Can a bias adjust be added to SX-580 at the 2SA798 pair?

[dlucy]

I've read more than a few threads here about SX-580's (and SX-680's, too, I think) where someone asks "where is the bias adjust" and several chime in "there is none, that's what you get in a BOTL". Looking at the schematic, isn't it possible to add a bias adjustment to the two dual-transistor 2SA798's? Just like adding an adjustment is possible (and works nicely, I'd say) for the BOTL SX-450 like here http://www.audiokarma.org/forums/index.php?threads/pioneer-sx-450-dc-offset.676140/#post-9022861 ?

The SX-580's two 2SA798's (Q301 and Q302) are shown here:



I'm a novice at all this stuff, but isn't R307/R308 or R301/R302 the bias resistor? Couldn't the same thing be performed here as was done above for the SX-450? Keep a low value resistor in line but add a trimmer pot in series to be the adjustable value?

Don't flame me if this is a terrible idea; it just seems like a similar approach to get a desirable outcome for the SX-580.

 

[petehall347]

am thinking you have bias mixed up with dc offset .

 

[rcs16]

The output stage bias current is set internally to the stk-0029 module.

 

[mykidsbutler]

Let me ask dlucys’ question a different way. If one does their due diligence looking for bad components, bad grounds and other causes of dc offset would it be possible to adjust off set by adjusting the values on r301 and r302. I am asking this after i have changed them from 150k to 130k as an experiment. On the left it lowered offset from 62mv to 42mv, right side from 58mv to 36mv. Is there a risk doing this? I know, silly to ask after i already did it but i was in a adventuress mood today. Thanks for any constructive response.

 

[dlucy]

Let me ask dlucys’ question a different way. If one does their due diligence looking for bad components, bad grounds and other causes of dc offset would it be possible to adjust off set by adjusting the values on r301 and r302. I am asking this after i have changed them from 150k to 130k as an experiment. On the left it lowered offset from 62mv to 42mv, right side from 58mv to 36mv. Is there a risk doing this? I know, silly to ask after i already did it but i was in a adventuress mood today. Thanks for any constructive response.

There should be no risk in slowly upping or lowering any resistance and watching the results. This is part of what decade resistance boxes of old did.

 

[markthefixer]

Does the old decade resistance box go open circuit (infinite ohms) during the switching? If so, in the idle current circuit that can indicate: "go blooie"....

DC offset adjust in that circuit is not a magic resistor value. It is a non-trivial addition - experimental no less.
I DON'T know if a sx-780 base current stealer / booster (q5,q6) transistor could be grafted in without inducing oscillation.

 

[mykidsbutler]

Thanks MTF for your response. I believe i understand what your saying, the circuit in this area needs to be a bit more sophisticated to achieve what im trying to do. With that being said, am i hurting anything by changing r301 and r302 up or down by one value to achieve a lower offset? Im a 59yo retired utility equipment mechanic trying to learn a new hobby to keep the ol’ fingers and brain cells nimble, so be easy on me please.

 

[dlucy]

Does the old decade resistance box go open circuit (infinite ohms) during the switching? If so, in the idle current circuit that can indicate: "go blooie..."

You make a great point, Mark. I was assuming the user would properly power down before switching to a new value in the decade box. That's not obvious, so a good thing to remeber when using such a switch-based device.

 

[markthefixer]

Thanks MTF for your response. I believe i understand what your saying, the circuit in this area needs to be a bit more sophisticated to achieve what im trying to do. With that being said, am i hurting anything by changing r301 and r302 up or down by one value to achieve a lower offset? Im a 59yo retired utility equipment mechanic trying to learn a new hobby to keep the ol’ fingers and brain cells nimble, so be easy on me please.

I would not expect to see ANY significant change (good or bad) under those conditions.
R301 / R302 are setting an input impedance and referencing it to ground.

The change I would do, would involve making a resistor divider like the one on the awh-048 SX-1250 power amp card:
R4,R3,R6,R5,VR1,R7 and see if that DC injection would help.
I have NOT looked at the stability of the SX- 580's power amp's regulated voltage needs.
(the +13v zener, cap etc feeding R303, R304 - that HAS to be stable)
There are some places where they are ports for injecting noise into the amp, and some that are not a significant problem.
lots more than I can do now.

 

[rcs16]

There are a few ways to compensate for the offset, one is to inject a correction voltages into the input and the other is the feedback.
mykidsbutler, suggest you buy Bob Cordell new revised power amp design book.
This is one reason to use jfets for the i/p devices, since they do not draw input current so you do not have these offset voltages being created by transistor input bias current. You can use different resistors for input and feedback....
So, you can do it like Pioneer did in so many other designs using that bias offset adj circuit. You want short wires so that it does not pickup noise.
The other method is a DC servo, which is a deep slope low pass filter, that samples the DC and inverts the correction signal into the FB node. You can do this manually with a pot too.
You want a stable/regulated voltages for your bias offset corrector.
you want at least a 100K or more resistor in series tied into the feedback node.
You can look at this design that shows how to implement the servo
http://forum.cxem.net/applications/core/interface/file/attachment.php?id=515698

 

[mykidsbutler]

Thanks rcs16 for the suggestions. I was pleased to get the discrete STK replacement for the 780 to work on this 580. Almost like the sound better than a 780. So i’m alittle OCD about getting the best sound from it. Thanks everyone for the help!

 

[ToyDoctor]

Does the old decade resistance box go open circuit (infinite ohms) during the switching?

The decade resistor boxes shouldn't go open because they were additive. At least the ones I used were.Just make sure you're starting in a safe range.

 

[ToyDoctor]

I can see why you would think that because they are in series that you should have to break connection from one to the next, but the rotary action carried through.

 

[markthefixer]

I can see why you would think that because they are in series that you should have to break connection from one to the next, but the rotary action carried through.

There were MANY of them out there from many reputable and DIS reputable manufacturers, and in my career I HAVE run into boxes that went infinite ohms upon switching.