Shop OBEX P1 Docs P2 Docs Learn Events
Propeller and 3x NiMH - is there any danger to fry a chip? — Parallax Forums

Propeller and 3x NiMH - is there any danger to fry a chip?

pik33pik33 Posts: 2,398
edited 2012-05-19 06:03 in Propeller 1
3 NiMH gives 3.6 v and this is in official voltage range, but fully loaded NiMH gives about 1.35V x3 - it is about 4V

I read here about Propeller run powered by 4.5V (and extremally overclocked) and it didn't fry, but I still wonder if someone powered it from these batteries.

Comments

  • jmgjmg Posts: 15,183
    edited 2012-05-18 01:54
    Problems with direct battery powering, include things like
    * Battery reversal (direct reversal will put the batteries straight across the substrate diode, something will die)
    * Battery removal when charging (dirty connections) and now that marginal 4V jumps to whatever the chargers kicks to..
    - CMOS processors also draw more current at higher voltages, so why not just use a regulator ? the batteries will last longer.
  • Mark_TMark_T Posts: 1,981
    edited 2012-05-18 04:13
    pik33 wrote: »
    3 NiMH gives 3.6 v and this is in official voltage range, but fully loaded NiMH gives about 1.35V x3 - it is about 4V

    I read here about Propeller run powered by 4.5V (and extremally overclocked) and it didn't fry, but I still wonder if someone powered it from these batteries.

    Everyone seems to think NiMH cells are 1.2V like NiCd cells, but they are not, they are 1.3V (and upto 1.35V freshly charged). So that's 4.05V for 3 cells. Add a 1N4001 diode to drop it a bit perhaps? They also decrease to about 1.2 to 1.25V as they discharge (whereas NiCd hardly change)
  • pik33pik33 Posts: 2,398
    edited 2012-05-18 04:18
    Maybe Shottky diode can be even better, it has lower voltage drop. And a diode solves battery reversing problem, too.
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2012-05-18 06:04
    Just use a normal LDO regulator as the better ones will only drop a fraction of a volt under load. I use the TC1185 a lot, it's an SOT23-5 pack and typically drops less than 200mv running the Prop. Of course there are many other regulators to choose from, just pick one.
  • Mark_TMark_T Posts: 1,981
    edited 2012-05-18 15:48
    Just use a normal LDO regulator as the better ones will only drop a fraction of a volt under load. I use the TC1185 a lot, it's an SOT23-5 pack and typically drops less than 200mv running the Prop. Of course there are many other regulators to choose from, just pick one.

    Be careful, some LDO regulators start sinking significant currents if the input voltage drops too low and might over-discharge the battery (most rechargeables cannot tolerate this without damage, note). If you handle brown-out correctly the circuit should stop draining the battery when its discharged, you don't want the regulator to spoil this nice behaviour. This sort of thing is not normally mentioned on the datasheet either...
  • jmgjmg Posts: 15,183
    edited 2012-05-18 17:33
    Mark_T wrote: »
    Be careful, some LDO regulators start sinking significant currents if the input voltage drops too low and might over-discharge the battery (most rechargeables cannot tolerate this without damage, note). If you handle brown-out correctly the circuit should stop draining the battery when its discharged, you don't want the regulator to spoil this nice behaviour. This sort of thing is not normally mentioned on the datasheet either...

    Indeed, the ideal regulator should have these features

    ** Inbuilt Reverse Battery tolerance
    ** Flat Icc vs Vcc
    ** Inbuilt Voltage Disable, that removes output loads, and drops Icc to Voltage Monitor levels
    A PGood flag is a natural extension of this.
    Doing this can also simplify the down stream reset circuitry
    ** Optional WDOG, that works on a power-removal basis
    Too many devices these days, have only a 'reset request' pin


    None of this is rocket science, and many regulators have some of these features, (so there is clearly no technology brick walls) but finding one that has all is elusive.

    and I'll add a more modern twist to the wish list

    ** Trim or Vcc option, to allow 5.1V USB chargers to power nominally 5V circuits.
    Here the LDO should have lowest drop, and is almost running saturated. Vo can be 4.83V to still work 5V devices in spec.
    ** Good Vmax - yes this is a trade-off, but car battery is a large market.

    Digikey shows close to 37,000 (!) LDOs, so the market really has too many slight variants, and no ideal solutions.
  • pik33pik33 Posts: 2,398
    edited 2012-05-19 01:53
    It seems in battery powered environment without a power supply, a diode is the best solution

    It drops these excessive 0.4 v
    It don't generate noise like these pulse regulators - it is important with sound synthesis
    It protects against reverse polarization
    it is cheap.

    Using Schottky diode, when insert full loaded batteries, I will get something about 3.6-3.7 V. It seems to be safe for a propeller
    With normal battery voltage - 3.6V - it will be something about 3.2 V, sufficient for a propeller to work.

    There should be a voltage detector in such a system, to alarm when batteries are too weak. One propeller pin and 2 resistors?
  • evanhevanh Posts: 16,112
    edited 2012-05-19 05:22
    Mark_T wrote: »
    Be careful, some LDO regulators start sinking significant currents if the input voltage drops too low and might over-discharge the battery (most rechargeables cannot tolerate this without damage, note).

    While having excessive draw is undesirable a NiMH, unlike Li-Ion or Lead-Acid, isn't going to care if it's discharged to zero or not.

    Reversing it's voltage on the other hand, ie: reverse charging, is meant to be damaging. Although, I've never seen an explanation as to how that damage might occur, eg: driving it beyond it's thermal limits or unmanageable amounts of gassing - both of which would require a decent reverse current to achieve.
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2012-05-19 05:43
    All the reasons have been pointed out why you shouldn't use a regulator because it will never be perfect and the only thing you get with the diode is the voltage drop (always) and reverse protection (woohoo). While reverse protection is "nice" is it really such a problem for you? There are solutions to this but do you have lightning protection as well? Perhaps a lightning rod and spark gap with a good ground for you know.....just in case.

    If I were designing for the mass market I would consider a lot of these "just in cases" but that doesn't help when you want a one-off practical solution. Use the LDO such as I mentioned (I also use MCP1700s and others) as even when the supply drops they draw very little current even at 1.5V input with a max of 80ua. What about low-voltage cutout? You find that the Prop will go into brown-out reset and the operating current will drop down to microamps but this doesn't help with any external components that may still draw power. Simple fix is to tie the /SHUTDOWN (or ENABLE) pin of the regulator with a resistor to ground and have a Prop pin keep this pin high which it can do when it's operating but will relinquish during brown-out (low voltage) and reset. But this could happen when you reprogram the Prop as well so by adding a capacitor in parallel to the pull-down resistor on the regulator you can keep it on long enough for the Prop to reassert the ENABLE pin. How do you get it started from cold? A push button momentary ON switch can connect the battery volts to the ENABLE pin but this means the Prop pin should be isolated a bit with a current limit resistor.

    Here's a sample circuit. R1 could be much higher as could C1 to increase the time that the circuit will stay on for during reprogramming (or else just hold the ON button in).

    prop battery regulator.jpg


    Come to think of it there have been some battery circuits presented in this forum too.
    1024 x 587 - 78K
  • Mark_TMark_T Posts: 1,981
    edited 2012-05-19 05:52
    evanh wrote: »
    While having excessive draw is undesirable a NiMH, unlike Li-Ion or Lead-Acid, isn't going to care if it's discharged to zero or not.

    Reversing it's voltage on the other hand, ie: reverse charging, is meant to be damaging. Although, I've never seen an explanation as to how that damage might occur, eg: driving it beyond it's thermal limits or unmanageable amounts of gassing - both of which would require a decent reverse current to achieve.

    To quote wikipedia on NiMH:
    Over-discharging

    A complete discharge of a cell until it goes into polarity reversal can cause permanent damage to the cell. This situation can occur in the common arrangement of four AA cells in series in a digital camera, where one will be completely discharged before the others due to small differences in capacity
    So unless you have all the cells in parallel discharging a battery completely will damage some of the cells unless they are _extremely_ well matched.

    And capacity and lifetime _are_ reduced if you discharge individual NiMH cells to zero, ask Toyota's hybrid vehicle designers!
  • evanhevanh Posts: 16,112
    edited 2012-05-19 06:03
    Mark_T wrote: »
    So unless you have all the cells in parallel discharging a battery completely will damage some of the cells unless they are _extremely_ well matched.

    That's prolly current draw related. It's unlikely this project is going to be drawing large currents at shutdown voltages.
    And capacity and lifetime _are_ reduced if you discharge individual NiMH cells to zero, ask Toyota's hybrid vehicle designers!

    AFAIK it's considered good conditioning to fully flatten NiMH's. The problem with large scale batteries is exactly what you highlighted above - they can easily enter reverse charge at high current.
Sign In or Register to comment.