Shop OBEX P1 Docs P2 Docs Learn Events
Misbehaving Voltage Regulators — Parallax Forums

Misbehaving Voltage Regulators

Private19872Private19872 Posts: 61
edited 2014-12-29 21:09 in Propeller 1
Recently, I got the Propeller DIP Plus Kit (http://www.parallax.com/product/130-32305) and I started to assemble it on my breadboard. I didn't get too far however, when I noticed that neither of the voltage regulators were working as expected. When powered by an 8.1v source, the 5v regulator gave off 6.85v. I found a 7805 I had lying around and substituted it in for the first regulator and got a steady 5v. Then I added the 3.3v regulator. Instead of limiting the voltage to 3.3v it only went down to 4.36v. Does anyone know what may be going wrong? I found a schematic on the same page as the link, and it appears that the regulators are in the same way. I don't have anything besides the regulators on my breadboard. Any help on this would be appreciated.

Comments

  • LeonLeon Posts: 7,620
    edited 2014-10-06 13:02
    Schematic?
  • jmgjmg Posts: 15,173
    edited 2014-10-06 13:06
    ... I don't have anything besides the regulators on my breadboard. Any help on this would be appreciated.
    No decouplnig caps ?
    Many (most?) regulators need decouplng caps, and some (like LM317) also have minimum load requirements - usually met choosing the Adj resistors.
  • RaymanRayman Posts: 14,652
    edited 2014-10-06 14:31
    Yes, you need capacitors at input and output of regulators, in general. For 7805 types, you'll want ~10uF of tantalum or electrolytic.
    Also, make sure you put the capacitors in the right way. They are polarized and putting in backwards with break them.
  • YanomaniYanomani Posts: 1,524
    edited 2014-10-06 16:21
    Recently, I got the Propeller DIP Plus Kit (http://www.parallax.com/product/130-32305) .................................... I found a schematic on the same page as the link, and it appears that the regulators are in the same way. I don't have anything besides the regulators on my breadboard. Any help on this would be appreciated.

    Hi Private 19872

    IF you are refering to the schematic drawn on page 27 of the Propeller Education Kit Labs Fundamentals Text whose download link is available at product's page, you should verify your setup against figure 3-4.

    There are two 1000uF x 6.3V electrolytic capacitors in the kit that must be correctly wired (double check them and also their polarity marks), one between LM2940-5.0 OUPUT terminal and GROUND, and the other, between LM2937-3.3 OUTPUT terminal and GROUND.

    Be sure to properly assemble the 100 Ohm resistor and the green led at the LM2937-3.3 output circuit.

    IF you are NOT USING a 9V battery, but an external 8.1V source, as you stated, be sure to mount the third electrolytic capacitor, 0.47uF x 25V between LM2940-5.0 INPUT terminal and GROUND.

    When assembling the electrolytic capacitors, keep the connections as short as possible, and also closer as practical to the regulators pins.

    The same advice do applies for the connections between the 3.3V regulator output, the Propeller and all the other circuits involved.

    After you ensured proper connections between all parts, apply power to the circuit and double check all the voltages.

    As a last recommendation, when you have finished assembling the parts that comes in the kit, do follow the Improve PE Kit Supply Voltage Stability topic at page 38 of the same manual.
    You will need two 0.1 uF ceramic capacitors, of suitable voltage rating (between 10V and 25V is better), to accomplish it, but it realy worths the effort.

    For short:

    The capacitors will ensure minimal noise from the regulator circuits to reach each other, and sure, the Propeller and the other logic circuits;

    The led (and current limiting resistor) provides enough current consumption to avoid unreliable regulator operation and also warns that the circuit has power connected to it.

    Hope it helps a bit.

    Yanomani
  • Private19872Private19872 Posts: 61
    edited 2014-10-06 16:47
    Thanks. I'll try that and get back to you all on it. I started with just the regulators and nothing more and was going to copy the capacitors and everything else around it, but if the regulators need a load and capacitors I'll add that and retest.
  • Mark_TMark_T Posts: 1,981
    edited 2014-10-07 05:57
    Thanks. I'll try that and get back to you all on it. I started with just the regulators and nothing more and was going to copy the capacitors and everything else around it, but if the regulators need a load and capacitors I'll add that and retest.

    Regulators without the appropriate capacitors may oscillate and destroy the chips they power. Similarly
    paralleling regulators can do the same thing. A regulator is a high gain amplifier combined with a feedback
    network to hold the output voltage constant. Without the right frequency response of the feedback network
    a high gain amplifier can become a high power rail-to-rail oscillator.

    Each particular regulator will say in its datasheet the minimum (and maximum) capacitance on input and output
    and may have requirements on the ESR (effective series resistance) of those capacitors.
  • mmowenmmowen Posts: 38
    edited 2014-12-26 23:23
    One item worthy of mention which appears missing from the foregoing threads is that of the minimum load current for the LM2937-3.3. I have also experienced this regulator settling around 4.3 volts if the minimum load current is not preovided. Try adding a 660 ohm resistor from the output to ground which will provide 5mA of load current at all times, and the problem will disappear. (Refer to datasheet National Semiconductor "LM2937-2.5, LM2937-3.3 400mA and 500mA Voltage Regulators" August 2005 page 3 noting the variuous references to Iout=5mA).

    Note: This comment has been added to assist others in future who are searching through these forums for solutions to this issue. Also note that similar requirements but with different values are present for many other linear voltage regulators.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-12-26 23:35
    Just a bit of overview...

    Linear regulator problems
    A. no capacitors OR the wrong value capacitors

    >> The 7805 has different capacitor requirements than the LM2940-5 and so on. MUST READ the data sheet.

    B. high power draw shuts down the thermal protection

    >> A short circuit in your project or the lack of a heat sink will shut off the regulator until it cools down.

    C. some regulators cannot tolerate electricity running backwards through them..

    >> This is true of the 7805 espcially. If one puts a big filter capacitor on the Vout and a smaller filter capacitor on the Vin. When you unplug or short the Vin, you may damage the 7805. This is in the data sheet. Other regulators may survive this snag, but again one must read the data sheet.

    D. unpredictable behavior in the Dropout region

    >> If one insists on running their voltage regulator below the recommended Vin range, it simply doesn't regulate. It may work sometimes and not others. MUST READ the data sheet.

    E. minimum load requirements

    >> This is new to me. May or may not always be true with certain regulators and not others.
  • DomanikDomanik Posts: 233
    edited 2014-12-27 12:53
    Lots of good advice in these posts and the observations about oscillation due to lack-of or wrong value caps was interesting. Along the same lines, sometimes electrolytics dry out and the values drop significantly.
    If I'm using a DVM I measure using DC then switch to AC. On AC I'm looking for 2mv up to 100mv. Much more than that spells oscillation. In the distant past I once measured DC of almost exactly 5 volts, called it good and went on looking for the problem. Later I looked at it with a scope and found a 2.5 volt sawtooth. The 5V DC was an average.
    In your case the 4.36 volt DC might not be strictly DC.
  • LawsonLawson Posts: 870
    edited 2014-12-29 20:28
    I think I was using an LM2937 but in one circuit that actually ran for 2-3 years, the 3.3v regulator had a ~400mV peak to peak saw tooth on it and measured 3.3v with a DMM. The sawtooth was also ultra-sonic so I didn't hear any audio from the capacitors. When I found the problem, the cause was too low an ESR output capacitor. A 10uf ceramic cap would oscillate like mad, while a 10uf ceramic in parallel with a 47uf electrolytic was fine. (and all through this abuse the propeller worked perfectly!)

    Marty
  • kwinnkwinn Posts: 8,697
    edited 2014-12-29 21:09
    Early on I serviced a lot of instruments that used electrolytics in the 1000 to 55,000 uF range and found a multimeter was all one needed to check the capacitors. The rule of thumb was to replace it if the AC voltage measurement was 10% or more of the DC measurement. Worked fine for those linear regulators where the typical input voltage was 50+% of the regulated output voltage. Won't cut it these days though.
Sign In or Register to comment.