Hi Arif,
Give it a try, but I do not think you will like that. Too much drop on the emitter resistors will generally give you a sound that degrades quickly at higher volume levels.
Nothing wrong with trying that. Generally it's best to stay away from high voltage drops across the emitter resistors. If you match the outputs well, and the heatsinks are adequate, you can then reduce the value of the emitter resistors. Just watch you don't end up with a positive temperature coefficient with bias current.
-Chris
Give it a try, but I do not think you will like that. Too much drop on the emitter resistors will generally give you a sound that degrades quickly at higher volume levels.
Nothing wrong with trying that. Generally it's best to stay away from high voltage drops across the emitter resistors. If you match the outputs well, and the heatsinks are adequate, you can then reduce the value of the emitter resistors. Just watch you don't end up with a positive temperature coefficient with bias current.
-Chris
Whoa - I thought things would slow down over the weekend!!
"So, tag - you're it!" - hehe, ok Chris - I asked so here it is:
The currents were based on the voltage drop across the resistors. I didn't ask the actual voltage readings (he probably wouldn't remember anyway), but I trust him to handle the ol' I=V/R thingy...
Arif, many thanks for the schematic...
Gotta run off to class - work is getting in the way... !
Cheers!
"So, tag - you're it!" - hehe, ok Chris - I asked so here it is:
The currents were based on the voltage drop across the resistors. I didn't ask the actual voltage readings (he probably wouldn't remember anyway), but I trust him to handle the ol' I=V/R thingy...
Arif, many thanks for the schematic...
Gotta run off to class - work is getting in the way... !
Cheers!
Hi Clem,
That might work until you get near the E-B potential Q104, Q105 and Q106 are active and therefore drawing current as well. So a straight measurement will be flawed. In other words, the measurement method lied to your brother.
I do trust he got the I=V/R thingy, he just didn't know what the R thingy was. Don't feel bad, I don't know what the effective R thingy is either.
This is just too much fun .... gotta stop
-Chris
That might work until you get near the E-B potential Q104, Q105 and Q106 are active and therefore drawing current as well. So a straight measurement will be flawed. In other words, the measurement method lied to your brother.
I do trust he got the I=V/R thingy, he just didn't know what the R thingy was. Don't feel bad, I don't know what the effective R thingy is either.
This is just too much fun .... gotta stop
-Chris
Chris,
I think I got lost there - the base currents associated with Q104 and 107 should be "smallish": Q104 doesn't probably pull any current until right before clipping; Q107 should be doing less then 10mA on its collector (based on constant-current source Q108), so its base is probably drawing in the sub-mA range, i.e. 0.2mA? Q105 isn't part of the 'normal' amplifier as its just an overheat shutdown. Please explain what you mean, thanks!!
Cheers
Clem
I think I got lost there - the base currents associated with Q104 and 107 should be "smallish": Q104 doesn't probably pull any current until right before clipping; Q107 should be doing less then 10mA on its collector (based on constant-current source Q108), so its base is probably drawing in the sub-mA range, i.e. 0.2mA? Q105 isn't part of the 'normal' amplifier as its just an overheat shutdown. Please explain what you mean, thanks!!
Cheers
Clem
Something from the man with the 555-II
Got this e-mail just now:
--------------
Anyway, looking at the diffamp, the ones in my version are 820 and 300 ohms for the collectors. The ones in the schematic show 825 and 365 ohms. I'm sure of my values because I read them off the resistor (precision type) and confirmed with the Fluke tester.
The voltage drop across the lower valued resistor is very
near 1.0 volts, which corresponds to about 1.2mA, plus
that which is the base current of the VAS. The other
resistor is 300 ohms and has 0.77 volts across it, or
about 2.5mA. If you use the 365 ohms specified in the
schematic, that's still about 2.1 mA.
Another observation... if you take off the active servo
IC, you effectively have a positive feedback loop in the
ampli. It's a total series resistance of 6.2 Megs, and
a 0.1 uF capacitor feeding the input resistance of the
diffamp, which is 100K, for a loss of only 62:1. I
wonder if the power amp will break into oscillation!
----------------
Got this e-mail just now:
--------------
Anyway, looking at the diffamp, the ones in my version are 820 and 300 ohms for the collectors. The ones in the schematic show 825 and 365 ohms. I'm sure of my values because I read them off the resistor (precision type) and confirmed with the Fluke tester.
The voltage drop across the lower valued resistor is very
near 1.0 volts, which corresponds to about 1.2mA, plus
that which is the base current of the VAS. The other
resistor is 300 ohms and has 0.77 volts across it, or
about 2.5mA. If you use the 365 ohms specified in the
schematic, that's still about 2.1 mA.
Another observation... if you take off the active servo
IC, you effectively have a positive feedback loop in the
ampli. It's a total series resistance of 6.2 Megs, and
a 0.1 uF capacitor feeding the input resistance of the
diffamp, which is 100K, for a loss of only 62:1. I
wonder if the power amp will break into oscillation!
----------------
Hi Clem,
It's not that easy. Both sets of transistors are conducting across the collector resistor. If the pair is in balance, the base to ground voltages, or base emitter if you prefer, will be within a couple mV. That is if the gains are equal.
If you disconnect R126 (?) from the op amp output, the amp's DC offset will rise to some figure. It should be less than |0.5V|. This will depend on the circuits designed in offset and the matching of the diff pair.
Since the base resistances are unequal, it makes it more difficult to calculate. A figure of -0.5 VDC off the top of my head, not taking into account the different base resistances.
-Chris
It's not that easy. Both sets of transistors are conducting across the collector resistor. If the pair is in balance, the base to ground voltages, or base emitter if you prefer, will be within a couple mV. That is if the gains are equal.
If you disconnect R126 (?) from the op amp output, the amp's DC offset will rise to some figure. It should be less than |0.5V|. This will depend on the circuits designed in offset and the matching of the diff pair.
Since the base resistances are unequal, it makes it more difficult to calculate. A figure of -0.5 VDC off the top of my head, not taking into account the different base resistances.
-Chris
Hi Arif,
Yes, that hold true as long as the base to "reference" voltages are the same on both sides of the diff pair. This is because the diff pair may be trying to compensate for imbalances in the rest of the amplifier.
You will find that IFF the betas are the same, the transistors are thermally coupled and the same temperature, the base resistances are the same and the rest of the amplifier is in balance, then the base-emitter voltages will be very close. ie the diff pair does not have to correct for the circuit balance and just does signal error correction. Normally, a designer will attempt to make the base resistances equal to minimize DC offsets. Just like an op amp.
I find that matching the transistor gains throughout many amplifiers improves the sonics, that's what many customers have reported over the years. Especially amps like the Marantz 500. DC conditions are also then much more stable over temperature.
-Chris
Yes, that hold true as long as the base to "reference" voltages are the same on both sides of the diff pair. This is because the diff pair may be trying to compensate for imbalances in the rest of the amplifier.
You will find that IFF the betas are the same, the transistors are thermally coupled and the same temperature, the base resistances are the same and the rest of the amplifier is in balance, then the base-emitter voltages will be very close. ie the diff pair does not have to correct for the circuit balance and just does signal error correction. Normally, a designer will attempt to make the base resistances equal to minimize DC offsets. Just like an op amp.
I find that matching the transistor gains throughout many amplifiers improves the sonics, that's what many customers have reported over the years. Especially amps like the Marantz 500. DC conditions are also then much more stable over temperature.
-Chris
Hi Arif,
If they are facing it works well, but it works fine as long as the cases are touching. Back to back, side to side whatever. The heatshrink isolates the parts from ambient a bit and traps whatever heat is generated inside. This helps equalize the temperatures between the transistors. Just don't short the leads!
You have probably seen this done on late seventies Marantz units and many others. If it's possible to do on a good amp I've repaired you will see it done. Even with cheaper units I will put a touch of grease between the transistors if I can get them to touch. This takes seconds to do.
-Chris
If they are facing it works well, but it works fine as long as the cases are touching. Back to back, side to side whatever. The heatshrink isolates the parts from ambient a bit and traps whatever heat is generated inside. This helps equalize the temperatures between the transistors. Just don't short the leads!
You have probably seen this done on late seventies Marantz units and many others. If it's possible to do on a good amp I've repaired you will see it done. Even with cheaper units I will put a touch of grease between the transistors if I can get them to touch. This takes seconds to do.
-Chris
.Chris, thanks for the explanation a page ago!
Sticking the dif-pair together is (was) also done on the vertical amplifiers of oscilloscopes, before the day of chopper-stabilized designs. Despite thermal grease and a copper strap, these older oscilloscopes still had a fairly notorious drift with temp, so I wonder how good the thermal tracking can actually get with discretes. Anyone ever try replacing the diff-amp with one of those single-package duals? Sonic improvements worth the effort/cost?
Thanks!!
Clem
Sticking the dif-pair together is (was) also done on the vertical amplifiers of oscilloscopes, before the day of chopper-stabilized designs. Despite thermal grease and a copper strap, these older oscilloscopes still had a fairly notorious drift with temp, so I wonder how good the thermal tracking can actually get with discretes. Anyone ever try replacing the diff-amp with one of those single-package duals? Sonic improvements worth the effort/cost?
Thanks!!
Clem
Hi Chris, thanks for the info - will keep that in mind. Next time I go parts shopping, I'll try to see if I can find any - living here has its ups and downs - one of the 'ups' is sometimes you get to find things that are genuine and long out of production, simply because no one else knew it was good stuff!!
Cheers!
Clem
Cheers!
Clem
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.