I think you could still get it lower..... You're right on about blowing on the input pair and getting drifting readings! I lucked out and mine was very low at power up and was able to be brought down to a very low level. My input pair measuers 3.7900 and 3.7907. Q2 was 4.2464, and Q11 was 4.2469. Although Nelson says it only matters to two places mine being all the same out to three places may have made some difference. At any rate it was fun matching them up....
I stuck a 4" fan at one end of the sinks I'm using for testing so the temperature stays managable...the fins are face down right now with the board mounted to the flat back.
I haven't made any measurements except for DC offset at this point. My rails are plus and minus 15 volts. Will make more measurements tommrrow.
Now I need to get the second channel done!
Mark
I stuck a 4" fan at one end of the sinks I'm using for testing so the temperature stays managable...the fins are face down right now with the board mounted to the flat back.
I haven't made any measurements except for DC offset at this point. My rails are plus and minus 15 volts. Will make more measurements tommrrow.
Now I need to get the second channel done!
Mark
Hi!
I admit - probably it's a stupid question - but nevertheless - is it also important that the physical look af the amp is as symetrical as possible or not?
I mean - I've intended to build A5 in the first place - with a chassis similar with original Aleph series - 4 heatsinks as sides.
But now I just got lucky
and some guy I know had additional pair of Chad's PCBs (btw - great work, Chad!!!) - but if I use 4 heatsinks as sides - it won't "look" simetrical ...
Well - I know - like I allready said - it's a bit stupid Q - but I just had to ask ..
Thanks in advance
I admit - probably it's a stupid question - but nevertheless - is it also important that the physical look af the amp is as symetrical as possible or not?
I mean - I've intended to build A5 in the first place - with a chassis similar with original Aleph series - 4 heatsinks as sides.
But now I just got lucky
and some guy I know had additional pair of Chad's PCBs (btw - great work, Chad!!!) - but if I use 4 heatsinks as sides - it won't "look" simetrical ... Well - I know - like I allready said - it's a bit stupid Q - but I just had to ask ..
Thanks in advance
Heatsink Purchasing Thread has Moved
Titled Heatsink Sourcing (split from AXB thread)
http://www.diyaudio.com/forums/showthread.php?s=&postid=347702#post347702
Titled Heatsink Sourcing (split from AXB thread)
http://www.diyaudio.com/forums/showthread.php?s=&postid=347702#post347702
Ok guys you got the DC offset done, time to check the AC current gain now.
I mentioned this time and time again. Be sure to box-up the amp in such a way as to be able to reach the various trimpots for final adjustment when the amp T is equilibrated. The process is iterative.
I'll add my YAHWN in here too.
I mentioned this time and time again. Be sure to box-up the amp in such a way as to be able to reach the various trimpots for final adjustment when the amp T is equilibrated. The process is iterative.
I'll add my YAHWN in here too.
Huh!
You pups think you've got it rough, gettin' the DC offset to calm down, what with blowin' on the differential an' all...ain't ya got nuthin' better to blow on?
When I started this whole thing, the prototype had the two IRF9610s an inch or two apart, an' no case around 'em, an' the draft from the main heatsinks was drifting over 'em on acounta the fact that the majority of the circuit was sittin' on top o' the main heat sinks with the outputs right underneath...
And you guys think you've got it bad!
Harrumph!
Worse yet, when I was a sprout I had to walk five miles in the snow, uphill both ways...
Grey
P.S.: Although it won't really help with the thermal issue, don't forget the other two pots. They come in handy for fine tuning the DC offset.
You pups think you've got it rough, gettin' the DC offset to calm down, what with blowin' on the differential an' all...ain't ya got nuthin' better to blow on?
When I started this whole thing, the prototype had the two IRF9610s an inch or two apart, an' no case around 'em, an' the draft from the main heatsinks was drifting over 'em on acounta the fact that the majority of the circuit was sittin' on top o' the main heat sinks with the outputs right underneath...
And you guys think you've got it bad!
Harrumph!
Worse yet, when I was a sprout I had to walk five miles in the snow, uphill both ways...
Grey
P.S.: Although it won't really help with the thermal issue, don't forget the other two pots. They come in handy for fine tuning the DC offset.
...i got the little critters settled down and acting right. Once
they're snug as bugs in their case they should be fine.
funny thing was, i was working on the amp during a warm spell
here in FL, and the air conditioner was on... i noticed the readings
were changing after the amp had been running for over an hour...
turns out the AC duct overhead was blowing on my table! That's
when i figured out the blowing on 'em thing... (i don't think i'm
ever going to type blow in a post again)
((yes we run the AC in early March here in Florida))
(((Global Warming??))) ((((must be the influx of DIY Alephs))))
they're snug as bugs in their case they should be fine.
funny thing was, i was working on the amp during a warm spell
here in FL, and the air conditioner was on... i noticed the readings
were changing after the amp had been running for over an hour...
turns out the AC duct overhead was blowing on my table! That's
when i figured out the blowing on 'em thing... (i don't think i'm
ever going to type blow in a post again)
((yes we run the AC in early March here in Florida))
(((Global Warming??))) ((((must be the influx of DIY Alephs))))
I haven't had any major DC offset problem with the first one. Will assemble the second this weekend and see how that one goes. I think 20mv is darn low! My input pair is fastened together with a thermal pad in between them.
Since I'm limited on heatsinks I will probably build a wind tunnel cooling setup for these two channels. I have a couple of 12VDC Pabst fans that ere really quiet running at about 8 volts.
Mark
Since I'm limited on heatsinks I will probably build a wind tunnel cooling setup for these two channels. I have a couple of 12VDC Pabst fans that ere really quiet running at about 8 volts.
Mark
Luxury... Bleeding Luxury.. I had to crawl to school with my sisters on my back in 3 feet of snow, uphill both ways...of course.
Actually I did have a question, the two IRF9610 trransistors used for DC Trim is it? Sgould they be back to back on a heatsink to thermally stabilize them to reduce drifting?
Anrthony
When I first brought up the second channel of my X the bias was not nearly enough to get things running properly, I could only adjust to a max. of .180 volts across R-5. I could only get about 3.5 volts across both R23, 25 with VR-2 at max, the other amp went way over 5 volts at this point with VR-2 at max. so this led me back to the CSS department. Made some checks and found all to be ok there, even the resistir values were right on. It was as though the 200 ohm pot was not 200 but a bit higher value, although it checked ok. So for kicks I paralleled another 330 ohm across R-26 and viola it works fine. Thus I ended up with Greys value of 150 ohms for the optional CSS, not the normal 330 ohm value the "stock" 9610 CSS uses. Is it possible that some parameter of this 9610 might be way off from the one in the other channels board? The suspect 9610 tests fine and is from the same batch as the first channels. Other wise with R-26 set at 150 all seems fine. Has anyone else experienced this?
Also is there a preferred way to make the adjustments.......
Should we adjust VR2 by setting the current draw of the diff pair for 20ma? Or just set it for 5 volts across R23,25....?
Should we adjust the bias to .5 volts by measuring across R5, R40, or across R6, R41?
Ultimately I think I'd do all of this with a distortion analyzer but it would be nice to have a firm place to adjust things to.... In reading back in posts there doesn't seem to be any definite way established........
Thanks!
Mark
Also is there a preferred way to make the adjustments.......
Should we adjust VR2 by setting the current draw of the diff pair for 20ma? Or just set it for 5 volts across R23,25....?
Should we adjust the bias to .5 volts by measuring across R5, R40, or across R6, R41?
Ultimately I think I'd do all of this with a distortion analyzer but it would be nice to have a firm place to adjust things to.... In reading back in posts there doesn't seem to be any definite way established........
Thanks!
Mark
Let's see...heat stability...what you want is mass, not surface area. Hence, a heat sink will actually increase your sensitivity to air currents because it'll act like a sail, catching air currents from the four corners of the Earth. For a while I was considering screwing the two front end MOSFETs to a block of aluminum or copper to raise the thermal mass. Just a rectangular block, mind you, not a heat sink. All bets are off if you're increasing the rail voltage in search of a 200W amp...you're going to need actual heat sinks. But the 'stock' Aleph-X with rails around +-15V or so is quite safe, as far as heat goes.
DC offset:
1) Absolute. To set the front end current source, run a meter from the output to ground. Adjust the pot until the output reads as close to 0V as you can get it. It'll drift a little. Go get a beer. Tweak it after finishing the beer. Get another beer. Tweak. Beer. Tweak. Beer. After a while, you will find that one of two conditions prevail. Either you've got the absolute DC pretty close to 0V, or you're inebriated enough that you no longer care. As long as you're somewhere close to 0V, your speakers won't care, either. And you didn't even have to buy them a beer.
2) Relative. Now measure the DC from the + output to the - output. If you used matched devices across the amp, that is the Aleph CCS devices match on each side and the gain devices match on each side, then this should be pretty low. However, for those of us who feel the need to obsess over something, the pots on the outputs will allow you a wee bit of control over the relative offset.
Or you can simply jack up the bias if that's what you want to do.
When you're done diddling the relative DC offset, go back and check the absolute offset again, as they can interact.
3) Neither absolute nor relative DC offset is going to have much effect on distortion until they get pretty far out of whack. Have another drink and don't worry about it. Or as Charlie Papazian puts it,"Relax, don't worry. Have a homebrew."
Grey
DC offset:
1) Absolute. To set the front end current source, run a meter from the output to ground. Adjust the pot until the output reads as close to 0V as you can get it. It'll drift a little. Go get a beer. Tweak it after finishing the beer. Get another beer. Tweak. Beer. Tweak. Beer. After a while, you will find that one of two conditions prevail. Either you've got the absolute DC pretty close to 0V, or you're inebriated enough that you no longer care. As long as you're somewhere close to 0V, your speakers won't care, either. And you didn't even have to buy them a beer.
2) Relative. Now measure the DC from the + output to the - output. If you used matched devices across the amp, that is the Aleph CCS devices match on each side and the gain devices match on each side, then this should be pretty low. However, for those of us who feel the need to obsess over something, the pots on the outputs will allow you a wee bit of control over the relative offset.
Or you can simply jack up the bias if that's what you want to do.
When you're done diddling the relative DC offset, go back and check the absolute offset again, as they can interact.
3) Neither absolute nor relative DC offset is going to have much effect on distortion until they get pretty far out of whack. Have another drink and don't worry about it. Or as Charlie Papazian puts it,"Relax, don't worry. Have a homebrew."
Grey
Yup.
One might go so far as to say he's the Nelson Pass of the brewing world. Knows all. Tells all. And does it with style, which is the difficult part.
He's got another book these days, slightly more advanced than the first one. Same accessible writing style. Same humorous approach. Seems to be a great guy. Accord him the same respect you would NP, John Curl, Charles Hansen, et. al. He's an original.
'Tis a pity there aren't more like him in the world.
Grey
One might go so far as to say he's the Nelson Pass of the brewing world. Knows all. Tells all. And does it with style, which is the difficult part.
He's got another book these days, slightly more advanced than the first one. Same accessible writing style. Same humorous approach. Seems to be a great guy. Accord him the same respect you would NP, John Curl, Charles Hansen, et. al. He's an original.
'Tis a pity there aren't more like him in the world.
Grey
Ok you beer drink'n dudes....
My whole problem stemmed from the fact that R32 at 332 ohms is way to high for either of my amps CCS's to bias on properly. I simply couldn't reach the 0 volt adjustment of VR2 at all, in fact one amp would barely turn on the devices and while the other would it would not turn them on hard enough to acheive the absolute 0 volt level, in fact it only made it down to about 1/2 rail voltage. I lowered R26 to 175 ohms and all is working peachy kean. One amp is running about 16mv and the other 22 mv offset.
Thanks Grey!
Mark
My whole problem stemmed from the fact that R32 at 332 ohms is way to high for either of my amps CCS's to bias on properly. I simply couldn't reach the 0 volt adjustment of VR2 at all, in fact one amp would barely turn on the devices and while the other would it would not turn them on hard enough to acheive the absolute 0 volt level, in fact it only made it down to about 1/2 rail voltage. I lowered R26 to 175 ohms and all is working peachy kean. One amp is running about 16mv and the other 22 mv offset.
Thanks Grey!
Mark
Grey,
These two boards are built as "stock" X amps. I used IR 9610's in the front end CSS. Since neither amps output would reach 0 volts I found that dropping fomr 332 ohms to about 175 ohms puts VR2 back in the center of its range. I might fiddle around with the ZVP3310's as I also have some of those floating around here... but for now they set up perfectly. I may have uncovered the same problem that another builder is also experiencing..... he could only zero out to about half the rail voltage.
Mark
These two boards are built as "stock" X amps. I used IR 9610's in the front end CSS. Since neither amps output would reach 0 volts I found that dropping fomr 332 ohms to about 175 ohms puts VR2 back in the center of its range. I might fiddle around with the ZVP3310's as I also have some of those floating around here... but for now they set up perfectly. I may have uncovered the same problem that another builder is also experiencing..... he could only zero out to about half the rail voltage.
Mark
I've got to wonder if potting the two input devices together into a small package would help the stabilization of the amplifier. Just blowing on these devices causes drift and the potting thought came to mind... create a larger mass that would be less affected by short term ambient temperature changes. Of course going too large would only add to the initial stabilization time, so some experimentation with this would be needed. Also, putting these devices in something like a crystal oven might be the real answer to fast stabilization and any long term drift.
Any thoughts on this.....
Mark
Any thoughts on this.....
Mark