• State Resolved
  • Locked Locked
  • Replies 14 replies
  • Subscribers 64 subscribers
  • Views 2634 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Improving load current of TPS22860 (eg 2 in //)

jyl
Intellectual 745 points
Other Parts Discussed in Thread: TPS22860, TPS22915, TPS27081A, TPS61085

Hi,

In a design where the quiescent current is very critical, I need to switch a voltage up to 5 V and TPS22860 appears to be the right choice.

However, the current limitation is a bit short...

Does it sound reasonable to put 2 devices in parallel?

During switching we may have a slight unbalance between the 2 for a short duration, but since the load will also be switched at the same time (could even be switched a little later on, if required) this should not be so problematic....

Any advice?

Thanks

JYL

  • 0
    TI__Genius 15760 points

    Hi JYL,

    I have contacted the appropriate engineer and they should get back to you soon.

    Thanks,

    Alek Kaknevicius

  • 0 in reply to Aleksandras_Kaknevicius
    Intellectual 745 points
    Thanks Alek for your prompt update.
    BR
    JYL
  • 0 in reply to jyl
    TI__Mastermind 24975 points
    Hello JYL,

    Although I have not done it, it should possible to put two of these devices in parallel. As you mentioned, the two may be unbalanced during the time they are turning on if one turns on faster than the other, but depending on how much over a single device your output current is, may be ok.

    Also, when both are completely on, they may not be perfectly balanced either since the Rds(on) can be different between the two devices. Also, if the quiescent current is critical, you are going to double that quiescent current by using two devices in parallel.

    It really depends on how much current you are pulling and if its consistent or just for a short period of time? What are your loading conditions?

    Best,
    Michael
  • 0
    TI__Genius 15760 points

    Hi JYL,

    It may be simpler to use a load switch with that current carrying capability rather than putting two switches in parallel. 

    The TPS22929D is a leaded, 5.5V load switch with a current carrying capability of 1.8A.  If a WCSP device can be used, then the TPS22915 has a 5.5V / 2A capability which should suit this application.  Let me know if either of these solutions could work.

    Thanks,

    Alek Kaknevicius

  • 0 in reply to Michael J Schultis
    Intellectual 745 points
    Hi Michael,
    Many thanks for your feedback.
    I know the quiescent current will double in that case, but it will stay far below what stronger switches are able to offer (typically 2 x 10nA versus micro Amps).
    The load requires about 300 mA, may be a little more in some cases.
    I also understand that the Rds(on) may not match perfectly , but I assume that once the one with the smaller Rds will reach its limit, the Rds will increase up to a balance between the two devices...

    BR
    JYL
  • 0 in reply to Aleksandras_Kaknevicius
    Intellectual 745 points
    Hi Alek,

    Thank you for your suggestions.
    However they can't suit our needs due to the quiescent current: they are respectively 2 to 7 uA for TPS22929 and 7 to 10 uA for TPS22915 which is almost 3 decades above what we can tolerate...
    Thanks anyway
    BR
    JYL
  • 0 in reply to jyl
    TI__Mastermind 24975 points
    JYL,

    To follow up on the following point:

    I also understand that the Rds(on) may not match perfectly , but I assume that once the one with the smaller Rds will reach its limit, the Rds will increase up to a balance between the two devices...

    This is not necessarily the case. If one reaches its current limit, it won't necessarily just stop allowing current to flow through it.
    In the case where the current increases above 200mA for a given load switch, the device is then operating beyond the absolute maximum specification and may simply fail (I can't guarantee it will fail, but I can't guarantee it wont).
    While I would expect the current to be shared closely between the devices, it would be hard to guarantee that the absolute maximum wouldnt be exceeded.

    Lets look at some of the datasheet specs. For worst case on a device at 3.3V and full temp range the resistance can be up to 1.3 Ohm. From the typical characteristics for R(on) the resistance is in the range of 0.63 typical (and its not guaranteed that it won't be lower. In this case, if the load current were 300 mA, then the current through the 0.63 would be just barely north of 200mA (which is the absolute maximum meaning the device may fail.)

    I think this example above is quite extreme and you are unlikely to see this kind of variation in R(on) between two devices in parallel. So this gives me confidence that the current will be shared between the devices in such a way that the absolute maximum will not be exceeded (assuming current loading of 300 mA). However, I can't absolutely guarantee this to be the case, despite being pretty confident that this setup should work.

    Does this make sense?

    Best,
    Michael
  • 0 in reply to Michael J Schultis
    Intellectual 745 points

    Thanks Michael for you feedback.

    Since Rds(on) increases with temperature, I'm assuming that the device with the smallest Rds(on) will carry the more current and then warm up a little, hopping it will balance with the second device before burning down...

    In our design , the switch should be On for only couple seconds a day, meaning that it won't be dreadfully stressed.

    Anyway, I'm just building a prototype so far. I'll give a try and keep the post updated with the first results.

    In the mean time, if you know other references able to carry a little more current with a quiescent current on Vin in the Off state in the range of nano Amps, simply let me know.

    BR

    JYL

  • 0 in reply to jyl
    TI__Mastermind 24975 points
    Hello JYL,

    This may be true, but it may not balance perfectly. Even still, with some quick calculations from my previous post, I am pretty confident it should work for this application, but it can be difficult to guarantee under every corner case.

    Only other device i might suggest is the TPS27081A. This is not as low leakage, but can handle higher currents.

    Please let me know how your prototype goes. I would definitely like to hear how well it works and any data you take in this application.

    Best,
    Michael
  • 0 in reply to Michael J Schultis
    Intellectual 745 points
    Hello Michael,

    The prototypes have now been mounted and used for a while and nothing special has been noticed regarding the TPS22860.
    It's true that I didn't try to overload them and also did not attempt to measure precisely how the current split between the two (not so easy to carry out without hacking the board), but I would say: so far so good...

    Thanks again for your educated comments.
    Best Regards
    JYL
  • 0 in reply to jyl
    TI__Mastermind 24975 points
    That is great to hear!

    Very happy to help,
    Michael
  • 0 in reply to jyl
    TI__Genius 15760 points

    Hi JYL,

    Would it be possible for you to share any end equipment / application details here?  It would help us to understand the need for a higher current with our low leakage devices.

    Thanks,

    Alek Kaknevicius

  • 0 in reply to Aleksandras_Kaknevicius
    Intellectual 745 points
    Hi Alek,

    The design concerns an IOT device, battery powered, for which we need to switch, once a day or so, a 15V power supply to feed our sensors but maintains the global leakage as low as possible to ensure years of operation with a single LiSoCl2 battery.
    The current on the 15V supply is about 60 mA, leading to about 300mA at the input of our DC/DC converter built around a TPS61085.
    The leakage of the TPS61085 is too high to only use the enable pin to switch the power (about 10 uA).
    This is the reason why we decided to add an efficient switch (in terms of leakage) in front of the DC/DC converter.
    The only (very) efficient one we found is the TPS22860, but it appeared a little to short in term of max current.

    Best regards
    JYL
  • 0 in reply to jyl
    TI__Genius 15760 points

    Hi JYL,

    Thank you for the additional information, this is very helpful!

    Thanks,

    Alek Kaknevicius