EMRFD Message Archive 4208

Message Date From Subject
4208 2010-02-06 04:46:02 Weddig, Henning-C... AW: [emrfd] Re: Crystal Test Set Designed by Chris Trask
Phil,
the starting and dying of ocillations just tells You that there is just not enough "loop gain" within the circuit to sustain stable operation, i.e. the loop gain ist just a few per mille or whatever below 1. For start up the gain must be higher than 1 in order to be able to start from noise until a stable oscillation is reached Then the loop gain mus be exactly 1, but beware the phase condition (loop phase must be 0 deg or 360 deg)!

For my circuitry I noticed that there was just not enough gain even if the output of the opamp slew to maximum, that is why I increased the resistor from 100 ohms to 330 ohms on the collector of the differential amp.

I found this too low gain condition by inserting a signal from a signal generator to the "left" side of the x-tal socket and measuring the signal at the right side with an oscilloscope. The amplified signal should be higher than the injected one, which wasn´t.

Another method would b have been to shrot the x-tal socket leads and see if theis oscillator would work on whatever freqeuncy (defined by the circuitry itself)
By then introducing a variable resistor and changing it´s value would give some information about the gain margin of the oscillator.

Regards
Henning Weddig



________________________________

Von: emrfd@yahoogroups.com im Auftrag von Phil Sittner
Gesendet: Sa 06.02.2010 04:58
An: emrfd@yahoogroups.com
Betreff: Re: [emrfd] Re: Crystal Test Set Designed by Chris Trask




Chris-

I found the cause of my problems with the test set-a poor choice of transistors. I revisited my breadboard and used a CA3086 array and your circuit performed flawlessly. It's pretty obvious to me that there just wasn't sufficient gain for the transistors I tried. The issue I had with the circuit starting and then dying out is consistent as well. The CA3046 is available from Jameco for less than a buck , the CA3086 is lesss than $2, and either seems to be a good choice. Thanks again for the support you gave me throughout this effort and keep up the good work.

Phil kd6rm
----- Original Message -----
4210 2010-02-06 08:58:11 Phil Sittner Re: Crystal Test Set Designed by Chris Trask
Henning-

Thank you for explaining how you evaluated the circuit. I spent a fair amount of time trying to get the unit to oscillate and finally did a temporary fix by using a pot to vary the bias to Q3(T2). This should have been a big hint!

I love your method of determing loop gain and will have to try it. The CA3086 worked great when I used the 330 ohm resistor you recommended (I didn't try the 100 ohm value).

Again thank you for explaining your methods.

regards,
Phil kd6rm

----- Original Message -----
4214 2010-02-06 12:57:44 Robert Cerreto Re: Crystal Test Set Designed by Chris Trask
Phil and Henning,
 
I performed the same experiments that Phil did and duplicated Henning's test methods and came to the same conclusions. I pulled a CA3086 from another project and tried that using a 330ohm collector resistor. That did the trick. However, I want to look into this further to see how high I can go in frequency. I ordered some CA3086's and CA3040's. Seems like they will make nice balanced mixers as well. May not be state of the art, but, I enjoy working with discrete transistors and appreciate well matched pairs in some of the "older" IC's. I also have some dual 2N3904's in a SMD package and they are very nice to work with for similiar applications.
 
Instead of the small resistance in series with the buffer base leads, I used some ferrite beads. This has been a habit with me when the application willl permit it. I once built a frequency counter from discrete stuff and learnd how well emitter followers can oscillate!!! That was long ago when I had hair on my head.
 
What a great list this is!!! I always complained about not having someone nearby to rap with. Well, this list has taken that complaint away!!!
 
73, Bob WA!FXT
 


4215 2010-02-06 15:45:31 Phil Sittner Re: Crystal Test Set Designed by Chris Trask
Bob-

I tried some 5th overtone 100mhz crystals and the unit oscillates fine at the fundamental frequency and I had no difficulties with other crystals ranging from 3.5 through 19mhz. I think Chris hid the ball out of the park with this one and the CA3086 did the trick. These days it's difficult to find rf signal transistors and we find an abundance of switching types like the 2n3904 and 2n2222a. Thank goodness that the arrays like the CA3086 are still readily available. Keep up the experimenting.

regards,
Phil kd6rm
----- Original Message -----