Shop OBEX P1 Docs P2 Docs Learn Events
Noise considerations for propeller power supply using charge pump regulators — Parallax Forums

Noise considerations for propeller power supply using charge pump regulators

Sleazy - GSleazy - G Posts: 79
edited 2008-06-06 12:48 in Propeller 1
I was intending on using a charge pump regulator (specifically the Texas Instruments TPS61025) as a power supply for the propeller.· The·TPS61025 charge pump regulator has the advantage over the LM1086·from the protoboard in that it can produce regulated 3.3v from· 1,2, or 3 NiMH NiCd Alkaline or Li-Ion batteries·anywhere between 5.5v to .8v .··With the LM1086 you'd need to use·at least·3 regular alkaline or 4 NiMH·batteries·in order to produce the·required·input voltage, wheras the TPS61025 you'd only need one AA·(and drain it quite a bit operatively).· However, the charge pump regulator has voltage ripple noise at its output (since its pumping) wheras the LM1086 does not.

Im concerned·that·using the TPS61025·might not be a viable option due to the·prop's sensitivity·to EMI and noise (Try lighting a piezo spark lighter about a foot away from the protoboard and "POOF").· I do understand that this is a bit extreme of an example (since the piezo sparkers can generate upwards of 20kV, almost EMP'ing of sorts), but i have also noticed "hard reboots"·when switching devices which share a burdened·conduit with the protoboard's wall wart·transformer supply

·With regards to·the TPS61025's continuous operation mode·output ripple swing of·20mV peak to peak, do you think this could be a problem?· I havent looked at the LM1086's ripple (if any) but i imagine that it's on the protoboard due to it's relative quietness, since theres really no "pumping" of charge, only stepdown.· I noticed that the prop will work from 2.7 to 3.6 V, but what kind of slew can it handle?·

What do you think?· Im going to give it a try no matter what , but PLACE YOUR BETS......

·

Comments

  • Sleazy - GSleazy - G Posts: 79
    edited 2008-06-06 12:16
    Well my sleazy self just tested the LM1086 from the Protoboard and found out its got the same 20mV ripple that the TPS61025 has in constant op mode.· You guys should really look into that regulator, it seems like it could be a good fit for the prop with wide battery configurations.·



    ····························doin the bulldance, feelin the flow·· yeah.gif
  • DroneDrone Posts: 433
    edited 2008-06-06 12:29
    Hi Sleazy,

    My gut feeling is that with 20mV ripple you won't have a problem. There are boost regulators out there that will do 3.3V and 5V from a Li-Po battery plus they have a buck stage to charge the Li-Po from 5V (typically from USB power). These parts are designed to go into MP3 players and the like. Don't remember part numbers, check-out www.linear.com and www.maxim-ic.com; maybe they have something. If you can find a part from Linear, it is likely you can model it in their free LT-Spice program.

    I'll try the piezo-lighter experiment - interesting.

    When you checked the proto-board ripple, were you powered from a battery or from a wall-wart? Methinks you should test when battery powered only.

    Good Luck,

    David
  • AleAle Posts: 2,363
    edited 2008-06-06 12:32
    I used one of them, the TPS to power at 5V an atmega, no problems smile.gif. Regarding the noise sensitivity of the propeller I think from my experience, 500 mV ripple does not bother the prop at all. I thought I had better power distribution and filtering and decoupling caps... well was not the case, the propeller does not reboot or has strange behavior. The buck/boost regulators by TI have a better efficiency than the charge pumps but greater ripple.
  • Sleazy - GSleazy - G Posts: 79
    edited 2008-06-06 12:37
    I just checked the LM1086's data sheet.· The spikes caused by load transient response in the LM1086·are practically equivalent to the TPS61025's 40mV peak to peak spikes, however the pulse width in the LM1086 is 20uS wheras the TPS61025 is only 400uS, so it actually slews less in the TPS61025 (a.k.a. less noisy).·· I think this should end up working.· Ill tell you what happens.
  • Sleazy - GSleazy - G Posts: 79
    edited 2008-06-06 12:48
    Yeah, those buck/boost in the bootstrap configuration can have high power capability, but they can be unstable·in the bootstrap config ( i think they parasitically·spike easier·with·bootstrapping if i can remember·), and in non-bootstrapped configurations the·voltage is more stable, but·the power output isnt up to par with what i need.· It seems those TPS6102x models have about the highest power in any regulator (without a honker transformer coil [noparse][[/noparse]bs]) with a wider input than anything else ive seen.· I guess that the TPS61026 can get about 2.7 watts through at most (which is nice cause i need 2.45W max, just about perfect)

    i think the buck/boost MAX669/668 was a decent chip , but just a bit less power out than the TPS61026, plus you had to bootstrap it.· The TI stuff is nice cause the TPS has the integrated battery latching if things get hairy. im not sure if the MAX chips had the latching, i just remember that the bootstrapping had some limitations.·

    I went thru the ENTIRE power supply cookbook that MAXIM puts out, and there wasnt anything that seemed better than the TPS6102x models, i think anything close was bootstrapped in the MAX circuits.·
Sign In or Register to comment.