Does my perseverance pay off?

[keantoken]

Okay, I've had a few... umm... embarrassing moments with my recent ideas on this forum so I'll try to ask more questions than I attempt to answer.

I have had this idea for a little while and have had some comments on it that are not necessarily positive. But on the other hand others don't seem to think too negatively of it. So I want a final opinion. the attached topology was inspired by some perseverance I had about the effects of an unstable load on the emitters of transistors.

I thought that an unstable or reactive load on a transistor's emitter would cause unintended fluctuations of the bias voltage, so if we made the collector the output point, then problem solved. would my perseverance pay off?

If the answer varies between situations, please be specific.

[CBS240]

Originally posted by keantoken


I have had this idea for a little while and have had some comments on it that are not necessarily positive. But on the other hand others don't seem to think too negatively of it. So I want a final opinion. the attached topology was inspired by some perseverance I had about the effects of an unstable load on the emitters of transistors.

I thought that an unstable or reactive load on a transistor's emitter would cause unintended fluctuations of the bias voltage, so if we made the collector the output point, then problem solved. would my perseverance pay off?

If the answer varies between situations, please be specific.

First, what about the AB bias emitter folower or Darlington EF didn't work for you? Was the load dependant instability related to bias or to the amp driving it?
I haven't had many problems with EF outputs if biased properly.

IMO, this circuit could work but the base of Q3 will be -V2-1.2V-VR2, and base of Q4 is V1-1.2V-VR1. I'm thinking you could use darlington outputs for Q1 & Q2 so Q3 & Q4 can be small, drive the bases of Q3 & Q4 with two complementary differentials. If you could include a Vbe multiplier related to bias currents in the differentials to adjust bias for thermal stability, you might just have an amp. IMHO EF circuit is better for driving lower impeadance loads.


There is an output topology I know works similar to this but base of Q2 is driven with collector of Q3, same for Q1 & Q4, emitters of Q3/Q4 tied together, feedback from output to emitters with voltage divider...or something like that. Vbe multiplier in there somewhere. Point is that the bases of Q3/Q4 are biased near 0VDC and was driven with an op-amp with GNF because the Av comes from the collectors of the outputs. I'm sure it is on this forum somewhere.

[keantoken]

Thanks, CBS

I got a comment that it wasted a lot of power through the bases, so if we used high-gain devices, we could cut that down. But I haven't heard much about Hi-Fi amps using high-gain outputs.

I had an idea about feedback between the biasing system and the output. It seemed to work well for correcting output stage distortion problems, but didn't do too well for phase shift vs. frequency. This is just my simulation, though. I have attached the idea in this post. I just want an opinion, I would rather have someone else take this idea further if it's worth something.

[ilimzn]

Although perseverance is a virtue, persevering stubbornly in all the wrong places makes become it's own oposite.

Here's a couple of facts and questions... you might find reading and answering them helpful:

1) Any change in collector or emitter current changes the voltage between B and E, as well as the base current. This is because Ie = Ic + Ib.

2) Changing voltage on emitter with base being held at constant coltage, by definition changes Vbe directly, by the same amount, forcing the transistor to adjust emitter current through the load to opose this change. Since base current changes exponentially with Vbe, and emitter current is Beta times larger, given that typical beta is double digit valued, changing of the emitter voltage by xternal means while sufficient base current is available, even a little bit, changes emitter current a GREAT AMOUNT. Because of this, a common collector stage (or emitter follower) is the one connection of the transistor that is LEAST susceptible to load variation.
On the other hand, eitter and collector current stays remarkably constant with change of voltage on the collector by external means - unless you try to use such a voltage that forward biasses the B-C junction.

Now the questions:

1) Please explain what you mean with 'unstable load'?

2) Please explain what you think the undesired effect of said load on emitters is, and why you think it is bad?

3) See if you can derive the base to collector voltage for either of your output transistors.

4) What limits the maximum voltage output of your output stage, and what happens when it's reached? Hint: check base currents of output stage transistors

Comments:

Just because you remove one problem by shifting to a different topology, it does not automatically mean you did not introduce other problems. In your case, several more than you 'solve', actually. If you answer the questions above, you will figure out what they are. Hint: you don't need to use simulator.

A bit of advice (again): simulators tend to sometimes produce garbage even when proper circuits are input. They definitely do work by the 'garbage in - garbage out' principle. Just because the input and output look fine, it does not mean the other currents and voltages in the circuit are within realistic ranges for a real world application. It all depends on where you look, and in electronics, you can't afford not to look pretty much everywhere and simulate for all sorts of overload, clipping and similar conditions. Given what you have posted (again!), it seems that you still need to, first and foremost, look at some circuit basics WITHOUT a simulator. Otherwise, it will take you years (perhaps more than you have at your disposal!) before you figure them out from countless simulations. Of course, if you want to persevere in making it difficult for yourself, be my guest... it's just that you will end up getting less and less (and finally zero) comments from people who could teach you the most. They just don't have that kind of time on ther hands...

PS the bias feedback idea has merit, and has been used before - just not in your circuit.

[CBS240]

Originally posted by ilimzn
(snip) On the other hand, eitter and collector current stays remarkably constant with change of voltage on the collector by external means - unless you try to use such a voltage that forward biasses the B-C junction.

Makes sense to me considering the collector is the dependent current source.



Kean

In your circuit the collecor/base of Q5 is 0.6V above the DC output which should obviously be 0V. The emitter of Q4 is 1.2V+VR3 above the negative V rail. All transistors seem to be saturated.

[keantoken]

I'm going to be entering high school next year, so maybe I can find a tech class. I will try my best to refrain from posting until then.

Thanks, everybody for being as patient as you have been, and sorry for being a nuisance. Maybe electronics just isn't my thing.

- keantoken

[ilimzn]

Originally posted by keantoken
...I will try my best to refrain from posting until then.
Thanks, everybody for being as patient as you have been, and sorry for being a nuisance. Maybe electronics just isn't my thing.

Only if you insist on learning about it from a simulator!
In the mean time - you can read, you can write, you can use Google. So, sit down and DO - there are several electronic courses on line...

[CBS240]

Originally posted by keantoken
(snip) Maybe electronics just isn't my thing.

- keantoken

I think what ilimzn is getting at is you must understand how and what makes a transistor work, starting from a silicon crystal, to the doped layers, depletion regions, ect. All this determines the properties and parameters a transistor. Then you may have a better understanding of how it will function simply by looking at a drawing on paper. What good is sim information if you know not what it means? I should probably use a simulator more often, but working with the real components is much more fun. At least I know the results are real. Seriously, don't take it so hard, when you have as much knowledge as ilimzn, I suppose you can be short with people.
If you are in HS, you should have plenty of time to learn the 'basics', but be patient. There is a lot more to learn than there is just to build a speaker box. If you remain interested, there are many schools of higher learning that can teach you all the lovely math that prove the basics. There is lots of info on the net but I don't know how reliable it is.
IMO, it is better when someone with knowledge compiles the info into a form that can be understood easier(classroom)...(diyaudio forums )