Post Go back to editing

Is it possible to parallel two LTC1550 for increased output current?

Category: Hardware
Product Number: LTC1550IGN
Software Version: NA

Hi,

I'm running a LTC1550 on a +5V supply (which cannot be changed). My design specs are for transients up to 13mA output @ -4V and a constant 10mA.  This doesn't give me much of an overhead, but I can't use another device due to the required low output ripple.

Are their any issues with placing two of these in parallel?  Any isolation required on the common rail between the devices (diodes etc), or any filtering concerns due to the charge pumps?

Thanks in advance!

Parents
  • Hi ARL,

    We never tried to connect the LTC1550s in parallel. Typically I would say no, there is no way to assure current sharing or to sync the frequency. Also each have their own control loop and will both try to regulate to 4V separately. However each have a high open loop impedance and this might be OK. The LTC1550 that regulates at the highest voltage will be supplying the current. If VOUT does start to droop because of IOUT*RDSON, the second LTC1550 will start supplying the current and keep the output in regulation. The only problem that might occur is the ripple voltage might periodically line up and provide a little higher ripple at some point because the frequencies can not sync up.

    I think this should work fine but you should try this on the bench and make sure the output ripple and performance is acceptable.

Reply
  • Hi ARL,

    We never tried to connect the LTC1550s in parallel. Typically I would say no, there is no way to assure current sharing or to sync the frequency. Also each have their own control loop and will both try to regulate to 4V separately. However each have a high open loop impedance and this might be OK. The LTC1550 that regulates at the highest voltage will be supplying the current. If VOUT does start to droop because of IOUT*RDSON, the second LTC1550 will start supplying the current and keep the output in regulation. The only problem that might occur is the ripple voltage might periodically line up and provide a little higher ripple at some point because the frequencies can not sync up.

    I think this should work fine but you should try this on the bench and make sure the output ripple and performance is acceptable.

Children
No Data