Hi,
A couple of questions regarding the small, low-TIM Leach. Schematic is here:
http://users.ece.gatech.edu/mleach/lowtim/graphics/ckt.pdf
1. R39/C19 and R40/C20 are used to sense voltage drop across the emitter resistors and shutdown the drive if the voltage drop gets too high. How come the other output transistors do not have the cap in parallel with the sense resistor?
2. For a configuration with separate supplies for driver and outputs, what should the driver +/- power supply voltage be?
3. I have a nice transformer, which after rectification yields +/- 48Vdc and +/- 42Vdc. What should I expect if I use this transformer?
A couple of questions regarding the small, low-TIM Leach. Schematic is here:
http://users.ece.gatech.edu/mleach/lowtim/graphics/ckt.pdf
1. R39/C19 and R40/C20 are used to sense voltage drop across the emitter resistors and shutdown the drive if the voltage drop gets too high. How come the other output transistors do not have the cap in parallel with the sense resistor?
2. For a configuration with separate supplies for driver and outputs, what should the driver +/- power supply voltage be?
3. I have a nice transformer, which after rectification yields +/- 48Vdc and +/- 42Vdc. What should I expect if I use this transformer?
I'm not sure about the first one - I can only guess it is some kind of compensation.
For the frontend voltage it is usually best if it is a little higher than the output stage voltage, although that does depend on the topology used. The main benefit is that it should be regulated. If you're not planning to use Leach's PCB's, I would suggest powering the predriver transistors Q14/Q15 from the frontend voltage also. If you are, then simply leave out R32 and R33 and connect the frontend voltage there.
With 42V DC rectified you can expect about 90W of power into 8 ohms. You may need to recalculate some parts of the circuit for this voltage. Leach has some guideline to this on his site. You may need to substitute some lower voltage zeners in the diff pairs.
For the frontend voltage it is usually best if it is a little higher than the output stage voltage, although that does depend on the topology used. The main benefit is that it should be regulated. If you're not planning to use Leach's PCB's, I would suggest powering the predriver transistors Q14/Q15 from the frontend voltage also. If you are, then simply leave out R32 and R33 and connect the frontend voltage there.
With 42V DC rectified you can expect about 90W of power into 8 ohms. You may need to recalculate some parts of the circuit for this voltage. Leach has some guideline to this on his site. You may need to substitute some lower voltage zeners in the diff pairs.
"With 42V DC rectified you can expect about 90W of power into 8 ohms. "
Not even with a regulated supply, the circuit cannot drive rail-to-rail. Depending on how 'stiff ' the supply is, I would guess more like 60W. Hooking the front end up to the ±48V would help a lot. I would add a Baker Clamp if I did this.
Not even with a regulated supply, the circuit cannot drive rail-to-rail. Depending on how 'stiff ' the supply is, I would guess more like 60W. Hooking the front end up to the ±48V would help a lot. I would add a Baker Clamp if I did this.
that means the 80 to 90W into 8r0 becomes 160 to 180 W into 16r0.
Why don't you just use a 40Vac transformer and get 170W into 8r0?
Thanks for this tip. I'm re-using an existing amplifier chassis, and the standard PCB's won't fit for what I want to do. I'll probably need to look at how that thing is biased.
A very good question. The answer, in short, is cost as money is an issue at some point.
One of my project criteria is to recycle as many parts as possible from an existing (dead) amp. The subject amplifier was originally sold as a bridge amp, and was rated at 200W into 8 ohms. Ideally, I have a chassis, heatsinks, transformer and power supply that can be re-used.
A lot of assumptions, but I'm hoping to replace only the PCB's and perhaps the output devices.
Thanks everyone for the helpful replies!
I went back and re-read some of those files. I learned that at one point a student had a problem with the protection circuit engaging. Removing one pair of these caps resolved the issue, but Dr. Leach indicated he didn't fully understand the issue. I'll layout the board to accommodate the caps - I can always omit later.
OK, I went and researched this a bit; please pardon my ignorance. I presume you mean you would add the Baker Clamp to the predrivers, like I saw on a post regarding the APT 1.
I understood from this that the Baker Clamp in effect provides a soft clipping feature but also helps avoid transistor saturation, and therfore aids turn-off.
Wouldn't it be better to put the Baker Clamp on the output devices, which are 500 - 1000 pF? Or is it effective enough to Clamp the predrivers?
Hypothetically, if the desired objective is to decrease the turn-off speed, wouldn't cascoding the output devices address this? I thought one of the benefits was an apparent reduction in device capacitance. I believe the collective wisdom is that cascoded outputs don't sound good, but one can't help but wonder if this has been thoroughly studied.
I went back and re-read some of those files. I learned that at one point a student had a problem with the protection circuit engaging. Removing one pair of these caps resolved the issue, but Dr. Leach indicated he didn't fully understand the issue. I'll layout the board to accommodate the caps - I can always omit later.
OK, I went and researched this a bit; please pardon my ignorance. I presume you mean you would add the Baker Clamp to the predrivers, like I saw on a post regarding the APT 1.
I understood from this that the Baker Clamp in effect provides a soft clipping feature but also helps avoid transistor saturation, and therfore aids turn-off.
Wouldn't it be better to put the Baker Clamp on the output devices, which are 500 - 1000 pF? Or is it effective enough to Clamp the predrivers?
Hypothetically, if the desired objective is to decrease the turn-off speed, wouldn't cascoding the output devices address this? I thought one of the benefits was an apparent reduction in device capacitance. I believe the collective wisdom is that cascoded outputs don't sound good, but one can't help but wonder if this has been thoroughly studied.
What parts have you got? The Leach might be an overkill circuit in this case... but with that spec transformer you can still do something good
I have the following:
- Chassis (looks tired, but hey, it's a box)
- 2 Heatsinks approx. 160mm x 210mm x 50mm, predrilled for TO-3's, that came with the above chassis
- 8 pair of vintage Sanken TO-3 BJT's in unknown condition
- ~1.2kVA transformer (it's big) with 1 higher voltage, low current winding and 1 pair of lower voltage, high current windings
Peripheral things:
- O-scope, function generator, and VOM
- TinyCad
- some free time
- EE degree
"OK, I went and researched this a bit; please pardon my ignorance. I presume you mean you would add the Baker Clamp to the predrivers, like I saw on a post regarding the APT 1."
I would do it just like the APT1, between the Vas and predrivers.
Non-inductive wirewound are specified, digi-Key has some suitable bulk-foil Vishay for around $1.
I would do it just like the APT1, between the Vas and predrivers.
Non-inductive wirewound are specified, digi-Key has some suitable bulk-foil Vishay for around $1.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.