|
Post by PinkFloyd on Nov 21, 2009 21:43:55 GMT
Here's how to do it directly on the chip.... just thought I'd take a few pics of the procedure as there don't seem to be any on the net. First off remove pin 6. Carefully bend over pin 5 to substitute pin 6: That's the OB. To perform the CA you will need a 1/4 watt 2k2 resistor (2.2K) You simply solder this between pin 5 and pin 7. That is now the chip in OBCA mode. Bend over the resistors, mount the AD-744 in the B'Dog and enjoy the music! Please do excuse my untidy looking way of mounting the resistor.... thing is, it doesn't have to look good folks it just has to be soldered into position
|
|
|
Post by prawn on Nov 21, 2009 22:31:35 GMT
Mike Using the output pin 5 (the comp pin) on the AD744 has been a standard mod for many years; it bypasses the output transistors on the chip and utilises the intermediate stage transistors which are biased internally into pure class A by an onboard 2ma constant current source. When the pin 5 is used there is a limited output current of about 2-3ma of pure class A (which is a DIYers dream) and often more than enough to drive a following buffer stage - if it's a very high impedance device such as the LME49600 this is about 30 times as much as needed to drive it fully. By loading the AD744 with a resistor from the positive rail as you indicate, you will (depending on your rail voltages) be adding another 5 to 10ma of bias to the internal transistors. This will put these intermediate transistors into tremendous strain and compromise the chip's life severely. By all means use pin 5 but forget the resistors mate!
|
|
|
Post by PinkFloyd on Nov 21, 2009 22:56:25 GMT
Hell Prawn, I am just following the "Xin" guidelines "blindly" I always used my AD-744 without the resistor but must confess it does sound slightly better with it in situ (more air around instruments)..... just how much are we straining the chip with this guy fitted? I always believe in living life to the max and if a "strained chip" sounds better then let's strain the little bastard to the limits
|
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on Nov 21, 2009 23:20:02 GMT
Prawn would be right about that observation, although I would keep the load current well under the 2mA (not that I would do this i have already passed comment on this subject) you effectivelly have a single ended common emitter stage with constant current load. and in this modded state some baggage hanging off the collector as well. Adding a resistor well attempt to increase the bias current but would vary over the load line, if you really must use another constant current sink on the collector.
|
|
|
Post by PinkFloyd on Nov 21, 2009 23:34:05 GMT
Bottom line guys, with the resistor in situ it "sounds" better.... not night and day but subtle things like more air, presence and a lovely detached walk around bassline....
Just how much are we straining the chip..... within max limits? I mean, on a scale of 1 to 100 is she on the 80% scale or the "she won't hold for much longer captain" scale?
In the Aune the chip is cold to the touch (which is always a good sign) and seems quite happy......
I'll gladly remove this thread if "safety" is compromised but, in the search of top quality sonics I will, personally let it ride in my amp and see where it takes me.... if she blows a gasket it's a simple case of replacing the chip minus the resistor....
Hell, the starship enterprise wasn't "designed" to go at some of the speeds it sometimes does but that doesn't stop the crew "flooring it" occasionally ;D
|
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on Nov 22, 2009 0:50:42 GMT
Bottom line guys, with the resistor in situ it "sounds" better.... not night and day but subtle things like more air, presence and a lovely detached walk around bassline.... Just how much are we straining the chip..... within max limits? I mean, on a scale of 1 to 100 is she on the 80% scale or the "she won't hold for much longer captain" scale? In the Aune the chip is cold to the touch (which is always a good sign) and seems quite happy...... I'll gladly remove this thread if "safety" is compromised but, in the search of top quality sonics I will, personally let it ride in my amp and see where it takes me.... if she blows a gasket it's a simple case of replacing the chip minus the resistor.... Hell, the starship enterprise wasn't "designed" to go at some of the speeds it sometimes does but that doesn't stop the crew "flooring it" occasionally ;D So what your saying is that the headphones are driven directly by pin 5 of the AD 744 ??
|
|
|
Post by prawn on Nov 24, 2009 21:39:25 GMT
Mike I'm running up an AD744 biased as you indicate into a very old test bed - so if it blows san fairy Ann. For a while I was getting 1.8mv offset now after 48hours or so some bizarre form of thermal runaway has started and the offset is now 41.2mv and rising. I think the worst case scenario is that the chip will lock up onto the neg rail and pulse a full rail voltage of +/- 18VOLTS (in my instance) down the chain taking everything with it and blowing out headphones if they are in use. Hopefully the chip would fail before this eventuality, but you never know! I really don't like this idea of Xin's. By the way wasn't he seriously discredited on another forum somewhere? If I've got this wrong (and I'm getting a real old guy now and get stacks wrong) then sincere apologies to all Xin afficianados.
|
|
|
Post by PinkFloyd on Nov 24, 2009 22:41:30 GMT
Mike I'm running up an AD744 biased as you indicate into a very old test bed - so if it blows san fairy Ann. For a while I was getting 1.8mv offset now after 48hours or so some bizarre form of thermal runaway has started and the offset is now 41.2mv and rising. I think the worst case scenario is that the chip will lock up onto the neg rail and pulse a full rail voltage of +/- 18VOLTS (in my instance) down the chain taking everything with it and blowing out headphones if they are in use. Hopefully the chip would fail before this eventuality, but you never know! I really don't like this idea of Xin's. By the way wasn't he seriously discredited on another forum somewhere? If I've got this wrong (and I'm getting a real old guy now and get stacks wrong) then sincere apologies to all Xin afficianados. The resistors have been in situ for a good 100 hours and the SQ is SUPERB...... Offset has remained consistent in this application (Aune amp) at LC 0.07mV and RC -0.09mV.... I have no need to call the fire engines out yet I'll continue with this 24/7 for the next few weeks and will report back..... I don't like to see anybody discredited without recourse to a second / third opinion.
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Nov 24, 2009 23:28:10 GMT
Forget my first PM MIke, I'd got my head up me bum If this was the usual way of adding the resistor to the op-amps output stage I'd try it but adding it onto the Comp pin as well as using comp as an output, it might run it too hard tbh Of course if its sounding ok thats fair enough, I think I'd only try this on an amp with output coupling caps though just incase it goes after a while
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Nov 24, 2009 23:29:10 GMT
BTW This mod should certainly give that TUBEY sound
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Nov 24, 2009 23:32:09 GMT
Bottom line guys, with the resistor in situ it "sounds" better.... not night and day but subtle things like more air, presence and a lovely detached walk around bassline.... Just how much are we straining the chip..... within max limits? I mean, on a scale of 1 to 100 is she on the 80% scale or the "she won't hold for much longer captain" scale? In the Aune the chip is cold to the touch (which is always a good sign) and seems quite happy...... I'll gladly remove this thread if "safety" is compromised but, in the search of top quality sonics I will, personally let it ride in my amp and see where it takes me.... if she blows a gasket it's a simple case of replacing the chip minus the resistor.... Hell, the starship enterprise wasn't "designed" to go at some of the speeds it sometimes does but that doesn't stop the crew "flooring it" occasionally ;D So what your saying is that the headphones are driven directly by pin 5 of the AD 744 ?? I'd have thought its driving an additional output stage? I'd be amazed if its headphones direct
|
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on Nov 25, 2009 14:02:09 GMT
|
|
|
Post by clausdk on Nov 25, 2009 14:08:53 GMT
Auch that must be painfull
|
|
|
Post by prawn on Nov 25, 2009 16:48:26 GMT
In the instance I was citing Mike I was running the chip at its' max permissible voltage and taking the max I could from pin 5 into an elementary complimentary pair with a dummy load. Also to be fair this chip had come out of my "seconds" box so it may have been subject to some previous hideous torture treatment that I can't remember. So in essence I was giving it an absolute test (I've now binned it) and in the Aune set up it may well hold on considerably longer - if not ad infinitum (but I still don't fancy the idea). Regarding the good doctor Xin there is a 260 page (!) complaints thread running over at Head Fi alleging all sorts of nefarious things about his activities, for anybody having a few hours to trawl through it.
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on Nov 25, 2009 18:13:22 GMT
You asked a good question Rob, I just mean't it as I'd be surprised if it was driving the cans direct , maybe it is I also don't have any first hand knowledge of the Aune, just trying to work things out by the pics posted
|
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on Nov 26, 2009 14:10:22 GMT
Ok after a bit more reading on Mikes other post, looks like there is an buffer stage between the AD 744 ( I missed the posts but read through them) Still a diagram would be nice ;D might all work out in the wash after all,..
|
|