anatech, I posted in reply to Andypairo's asking if there were any 100w builders --just to say I was interested.
Since you asked, it's because my whole audio system operates in balanced mode -phono and cd players out and amplifiers in. Amps are Atma-sphere M60 balanced differential OTLs. They put out a lot of heat. A balanced krill would be nice for summer listening. When Steve mentioned his 400 w design he said it operated in balanced mode and that balanced delivers better results. He also mentioned that putting balanced input on the phase splitter wouldn't work. I don't think I could do a 400w amp. I hoped to use its balanced input arrangement on a 100w. Almost gave up hope when Steve left. Now that he's back...
ostripper said:
Hi OS. Nice looking board. I just wanted to point something out in case it has been accidentally overlooked. I notice that your drivers are different transistors from the outputs. I also notice that SD's schematics always use the same transistors for both. Your board layout won't allow the latter.
Brian
If you look at the Ft and gaincurves of the mje15032 (to-220)VS.
the njw -0281 (to-3p) they are identical. So, why waste another
big tranny?
Same with the other suggested pair.. FJP5200(to220)/Fjl5200
(to264), same 30mhz/ Ft 100Hfe over most voltages.
Do you want a to-3 package version? 15 minutes would be it..
to modify the modern version. The MJ15023/24 are lower
gain/ft , I would match them with 2sc4793/2sa1837. all
that drilling(to-3)..is it worth it?
http://focus.ti.com/lit/ds/symlink/opa445.pdf
IC can do +-40v , I want to get the SMD version and put it
on a little 2" X 1" card that "plugs in" to the OPS board.
They are expensive 5$ for SMD/ 7$ for DIP.
OS
You guys now have two threads. One on-topic and one off-topic here:
http://www.diyaudio.com/forums/showthread.php?threadid=140235
I'd be very glad if both could be used for their intended purpose.
And... be nice to each other.
/Hugo
http://www.diyaudio.com/forums/showthread.php?threadid=140235
I'd be very glad if both could be used for their intended purpose.
And... be nice to each other.
/Hugo
Yeah ,that is a old one (opa445) OPA453 is cheaper AND newer!!
with a slew of 35+ V/uS.. nice little 7-pin flatpack,
http://focus.ti.com/lit/ds/symlink/opa453.pdf
The 445 does about the same as a LM3886, but the 453
is X2. What a dandy "chipamp booster" the krill would make.
OS
I'm waiting to put a tube front end onto it - expecting great things from it. So come on guys let's get this thing built, tested, debugged, analysed, simmed & improved (if possible) - I have to stop drooling over the laptop it's efffeecttinng thhe keeybboaaaard!!!!!!!!ostripper said:What a dandy "chipamp booster" the krill would make.
OS
http://www.fairchildsemi.com/ds/FJ/FJP5200.pdf
nice little device,seems to be perfect "little bro" to fjl5200
same Ft/cob/hfe , so it should work the same as 2 TO-3P's
as the darlington. It is also a direct sub for MJE15032, but
with a better model (saturation data)."ultimate driver"
OS
Hi OS,
Hey, good find!
I was nervous about Fairchild's new found transistors being any good. I guess they are and that I should have a closer look at them.
Thank you for posting the link. These do answer the need for better drivers. Now, they even seem to have metal tabs, so here is a possible replacement for those 2SA968B and 2SC2238B devices.
Have you tried any Fairchild products yet?
-Chris
Hey, good find!
I was nervous about Fairchild's new found transistors being any good. I guess they are and that I should have a closer look at them.
Thank you for posting the link. These do answer the need for better drivers. Now, they even seem to have metal tabs, so here is a possible replacement for those 2SA968B and 2SC2238B devices.
Have you tried any Fairchild products yet?
-Chris
I've started the assembly for the other channel, to see if this one performs differently.
A side note, maybe useless but who knows...
I used for the diff pair a dual BJT (2SC3067)
The other small signal BJTs were 2SC2240/2SA970
The TO220 BJTs were KSA1837/KSC4793 from Steve
The driver and outputs were MJL1302A-3281 from ONSemi.
All transistor were HFE matched.
Ciao
Andrea
A side note, maybe useless but who knows...
I used for the diff pair a dual BJT (2SC3067)
The other small signal BJTs were 2SC2240/2SA970
The TO220 BJTs were KSA1837/KSC4793 from Steve
The driver and outputs were MJL1302A-3281 from ONSemi.
All transistor were HFE matched.
Ciao
Andrea
Hi Andrea,
Those are all good devices. Nothing wrong with any of your choices except I am not sure about the KSA1837 and KSC4793. I have the 2S versions, but their are only good for 1 ampere. You don't need that high voltage, so maybe C4381 and A1667 or FJP5200 and FJP1943 might be a (much) better choice for drivers.
I also have some 2SC3067 on hand. Some early testing shows they seem to be decent parts. I have yet to use any yet, but their time is coming.
Where did you get yours?
-Chris
Those are all good devices. Nothing wrong with any of your choices except I am not sure about the KSA1837 and KSC4793. I have the 2S versions, but their are only good for 1 ampere. You don't need that high voltage, so maybe C4381 and A1667 or FJP5200 and FJP1943 might be a (much) better choice for drivers.
I also have some 2SC3067 on hand. Some early testing shows they seem to be decent parts. I have yet to use any yet, but their time is coming.
Where did you get yours?
-Chris
c2cthomas said:
Should be OK - I did the matching myself with Steve sitting next to me.
On the other hand - two wrongs don't make a right....
Thomas,
I have no doubt they are good transistors.
This evening I completed the second board and, after tapping the heatsink (and breaking a tap
) I fired it up.To be quicker I didn't use the usual mica-thermal grease insulators but some sil-pads.
Unfortunately the effect is almost the same.
Almost because the higher thermal resistance of the silpads makes the output BJTs (slightly) warmer so they pass some more current.
It looks like I'm using a too good heatsink
To get a confirmation I tried lifting one of the output BJTs from the heatsink and its current started increasing.
So the ouputs are running too cool relatively to Q7 and Q10 that are
like hell (cannot keep my finger on them).It seems I can't do anything different than either heatsinking Q7 and Q10 or "downgrading" the amplifier to a 50W amp (or something in between 50 and 100W)....
Open to suggestions....
Ciao
Andrea
Hi Andrea,
I'm at a loss - Steve is the designer and I too am but a mere builder with a technical background - not an engineering background. I know that Steve will stand by his design and his amps so this will get resolved one way or the other - we just need to wait for him to recover from his tests at the hospital today. I'm hoping that he will be back on-line soon and that everything is OK with him.
I have just set my work area back in order and will start back on my construction tomorrow. Mine amps are 50W but perhaps I will stumble across something helpful.
I'm at a loss - Steve is the designer and I too am but a mere builder with a technical background - not an engineering background. I know that Steve will stand by his design and his amps so this will get resolved one way or the other - we just need to wait for him to recover from his tests at the hospital today. I'm hoping that he will be back on-line soon and that everything is OK with him.
I have just set my work area back in order and will start back on my construction tomorrow. Mine amps are 50W but perhaps I will stumble across something helpful.
unmibh said:
I too was wondering about the temperature compensation circuit and if it is properly hooked up and attached to the heat sink?
Looks as if you beat me to that one unmibh. How is your amp running now? What do you think about it's sound? Can you provide any measurements? I don't intend to be a pest - just curious.