after the amp's working.
good luck,
m.
I agree the rails are a wee bit low but mine starts to function at about 10 volts or so.... pay particuluar attention to which way you installed D1a as I had one in backwards on one of my boards and the amp will do absolutely nothing....ok, if you had enough beers that afternoon it'll sit there and smile at you!
You should be able to measure 9.1 volts across D1a! Even with low rail voltages it should be doing something assuming all parts are correct and are in correctly, and that the semis were not damaged upon install.
I made some measurements on my amp which runs 16 volt rails. I get exactkly 40 watts RMS into 8 ohms at clip. Hope it might be of help....
Q6a...G 6.8, D 3.83, S 10.92
Q1... G5.5, D 16.0, S .7
Q10... G 5.56, D 16.0, S 5.37
Q2... G -10.84, D .2, S-15.7
Q11... G -10.9, D.3, S-15.7
My absolute offset starts somewhat high, about .7 but settles down to about 23 mv or so after about a half hour to 45 min .... and sometimes running past into the negative voltage offset realm. This all depends on room temp, air flow over the amp, sunspots, close asteroid flybys, and how many beers you've had that afternoon.
Also if you cannot get the absolute offset down to 0 volts DC then you may need to lower the value of R-26 to about 150 ohms. I experienced this on both my boards and I've found a couple of other builders that had this same problem.
I moved my X boards off my computer desk and temporarily out onto my Dynaudios yesterday and was absolutely blown away by the sound of the X. Much more inner detail and timbre can be heard than with a normal Aleph amplifier. There is also much more slam and a better defined bottom end..... and all this from a 40 watt amp..... most impressive to say the least. I don't think anyone building them will be let down in the least.
Mark
You should be able to measure 9.1 volts across D1a! Even with low rail voltages it should be doing something assuming all parts are correct and are in correctly, and that the semis were not damaged upon install.
I made some measurements on my amp which runs 16 volt rails. I get exactkly 40 watts RMS into 8 ohms at clip. Hope it might be of help....
Q6a...G 6.8, D 3.83, S 10.92
Q1... G5.5, D 16.0, S .7
Q10... G 5.56, D 16.0, S 5.37
Q2... G -10.84, D .2, S-15.7
Q11... G -10.9, D.3, S-15.7
My absolute offset starts somewhat high, about .7 but settles down to about 23 mv or so after about a half hour to 45 min .... and sometimes running past into the negative voltage offset realm. This all depends on room temp, air flow over the amp, sunspots, close asteroid flybys, and how many beers you've had that afternoon.
Also if you cannot get the absolute offset down to 0 volts DC then you may need to lower the value of R-26 to about 150 ohms. I experienced this on both my boards and I've found a couple of other builders that had this same problem.
I moved my X boards off my computer desk and temporarily out onto my Dynaudios yesterday and was absolutely blown away by the sound of the X. Much more inner detail and timbre can be heard than with a normal Aleph amplifier. There is also much more slam and a better defined bottom end..... and all this from a 40 watt amp..... most impressive to say the least. I don't think anyone building them will be let down in the least.
Mark
ok...
everything seems to be in place correctly, the only one i'm not sure about is the D1A... its hard to see on the silkscreen layer so... here comes the schematic... ok seems to be in right...
lets see..
the (yeah i know the cathode is the + and the anode is -, but i forget on diodes) black ring side fo the diode measures...11.72vdc to ground... and the other side measures 2.594 vdc to ground... which is a 9.126 across it... so thats not wrong unless its in wrong, but it looks right to the schematic
i have an irf9610 in the place of q6a and am using J1a... should i use j1b? i just tryed j1b as wel, and it didnt help...
i'm feeding it a signal from a Telulex SG-100 function generator, sinewave @ 1,000 Hz 0.0 dBm, and i increase the dBm to +18, and nothing still happens... (fets dont seem to turn on), but i hear the 1khz wave through any speakers connected
MPSA18's look like they are installed correctly (or the halfcircles are on the half circile side of the transistor)
everything seems to be in place correctly, the only one i'm not sure about is the D1A... its hard to see on the silkscreen layer so... here comes the schematic... ok seems to be in right...
lets see..
the (yeah i know the cathode is the + and the anode is -, but i forget on diodes) black ring side fo the diode measures...11.72vdc to ground... and the other side measures 2.594 vdc to ground... which is a 9.126 across it... so thats not wrong unless its in wrong, but it looks right to the schematic
i have an irf9610 in the place of q6a and am using J1a... should i use j1b? i just tryed j1b as wel, and it didnt help...
i'm feeding it a signal from a Telulex SG-100 function generator, sinewave @ 1,000 Hz 0.0 dBm, and i increase the dBm to +18, and nothing still happens... (fets dont seem to turn on), but i hear the 1khz wave through any speakers connected
MPSA18's look like they are installed correctly (or the halfcircles are on the half circile side of the transistor)
the black stripe on D1a should be to the left.
you should have a jumper across Q12a
a jumper across R48
a jumper across J1a
make sure Q5 and Q7 (diff pair) are electrically isolated.
MPSA18's can be a little delicate, you don't want to over heat
them, causing failure...
Be sure to double check your Power Supply, that you're getting a
good + and - rail voltage...
if none of this works, post a picture...
m.
you should have a jumper across Q12a
a jumper across R48
a jumper across J1a
make sure Q5 and Q7 (diff pair) are electrically isolated.
MPSA18's can be a little delicate, you don't want to over heat
them, causing failure...
Be sure to double check your Power Supply, that you're getting a
good + and - rail voltage...
if none of this works, post a picture...
m.