ARRL

Secure Site Login

Forum Home - Rules - Help - Login - Forgot Password
Members can access, post and reply to the forums below. Before you do, please first read the RULES.

MPF102 oscillating at 60 Hz

Nov 7th 2020, 15:21

henryrichter

Joined: Oct 14th 2018, 20:05
Total Topics: 0
Total Posts: 0
I am trying to create a simple Hartley type oscillator circuit using a MPF102 JFET. I am reading 60 Hz wherever I touch the oscilloscope probe to the circuit.
I have rebuilt this circuit, changed the ground spots on the copper clad board changed the components etc. If I take the coil out of the circuit it still oscillates- it appears to be the transistor itself.
For what it is worth with the coil attached the tank circuit generates 20 volts P-P of 60Hz with a 12 volt supply. So that is "working"?
I am perplexed. I'd guess I have a ground loop I can't find?
Thanks
Nov 7th 2020, 19:06

WA0CBW

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
What are you using to power the circuit?
Nov 8th 2020, 08:20

henryrichter

Joined: Oct 14th 2018, 20:05
Total Topics: 0
Total Posts: 0
LM317 regulator power supply. I thought about ripple passing thru the circuit and then used a 9 V battery and it still occurs.
.
Nov 8th 2020, 10:22

henryrichter

Joined: Oct 14th 2018, 20:05
Total Topics: 0
Total Posts: 0
In fact, removing the coil , using the circuit as a common source amplifier or even leaving the source lead free in the air, I still get 60 cycle, even touching the copper clad board.
I've turned off all the lights in the room, that didn't help.

I have a version of this circuit in a die cast box with two variable caps which works between 20-30 MHz. There is a lot of RF energy in the air as passing my hand over the open box shifts the freq and/or amplitude.

The current effort is to experiment with components of this circuit and a following 40673 amplifier to achieve +7dBm at a diode ring mixer

Nov 8th 2020, 18:01

WA0CBW

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
What kind of scope are you using? Are you common grounding the scope and your circuit?
Bill
Nov 9th 2020, 04:51

W1VT

Super Moderator

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
You can have magnetic coupling between your AC wiring and the oscilloscope. A loop of wire can act as an "antenna" on the oscilloscope end, even though you use a shielded cable and "ground" everything. The coupling can be minimized by using twisted pair.

https://www.picotech.com/library/oscilloscopes/rejecting-common-mode-noise-in-oscilloscope-measurements

Zak W1VT
ARRL Senior Lab Engineer
Nov 11th 2020, 16:59

henryrichter

Joined: Oct 14th 2018, 20:05
Total Topics: 0
Total Posts: 0
The scope is a Textronix about 6years old in a plastic case.
The PS is an Elenco 317/337 kit with floating output ground.
I should mention that all the above 60 cycle detections were done with the positive scope probe only. Touching the ground clip virtually anywhere flatlines the output.
I have a conductive mat on my bench top connected to house earth, thus connected to scope ground. I ran a jumper to the circuit from a plug on the mat and the output again flatlines.
I have tried a twisted pair scope lead ( very noisy) and a twisted pair DC power lead to no effect. I simplified my 115V power to one power strip for the PS and scope only. I put the PS and scope on a grounded metal sheet and ran a jumper to the PS chassis. None of these efforts eliminated the 60 cycle.

I built a different oscillator, 2N4416 based, with the gate diode (EMRFD fig 4.4) I got smart and put a BNC connector on the output winding of the coil and ran a BNC cable to the scope. What do you know, it works fine.
This second oscillator has no drain or source resistor so perhaps it doesn't have a pickup point for AC currents?
Speculating
These are only the second and third oscillators I've ever built, and now achieving higher tank voltages, perhaps nothing i have built before had a strong enough RF field to couple with an AC field.

I don't know where I am. If there is a systemic problem at my workbench I'd like to fix it.
I didn't know about differential probes till now; they seem to be mainly for high voltage? and at any rate are not cheap.
Or maybe the problematic circuit is just borderline unstable and thus reactive.
Nov 12th 2020, 11:48

WA0CBW

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
Is the probe "loading" the circuit? Are you using a direct x1, or a x10 probe?
Nov 12th 2020, 12:48

henryrichter

Joined: Oct 14th 2018, 20:05
Total Topics: 0
Total Posts: 0
A 10x probe.
I've had the probe load a circuit before. It dampens the signal but does not induce 60 cycle.
Nov 13th 2020, 15:43

WA0CBW

Joined: Apr 4th 1998, 00:00
Total Topics: 0
Total Posts: 0
Thinking outside the box....... Is the scope's 3rd wire safety ground intact and connected to a 3-wire grounded receptacle? Have you measured any AC voltage between the scope ground lead and the circuit ground?
Dec 22nd 2020, 10:56

henryrichter

Joined: Oct 14th 2018, 20:05
Total Topics: 0
Total Posts: 0
Well, some of us unfortunately have to travel for work. Delving back into the project, I have determined that better construction techniques solves the oscillation problem. Specifically layout and better coil winding techniques- higher Q.
Using the Low Noise oscillator (figure 9.15 ARRL 2021 handbook) I now have 4 coils supplying a minimum of +15dBm from 2.7-29Mhz. Specifically on the 80-10 meter bands the circuit supplies 18dBm with a noise floor of -55dBm and 3dB bandwidth of 100K or better.

I still have the issue of band switching the coils attached to the tank capacitor. I had hoped to use pin diodes but they disrupt the tank. It seems I have only two options: use DB9 connectors to which the coils are soldered, or build 4 oscillator circuits, and switch ON-OFF the Power, Emitters ground and the hot end of L1. Otherwise the coils of the turned off circuits could act as antennas?
Also, I would be switching a coil lead with RF energy. I had hoped to avoid the extra lead length of the wire from the coil to the switch and back.
I will have the same problem on the front end, with 3 tuned circuits, each requiring3 or 4 coils.
Reading several books on design I haven't found this problem covered, or seen a circuit example. Maybe I am missing something.
Any help would be appreciated.
Thanks


Back to Top