Aleph-X builder's thread.

Jump to Latest
SGregory

SGregory

Member
Joined 2003
Incresaed Bias on the Diff Pair.

Forgive me for not bookmarking the post and linking it here, but I was not thinking when I saw it. And I havn't been able to find it again. I can't remeber what I used for search topic.

In my searching, I have found that there has been quite a bit of discusson about increasing the bias. I also saw a short post contemplating why Nelson originally chose the 400 ohms of the load resistor. Paraphrasing the comment from I believe Mr Thurgard, '400 ohms seems to give the maximum amount of gain for a given bias to achieve proper bias on the gain devices' he went on to say something well beyond my ability to understand that lowering the value of the load resistor and increasing the bias is exceeded by the change in transconductance.:xeye:

If that is true, then the 400ohm would be close to optimal for the characteristics of the devices used in this circuit.

Could anybody with a more fundemental grasp on this help to clarify?

Thanks
Scott.
 
hifiZen

hifiZen

Member
Joined 2001
moe29 said:
... Of course this means they're not mounted
back to back during testing...
Click to expand...
Hmm. Not to put a fly in the ointment, but they really ought to mounted together if you can manage it... junction temperature affects a MOSFET's behaviour, so you need to be sure they're at exactly the same temp when you do the matching. This is the reason they're mounted back to back on the PCB - the tight thermal coupling keeps the die temperatures very close, and their Vgs drift matched as closely as possible during operation.

You can also do as Grey says, and swap Q2 and Q11 around to take advantage of their Vgs mismatch...
 
M

Mark A. Gulbrandsen

Member
Joined 2001
Ahhh, The things we do when we get tired... I mistakenly posted this under the Electronics and parts heading......:bigeyes: So I',m moving it to where it belongs.....

I'm about to run up my first X board on the variac and was wondering whats the proper procedure for setting the input pair current source adjustment???

Thanks,
Mark
 
G

GRollins

Member
Joined 2001
Mark,
Tired is a topic I know something about. I can sympathize.
The front end current source adjustment is to set the absolute DC offset at the output. If you're using either of the resistor sets I posted in the original thread, you should be fairly close. As I recall, the original set was roughly +-10% from the nominal current, and the alternate was for something a little tighter (+-5% perhaps? I don't remember exactly). You're not going to be really far off the mark either way. Hook the booger up and twist the little feller's tail until the absolute DC offset--measured from output to ground--is as close to 0V as possible. Let it cook for a bit, then check it again.
Use a load resistor at first, just to be on the safe side.
(Yawwwn...)

Grey
 
K

koolscooby

Member
Joined 2003
bradleyehughes.com
GRollins said:
Mark,
Tired is a topic I know something about. I can sympathize.
The front end current source adjustment is to set the absolute DC offset at the output. If you're using either of the resistor sets I posted in the original thread, you should be fairly close. As I recall, the original set was roughly +-10% from the nominal current, and the alternate was for something a little tighter (+-5% perhaps? I don't remember exactly). You're not going to be really far off the mark either way. Hook the booger up and twist the little feller's tail until the absolute DC offset--measured from output to ground--is as close to 0V as possible. Let it cook for a bit, then check it again.
Use a load resistor at first, just to be on the safe side.
(Yawwwn...)

Grey
Click to expand...


how big of load resistor? (See i'm not close to having one built, actually will be in 2 weeks, ordering more parts today)... so i'm asking questions so when i'm not near a computer when i test at the lab, i'll know some stuff i need to :-D
 
G

GRollins

Member
Joined 2001
I've got a pair of those big, honkin' Dale 8 ohm 250W beasties, but I also use your average rectangular ceramic 8.2 ohm 20W resistors, too. Even if something goes seriously awry, they can take punishment long enough for you to turn the amp off. For that matter, they're cheap enough that if you manage to crisp one you don't have to take out a second mortgage on ye olde homestead to buy another.
Don't forget that resistors can take a lot more abuse than gain devices.
As a practical matter, I've never managed to kill a load resistor. That said, I'll probably do so tonight when I get home.

Grey
 
Netlist

Netlist

Moderator
Joined 2003
Paid Member
Keep in mind that the value entered in cel B64 of AXE1 should be the total dissipation for one monoblock or one channel of AlephX. So if you plan to use two big heatsinks with a Rth of 0.3°C/W each, you should enter 0.15 in cel B64.
I’m just pointing this out because I was trapped with this.
 
M

Mark A. Gulbrandsen

Member
Joined 2001
I finally got around to firing up my first "stock" X channel and after finding that I had installed D1a in backwards all is now working just fine and with very low DC offset of about 20mv! Perhaps matching the FETs with my Racal 5003 paid off! We'll see how low I'm able to keep it after a few hours of warm up. This amp runs VERY :hot: .

Thanks again Chad for all your efforts with the PCB's and to Grey and Nelson for answering some of my stupid questions and sharing their designs with us!

Mark
 
moe29

moe29

Member
Joined 2002
Congrats Mark!

I got mine down to around 40-50mV with a diff pair that was
a better match. Now i'm working on channel 2.

try this: while measuring DC offset, blow on the differential pair
and watch the reading change! I'm afraid that once i have the
chassis built around the board, that my test bench readings
won't be the same....
 
You must log in or register to reply here.
Top Bottom