The diode at the gate (or the gate-channel diode in a JFET) rectifies your oscillation signal amplitude and charges the series capacitor so that the gate will be negative. The larger the amplitude is, the more negative the gate gets, and the drain current gets smaller. This is the AGC which keeps the signal reasonably clean.
The trick is from the tube era. A grid-dip meter functions by measuring the grid/gate voltage by measuring the current in the grid resistor.
-- Tauno Voipio
So you're essentially suggesting to convert to a Colpitts. I have built Colpitts oscillators before and they seem easier to understand. I thought I'd take a shot at this strange Hartley and the use of tapped inductors.
My use of R5 was to test the effect of an external load to ground taken from the summation point. I wanted to see how low my load resistance can go.
Couldn't find anything in the electrical spec related to input capacitance. It seems they don't specify it.
Could you explain a bit more why? According to the book, L is at DC ground hence the average will always be zero.
Now I'm going crazy. My multimeter connected to the tank shows 100mV in the AC setting when my scope is showing 9V p-p. I thought the Ac setting measures RMS!!
When the circuit is operating properly it is desirable to keep the gate from going positive WRT the source as forward biasing the gate source junction is very bad news for circuit linearity and JFET operation. The cap charges up against the diode to accomplish this.
Jan, that is a Colpitts. Please do not leave the gate capacitor and resistor out. They provide the amplitude control, see my previous mail.
-- Tauno Voipio
The multimeter AC measurement is not reliable on these frequencies, the calibration is for the line (50 / 60 Hz).
-- -Tauno
This sounds intuitive but I noticed things are not so linear.
I used two probes on my scope to subtract source from gate voltage (VGS). T here are 3 phases of operation:
- Supply is less than about 2.9V. In this case there is no oscillation. Sup ply is drawing 20 mA
- Supply is from about 2.9-3.9, oscillation starts and as you increase/decr ease supply voltage, VGS will go more/less negative. Supply current also go es from 20mA to almost zero very rapidly. You see there is a very small win dow where this happens. Also there is some kind of hysteresis going so the tripping points will change depending on whether you go up or down.
- As supply goes higher than about 3.9, current almost vanishes, and the ch ange in the negativity of VGS changes very little as you increase your supp ly.
This reminds me of CMOS inverter action.
The circuit is much more complex than I thought and very ecucational.
What frequency. What multimeter? You're lucky if you have a meter that covers more than a few kHz.
My (expensive) Fluke 87V only goes to 20kHz on *some* ranges. On others it's 5kHz.
My even more expensive HP bench multimeters likewise.
--
"For a successful technology, reality must take precedence
over public relations, for nature cannot be fooled."
(Richard Feynman)
Yes. It was a dumb thing to expect :) What was I thinking?!!
Something like an HP3400A, HP3403A, or Boonton 93A can easily do that. All pretty cheap on eBay if you're patient. (I just got a 93A in excellent shape for $120 including shipping.)
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 USA +1 845 480 2058 hobbs at electrooptical dot net http://electrooptical.net
On a sunny day (Sat, 20 Jul 2013 20:53:24 +0300) it happened Tauno Voipio wrote in :
Is it really American to give every freaking circuit a name? I do not even remember those names.
An oscillator is basically an amplifier with feedback so big and a phase so that the thing oscillates. To give every version of amplifier with feedback > 1 a name is insane. There are an almost infinite number of ways to make oscillators, some wanted, some unwanted (parasitic).
WTF is it to you? I provide a better, simpler, cheaper circuit, do you REALLY think the amplitude of this one CHANGES? For WHAT?
If you want amplitude control add a potmeter in the output, and in my case take the signal from the source. ;-)
If you want to poke around an RF circuit and take measurements of the voltages therein, you want an RF voltage probe.
It's really only necessary if your oscilloscope bandwidth is lower than the voltages in your circuit.
-- Tim Wescott Control system and signal processing consulting www.wescottdesign.com
On a sunny day (Sat, 20 Jul 2013 10:12:25 -0700 (PDT)) it happened "M. Hamed" wrote in :
I try to avoid tapped inductors, as well as tapped capacitor, and come to think of it tapped resistors. See my reply to Voltpro or whatever for names of amplifiers with feedback > 1.
Zero, as I explained before.
Measure it?
say the amplitude is 10Vpp, that would be 5 V below zero, and awfully cold. And it is not always a good idea to grab from L, ads that loads it (Q that was so precious to you), in my example take signal from source,
On a sunny day (Sat, 20 Jul 2013 10:25:47 -0700 (PDT)) it happened "M. Hamed" wrote in :
Does you multimeter go to 1 MHz? most hardly to 100kHz.
Americans, Brits, and (I believe) Australians, Germans, and Frenchmen honor designers by naming their designs after them. Even your idols, the Soviets did.
I'm sure I could name some Nederlanders, too, given time.
--
"For a successful technology, reality must take precedence
over public relations, for nature cannot be fooled."
(Richard Feynman)
On a sunny day (Sat, 20 Jul 2013 12:42:37 -0700) it happened Fred Abse wrote in :
A lot of time, I would not call any amplifier with feedback > 1 an 'invention'. I once called a circuit he designed after my boss, but after a while he told me not do do that.
With these things and inventions: Edison invented the light bulb, but IIRC an Englishman was there first (Edison lost the patent), and a while back I was surprised to read a German invented the computer, etc etc... National pride, egos, money, it is all in the game. I agree like with naming stars XYZ1232132 from some catalog, naming circuits with some alpha numeric would make it more difficult perhaps, but why bother, we can always show the diagram.
Here are 2 other oscillators that may make your head spin:
Do you see those 2 round white ceramic disks? That is what I happen to be playing with these days.
Ok, now it's starting to sink in! Thanks.
Yes, I see them. Tell me more.
--
"For a successful technology, reality must take precedence
over public relations, for nature cannot be fooled."
(Richard Feynman)
It is frequently convenient to refer to standard circuits by name to assist communication between discussers ;-)
The Hartley can be constructed using two inductors rather than one that is tapped. In my opinion, that makes it sort of a Colpitts anyway. I don't like tapped inductors because it is harder to change for experiments. YMMV.
While I'm thinking about it, I had some problems getting a Colpitts to start and run properly. I learned that my ceramic caps had high (for ceramic) dissipation factors (low Q). Replacing them with NP0 cured the problem. You may not run in to this problem, but it might be worthwhile to remember.
You're doing great. Keep up the good work.
Cheers, JohnS
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.