I saw the same article from EDN Europe and I had to see how it worked. See attached drawing for what I simulated. I saw nothing special about it. Distortion looked good but clipping behavior was funky. Negative peaks clipped before positive peaks. There are other amps on the DIYAudio site that are much better. Simulation is a trap sometimes since all the same numbered devices are identical. Any thoughts on how the amp behaves as things heat up?? Probably badly unless all relevant transistors share the same heatsink.
"That which cannot be controlled, must be made irrelevant."
I worry that the base currents are not controlled. In the Figure 1 schematic http://www.edn-europe.com/cmsimages/1209/35f1.jpg , it is assumed that Q1's base current is exactly equal to Q3's base current. It's also assumed that Q2's base current is exactly equal to Q4's.
Or as the article puts it, Assuming all four transistors (two NPNs, two PNPs) are perfectly matched ...
This assumption worries me. In my opinion, base current cannot be controlled. You can attempt to find matched pairs (in this case, matched quads), but how well do they match as the bias point shifts, as temperature varies, and over time (matching "drift")?
Thanks but no thanks.
I worry that the base currents are not controlled. In the Figure 1 schematic http://www.edn-europe.com/cmsimages/1209/35f1.jpg , it is assumed that Q1's base current is exactly equal to Q3's base current. It's also assumed that Q2's base current is exactly equal to Q4's.
Or as the article puts it, Assuming all four transistors (two NPNs, two PNPs) are perfectly matched ...
This assumption worries me. In my opinion, base current cannot be controlled. You can attempt to find matched pairs (in this case, matched quads), but how well do they match as the bias point shifts, as temperature varies, and over time (matching "drift")?
Thanks but no thanks.
Last edited:
I can assure you that it does work, and the base current is a minor issue.
After all, when you design a circuit intended to work with unsorted transistors like BC549 (without suffix), you have to cope with Hfe's in a ~10 to 1 range, with the base current similarly undefined, yet it can be made to work.
That is certainly not the main problem.
The main problem is that you have to ensure a free equilibrium between the bias control transistor and the transistor under control.
This more or less rules out low value B-E resistors, darlington configurations, etc.
This means that the actual usefulness of the topology is quite limited. Limited in fact to low-power, single transistor pair output stages.
I have some authority to say so, since I have a bit of "prior knowledge" about that circuit...
The topic has already been discussed here:
http://www.diyaudio.com/forums/solid-state/160285-class-b-w-o-crossover-distortion-1975-a-8.html
J. Broskie has also written about it here (near the bottom of the page):
CCDA & Class-AC
And then here (bottom of the page):
Tringlotron Deconstructed
Very recently, some other variants have been shown here and in subsequent posts:
http://www.diyaudio.com/forums/soli...le-mini-transistor-amplifier.html#post2895329
No, unusually the output of the control transistor(s) is the base current, and that base current controls the output transistor(s).
The quiescent current will be determined by the VAS current and the ratio the Hfe's
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.