Any way to make a 2N3904 Transistor work like a 2N3704? Need help within 4 hours!!Ah!

Discussion in 'General Electronics Chat' started by crewxp, Apr 21, 2008.

  1. crewxp

    Thread Starter New Member

    Apr 21, 2008
    5
    0
  2. Caveman

    Active Member

    Apr 15, 2008
    471
    0
    Just try 'em out.

    Whether they have any significant difference all really depends on what you are doing. In most situations that you encounter in school, it doesn't matter as long as your circuit doesn't get near the maximums.
     
  3. crewxp

    Thread Starter New Member

    Apr 21, 2008
    5
    0
    We're building a robot with 2 motors. When we use the 3704, it works fine. When we use the 3904, the motor goes really slow... like it's not enough voltage.

    http://www.crewxp.com/lab.doc is our lab manual for this.
     
  4. Caveman

    Active Member

    Apr 15, 2008
    471
    0
    It sounds like your transistor has less gain than the original one. How big is V+?
     
  5. crewxp

    Thread Starter New Member

    Apr 21, 2008
    5
    0
    the voltage coming out of the battery is 7.5V.
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The 3904 can't carry the current that the 3704's can. 3704 is rated for up to 500mA. 3904 is rated 200mA, but don't expect more than 100mA from it.

    If you look at the datasheet for the 3904, the HFE starts dropping off rapidly @ 50mA.
    Oh, you're looking at Fairchild datasheets. Go get the datasheets from National Semiconductor; they have dynamic hFE curves over temperature and collector currents. Oops, National Semiconductor took those datasheets off their site. Alldatasheet has one for the 3704 by "mE", it at least has gain curves in it.
    http://www.alldatasheet.com/view.jsp?Searchword=2N3704
    http://www.alldatasheet.com/view.jsp?Searchword=2N3904 <- grab the National Semiconductor sheet, 2nd to the last in the list.

    What's your Ic (collector current)? It's not going to work too well if you're trying to sink more than 100mA through it, as your hFE will go in the toilet.

    Do you have some 2N2222's around? hFE is lower, but more current capacity.

    What are you using for a base resistor?

    Do you have ANY other NPN's around?
     
  7. crewxp

    Thread Starter New Member

    Apr 21, 2008
    5
    0
    The lab manual posted shows how we designed it. We're using 2 different resistors, and a potentiometer.

    Here's a picture of it too http://www.crewxp.com/IMG_0111.JPG
     
  8. Caveman

    Active Member

    Apr 15, 2008
    471
    0
    Actually, just found it in the book. I would be worried about pulling more that about 4mA on the output of the comparator when low. So you can drop the resistor to 7.2V/4mA to 1.8kOhm. So drop the pullup resistor to about 3-5kOhms to see if it gets enough current.

    If that doesn't work, you can use two transistors in what is called a darlington configuration. Basically you connect the emitter of the first to the base of the second and connect the collectors together. Now your base is the base of the first, your emitter is the emitter of the second, and your collector is the shared collector. This will basically square your gain.
     
  9. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Just checked the specs on the LM393; it'll sink up to 6mA with the output rising to 1.5v.

    The lab says 10k for the pullup resistor in the text, but shows a 1K pullup resistor in the schematic instead. Caveman's estimate of 1.8k sounds good to me. You might be able to go as low as 1.5k, but that might keep the 2N3904 biased on a bit.
     
  10. crewxp

    Thread Starter New Member

    Apr 21, 2008
    5
    0
    Wow, thanks a lot guys!

    Quick question though, when you say drop the resistor to 7.2V, you mean change the potentiometer to 1.8kOhm?

    Could you explain that a little better, haha... to someone who's just starting out?

    Thanks bud!

    ps: The way we have it setup right now, the potentiometer with the 2N3704 is at 3.437 M Ohms. It works so that the circuit kicks on after like 3 seconds. when it's changed to 3.237 M Ohms, the circuit kicks on after 2 seconds. That's how it's supposed to work.

    So, The circuit we're building with the 2N3904's will work!? But we just need to change the potentiometer to 1.8kOhm?

    Thx, sry if this is confusing, kinda don't understand myself.
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Wait - look at the output of the LM393 - it's on pin 1.

    It's connected to both the base of the 2N3704, and in the text, a 10K pullup resistor (that's shown on the schematic as a 1K resistor)

    10k is too high of a resistance. It needs to be lowered. Use a value between 1.5k and 2.2k. 1.8k should be just about right. If you don't have a single 1.8k resistor, you can use two 910 Ohm resistors in series, or two 3.6k resistors in parallel.

    If you don't have any of those value resistors, use this page:
    http://www.qsl.net/in3otd/parallr.html
    For resistance, put in
    1770
    Then change E12 to E24. It will give you a bunch of series/parallel combinations you can use to get 1770 Ohms. (You could put in 1800, but since that is a standard E12/E24 value, that's the only value you'll get back)
     
  12. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The 2M3904 transistor is too weak to drive the motor.
    A base resistor with only 4mA in it (1.8k) doesn't provide any transistor with enough base current to saturate properly when its load is hundreds of mA.
    The LM393 doesn't have enough output current to drive a pullup resistor with a value lower than 1.8k.
    It is a poorly designed circuit.
     
  13. Caveman

    Active Member

    Apr 15, 2008
    471
    0
    crewxp,

    How'd it go? Did you get something useful out of it?
     
  14. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I'd be interested in hearing the results myself.

    Agreed that 4mA base current has no chance of putting the transistor into saturation.
    It certainly is better than the 0.75mA base current it was receiving before - but the 2N3904 just doesn't have near the collector current capacity that the 2N3704 does.

    Too bad they didn't pick up a TIP120 instead.
     
  15. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
  16. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
  17. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    The few measurements I've made tend to confirm their claims.
     
  18. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    If measurements of a transistor are different from the ordinary datasheet then the transistor is not that part number but is another one. You won't know which other spec's for it are different from what is expected unless you measure them and make your own datasheet.
     
  19. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    What exactly are you trying to say?

    I bought a volume batch of transistors, labelled 2N3904/6 from a far eastern manufacturer, whose spec sheet I published in my link, claims higher voltage and current handling capability than my tables for these transistors.
    My tables go back to 1970 and I keep several subsequent revision issues because specs change over the years.
    Later versions have better specs for the 2N390X series and many others, but the far eastern version has better yet.

    My 'few measurements' means that I sacrificed a handful of my batch in destructive testing -yeah I deliberatly blew a few up.
    'Tends to confirm' means that I did not kill enough to be statistically significant.

    Incidentally I have also had a batch of '2N3704' (=labelled) transistors from India also with enhanced specs.

    I do actually posess a Phillips PM6505 transistor analyser, but it is under repair at the moment so I can't currently do the 4 terminal dynamic measurements.

    I did comment in another thread that modern far eastern sources seem to be supplying enhanced spec semiconductors at quite low prices.
     
  20. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    If the spec's are different than all the other manufacturers then it is a different transistor. If it is new then it should have its own new part number. I suspect it is just another transistor (maybe a Japanese one) with an American number printed on it.

    I didn't see a datasheet. Just a few undetailed claims.

    The lousy old 741 and LM324 opamps are still made with the same lousy performance.
    A Japanese manufacturer "enhanced" their copy of Texas Instruments' TL071 opamp, it oscillated all the time and they were all recalled.
     
Loading...