• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Should bias be equally distributed over emitter resistors?

Scott75s1974

Scott75s1974

Well-Known Member
I've replaced the bias pots in my Scott r75s with multi-turn Bourns units. After searching a long time for the specified 275 ohm units and settling for 250, I found the removed trimpots were in fact 325 and 500 ohms. So apparently not crucial. (The pots were in series with another resistor 47ohms to the NPN side. Being able to only find a 250 ohm pot, I planned on replacing the 47ohm with a 75 ohm series resistor.)

After I replaced the pots, before messing with the other resistor, I set the bias and then measured each separate emitter resistor to the output. Each came in nicely balanced at 50/50. (e.g. .026v bias, +.013 and -.013). This seems perfect to me.

I just wanted to verify that this is proper. Namely, that I'm not ignorant of some requirement that the NPN side needs to be more or less biased than PNP side or something.

Thanks!
 
Register to hide this ad
Thanks! So, if I read you correctly: the emitter resistors set the 50/50 split regardless of the relative bias on the driver transistors. So my measurement of each side is not particularly diagnostic.
 
One of the tests specified in some amp service manuals is exactly what you did, a current sharing test. Basically you run the thing to some output power and make sure voltage drop across the emitter resistors is within some % to make sure the transistors are sharing the load. Of course it relies on the resistors being on value too. Its also a quick and dirty test to see if you've got a weak, dead, or poorly matched output transistor in an amp with a bunch of them in parallel. The bad one won't show as much voltage drop, or a current hog will show too much.
 
gadget73 said:
One of the tests specified in some amp service manuals is exactly what you did, a current sharing test. Basically you run the thing to some output power and make sure voltage drop across the emitter resistors is within some % to make sure the transistors are sharing the load. Of course it relies on the resistors being on value too. Its also a quick and dirty test to see if you've got a weak, dead, or poorly matched output transistor in an amp with a bunch of them in parallel. The bad one won't show as much voltage drop, or a current hog will show too much.
Click to expand...
This is good practice, but only applies to parallel outputs. The Scott only has one PNP and one NPN output per channel. If the 0.47 ohm resistors are same value, then the currents have to match. Difference in voltage would be due to differences in resistance....
 
Yes the current across each emitter resistor, when the amplifier is in a quiescent state (switched on with no signal and no load), should be approximately equal. Usually the manufacturer gives a desired volt drop across one of the emitter resistors as a setting point for the quiescent current at idle. This is set to a value which is just above the output transistor switch off threshold. The setup is what makes a class B amplifier into a class AB.
If you think there is a problem it is worth setting the volt drop as specified and then measuring both sides for comparison. Don't be too critical though as the actual current passing through the emitter resistor (and consequently the measured volt drop) will be dependant upon the precise gain of the transistors, the precise gain of the driver stage, the junction temperatures and the precise value of all the resistors in the local chain. The value stated is usually a minimum and if it is slightly above this it won't have any noticeable effect.
 
Scott75s1974 said:
I set the bias and then measured each separate emitter resistor to the output. Each came in nicely balanced at 50/50. (e.g. .026v bias, +.013 and -.013). This seems perfect to me. I just wanted to verify that this is proper.
Click to expand...
Without a load connected to the amplifier the current flowing through the NPN transistor goes through its emitter resistor, through the other emitter resistor and through the PNP transistor - there's nowhere else for it to go. (Ok, some could go through the feedback resistor, but there such a tiny amount going that way that it can be ignored.) So as long as the two emitter resistors have the same value they will have the same voltage drop across them - they have to because the same current flows through both.
 
If you want to know how the transistors are matched to each other, measure the emitter to output DC at idle. Compare all the NPN's, then all the PNP's. You will see differences, but as long as the differences aren't huge, say more than 25%, don't worry about it. The emitter resistors help them share the current better.
 
Thanks for all of these responses. So, I'm barking up the wrong tree to measure relative voltage on the two emitter resisters. Let me try to explain what I'm trying to figure out. I have it dialed in for the proper bias measurement in the SM across the two output emitter resistors. I'm wondering if things are ok for the biasing of the drivers. Why am I asking this? I'm afraid my replacement of the trimpot might result in underbiasing of the negative side driver.

I had to replace the bias pots with lower ohm versions. Spec. was 275, but actual on removed pots was 350 and 500 ohms, closest I could get was 250. There was also a resister (R9) in series with the track leading to the base of one of the drivers. That seems to be part of a bigger voltage dividing circuit. Since I've dropped the resistance between the bases of the two drivers by 100-250 ohms, I want to make sure that biasing of the drivers is not out of wack on that side (TR7). I'm wondering if I need to increase R9 to compensate to make sure the positive side isn't more biased than the negative. (There's also another resistor in series R11 390ohms).

I'm assuming the bias pot set a voltage divider that biases the two driver transistors (TR4 and TR7) and these in turn bias the two output transistors where the measurement is taken.

I could just go and see if the driver bases are at +.55 and -.55 as per the schematic, but those readings swamp the bias adjustment, and I understand that the relevant bias is the output transistors, since that's where the SM says to measure.


driver.jpg
 
Last edited:
triode17 said:
If you want to know how the transistors are matched to each other, measure the emitter to output DC at idle. Compare all the NPN's, then all the PNP's. You will see differences, but as long as the differences aren't huge, say more than 25%, don't worry about it. The emitter resistors help them share the current better.
Click to expand...
This is just a single NPN/PNP output A/B amp. So no worry about matching output loads on ganged transistors. (I wish I had that problem!)
 
Last edited:
Don't worry about getting exactly .55 volts. This will depend on the batch and manufacturer and is not a written in stone value. Also on the HFE of both driver and outp[uts, which can vary greatly.The driver and the output work together in a loop, so getting the correct bias voltage across the emitter resistors will put both in the correct operating point. Adjustment affects driver and output - there is no separate adjustment for the driver.
 
A few weeks ago I found a Heathkit AA-1640 amp with the meters. It has no amp setting adjustments. Nothing. The DC Balance is under 10mV' per side. I replaced the mixed matched quasi-comp NPN outputs with the MJ21196's and she run's perfectly.
If you ever get a chance to try one out, take it. Very impressed with it's simple circuit.
The scary thing is, it runs 91 volt rails with 100v filters. :cool:
 
If the voltage reading across the emitter resistors is right, then its right. Base voltage is going to vary based on specific transistor characteristics. Its not the more important number by any stretch.

If you're really concerned about getting the original 275 ohm pot value, take your 250 ohm pot and add a 12 ohm resistor at each end, or a single 22 ohm and just realize that the pot will be further from mechanical center. I seriously doubt it matters all that much.

zebulon1 said:
The scary thing is, it runs 91 volt rails with 100v filters
Click to expand...

My Phase Linear was running 108v rails on 100v filters. Spec is 100v rails with 100v caps. New caps are 160v.
 
zebulon1 said:
A few weeks ago I found a Heathkit AA-1640 amp with the meters. It has no amp setting adjustments. Nothing. The DC Balance is under 10mV' per side. I replaced the mixed matched quasi-comp NPN outputs with the MJ21196's and she run's perfectly.
If you ever get a chance to try one out, take it. Very impressed with it's simple circuit.
The scary thing is, it runs 91 volt rails with 100v filters. :cool:
Click to expand...

There are some amps which have no bias adjustment because they use a bias servo circuit which provides a constant bias.
 
You must log in or register to reply here.
Back
Top Bottom