VA load - simple resistor or bootstrap

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm studying a a minimalist design (DIGI 125 ) that uses a simple resistor as the current source to load the VA transistor. Most circuits here and elsewhere use a more elaborate resistive divider with the midpoint bootstrapped to the output node or an active current source.

Can someone knowledgeable please summarize the main tradeoffs between the simple resistor load and bootstrapped load? I expect that the current sourced by the bootstrap will be more consistent over a bigger voltage swing BUT.... what of capacitor distortion? Also.... the biggest voltage swings in real music will occur at bass frequencies where loop gain is greatest to mask the distortion.
 
Here is the schemat .....
 

Attachments

  • DIGI125_schematic_small.jpg
    DIGI125_schematic_small.jpg
    66.5 KB · Views: 210
A simple resistor load means you're trading off loop gain for VAS current. Once you're got a decent amount of current, you may find that loop gain only starts rolling off well above the audible range. Gain bandwidth may well be OK, but the maximum amount is "capped", know what I'm saying?

Hence why people commonly use a current source instead. Very high dynamic impedance, probably to be swamped by output stage input impedance (= load impedance / accumulated beta; note effect of beta droop). You could use a big honkin' inductor, too, but linearity of those is notoriously crummy in the required values, not to mention their price tag and size.

The bootstrap dynamically increases load impedance, too. Attaching the upper end of the lower resistor to the output (voltage gain 0.9something) makes it look much bigger than it actually is. As a bonus, VAS output voltage can actually exceed the positive supply for maximum output stage driving, which is why many older low-voltage (e.g. car audio or similar) power amp ICs have a bootstrap cap connection. (Those usually bootstrap the VAS + driver stage supply.) The downside? Now the output stage has already "given up" some of its transconductance for the bootstrap, so output impedance increases and load immunity worsens when compared to a current source solution. Performance does tend to get much better over the un-bootstrapped resistor loaded VAS though.

BTW, the last time people thought a driver stage run on output transistor bias current was a good idea was in the 1970s. (I saw this in a Yamaha receiver, the CR-640 I think it was. The next model up had beefier drivers running more current.) It works fine in class A, actually, but is a bad idea in a switching amp. Without current of its own, the driver stage is not able to pull base charge out, resulting in high-frequency cross-conduction problems that at the very least can cause excessive heating (and where that is, secondary breakdown is never far away).
 
Thanks for a great summary. as far as the driver stage biasing is concerned, Ian Finch and others have been making this point to me in another thread I started. I'm going to substitute some beefy TO220 drivers and bias them hot into class A with a low value emitter resistor. I'm not convinced that the charge suck out cap is a good idea and will potentially upset the biasing when the cap unloads its stored charge at the "wrong" time with a varying frequency and amplitude waveform. Any comments?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.