Re: [Hive 76 Discussion] Fixing the green strobe on the accent quizzor

[Sean McBeth]

Wow, thanks Don. This looks very helpful.

On Sep 6, 2012 1:47 PM, "DonTrackbiker" <d...@donklipstein.com> wrote:

To the regulars working on this project,

 

  I want to add (even if repeating some already said):

 

  The green strobe was not working well, apparantly because a transistor was

not passing sufficient current.

 

  As was discussed in the space, 39K-ohm "base resistors" were paralleled

with 1K-ohm resistors so that the associated "2N3904" transistors can pass

the ~150-200 mA of current needed to well-power the strobes.

 

  The red strobe got working, and the green one was not as of a few minutes

before midnight.

 

  I would like to propose a troubleshooting proceedure based on my

esperience with similar electronics:

 

[All voltages to measure here are with respect to known good ground unless

mentioned as otherwise]

 

1)  Measure both ends of the 1K-ohm resistor that was added to get the

  green strrobe working.  The "Arduino end" should be at 5 volts or not

  much less, and the "transistor base end" should be at around .75-.8 volts.

 

    At a point Wednesday evening, I heard a report that both ends of the 1K

  were at close to 5 volts.  This suggests to me that the "transistor base

  end" of the 1K resistor was failing to conduct current through the base of

  the associated transistor.

 

    There are many reasons why I think this could occur, by my favorite is

  that there is a break between the "transistor base end" of the 1K resistor

  and the base lead of the transistor.

    This leads to follow-up testing as such, if I was doing that:

 

2)  Measure the voltage at the physical base lead of the transistor - on

  the lead, as opposed to at a solder joint.

 

    That voltage should be ~.75-.8 volts if everything is working right.

 

    If it's around ~.6-.7 volt, that probably means the 39K base resistor is

  feeding current to the transistor's base but the 1K one is not.

 

    It it's much less than .6 volt, then probably the transistor's base is not

  being fed current.  Verify base resistor lead voltages to check for breaks

  upstream of the transistor's base lead, a short from base to ground, or

  lack of "high signal".  (Base voltage less than .6 volt will be

  accompanied by "transistor end" of base resistor having similarly low

  voltage and "arduino end" at ~4-5 volts if a short from the base

  connection to ground is the only problem.)

    If the transistor's base lead itself is at more like 4-5 volts, then probably

  either the transistor is bad or its emitter is not contacting ground.  At this

  point, then:

 

3)  Measure the voltage on the transistor's physical wire lead.  If that is

  low and the base lead's voltage is around 4-5 volts, then almost certainly

  something is wrong with the transistor.  If the transistor's physical

  emitter lead is is at around or over 4-5 volts, then the problem is

  probably at-least-in-part failure of the transistor's emitter lead to contact

  ground.

 

==================

 

4)  If all above refuses to compute, make sure the add-on paralleled base

  resistor is a 1K-ohm one and not failed-open.  Use an ohmmeter if things

  come to such a desperate situation - and have one lead of the resistor

  connected to nothing but the ohmmeter.

    Keep in mind that brown-black-red is usually not 1.0K when tolerance

  is tighter than 5%.    For tighter tolerances, 1.0K is usually

  brown-black-black-brown.

 

    Also, check for strobes to not stress 3904-type transistors.  Strobes

  typically conduct DC unsteadily on both 10's-of-microseconds and

  100's-of-milliseconds time scales.

    I give slight chance that transistors overheating and failing can be fixed

  by using even-lower 470 ohms instead of 1K - as long as Arduino outputs

  *comfortably* source ~10 milliamps.  Otherwise, I have some favorite

  "power MOSFETs" that can greatly improve over 3904s.  4.5 volts can

  switch an amp or two within a microsecond through a 1K resistor, with

  transistor voltage drop less than .4 volt.

 

=================================

 

  If the transistor is barely conducting sufficient average current:  It may

help to put significant capacitance across the strobe.  Use the maximum

that won't blow the transistor from current surge during charging - likely a

few to around a hundred microfarads.

 

 - Don Klipstein (d...@donklipstein.com)

--
To post to this group, send email to hive76-d...@googlegroups.com
To unsubscribe send email to hive76-discuss...@googlegroups.com
For more awesome goto http://groups.google.com/group/hive76-discussion?hl=en

[Daniel Toliaferro]

Don, you rule.

[Brendan Schrader]

Thanks Don. Once the transistor was replaced all was well.  I appreciate your troubleshooting the 39k vs 1k resistor for the strobes.

Brendan