Calling all amp theory gurus

[Brion55]

I own a 100 Watt rack mount Traynor (PA system) amp that I've used for years. I'm thinking of building two just like it for a project. In looking at the design, there's something that I haven't been able to figure out what it's for. I first thought it might be an error on the drawing, but in looking at the amp, it is indeed there.

I understand the circuit around the 2N6553 is a current sink, but what in the world does the zenor (1N4745 going to ground) in series with the 22MF cap do? Is it some kind of limiting circuit? I've not seen anything quite like this before.

[Allexx]

It looks like this amplifier has no relay based protection circuit that delays loudspeaker connection to amplifier after powering it up. Usually without that circuit there is uncontrollable DC pulse applied to the speaker duration of which depends on time costants of current sources.
In that circuit diodes of current source are immediately biased by charge current which flows via zener diode and capacitor in series. This capacitor limits the current. This way capacitor in parallel to diodes is being charged much faster then via 33kOhm resistor to the ground. After that DC biasing current is determined only by 33k resistor.
It limits duration and amplitude of the click in the speaker.

[Nelson Pass]

There is at least one major error in the circuit, but setting
that aside, IMHO the current source biasing for the front end
diff pair and VAS is not very good, and the circuit you ask about
appears to be an effort to patch it. I suggest you borrow
a current source design from another amplifier.

[GRollins]

It might be worth your while to review the entire schematic against the actual circuit before attempting to build more amps like it. If the schematic doesn't agree with the physical circuit, then scribble in corrections.

Grey

[Miles Prower]

It looks like a deglitcher to keep the CCS transistor out of cutoff while the 22uF capacitors charge up. The zener has a V(z) of 16V, and has 43.6V between that point and ground. It looks like a hack to correct some problem they had with adding those capacitors to the CCS there. However, that isn't what concerns me about this. See #1 of attached schematic: that capacitor has the wrong polarity. That is just asking for trouble, even if the voltage across is just 1.4V.

[moamps]

Originally posted by Miles Prower
.....even if the voltage across is just 1.4V...

I'ts just 0,7V. The emitter resistor is missing.
VAS CCS doesn't work.

[Miles Prower]

I don't know where that schematic came from, and didn't bother to check past that backwards capacitor to see if the CCS transistor was an internal Darlington or not. If it isn't, then that missing emitter resistor will, indeed, cause it to go into hard saturation and cease to function as a CCS anyway. Makes me wonder if that isn't a sample schematic from a kit. These are often deliberately wrong so as to result in a non-functional circuit so that you will be forced to buy the kit to get the correct design.

Still, WTF are all those capacitors doing down there anyway? I did a solid state design and didn't have a single one in any CCS in that design.

[Brion55]

I went back and looked at the schematic and my hand drawings from the actual amp and found two big errors on my drawing!

The polarity of the cap in question was indeed wrong, and I left out the emitter resistor. I have gone back through the drawing and believe this is now correct.

The amp does not have any kind of relay power on/off pulse protection. I agree the zener and cap are probably an attempt to limit that. I intend to add relay protection if I build any of these. I also wonder if making the bias adjustable would be a good idea.

The amp works pretty well, but if anyone has one around 100 watts that would be better, I'd appreciate seeing the schematic.

Here is the corrected drawing:

[amplifierguru]

Hi Brion,

glad I came in late - you sure you have the drawing right?

The 22MF on the input tail current source as a filter is doing the wrong thing! REMOVE IT

The 16V Zener and 22MF you speak of REMOVE BOTH.

The 15pF should be on the Vas not it's current source. It needs a source impedance - not two diodes. MOVE IT

The input stage RC comp (series 1n5, 470R) is useless for a single ended output. You can move it to across the 2K2 that's feeding the Vas or probably LEAVE IT OUT.

CHANGES

Take the base of your BC550 current source to the top of the two diodes for the Vas current source and change the emitter 15K to 560R . Put a 2K2 between the same diodes and the Vas CCs base with a 22MF to the neg supply - that will do the dethump. Change the emitter resistor to roughly 56R - should give more symm clipping and slightly lower THD.

For starters.

Cheers,
Greg

[Brion55]

Greg et al,

I often wondered why this amp seemed more complicated than it needed to be. I won't change the original Traynor amp as it works fine (even with the extra junk), but I appreciate the help with designing something better and less complicated that I can build for my project.

I think this drawing accurately reflects the changes you have suggested. It certainly simplifies the design, and makes it look like many other amplifiers I have schematics for. If you see any other glaring errors (or even some that aren't so glaring) please let me know.

Thank you all!