DIY Audio Projects Forum

I said it didn't work unfortunately. I will look again as it may be a misplaced connection, I ran out of bench time to check why, it seems it was oscillating That's why I said I needed to do more research.

I like this one and would like to get it to work because it could possible to drive it DC coupled from my detector (I'm out at the moment, but when I get a chance I will try to post the schematic of it, it may give more context.) Also, I think that your circuits current consumption would be dependant on its output, and satisfy the specific requirement of battery economy Like class B push pull???



I want to do it as well as I can. I'm competitive, so shoot me!

Thanks Alan, sweetie! (it is Alan you call yourself isn't it?) You're too kind!

Luv, Ami.

Here is the sliding bias that should work on 3V, I did calculation and change the resistor values. The output swing is limited to about 1.1V peak to peak without large distortion. Of cause it doesn't have the swing of the one with MOSFET, but it should work.

Make sure you check the voltages at different points I wrote in RED. You don't have much room for error. Report back and we can do some minor tweak.



School project and real product design is totally different. We go through a lot of verification to check whether if we build 1000 of them, they all behave the same, we worry about reliability, predictability. This sliding bias in quite advance. The Aleph power amp designed by Nelson Pass using this. He is very famous in high end world. His brands are Threshold and Pass Lab. That's the reason I am interested in this sliding bias. This is one of his smaller amp design:http://www.kk-pcb.com/a3serv.pdf

You can see the schematic. The main difference is he use MOSFET as common source output. In my circuit, I use Q1 as emitter follower. Using common source presents a lot more complication. You will need closed loop feedback to stabilize the output so it won't swing to rail. Like the second circuit that I use two transistors with gain. You have oscillation because it's a closed loop feedback to stabilize the output and unless you do it right, it will oscillate.

Personally, I bought one of Nelson Pass's Stasis amp, I am not impressed, I don't like the way he use common source as the output impedance is governed by the negative feed back and it's a bad practice to rely on negative feedback to get low output impedance. I actually rather use the circuit I have here, just use MOSFET as Q2........BUT again, back to production design, it's a whole lot more complicated than the simple circuit. real amp has to have many output transistors in parallel, eg. Q1 and Q2 is duplicated like 8 or 9 times to get the sound quality. That's when all hell break loose and that's where engineering is.

Anyway, I am long winded, I am very into amp design, once I get going, It's hard to stop talking.

Yes, My name is Alan.

I want to emphasize, I am pretty sure I can speak for other people that are helping you here. Don't take the circuit we gave you as a proven working circuit. It's just suggestions, it's up to you to breadboard, test. If it doesn't work, you note down DC voltage at different points like I labeled in my circuit diagram. You report back with information, what's not working, how's it's not working, label the voltages you see and give description.

DC analysis is about the most important thing in learning electronics.

then with the info, we can help you to modify it to make it work. It's an interactive process. So if you just come back and say the circuit doesn't work, we cannot help you any further. I am sure every one of the circuit will work if you go through the process.

Alan.

I get what you are saying, I'm not dismissing anything, that last circuit I just didn't get time to study it. I asked teach last week why such severe design constraints. He said that he chose the exercise as it wasn't really possible to just copy a working design from the net and get away with no thought input at all. At least we have a full term to sort it.

I found that r&ec magazine design, and some guys on another forum have suggested I use that verbatim as it fills the specific precisely. Though there is a problem of a unobtainium transformer, no details about it have turned up, and I'm not experienced enough to know how critical it is. One suggestions was to replace it in the design with a "choke and capacitor output" but no suggestions of how that is done (what kind of choke?), the guy suggested it then stopped talking, (should I be able to work that out by myself?) I know that a choke can be used as an output device load to improve efficiency but there must be a laid down method of calculating the value and size of the components.

Now who's rambling!

Here's my radio schemo:-

And this is the Sir Douglas Hall "sliding junior" 100mW amplifier thoughts?:-

R1= 1k
R2= 10k
R3= 220k
VR1=1k5
C1= 1nF
C2= 10nF
C3= 800uF
D1,2,3=1N4148
The only other problem I see with this is the output transistor has a higher Ic and Pwr dis than the ones we're allowed.

O.M.G. I think I'm turning into a nurd!

Time for bed!
Luv, Ami.

This sliding junior is WRONG in so many ways. I wrote down the values, put in some voltages and labeled 1 to 4


1) Input is shorted by 1nF cap. What if the driving impedance is high like a common emitter input, you short out the high frequency.

2) I wrote down the DC voltage. TR1 need the base to be at about +2.3V to turn off to start conducting, anything higher ( say +2.5V or higher), TR1 will remain OFF. You look at the voltage of the junction of D2 and R1. It's SET at about +1.6V. With D3, TR1 will NOT even start turning until VR1 is adjusted towards the bottom.

3) Once TR1 start turning on, this will turn TR2 on very fast as there is no other way to remove the current coming from the collector of TR1. TR2 will put the emitter to about 2V ( Vbe of TR2 plus saturation voltage of TR1). High current will drive through the primary of T1. You might burn something.

4) There is no control of sliding bias that I can see by adjusting VR1.

Also, if the input is not AV coupled, that is having Rin pulling to ground ( 0V) TR1 will be pulled full on as D3 will be reverse biased and cook the transformer T1

Others can take a second look. To me, this circuit is WRONG in so many ways it's not even funny.

This is how you look at the circuit. Before doing anything, analyze the DC voltages, make sure it makes sense. Then verify by measuring the circuit. If you can get the DC making sense, everything will come easy.

What's wrong of being a nurd? I am a nurd and proud of it. When I was in my late teens and very early 20s, I was the cool kid. I was in the band, winning competitions and performing on radio and tv. Then I grew up. That period of time is so so short when looking back. In college and in the outside world, who cares. Look at all the home coming king or queens, they were so cool, then have a kid, getting fat.....a lot of them become trailer trash!!.. Then look at Bill Gates, he's a nurd and looks like a nurd, He's the supper cool one now. Look at Teresa May, you can laugh she's old and fat. She's your prime minister!!! When you have a cool career, well established, then you are truly cool. Have all the home coming kings and queens working under you.

Thanks Alan.

I think the original design of the "sliding junior" was for a small radio too. My uninitiated thoughts were:
The capacitor C1 was included to filter rf.

It would definitely need DC blocking capacitor at the input, of say 1uf ?

I would put a resistance between the collector of tr1 and base of tr2, and another from the collector of tr2 to +3v for turn off and protection? Maybe 100ohm and 2ohm respectively?

The circuit seems to me to be a bit upside down The interconnection of the transistors makes them work as one high gain PNP, correct?

Sir Douglas Hall was apparently a respected electronic author, and a member of another forum said he built the sliding junior as a kid, he said it was temperamental to get working and the audio quality wasn't great on quiet to loud step level transitions, but it did work.

It may be worth breadboarding. If it works it would be interesting to find out HOW. If it doesn't work its still something to put in the project notes

At my school the nurds are as cool as the athletes, I always enjoyed helping Dad with his hobby stuff, he worked as an industrial control and instrumentation design engineer and made stuff like this look easy, so I thought it was easy! But its fun finding out that its not.

Can't wait to carry on later this week analysing your circuit too.

Laters.
Luv, Ami.

It's actually high gain PNP driving a NPN emitter follower. I spent more time on this circuit since it's a sliding bias that I am interested. The problems I mentioned still stand. This is my analysis:

What the circuit tries to do is to vary the current through the transformer according to the input signal. The larger the signal, the higher the current through the transformer to keep it in class A. But this is where the problem begin:

1) in order to vary the DC current through the transformer according to the signal, you need to drive the base of TR2 more positive with the amplitude of the signal.

2) Somehow, a voltage is set up across C2 which is the integrate capacitor that control the sliding bias.

3) D1 and D2 set up +1.6V for VR1. The circuit has to be ACV coupled. When the signal first goes positive, D3 turns on, the output see lower resistance of parallel of VR1 and R1 which is 1K//1.5K = 600ohm. This load down the positive going signal and clip the positive part of the signal.

4) When the negative half of the input signal. D3 turns off, C2 is discharged through R2. So the voltage across C2 decrease. This make the voltage at the emitter of TR2 increase, thereby increase the current through the transformer T1. This will increase the bias and keep the amp in class A. When the signal goes away, C2 charge back up through D3 and R2 and VR1.

Now, ignore the concerns I have in the last post. Just assume it all work. Still, in order for this circuit to work, the output impedance has to be high, so it will clip by the sliding bias circuit on the positive half of the signal. But the impedance through C2 and R3 can load the of the driving stage down.

You need an output transformer to block the DC current. That is one more component that is more expensive than transistors. Also, the big point of transistor amp is to do away with the output transformer that not only limit the frequency response, it also create more distortion.

You might get away with two transistors, but you end up having a more expensive circuit, sound quality is hand held transistor radio at best. Using transformer is so so old!!!

The circuit is so badly design, you have to tweak and tweak to get it to work. I bet when room temperature change, it will go hey wired.

I don't know about the author, but like Raj telling Sheldon in Big Bang Theory " Not all the stuffs come out of you are gold, some are GAGA!!!". This is engineering, it's science. You can analyze the circuit. Looks like some kid wiring it up in the garage.

Finally got back to this!

Just for interest I did build the Douglas Hall design.

You're correct, it was Kaka!

But...

I love a challenge, AND, pulling ideas together have managed to make it work with reasonable fidelity, bootstrapping the bias was the key, as well as giving it degeneration resistance for DC stability. Inductive load is still in there, but is a necessary evil for efficiency at such a low battery voltage and restrictive active component count.

A great old guy (Norman), wound the gapped audio choke for me, otherwise I'd have been sunk with the idea.

I used capacitor to couple the speaker to save an unnecessary transformer winding or tap.

Me thinks Teach was just trying to get us to think, you just wouldn't do it this way with transistors available at pennies each.

The set has been handed over with the overview paper for marking

Any way, for interest I've attached the final circuit.

Luv,
Ami.