Author Topic: stability of class A driver I have built for my YIG  (Read 1348 times)

0 Members and 1 Guest are viewing this topic.

Offline john23Topic starter

  • Regular Contributor
  • *
  • Posts: 207
  • Country: aq
stability of class A driver I have built for my YIG
« on: December 13, 2024, 08:38:48 pm »
Hello , I Have YIG with the following properties , I want to drive 20mA with input 50mV at most.
The following configuration was chosen.However I got a very serios oscilations.
I know that R3 is current sensing of the feedback.
In photos 1 and 2 I have a very bad oscilating driver, then i changed a little the configuration and got a much more stable responce as shown in photos 3 and 4.


Why the opamp has only a capacitor as a feedback loop, what is the logic in that?

Ltspice file of the non stable version  attached.
Thanks.

https://eu.mouser.com/datasheet/2/609/AD8033_8034-1502231.pdf
sesitivity: 450 KHz/mA
3dB bandwidth 2.2MHz
resistance 2Ohm
inductance 1.5uH

 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 9131
  • Country: us
  • Retired, now restoring antique test equipment
Re: stability of class A driver I have built for my YIG
« Reply #1 on: December 13, 2024, 10:05:31 pm »
Why did you use such a small feedback resistor on the op amp (100 ohms)?
That means the gain from the input to the MOSFET gate is only -1 V/V.
Also, why is the drain resistor on the MOSFET drain only 10 ohms?
 

Offline john23Topic starter

  • Regular Contributor
  • *
  • Posts: 207
  • Country: aq
Re: stability of class A driver I have built for my YIG
« Reply #2 on: December 14, 2024, 07:23:19 am »
Hello TimFox, this is a good  question.
As i see it this opamp is the Vgs source for the mosfet.
the mosfet is a sort of current duvuder between the load and the ground at the source.
I dont know why they put a capacitor in the feedback.
I chose the OPAMP model blindly and played with the values till i got a good opening of the mosfet.
What is the  analog logic for putting the capacitor in the feedback?
What is the mathematical logic behing this opamp structure which is supposed to be type A driver?
Thanks.

"Why did you use such a small feedback resistor on the op amp (100 ohms)?"
 

Offline mtwieg

  • Frequent Contributor
  • **
  • Posts: 409
  • Country: us
Re: stability of class A driver I have built for my YIG
« Reply #3 on: December 14, 2024, 01:33:08 pm »
It would probably be much easier to put the coil directly in series with the source or drain of the FET, like in the circuit described here: https://www.microlambdawireless.com/resources/driverappnote2.pdf

Why the opamp has only a capacitor as a feedback loop, what is the logic in that?
The feedback loop is not just the components directly connected between the output and inverting input of the op amp. The other components R2, R3, R4, M1, and L1 (in your second schematic) also act as negative feedback. Those components dominate at low frequencies (ensuring the output current accurately tracks your input voltage) while the feedback capacitor help stabilize the circuit at high frequencies (in exchange for slowing down the response time of the coil current).
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7532
  • Country: pl
Re: stability of class A driver I have built for my YIG
« Reply #4 on: December 14, 2024, 02:26:27 pm »
What's the point of this common source amplifier here?
Controlling inverting stages with opamps is not entirely trivial.
 

Offline bson

  • Supporter
  • ****
  • Posts: 2506
  • Country: us
Re: stability of class A driver I have built for my YIG
« Reply #5 on: December 14, 2024, 09:07:18 pm »
Consider what happens on a falling output.  L1 keeps going through magnetic inertia.  M1 Rds jumps.  U1 has high input impedance.  This means R3 is left to drain the voltage buildup in L1.  Perhaps R3 is too big?

You can't just reduce R3, and besides it will always show a similar pattern so you just trade quiescent current for a flatter output.

Fundamentally you need a complementary driver to M1 to pull the output low.

But for 20-50mA you may not need M1, you could just drive that directly with many modern opamps, assuming the supply voltages aren't out of spec.  At least up to 35mA or so, but probably also 50mA.  R3 is fine.  I'd suggest adding a flyback diode (Schottky rectifier variety) across L1, just to protect your circuit in case of an abrupt power-off.
 

Offline KE5FX

  • Super Contributor
  • ***
  • Posts: 2115
  • Country: us
    • KE5FX.COM
Re: stability of class A driver I have built for my YIG
« Reply #6 on: December 14, 2024, 10:27:05 pm »
For just tuning the FM coil, I'd use a high-power opamp by itself as the Micro Lambda appnote suggests.

Tuning the main coil is where things get interesting.  ADA4870 could handle that in many cases, but you absolutely do need feedback that represents the current through the reference resistor.  The Micro-Lambda app note skips that in the FM coil driver because they probably assume it will be operated in a PLL (will yours?) 

In this circuit, the lag introduced by the tuning coil needs to be compensated in the - feedback loop.  Look at the schematic of some of the old HP sweeper plugins like the 86222A, and you'll see multiple series RCs for that purpose.  You don't necessarily need to do the math, just drive it with a square wave and pick some RC components that minimize the ringing.

There should be a small resistance at the MOSFET gate as a parasitic suppressor, and there should be a larger resistance in the return line to the opamp's + input to keep from exposing it to high flyback current.
 

Offline hanakp

  • Regular Contributor
  • *
  • Posts: 124
  • Country: cz
Re: stability of class A driver I have built for my YIG
« Reply #7 on: December 16, 2024, 12:54:47 pm »
From theoretical standpoint, you're driving an inductive load from a current-output amplifier. This produces the same stability problems as when you drive a capacitive load from a voltage-output amplifier. Google for "driving capacitive loads with opamps", but this is a good start:

https://www.analog.com/en/resources/analog-dialogue/articles/ask-the-applications-engineer-25.html

If it doesn't help, then you may have to design a full-blown feedback loop compensator. A few years ago, I found these white papers as the most useful:

https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/38/slup340.pdf
https://www.ti.com/lit/pdf/slva662
« Last Edit: December 16, 2024, 01:01:45 pm by hanakp »
 

Offline john23Topic starter

  • Regular Contributor
  • *
  • Posts: 207
  • Country: aq
Re: stability of class A driver I have built for my YIG
« Reply #8 on: December 17, 2024, 12:45:16 pm »
Hello,I have made a new driver as shown in the LTSPICE files and  photos  by the manual you reccomended.
Its used for a PLL system  where I need to tune at a resonance  shown in the network aqnalyzer photos attached.
How do you reccomend me continue developing this driver so i will be good enought to tune the FM coil the this hight Q resonance?
Thanks.
https://www.microlambdawireless.com/resources/driverappnote2.pdf

YIG FM coil data:
sesitivity: 450 KHz/mA
3dB bandwidth 2.2MHz
resistance 2Ohm
inductance 1.5uH


 

Offline john23Topic starter

  • Regular Contributor
  • *
  • Posts: 207
  • Country: aq
Re: stability of class A driver I have built for my YIG
« Reply #9 on: December 18, 2024, 12:00:12 pm »
How can I know if my driver can tune to my resonance?
Thanks.
 

Offline khach

  • Contributor
  • Posts: 38
Re: stability of class A driver I have built for my YIG
« Reply #10 on: December 18, 2024, 02:06:03 pm »
For high stability YIG driver at constant current ( single frequency, not sweep) your need additional voltage feedback circuit ( in red mark). That circuit stabilize current feedback by additional voltage signal from coil inductance. If YIG switch to sweep operation mode that circuit may be disabled by external switch ( not shown on picture)
 

Offline john23Topic starter

  • Regular Contributor
  • *
  • Posts: 207
  • Country: aq
Re: stability of class A driver I have built for my YIG
« Reply #11 on: December 18, 2024, 09:21:53 pm »
Hello , COuld you post the full photo?
I wat to try and simulate it.
What about the resonance photos I posted  .
Given the tuning needed to lock on this resonance .
What propertion does the controller needs to exsibit?
Thanks.
 

Offline khach

  • Contributor
  • Posts: 38
Re: stability of class A driver I have built for my YIG
« Reply #12 on: December 20, 2024, 02:09:45 pm »
That's part of Gigatronics 610 YIG drv.
 

Offline john23Topic starter

  • Regular Contributor
  • *
  • Posts: 207
  • Country: aq
Re: stability of class A driver I have built for my YIG
« Reply #13 on: January 24, 2025, 01:42:52 pm »
Hello  ,I have built a PI controller shown below.
How can I convert the PI controller AC behavior into current driver?
Thanks.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf