Operating Frequency of NPN Amplifiers.

[Reitan]

Thanks,Ron! The diagram you posted was like the one I found, only one question- is it a vew from the top? If it is, I'm in trouble, 'cause thats how they're hooked up right now and I don't even get a click on my earphone when I turn the power on. But, if I'm hearing you right, if I can get these transistors going, I can take the frequency of this radio up pretty high with very few modifications.I'll take one more look at the circuit board,I might have overlooked something,I guess I've been at it over 12 hours straight.It's the first time I ever etched a circuit board,though. It makes it look downright pretty,even if it don't work.

 

[Roff]

<snip>is it a vew from the top?<snip>

No. :cry:

 

[Reitan]

Ha!Ha! I'll redo it immediately!

 

[Reitan]

It Works!

O.K! It's working! Note to anyone who might read this-lead diagrams are viewed from the bottom of the transistor. I am getting a strong local AM station on every tap of my coil though, so I'm guessing something is causing my tank to load up. But the good news is no oscillation, and the circuit is working. Could it possibly be my hi-Z earphone? Maybe I need to pick the voltage up from 8V? Both? I think I'm going insane, it's almost been 15 hours now and when I close my eyes, I see circuit board. Ron, if you read this, thank you. I don't think the 2N5179's are overkill, but then, I once re-wired up a clothes dryer with 8-guage solid conductor. The wires kept burning, now only the clothes do.

 

[Reitan]

O.K., I played with the 1k resistor by replacing it with a 15k variable, and I played a little with the value of R1. I'm not getting any results with unloading my tank. I think I'm changing the biasing for Q1 too much. My tank could be the problem also, so here's what I did to that; I replaced C2 with a variable 5-365pF(I'm guessing), Parallel with that I put a 470pF, then in series with those I put a 330pF. The 330pF is connected to the collector of Q1, as is the tap switch for my coil. The coil is 22 uH, and can be tap selected down to 5uH. In case coil construction makes any difference, I wound 19 turns of homemade litz wire on a pvc pipe, 2.75"(7cm) in diameter, spacing the turns evenly over 2.25"(5.7cm), it has taps with very short leads every other turn. If I am correct,my capacitor set-up runs from 195pF to 237pF. With 22uH in parallel with that,I should be tuning from 2.2MHz to 2.4Mhz.With the tap at 5uH, I should tune from 4.6MHz to 5MHz. However, this is not the case, I am pulling in a teeny-bopper station that seems to be the curse of my existence in this world, it's as if their transmitter were in my backyard.It varies in strength depending on the coil tap. Is it possible that the radio is working, but this AM station is wiping out my reception? Is the capacitance in my tank too high and loading down my tank? I think I could decouple the coil, but that would mess up the tuned circuit in this case,wouldn't it? Could I possiby put a coil in front of my antenna input(C1) to trap this insane AM station? I'm open to possibilities.

 

[Roff]

How do you know the value of your inductor? Is it calculated or measured?
Are you sure you don't have any shorted turn(s)? They will absolutely kill your coil's Q.
I don't think you should fool with the biasing much. It should be pretty much optimum. And you said that you have no oscillations. How do you know? They're like carbon monoxide - you can't hear them, you can't see them. etc. They will often show up as a change in bias conditions when you move your hand near the board, or touch a node with your finger. The finger on a node will frequently kill oscillations. I have often said I wish they made a finger component that you could ship with a product. For Q1, a good point to monitor is the top of the tank (junction with 1k resistor). If that node voltage changes when you touch the base of Q1, it's probably oscillating. Similar tests can be made on the other trannies.

 

[Reitan]

I checked everything out. My coil is just calculated, not measured. I wish I had an L-meter! But it can't be too far off. I didn't know oscillation was a silent phonomena. Now I'm thinking of all the radios I built that didn't work...Anyway, I checked the voltages, and I have a rock-steady 2.35V at the collector of Q1. Just to check, I pulled a 38ohm resistor out of my box and put it between the base of Q1 and C3. Now I think I'm geting somewhere, because I hear a squealing when I touch Q2.I also get it when I touch the 100 ohm RF gain pot.I did fid a spot where a coil lead could have been shorting to ground, so I replaced the coil leads, now there is no more AM bleed-over. You were absolutely right Ron,once again. Does this addition of the 38 ohms upset the biasing for Q2 and make it act funny?

 

[audioguru]

Now you need a tuned circuit at the input to get rid of strong local signals of just about any frequency before the transistor amplifies them.
A tuned circuit at the input of the common-base transistor would be heavily loaded down by the low impedance so its tuning wouldn't do much. Even a tuned trap wouldn't work well with the common-base's low input impedance.
Maybe the input transistor should also be a cascode pair that would have a much higher input impedance, then use a parallel tuned circuit at its input.

 

[Reitan]

I've played with it some more. I put the whole circuit on a circuit board except the tank and the pots. I'm wondering if the 2inch leads going to the RF gain pot are the cause of Q1's oscillation. I changed the base resistor of Q1 to 56 ohms, and the squeaing toned down a bit, but I notice when I adjust the RF gain pot that there is a peak in sound around the middle of the adjustment range, and there are a few spots where it goes real quiet,not quite like a regen, but close in sound. I'll try using copper strips instead of wires to connect my tank circuit and see if skin effect is an issue with the reception, and continue to try to deal with the base of Q1. I'm no engineer, so my interpretation of the schematic probably isn't the best, I can change that around a bit as well. This ain't over yet, not until I set every clock in my house to WWV. I also remember Ron had mentioned putting a ferrite bead on the base lead, I know I have a couple of those somewhere, just need to find them.I will try that as well.

 

[Nigel Goodwin]

I've played with it some more. I put the whole circuit on a circuit board except the tank and the pots. I'm wondering if the 2inch leads going to the RF gain pot are the cause of Q1's oscillation. I changed the base resistor of Q1 to 56 ohms, and the squeaing toned down a bit, but I notice when I adjust the RF gain pot that there is a peak in sound around the middle of the adjustment range, and there are a few spots where it goes real quiet,not quite like a regen, but close in sound. I'll try using copper strips instead of wires to connect my tank circuit and see if skin effect is an issue with the reception, and continue to try to deal with the base of Q1. I'm no engineer, so my interpretation of the schematic probably isn't the best, I can change that around a bit as well. This ain't over yet, not until I set every clock in my house to WWV. I also remember Ron had mentioned putting a ferrite bead on the base lead, I know I have a couple of those somewhere, just need to find them.I will try that as well.

Skin effect is a VERY over used term, and it's EXTREMELY doubtful that it's anything to do with your problems! - also, to reduce skin effect you don't use copper strips, you use Litz wire, multiple thin strands insulated from each other.

 

[audioguru]

The RF amplifier might be oscillating because its power supply isn't bypassed with a 0.01uF ceramic disc capacitor and a 100uF electrolytic capacitor to circuit ground right at the circuit. These caps are very important if you are using a little battery with a high internal resistance, or a power supply farther than a few inches away. :lol:

 

[Reitan]

Thanks, guys. I'm glad I haven't changed the wire out for copper strips,'cause I have homemade litz wire on it right now. I have a 50lb roll of enameled 30guage, and I just gather up 6 or 8 strands, tie one end to a doorknob,and a heavy book to the other end. I hold the wire close to the book with one hand and spin the hell out of the book with the other hand. 2 Minute litz wire. The guage could probably be smaller,but it's all I have on hand. I will bypass the power supply right away and see if it helps, I don't know why I didn't think of that.

 

[JimB]

Reitan, the circuit you have will never work very well.
You have one tuned circuit which will have a low Q due to the loading effects of Q1. As a result, the selectivity will be VERY poor, and any station with good signal strength will just pass straight through.
This is what you are seeing.
Can I suggest that you google "regenerative receiver" and look at some of the circuits that finds.

In a regenerative circuit, positive feedback is used to bring the circuit to the point of nearly oscillating, this results in (relatively) good selectivity and gain.

Give it a try, you probably have most of the bits to build a "regen" already.

JimB

 

[Reitan]

I tried adjusting everything all over again, and I get 2 AM stations and that's it. The bypass of the power supply quieted things down considerably, and I guess JimB has hit it on the head- this reciever isn't all that hot. But I did waste 20 dollars in parts, superglue my finger to my variable capacitor, burn myself several times with the soldering iron, and drip hotglue all over my new workbench. And I guess I even learned a little bit about the original question that started this all. It's not the transistor biasing that gets you recieving a specific frequency, it's the tuned circuits connected to them that does it. I have built a few easy AM regenerative recievers.Actually, I built the same design 3 times with different parts,and tiny alterations. I just have no idea how to change a design from AM to SW. I tried changing the tank on a good AM regen for shortwave and got nothing but static. That's why I thought the transistors had to be specifically biased for a certain frequency range. The regen I built was a popular one, and everyone probably has a schematic of it. If anyone out there has a novel regen idea that will pull in shortwave,maybe even with an audio amp for an 8ohm speaker, I will start a new thread and build it. The more sensitive and selective the better. I appreciate everyone's help with this one,and it did work.Thanks!