DIY Audio Projects Forum

Hello peeps!

I hope this is in the right place!

I've finally managed to make time to improve my little radio set that I built last year so that it could be entered in the schools science competition, but I'm struggling with the audio amplifier section, so that it can be used with a loudspeaker for more comfortable listening than at present, as it's only able to drive an earphone.

I want to keep things as simple as possible, but use only discreet components, (using chips or not being able to explain the circuit function will get marked down) and as the set is running on a single AA cell at the moment, keeping to that would be a nice feature too, but not absolutely necessary as we've been allowed to design for 3V. We've been allowed to use 3 transistors in the design, and in my case that means 2 can be used for audio. We can use BC 549, BC 559, BC327, or BC 337, or any combination. I've already used a BC 549 as the detector.

I know the power output would be low, but a few tens of mW from a 100mV input would be fine, and as the radio is medium wave AM, no pretence of Hi Fi is necessary.

Greg, our teach, says that I should look at a simple sliding bias design for battery economy, he said they are sometimes called "Gray amplifiers", but I've not found much suitable from searching the net, only big power amplifiers, and don't really know what they are, or how the principal works.

The audio at the moment is developed across the 300 ohm DC resistance earphones, and is 100mV on a strong signal.

I don't want anyone to do my work for me but a push in the right direction would be sweet.

Luv. Ami.

Maybe a common emitter then a common collector output buffer to drive your radio
Or two series common emitter for larger gain

Use capacitors to coupple stages from dc

This is a simple sliding bias circuit:

https://www.tubecad.com/2007/11/blog0126.htm

this is interesting, I learn something on sliding bias. Nelson Pass's famous Aleph amp is using sliding bias also, I never really stop and study it, this gives me a chance to understand how it works. I don't particular like it, but it's knowledge.

I modified to use BC337 NPN that you have. I have to change the values as this is bipolar transistor and you use battery ( 9V) and also much lower current.

I biased the Q1 base at +5.6V to give +5V at the emitter.
the idle current of Q2 is set by R7 and R6 through the diode to get 1.4V at the base of Q2
This gives about +0.7V at the emitter of Q2.
I set the idle current by the 27ohm resistor to about 26mA.

The theory is such when there is a signal at the "IN", emitter of Q1 will swing up and charge C2 up and voltage at base of Q2 increase. This increase the current of Q2. So at the negative half of the input waveform, the Q2 pull the output down harder.

When the signal is removed, C2 discharge back to idle level and back to 26mA.



You need to experiment with this, I did not spend a lot of time calculating, just give you a head start.

This circuit is just a emitter follower, there is not gain, if the signal is too small, you need to add a gain stage before this.

EDIT:

I updated the schematic. I forgot C3 the AC coupling cap.

Thanks for the tips KY and Yungman!

Yungman, you have certainly spent some time considering this, thank you!

Normally when we are given projects, the design constraints are given a score of points that are taken from 100 if we can't comply.

In your circuit you have no voltage gain but is it possible that I could use a small step up signal transformer as a load for the detector so I don't have to use more than three transistors for the whole design?

Also using 9volt battery instead of 3volts would loose points, I will have to see if I can get your circuit to work at the lower voltage.

(2×Vce saturation is the problem caused by the low supply isn't it?) But it would be a problem with a complementary, or stacked stage of any type?

Just a sanity check, would this seem a reasonable explaination of the circuits operation?

"The idea behind the circuit is that by sampling the output and raising or lowering the bias on transistor 2 and thereby making its collector current vary in sympathy, provides just enough current for transistor 1 to amplify the signal present at its base at any particular moment, providing better battery economy than with a fixed bias amplifier."

Thanks again.
Luv. Ami.

Going to 3V is much harder. You can change Q2 to MOSFET and gain 0.7V of head room as MOSFET can swing down to Vds=0V without saturation problem. But finding a through hole low Vgs N-MOSFET is NOT easy. I assume you need to build this. This is one example of MOSFET I am talking about. You cannot buy this anymore.

https://www.digikey.com/product-detail/en/vishay-siliconix/SUP85N02-03-E3/SUP85N02-03-E3-ND/2623115

But you can look at Mouser and other places in your area to see if you can find one that is through hole. You need to do your own work.

If you can find a MOSFET, you then recalculate all the resistor like how I explain it and post the circuit. Then we'll talk.

Current of Q2 is adjusted by the signal so it has enough current to pull the load voltage down.

What kind of class is this? This is not a beginners' class!!! I am surprised they offer this kind of class. I don't like this circuit, yes, it might be more efficient, but it takes quite a few cycles to pump up the current. So the there will be clipping at the beginning of an impulse signal.

Also, changing the sink current will change the Vbe of Q1 and this create distortion. Not that your project cares.

I feel a lot of people do things differently just to be different. I am into high end amplifiers, I rather improve an existing circuit that is good than to design an inferior circuit just to be different. Who doesn't want to have a circuit of his own invention, but one always has to ask themselves whether it is better or just want to be different.

Hello Yungman.

At my school we have some subjects that can be chosen at secondary (year 6), one that I chose was technology studies, and electronics practices can be taken as an extra within that course. This is now my second year, I have so far built a photography exposure timer, for my Mum who is a keen photographer and still likes to use her old-fashioned film cameras and develop the film herself. Also built an electronic metronome that I use for music studies.

I would agree with you that there is no reason why it should be made novel at the expense of performance.

Also I think I understand what you say about the MOSFET for transistor 2, and I wouldn't have a problem with smd part. I have good eyes and steady hands I used to mount these devices on adapter boards for my late Dad, but for this project we have been told what parts we can use, the teach just says, "Cathedra mea regulae meae." He did concede to the detector circuit I'd used as it works well.

I'd also like to get this to work with a speaker so I could use it in my bedroom as I've a favorite radio station that only transmits on MW and internet, and I don't really like the earphones.

The Gray amplifier was only suggested, if it were to difficult to make work then something else with TWO BJTs from the list would probably be OK if it could be shown to be better. Also if needed the battery could consist of two "F" cells to supply more current. The brief only says it should be able to run from 3volts and be portable

I won't give up yet.... I don't want to be moved to the home economic class (cookery and homemaking...yuk!)

Luv, Ami

OK, I gave it a try. It's not easy as you don't have headroom. I use a low threshold MOSFET. You have to look for one. There are plenty if you can use surface mount, but it's hard to find through hole ones.



I also have to change the diode to a schottky diode to lower the Vbe of the diode. You don't have head room.
I added a VR1 to trim the voltage as it's much harder to predict the Vgs of MOSFET. you just adjust so you get 0.5V across the 20ohm resistor R8. This will give you about 23mA of idle current.

Again, you have to double check and play with the values of the resistors R6, R7.

The bias is not stable as the bias will change when temp change of the MOSFET. But if you only have to keep it on for 2 minutes to show your teacher, it will work.

It's hard to use BJT for this, I don't have extra 0.7V to spare!!!

I think for your case you can find calculators online for common collector and common emitter schematics.
As suggested earlier I recomend you to capacitor coupple the input and in between stages to block dc but allow ac signal to flow.

I think these simple single stages will allow for plently of voltage swing while being simple.

I think what yungman has draw is a bit too far advanced and I think only advantage over simple common collector buffer is that lower distoriton which is not a requirement.

He mentioned the teach want him to look into sliding bias.