Sunday, June 2, 2013

FH020AA FET SHUNT vs D.I.Y (SCR Shunt & Series type)

Kebetulan saya sudah mencoba 2 macam "kiprok" D.I.Y, 
satu SCR Shunt Based (pernah saya posting sebelumnya) dan satu lagi Series type





Motor butut yang digunakan D-Tracker 150, sedang peralatan seadanya yang digunakan :
- Shindengen FH020AA FET Shunt, D.I.Y (SCR Shunt & Series)
- Hantek DSO8060 Oscilloscope
- Fluke 179 Multimeter
- Maynuo M9712 (300W Programmable DC Load)

Metode pengujian, beban ke lampu utama, rem dkk dinonaktifkan dan mesin diset ke kisaran 3500rpm 
gak berani lebih tinggi karena knalpotnya terlalu berisik :D

DC Load dipasang langsung ke aki dan beban divariasikan mulai dari 1A, 2A, hingga 6A untuk mengetahui seberapa stabil regulasi tegangan output yang terjadi
Multimeter dipasang pada kutub aki untuk memantau tegangan
dan Oscilloscope dipasang pada Output Spul sebelum masuk ke "kiprok" untuk melihat grafiknya

Dibawah ini hasilnya

1. Shindengen FH020AA FET Shunt

Tegangan output yang dihasilkan (pada kutub aki)
No Load = 14,37 VDC
1A = 14,25 VDC
2A = 14,12 VDC
3A = 14,00 VDC
4A = 13,83 VDC
5A = 13,71 VDC
6A = 13,62 VDC

2. D.I.Y SCR Shunt

Tegangan output yang dihasilkan (pada kutub aki)
No Load = 14,67 VDC
1A = 14,66 VDC
2A = 14,63 VDC
3A = 14,61 VDC
4A = 14,61 VDC
5A = 14,63 VDC
6A = 13,88 VDC

3. D.I.Y Series type

Tegangan output yang dihasilkan (pada kutub aki)
No Load = 14,23 VDC
1A = 14,22 VDC
2A = 14,23 VDC
3A = 14,22 VDC
4A = 14,22 VDC
5A = 14,22 VDC
6A = 13,85 VDC

Terlihat pada ketiga grafik diatas, pada 6A sudah tidak terjadi "regulasi" yang mengindikasikan kalau outputnya "mentok" (harap dimaklumi karena motor butut yang digunakan hanya mampu segitu saja) :)

Untuk hasil regulasi output vs variasi beban ==> dari ketiga hasil diatas, Anda bisa menyimpulkan sendiri :)

23 comments:

  1. nice work om suhu *jempol*
    kalo kendala berisik knalpot mgkn bisa pake motor/dinamo bekas tuk muter alternator ... dibikin variable speed, mgkn ...

    ReplyDelete
    Replies
    1. Kebetulan baru nyoba knalpot "racing"
      Ini sudah rencana dilepas lagi buat ngetes di 5000an rpm supaya lebih adem di telinga :)

      Untuk saran menggunakan dinamo, saya terlalu malas Suhu :)

      Delete
  2. Hi TJERET,

    You did a wonderful job on that difficult subject.

    Do you by any chance speak (read/write) english. I am from France and own a triumph. I am in the electronic field. We often have some "talk" about regulators/rectifiers on our forum (www.triumphall.com). On www.triumphrat.com as well.

    I have a few questions regarding the graphics. Would you be kind enough to take some time to answer them.

    2-I understand that the graphics are corresponding to the voltage at the stator output. Am I right?

    2-the oscilloscope time base shows 5ms per square (that is 60ms total). That would mean your engine is rotating at more than 8000rpm which is not consistant w/ the 3500 rpm indicated. Can you explain what is wrong w/ my calculation?

    3-I thought that the shindingen in addition to not getting hot because of the low FET voltage drop would show some heat saving for the stator. But the graphics tend to show that sometime the stator is short circuited right at the beginning of the wave (sometimes and especially for low amperage there is no any positive voltage or no any negative voltage). I would assume that whenever it's not necessary the FET wouldn't be triggered so that we would see the voltage going to 60Volts. Can you explain what is wrong w/ my understading?

    Thanks in advance for your replies.

    Fred

    ReplyDelete
    Replies
    1. Hi Fred, sorry for my poor English but i'll try my best to answer your question

      1. Correct, that Voltage at alternator output (i put my scope probe before R/R input)

      2. To be Honest, I don't know how to calculate :)
      i'm just guessing the engine rpm, based at idle engine about 1800 rpm (it showing about 105Hz with scope automeasuring tool) with highest load
      i'm taking those graphic about 220-250Hz fluctuating

      3. AFAIK, This FH020AA was Shunt type. So if Voltage had cross set point it will be triggering to short alternator not open circuit alike your describe.
      About heat, with small/low RDS On Mosfet (my guess about 20mOhm or so) it only draw less than 1Watt for shunting

      p.s
      My bike was single phase with 8 pole alternator

      Delete
  3. Thanks Tjeret for replying,

    2-Regarding the rpm I guess my mistake come the fact you have 8 poles so the frequency should be divided by 8 to obtain the rpm (and by 60).

    3-1W is the power dissipated in the FET, but in the mean time the power dissipated in the alternator (stator) is huge. Even higher than w/ a SCR RR. W/ the FET RR sometime the wave is not there (especially at low loads). That would tend to show that the FET stays triggered.

    An additional point regarding your DIY serie RR. It seems to me that it's a SRC one. The SCR seems to get triggered w/ a variable delay in relation w/ the output voltage. Is the SCR computer driven?

    Fred

    ReplyDelete
    Replies
    1. Agree with you, this FET Shunt will put a lot off heat to alternator at low load

      Yes it's SCR Based
      No its simple circuit just alike SCR Shunt
      (you can search with "ducati/rotax regulator rectifier" keyword)

      Regards
      La Ode

      Delete
  4. Thanks. Very interesting. I'll need some time to digest the schematic.

    So you did one?

    Fred

    ReplyDelete
    Replies
    1. Yes, basically its still the same topology with original
      i just remove some parts and make a little modification to meet my need

      Regards
      La Ode

      Delete
  5. I understand that G lines are connected to the stator and B+ is the battery +. Ground is battery -.
    What are C and L? Is C the output voltage measurement input (so connected to battery + as well)?

    Fred

    ReplyDelete
    Replies
    1. Correct, C connected to Battery+ as voltage sensor but should connect via ignition key
      L for Lamp

      Regards
      La Ode

      Delete
  6. I started to strip down the schematic. It should be possible to connect C to the battery providing we detect a threshold of say 13V. Some more components for sure

    What is the lamp for?

    Fred

    ReplyDelete
    Replies
    1. It's should be fine, but practically when engine off it waste a "lot" off current while do nothing

      I'm guessing for main headlamp
      One connect to Battery+ and the other one connect to this L pinout

      Regards
      La Ode

      Delete
  7. I have a few remarks on the schematic. May be you can explain.

    1-What's the point of having R3 and R5 in parallel?
    2-How D7 can inject some current into the A56 PNP transistor?

    Probably a few more after that.

    Thanks.

    Fred

    ReplyDelete
    Replies
    1. 1. Just to get a proper value of set point voltage
      2. http://pubpages.unh.edu/~aperkins/pdf/Misc-devices/unijunction.pdf
      Same as circuit 1A, except using PNP with put load at bottom side and put zener backward of course
      But you can replace them and put "shunt type" as regular use (as i did)

      Please, as long as i know i'll try to answer it :)

      Regards
      La Ode

      Delete
  8. 2-so you mean D7 is a zener? So why not putting a zener symbole and calling it Z7.

    What do you mean by :"But you can replace them and put "shunt type" as regular use (as i did)"?

    BTW I appreciate talking (writing) w/ you. Thanks for that.

    Fred

    ReplyDelete
  9. Yes it should be Zener

    This is what i mean, replace R3,R4,R5,C2 and D7
    http://en.wikipedia.org/wiki/Zener_diode (see at "uses")
    use this circuit to MPSA56 BASE

    You're welcome

    Regards
    La Ode

    ReplyDelete
  10. OK then, it's a zener.

    I think the transistor A56 should be kept. By amplifying the zener current it gets the transition sharper and uses a shorter portion of the zener curve. I used a similar system to trigger a relay a while ago. It was also voltage monitoring. But not a RR.

    ReplyDelete
  11. So as I understand it it triggers the thyristors for an entire half wave whenever the voltage drops below say 14V.

    If that's the case it's a little brutal given that on most of the bikes the stator voltage can raise to up to 200V!

    I would have though the the thyristor would have been triggered w/ a variable delay in regard to the difference between the wanted voltage and the measured voltage (closed loop).

    Fred

    ReplyDelete
  12. I must be wrong. It's not in accordance w/ your oscilloscope results.

    Fred

    ReplyDelete
  13. "So as I understand it it triggers the thyristors for an entire half wave whenever the voltage drops below say 14V.

    If that's the case it's a little brutal given that on most of the bikes the stator voltage can raise to up to 200V!"

    Your analysis is absolutely right, 600V Thyristor is common use
    On "big bike" maybe can produce this high voltage
    but remember my poor bike only 143cc with small alternator :)

    About the other analysis
    You think to deep about it, maybe you have to try it (it's simple circuit anyway)

    Regards
    La Ode

    ReplyDelete
  14. Yeah I could. I am laking time those days. But my asset is that I have a buddy who own an alternosta (variable transformer).

    Fred

    ReplyDelete
  15. http://www.aero-hesbaye.be/circuit_elec.htm

    Basically this guy says the command as shown on the schematic you gave (and tried out) gives a very irregular load on the stator. So it stresses the battery and the engine because it gives current bursts. Instead his advise is to manage a variable delay to trigger the SCR in regard to the voltage. The higher the voltage the bigger the delay until the point (say 14.5V) the SCR are not triggered. Still to determine a simple schematic to achieve that and that works out w/ a variable frequency as on our bikes.

    Fred

    ReplyDelete
    Replies
    1. Great link, thanks

      Like your previous comment, possibility this one will need kind a "sensor" to AC sinosuidal to make each Thyristor work independent (close loop)

      To complicated for my brain :)

      Delete