Announcement

Collapse
No announcement yet.

What is the purpose of the series diode?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Qiaozhi
    replied
    As Carl said, the theory seems to make sense, which is why I decided to investigate further.

    From simulation it is clear that the MOSFET output capacitance appears to get disconnected from the coil by the diode. You can see that by examining the shape of the coil current after TX-off. There is a small hump in the decay curve where the coil's parasitic capacitance initially becomes charged and then subsequently discharges. This hump is reduced in amplitude when the diode is present. Also, the amount of reduction is related to the ratio between the MOSFET's output capacitance and the coil's parasitic capacitance.

    But ... in practice (on a real circuit) the opposite appears to be true.

    I currently need to rebuild my test circuit, so it might be some time before I can do some in-depth testing.
    Please keep up this interesting discussion.

    Leave a comment:


  • Teleno
    replied
    Originally posted by Carl-NC View Post
    Back when I was doing PI design at PriorJob I played around with the series diode trick. Like Qiaozhi I found that the diode made settling slightly worse. The "theory" that it should help made sense, and I can run simulations that show an improvement, so I chalked it up to the rest of the circuitry not being fast enough to show an improvement. I never got back to look at it again. In any case, I've yet to build a working circuit in which the diode improves settling.
    300 ns advantage (what you get from the diode trick) doesn't make a difference if the amplifier cuts off at 200kHz.

    Leave a comment:


  • bbsailor
    replied
    Everyone thinking about this,

    I seem to remember Eric Foster commenting about the series diode trapping a high voltage between the MOSFET and the diode that could be bled off by using a resistor to ground whose value is based on the time constant of the MOSFET equivalent capacitance. Just observe the high voltage at the MOSFET-diode junction and reduce the resistor value at this junction until the voltage falls close to zero V. Then, based on that resistor value you can infer or calculate the MOSFET effective capacitance value. This test might even help evaluate various MOSFET candidates for the least capacitance.

    This will add another piece of data to include in the discussion about the diode's value in affecting early sampling.

    Thanks

    Joseph J. Rogowski

    Leave a comment:


  • Carl-NC
    replied
    Back when I was doing PI design at PriorJob I played around with the series diode trick. Like Qiaozhi I found that the diode made settling slightly worse. The "theory" that it should help made sense, and I can run simulations that show an improvement, so I chalked it up to the rest of the circuitry not being fast enough to show an improvement. I never got back to look at it again. In any case, I've yet to build a working circuit in which the diode improves settling.

    Leave a comment:


  • Orbit
    replied
    Green simulate the bvy 28 to see what it turns out ?

    Leave a comment:


  • green
    replied
    What Teleno said. Scope pictures of coil turn off with no damping resistor. Resonance a lot higher with diode in circuit. Damping resistance=pi*L*resonant frequency, the higher the resistance the faster the decay.
    Attached Files

    Leave a comment:


  • KingJL
    replied
    Originally posted by Teleno View Post

    What's not moedeled is the breakdown voltage of the body diode. I place a corresponding diode in parallel for simulation.
    That is universally not modeled in LtSpice. What is missing in some models is cdgmax and cdgmin. I found this out when comparing pi tx circuits where I was using a MOSFET that had 1/3 the coff of a IRF740, but could not see any difference in performance. When I found a subcircuit definition with all parameters modeled, I could see the difference... and the subcurcuit definition also modeled the diide.

    Leave a comment:


  • Teleno
    replied
    Originally posted by KingJL View Post
    Qiaozhi,
    I don't know what MOSFET your are using in your simulations, but some of the commonly used models (e.g. IRF 740, etc.) do not have the dynamic capacitance modeled. I found this out the hard way!

    Kind regards,
    J. L. King
    For this analysis an equivalent static capacitance (Cjo) would do. This is modeled in LTSpice.

    What's not moedeled is the breakdown voltage of the body diode. I place a corresponding diode in parallel for simulation.

    Leave a comment:


  • KingJL
    replied
    Qiaozhi,
    I don't know what MOSFET your are using in your simulations, but some of the commonly used models (e.g. IRF 740, etc.) do not have the dynamic capacitance modeled. I found this out the hard way!

    Kind regards,
    J. L. King

    Leave a comment:


  • Qiaozhi
    replied
    Originally posted by Teleno View Post
    Actually it's because the diode blocks the MOSFET capacitance during the fall part of the transient, reducing the average capacitance seen by the coil.
    Good answer.

    However, at the moment that's not what I'm seeing in practice. Perhaps there's a difference in the actual circuit I'm using for testing. I need to investigate this further.

    More later ...

    Leave a comment:


  • Teleno
    replied
    Originally posted by Qiaozhi View Post
    It is true that (with the diode inserted) you can use a higher value damping resistor. But that's because the diode has increased the decay time, and not reduced it. Therefore you have to compensate by increasing the damping resistor value.
    Actually it's because the diode blocks the MOSFET capacitance during the fall part of the transient, reducing the average capacitance seen by the coil.




    Attached Files

    Leave a comment:


  • Teleno
    replied
    Originally posted by Qiaozhi View Post
    The original Minipulse design did not have a diode in series with the MOSFET and the coil. So what is the purpose of the series diode in the Minipulse Plus?

    When the REV-D version was created, there were a number of requests to provide space for this diode. Supposedly, this allows earlier sampling. Up until now I have not questioned the reasoning behind this request, as others have reassured me that it really does make a difference. But recently I decided to test these claims further:

    I cannot remember exactly what the theory behind the diode was supposed to be, but I vaguely remember that it was either:
    1. The diode capacitance is in series with the output capacitance of the MOSFET, and hence the overall capacitance is reduced, allowing earlier sampling.
    2. The breakdown voltage of the MUR460 is 600V, which is above the IRF740's VDS of 400V, thus allowing the coil current to decay more quickly.

    From increasing the TX-on time such that the flyback voltage was above 400V, showed clearly that option 2 is incorrect, as breakdown still occurred at 400V. In fact, this should have been obvious from studying the circuit diagram. At TX-off, the coil current continues to flow in the same direction whereas the voltage across the coil changes polarity and increases dramatically in amplitude. Hence, although the voltage across the coil is inverted, the polarity of the voltage across the MOSFET remains the same. This of course means that the diode is always forward biased.

    So what about option 1?
    At first I suspected that the diode may be acting as a non-linear series resistance, which would effectively reduce the maximum coil current, allowing it to sample earlier, rather than the capacitance theory being the cause. However, I was somewhat surprised to find that in fact removal of the diode allowed the main sample to be adjusted to a lower setting. Which is completely opposite to what was being claimed.

    Basically, my conclusion is that the series diode does absolutely nothing to improve early sampling, and actually does exactly the opposite.

    My assumption here is that the person who came up with this theory had a PI circuit that was just reaching the breakdown voltage of the MOSFET at TX-off. When the diode was inserted, the maximum coil current was restricted, and the breakdown voltage was no longer as high as before. The result being that the design could now sample a little earlier than previously. The wrong conclusion was subsequently drawn from the results

    Discuss ...
    The voltage transient has two parts: a rising part before the HV peak (Vp) and a falling part where voltage drops back to zero after the peak.

    During the rising part the output capacitance of the MOSFET (Co) gets fully charged through the series diode. Once Vp is reached the voltage starts to drop, but Co remains charged to Vp because the series diode is reverse biased (blocked). In these conditions, the MOSFET capacitance gets isolated from the coil during the falling half of the transient, which speeds it up a little bit.

    Leave a comment:


  • Qiaozhi
    replied
    Originally posted by green View Post
    A chart from another thread awhile back. The diode reduces the capacitance.
    That statement does not appear to be correct, as the diode is never reverse-biased. At least, I can find no evidence that supports the theory.

    Originally posted by green View Post
    Higher damping resistance to critical damp allows earlier sampling.
    It is true that (with the diode inserted) you can use a higher value damping resistor. But that's because the diode has increased the decay time, and not reduced it. Therefore you have to compensate by increasing the damping resistor value.

    Originally posted by green View Post
    The peak current in the example was higher without the diode because of the diode drop when charging the coil. With equal peak current there would be a greater signal difference with the diode. Maybe not a big gain in signal strength but without the diode I couldn't sample sooner than 10usec.
    I have tried decreasing the maximum coil current with the diode inserted, in order to match the test without the diode. Best results were obtained without the diode. Both SPICE and real measurements agree with each other.

    Originally posted by green View Post
    The voltage avalanched with and without the diode. Have you tried to critical damp the coil in spice with and without the diode?
    In fact this observation tends to confirm what I said in my initial post. All the diode is doing is to reduce the time the MOSFET is in avalanche mode. Hence the [apparent] result is that it allows the current to decay more quickly.

    The bottom line is that you should never let the MOSFET avalanche. This has nothing to do with component reliability, but simply that it allows the coil current to decay as quickly as possible in combination with critical damping. If your MOSFET never enters avalanche mode, then the diode has the opposite effect. In fact, you would be better to just reduce the maximum coil current, either by decreasing the TX-on time and/or using a suitable series resistor.

    Leave a comment:


  • green
    replied
    Originally posted by homefire View Post
    If I Build one I'm going to use a Wireode LOL Jumper. Save a dollar on something that does seemingly Nothing but cause .7volt drop across the diode.
    Chart I posted awhile back. Sample delay time 6, 10, and 14usec. Sample time was adjusted to cancel ground signal so the sample time was longer for the 10 and 14usec delay times(GEB sample time 100usec). The diode was used for all of the tests. Just shows the effect delay time has on detection distance for different targets. The 4 grain gold nugget showing the most effect. Targets were swung from a pendulum and detection distance was recorded with GEB on and off.
    Attached Files

    Leave a comment:


  • homefire
    replied
    If I Build one I'm going to use a Wireode LOL Jumper. Save a dollar on something that does seemingly Nothing but cause .7volt drop across the diode.

    Leave a comment:

Working...
X