It has not been mentioned here whether the amp has
a capacitor in the feedback loop or not.
If not, then 100 mV is a good figure, although you can
trim it for better after the amp has reached ordinary
operating temperature. You can trim the offset in half
a dozen different ways, one favorite being to place a
high value resistor G-S on one or the other of the diff pair.
As to the hum, if the transformer secondaries are not
perfectly matched, then any net difference in the draw
from + to G and - to G results in some transformer saturation,
thus the hum. Having DC offset through a load results in
this difference.
If the transformer has dual secondaries, you can also
eliminate this by using dual bridges, one for each secondary.
a capacitor in the feedback loop or not.
If not, then 100 mV is a good figure, although you can
trim it for better after the amp has reached ordinary
operating temperature. You can trim the offset in half
a dozen different ways, one favorite being to place a
high value resistor G-S on one or the other of the diff pair.
As to the hum, if the transformer secondaries are not
perfectly matched, then any net difference in the draw
from + to G and - to G results in some transformer saturation,
thus the hum. Having DC offset through a load results in
this difference.
If the transformer has dual secondaries, you can also
eliminate this by using dual bridges, one for each secondary.
Klaus
"You can trim the offset in half a dozen different ways, one favorite being to place a high value resistor G-S on one or the other of the diff pair."
Listen to the master!!!
He is actually speaking about R8 in Aleph60.
I think a 100K precision 25-turn trimmer would be fine for the job and you won't have to replace it with a resistor.
You could go one step further and insert R1 (a 200K trimmer should be fine). This will help you set the voltage across R14.
And if you decide to take it even further, remove C4 and add R0.
Good luck,
Nick
"You can trim the offset in half a dozen different ways, one favorite being to place a high value resistor G-S on one or the other of the diff pair."
Listen to the master!!!
He is actually speaking about R8 in Aleph60.
I think a 100K precision 25-turn trimmer would be fine for the job and you won't have to replace it with a resistor.
You could go one step further and insert R1 (a 200K trimmer should be fine). This will help you set the voltage across R14.
And if you decide to take it even further, remove C4 and add R0.
Good luck,
Nick
Re: Klaus
The problem is that it´s not your choice wich resistor (R1,R8)
to add. It depends on the polarity of the offset!
Adding R1 is good for lowering a negative offset only.
Adding R8 is good for lowering a positive offset only.
R8 adds some kind of imbalance to the diff.pair.
These tweakings stem from mass-production. In DIY-mode
adjusting the bias-resistor for the diff-pair is much more elegant
as you cure the cause and not the symptom!
Uli
Hi Nick,Nickolas K. said:"You can trim the offset in half a dozen different ways, one favorite being to place a high value resistor G-S on one or the other of the diff pair."
He is actually speaking about R8 in Aleph60.
You could go one step further and insert R1
Nick
The problem is that it´s not your choice wich resistor (R1,R8)
to add. It depends on the polarity of the offset!
Adding R1 is good for lowering a negative offset only.
Adding R8 is good for lowering a positive offset only.
R8 adds some kind of imbalance to the diff.pair.
These tweakings stem from mass-production. In DIY-mode
adjusting the bias-resistor for the diff-pair is much more elegant
as you cure the cause and not the symptom!
Uli
sorry for posting so late but i was offline for some time.
iam not trying to run down any aleph amp
any misunderstanding may be caused by my poor english skills
i am in proces of building my aleph 5
iam sory for any misunderstanding and confusion i caused
i hope that i didnt do something very bad
and again i blame it on my english skills
and my real name is Boleslav Zaras
so i say sorry to everyone
iam not trying to run down any aleph amp
any misunderstanding may be caused by my poor english skills
i am in proces of building my aleph 5
iam sory for any misunderstanding and confusion i caused
i hope that i didnt do something very bad
and again i blame it on my english skills
and my real name is Boleslav Zaras
so i say sorry to everyone
xsnailx said:
Hi again,
matching FETs requires a very constant voltage supply source. Or a very constant current supply, of cause. Constant voltage is good given by "just" a car accu for 10mA, at 0,6A for the power Fets it could become critical - with unloading sinking voltage.
The voltage does not matter - you just have to set the right resistor value: e.g. 12V-4V (for the fet) = 8V to drive 10mA = 800Ohms. But the constance of the current becomes worse with lower supply voltage - the 4V are in reality 3,85-4V, so affecting the current value. For perfect matching the voltage should be much higher.
regards
Klaus
Klaus, thank you for claryfing this topic
i have acces to a good 15 dc but in work only and it is a little uncomfortable for me to do matching there nor i can borrow it home s i thought about some homebrew method with accu maybe 2 0f them so it will give 24 V ??? i am not afraid of the power draw as they will be 100 ah from my diesel
i have acces to a good 15 dc but in work only and it is a little uncomfortable for me to do matching there nor i can borrow it home s i thought about some homebrew method with accu maybe 2 0f them so it will give 24 V ??? i am not afraid of the power draw as they will be 100 ah from my diesel
CCS
Hi,
why not build a constant current source driven by the accu?
PS: a car accu delivers the needed current for hours!
PPS: I match my power Mosfets with 200mA.
You can (if you have too many fets) match them for equal
Transconductance too: match with 100mA - first step.
Then match this batch for equality at 1 A!
That tight matching makes no sense in practice!
matching for 100mV (GS) is enough for 1R source resistors.
Hi,
why not build a constant current source driven by the accu?
PS: a car accu delivers the needed current for hours!
PPS: I match my power Mosfets with 200mA.
You can (if you have too many fets) match them for equal
Transconductance too: match with 100mA - first step.
Then match this batch for equality at 1 A!
That tight matching makes no sense in practice!
matching for 100mV (GS) is enough for 1R source resistors.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.