Prop Plug NPN transistor value

TCTC Posts: 1,019
edited 2014-04-04 - 21:05:24 in General Discussion
Hello All

I have no idea on how to calculate what would be the best transistor to use for the Prop Plug. And I don't know where to start to learn. The datasheet does not give a value for the transistor. And all of parallax's products that use a "FT232RQ" USB to UART chip does not give a value for the transistor.

So I was wondering, Does anyone know what the value of the transistor should be?

And also, how can I learn to be able to design a transistor circuit? Like what current rating would I need, what supporting components would I need, etc...

Thanks
TC

Comments

  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2014-04-03 - 07:07:39
    Hi TC;
    TC wrote: »
    I have no idea on how to calculate what would be the best transistor to use for the Prop Plug. And I don't know where to start to learn. The datasheet does not give a value for the transistor. And all of parallax's products that use a "FT232RQ" USB to UART chip does not give a value for the transistor.

    So I was wondering, Does anyone know what the value of the transistor should be?
    That is because it is a very uncritical. In general, almost any small switching transistor will work in these applications.
    Of course NPN vs PNP must be observed.

    The venerable NPN 2N3904 and PNP 2N3906 have been standards for many years.
    There are numerous similar parts which all should work, some surface mount some in TO-92 packages.

    In general they should have:
    1. A gain of at least 30 at low collector current.
    2. A voltage rating of at least 20V.

    I call these "Penny Parts" as they are very low in cost and at least 100 of each should be in the junk box.
    (Along with 2N7000 small MOSFETs.)

    Duane J
  • TCTC Posts: 1,019
    edited 2014-04-03 - 07:56:02
    Wonderful, thank you.

    Do you know how I can learn how to use transistors?
  • xanaduxanadu Posts: 3,302
    edited 2014-04-03 - 08:12:03
    I keep this document handy, right alongside the datasheet for the specific transistor I'm working with.

    (PDF) -- http://www.pitt.edu/~qiw4/Academic/ME2082/Transistor%20Basics.pdf

    Parallax also has some great tutorials, you can find them in the Stamp books and Propeller books. Examples like this go a long way http://learn.parallax.com/propeller-c-simple-circuits/sense-light

    Build some small circuits on the side to get started. If you can find a cheap bag of through hole NPN and PNP assorted transistors that would be good too.
  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2014-04-03 - 09:06:16
    TC wrote: »
    Wonderful, thank you.

    Do you know how I can learn how to use transistors?

    Here is one Williamson Labs.
    But there are many others.
    Lessons in Electronic Circuits
    BIPOLAR JUNCTION TRANSISTORS

    Duane J
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-04-03 - 09:18:23
    The world of transistors, resistors, and capacitors seems very complicated at first; but once you get involved in digital projects you will find that there are really only about 6 of each (transistors, resistors, and capacitors) that you need for the majority of projects.

    Transistors are still being used for these simple low power solutions, but anything with a lot of power has gone over to MOSfets and they run cooler and waste less power.
  • Tracy AllenTracy Allen Posts: 6,443
    edited 2014-04-03 - 12:51:59
    In my estimation, the Prop Plug transistor circuit is an example of a problematic design. Here is the schematic:
    propPlug.png


    The problem is that the 10nF capacitor is connected directly to the base of the NPN transistor, and it drives the transistor much too hard. That happens when DTR goes from low to high to reset the Propeller. This would not matter except for the fact that the sudden load on the 3.3V regulator inside the FT232 causes it to go unstable, with the result that the RST\ pulse is a nasty oscillation rather than a nice clean pulse. There was a discussion about this a while back on the forum. It may not have any definite bad effect to have an oscillation on the Prop RST\ line, questionable, but why allow it?

    There should really be a resistor in series between the capacitor and the base. This is the circuit I use.
    PropPlugEME.png


    The FT232RL is basically the same as the FT232RQ, except for the packaging. The transistor Q1 is a "digital" transistor, also called "pre-biased" that integrates two 10kΩ resistors. (e.g. Digikey BCR133IN). I use a 1nF capacitor, which gives a RST\ pulse plenty long. Most importantly, the R and C in series between DTR and the transistor base limit the surge current so that the FT232 is not whammy punched and the RST\ pulse is nice and clean.
    627 x 386 - 55K
    429 x 238 - 23K
  • TCTC Posts: 1,019
    edited 2014-04-03 - 17:08:58
    Here is one Williamson Labs.

    OK, I really like that site. So much so, I book marked it.
    In my estimation, the Prop Plug transistor circuit is an example of a problematic design. Here is the schematic:
    propPlug.png


    The problem is that the 10nF capacitor is connected directly to the base of the NPN transistor, and it drives the transistor much too hard. That happens when DTR goes from low to high to reset the Propeller. This would not matter except for the fact that the sudden load on the 3.3V regulator inside the FT232 causes it to go unstable, with the result that the RST\ pulse is a nasty oscillation rather than a nice clean pulse. There was a discussion about this a while back on the forum. It may not have any definite bad effect to have an oscillation on the Prop RST\ line, questionable, but why allow it?

    There should really be a resistor in series between the capacitor and the base. This is the circuit I use.
    PropPlugEME.png


    The FT232RL is basically the same as the FT232RQ, except for the packaging. The transistor Q1 is a "digital" transistor, also called "pre-biased" that integrates two 10kΩ resistors. (e.g. Digikey BCR133IN). I use a 1nF capacitor, which gives a RST\ pulse plenty long. Most importantly, the R and C in series between DTR and the transistor base limit the surge current so that the FT232 is not whammy punched and the RST\ pulse is nice and clean.

    Thank you for the circuit, I have changed my schematic to match yours.

    I do have a question though, Why is it companies (not just parallax) provide a schematic that could potently be wrong?
  • Tracy AllenTracy Allen Posts: 6,443
    edited 2014-04-04 - 07:58:51
    I do have a question though, Why is it companies (not just parallax) provide a schematic that could potently be wrong?
    Not intentionally, usually anyway. It's more like your tag line, ""CONFIDENCE"
    The Feeling you get before you truly understand the situation.".

    The standard Prop Plug circuit is not wrong--demonstrably--it works. It is merely problematic. I happened to look at the reset signal carefully with an oscilloscope, saw anomalies. First it was one of own prop board designs using that circuit with the FT232RL, which was often having trouble accepting new programs, sometimes okay, sometimes "no prop found". To compare, I looked at the RST\ signal on the Prop plug and on the Quickstart board, which have basically the same circuit, and saw similar anomalies, but not as bad. Rarely any problems programming those. Traced down the cause, asked Chip questions about the internal reset circuit, and came up with a slight modification of the circuit (as above) that vastly improved the look of the signal. It also improves the tolerance for variation of parts, such as the specific transistor used. It does not involve a change of existing circuit boards, because pre-biased transistors drop into exactly the same footprint as a standard transistor. I do wish that Parallax would change the published circuit, but it is hard to give much priority to something that works 99.99% of the time.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-04-04 - 08:26:00
    In electronics, there is always room for improvement.
    You press something into production to get to market to gain revenue to survive. You rely on your best knowledge with the belief that you got it all right on the first try.

    Some times you win, some times you don't
    Engineering is an art of estimating how devices and structures yet to be built will perform.
    Once complete, your knowledge pays off, or it is 'back to the drawing board' for further learning.

    The main thing is that as a business Parallax stands behind its products to assure you are a satisfied customer. So all that drama is behind the scenes.

    Trying to keep customers informed so they can copy the changes is hard to keep up with. Only so much time in any day. But you can come to this Forum and ask.
  • Tracy AllenTracy Allen Posts: 6,443
    edited 2014-04-04 - 09:47:42
    Here is the thread where I first brought this up, Fuzzy-tail-on-RESET-pulse-from-FT232R. There are 'scope shots there, and highly informed commentary from forumistas. A good lesson on transistors circuits!

    The Parallax circuit exhibits lots of variation from board to board and from reset to the next. Variation of gain among transistors and variations in stability from one FT232 to the next could account for a lot of that. Situations that involve instability often have a chaotic characteristic.

    Yesterday I took a couple more 'scope shots to see if anything has changed.
    Red trace is the signal taken at RST\. The blue trace is zoomed in time and amplitude. Note changes in timebase.
    1) my current circuit, 10k-10k prebiased
    2) a QuickStart-A board I picked up at Radio Shack a couple of weeks ago. Fairly consistent output trace.
    3) a QuickStart-A board I received soon after they originally came out. This is the worst case trace. I had to reset many times to find one this bad. They were usually more like the one in (4). Despite the ugly trace, it never said "prop not found" during this session.
    4) a Prop Plug from 2013 purchase. This is the worst case after many tries.
    10k10k.png
    QS-good.png
    QS-worst.png
    PropPlugWorst.png
    640 x 480 - 9K
    640 x 480 - 8K
    640 x 480 - 8K
    640 x 480 - 7K
  • TCTC Posts: 1,019
    edited 2014-04-04 - 14:44:31
    But you can come to this Forum and ask.

    I agree with you there. For years, this forum has been a great help to me. I have learned a lot from everyone on here.
    Here is the thread where I first brought this up, Fuzzy-tail-on-RESET-pulse-from-FT232R. There are 'scope shots there, and highly informed commentary from forumistas. A good lesson on transistors circuits!

    The Parallax circuit exhibits lots of variation from board to board and from reset to the next. Variation of gain among transistors and variations in stability from one FT232 to the next could account for a lot of that. Situations that involve instability often have a chaotic characteristic.

    Yesterday I took a couple more 'scope shots to see if anything has changed.
    Red trace is the signal taken at RST\. The blue trace is zoomed in time and amplitude. Note changes in timebase.
    1) my current circuit, 10k-10k prebiased
    2) a QuickStart-A board I picked up at Radio Shack a couple of weeks ago. Fairly consistent output trace.
    3) a QuickStart-A board I received soon after they originally came out. This is the worst case trace. I had to reset many times to find one this bad. They were usually more like the one in (4). Despite the ugly trace, it never said "prop not found" during this session.
    4) a Prop Plug from 2013 purchase. This is the worst case after many tries.

    Thank you for the link. And that is insane how much oscillation there is, and how just changing one part can make all the difference.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-04-04 - 17:11:03
    Demonstrates why a good o-scope is needed for really locating problems.
  • TCTC Posts: 1,019
    edited 2014-04-04 - 21:05:24
    Demonstrates why a good o-scope is needed for really locating problems.

    Sometimes I wish I had one, but I cant afford it. :frown:
Sign In or Register to comment.