Why can't I use a 4N25 for an analog isolator

Discussion in 'Analog & Mixed-Signal Design' started by Babybird002, Mar 30, 2016.

  1. Babybird002

    Thread Starter New Member

    Mar 30, 2016
    9
    0
    I have a sensitive an unusual dash dimmer switch on my vehicle. 1999 Miata. Unlike most dimmers it does not simply connect the dash lights through a variable resistor to ground.

    Because of the complexity of the circuit I need to isolate the circuit optically in order to add an aftermarket radio and gauges [ to dim the lights in the radio and gauges along with the dash]

    I have researched this forum and seen posts that suggest that a 4N25 optical isolator can not be used for an analog circuit, just switching circuits. My question is: Why not?

    Attached is my proposed circuit, if I am way off base let me know. And if you have an alternate circuit design it would be appreciated.

    Babybird
     
  2. Lestraveled

    Well-Known Member

    May 19, 2014
    1,957
    1,215
    Hello @Babybird002 and welcome to the forum.

    I have looked at your circuit and I don't understand what you are trying to do. Perhaps if you were to talk about it some more I might get it. For now I am going to make a guess.

    Do you are think that you can vary the current to the led and the current through the transistor will proportionally vary? The answer is the area that you can linearly control the transistor is very very small. So small to make it unusable. There are other opto-isolators that can transfer analog signals. There is also the issue of how much current they can handle. Am I in the ball park??
     
  3. Bordodynov

    Active Member

    May 20, 2015
    637
    188
    Why can not it? Can!
    4N25_analog.png Look at my circuit.
     
  4. ericgibbs

    AAC Fanatic!

    Jan 29, 2010
    2,503
    380
    Hi B,
    Your circuit is very interesting, but the presence of R3 means that the 4N25 circuit is no longer isolated.

    Eric
     
  5. Babybird002

    Thread Starter New Member

    Mar 30, 2016
    9
    0
     
  6. Babybird002

    Thread Starter New Member

    Mar 30, 2016
    9
    0
    Yes that is it. I did not understand the non-proportionality aspect of the 4n25. While I was waiting for
    a response, I created a simplified simulated circuit. [Attached] You understand what I am trying to do.
    Can you suggest an optical isolator that will provide a proportional output?

    Or is that what post #3 is. That is a doozy. Not sure if I can do that one.

    And I can fix any problem with low power once I can get the basic circuit working.
     
  7. crutschow

    Expert

    Mar 14, 2008
    13,000
    3,229
    How can you do isolation if you only have one battery supply?
    For isolation you need two isolated supplies.

    Explain the problem you are trying to solve exactly.
    Give us the problem, not a proposed solution.
     
    atferrari likes this.
  8. Bordodynov

    Active Member

    May 20, 2015
    637
    188
    I fixed scheme.

    4N25analog.png
     
    absf and ericgibbs like this.
  9. Lestraveled

    Well-Known Member

    May 19, 2014
    1,957
    1,215
    Looking at your first drawing, it appears that the cars dimmer system uses pulse width modulation. I suspect the reason the car dimmer self destructs when cross connected, is that the car lights are connect to +12 and the radio lights are connected to ground.
    So, you don't need a analog opto-isolator. A DC solid state relay might work, like a ODC5. How much current does the radio circuit draw.

    It's late. I'm going to bed. Tomorrow.
     
  10. Babybird002

    Thread Starter New Member

    Mar 30, 2016
    9
    0
    Yes You've got it exactly. Me too. Gone to bed [about time]
    Don't know how much current but very little just a few leds [radio] and five small bulbs [gauges]
    I will research DC solid state relay. And look carefully at B's revised circuit.
     
  11. KeepItSimpleStupid

    Well-Known Member

    Mar 4, 2014
    1,143
    202
    You really haven't defined the inputs and outputs. You have said what the old style dim was: a high powered reostat in series with a lamp.

    I get the "impression" that what you have is a signal called "DIM" which is basically "The headlights are on"and this creates a PWM signal of lower intensity. Whether or not it's a fixed lower intensity, I don;t know.


    I think some aftermarket radios just use the "DIM" signal rather than to actually control the intensity.

    For "giggles" look at post #3 here: http://ls1tech.com/forums/stereo-el...ation-dimmer-wire-car-radio.html#post17290513

    So, do you really have 0-100% control of the intensity of the radio illumination AND the new digital instruments?

    Or, maybe just DIM for the radio and 0-100% for the new digital instruments?
     
  12. Babybird002

    Thread Starter New Member

    Mar 30, 2016
    9
    0
    Crutschow Thanks. Please find attached THE PROBLEM as best as I can describe it. And I do not know the
    answer about separate power supplies because this is an automobile circuit and there are no other power sources. I will say this, when I started researching this problem I thought I would find an answer in just a few
    minutes on the internet. It just seems to me that being able to track the DC voltage of one circuit to another while maintaining isolation between the circuits would be something many would want to do. I really thought
    I was going to find an integrated circuit that you could hook up a variable DC input and have that variable DC input tracked by the DC output while maintaining total isolation of the two circuits.

    Again thanks for your interest and I hope the attached "THE PROBLEM" helps.
     
  13. Babybird002

    Thread Starter New Member

    Mar 30, 2016
    9
    0
    Thanks KeepItSimple The illumination control to my radio and gauges are true variable controls, not just dim.
    100% for the radio AND digital instruments. The output of circuit one is some kind of signal impressed voa a rotary knob across the circuit one bulbs that control the circuit one bulb intensity. I do not know exactly what it is. It is described in some other literature as a pulse generator and LESTRAVELED figured it out more precisely
    than I can by examining the control circuit attached to my post #1 which I am attaching to this reply as well. [IT IS THE OSCILATOR CIRCUIT SURROUNDED BY WHITE IN THE ATTACHEMENT]
    I will be looking at the link a little later today.
    Babybird
     
  14. KeepItSimpleStupid

    Well-Known Member

    Mar 4, 2014
    1,143
    202
    I think i would take a totally different approach.

    Let's ASSUME that the illumination signals are positive voltages. e.g. 12 V is 100%

    Correct?

    Anyway, PWM or pulse width modulation is where the duty cycle of the power to the LEDS/BULBS is varied.
    SO, 100% is fully on. If the signal spends 50% of the time off and 50% of the time ON, you get an effective 50% of the available voltage, The advantage of this technique is that there is little wasted power.

    I did a very quick search and came up with this part: http://www.irf.com/product-info/datasheets/data/ir25604.pdf

    which would basically be the main part: http://www.digikey.com/product-sear...s/pmic-gate-drivers/2556427?k=IR 25 6 04 SPBF

    Logic levels seem pretty much compatible, so the only real thing is just how much current does the illumination signal need (non-OEM instrumentation and Radio) so the FET can be selected.

    The chip contains both a high and a low side driver, so you would only need one MOSFET for the high side driver.

    the worst thing that could happen is that the 0 and 100% is reversed, but I don't think that's gonna happen.

    And use something like http://www.digikey.com/product-detail/en/rohm-semiconductor/RSR020N06TL/RSR020N06TLCT-ND/3769400 for the FET.

    Even though this is an IGBT driver, I think it should work.

    The IC just offers a bit of protection and squaring up the input signals.
     
  15. Lestraveled

    Well-Known Member

    May 19, 2014
    1,957
    1,215
    @Babybird002
    Do you agree with this drawing?
    [​IMG]

    If you do, I think I have a simple solution. I need to check a few things first.
     
  16. Lestraveled

    Well-Known Member

    May 19, 2014
    1,957
    1,215
    I think this will work.

    [​IMG]
     
  17. KeepItSimpleStupid

    Well-Known Member

    Mar 4, 2014
    1,143
    202
    Les:

    I think you need to label your pics so they are a bit more understandable?

    Post #15: The right hand side +12 at the top right, re-label to "Illumination (Gauge, Radio)

    Post #16: Add a spur between the Collector (Tip 30) and the Lamp the same "Illumination (Gauge, Radio)"

    At least, I think that's your intention.
     
  18. Lestraveled

    Well-Known Member

    May 19, 2014
    1,957
    1,215
    @KeepItSImpleStupid

    KISS, in post #15, I wanted to show that the car lights were referenced to +12V and the radio lights were referenced to ground. This showed the nature of the incompatibly between the two wiring systems. The two +12V symbols are connected together.
    I considered your suggestions, but I would not change anything. I guess my way of keeping it simple is different than yours.
     
  19. Babybird002

    Thread Starter New Member

    Mar 30, 2016
    9
    0
    Dear Lestraveled,


    THANK YOU. THANK YOU. THANK YOU.

    I will consider this thread closed until I build and test your circuit and will only reopen it if I run into problems.

    By the way I purchased a Miata dimmer swich [with connector and pigtail attached] and a used instrument cluster so that I can breadboard and test everything on my bench. I will send you a pic when everything is working.

    Oh And did I mention?

    THANK YOU!
     
  20. M_anjom

    New Member

    Jul 19, 2016
    2
    0
Loading...