Charging capacitor with DC-DC boost converter

Hello All,

I am powering a latching solenoid valve that requires a 100ms pulse at 6V and 7W via a 1500uF capacitor with a transistor and a 555 timer controlling the pulse. Currently, the capacitor gets charged directly from a 9V battery through a 120Ohm resistor. At the moment the battery gets drained faster than expected and stops working when the voltage in the battery drops from use.

My proposition is that I put a DC-DC converter between the battery and the 120ohm resistor so that we can use the battery at its lower voltages by trickle charging the capacitor.

I found this IC(LM2574N-12GOS-ND) but maybe there is a better way to do this operation.

Thank you.

Entechbob

Hi Entechbob,

Yes, you can extend the battery life with a switching regulator. With a proper
choice of regulator you can actually omit the 120ohm resistor (the resistor just
dissipates power). Choose a switcher with relatively low current limit to ensure that
the instantaneous loading of the battery while charging remains sensible.

The LM2574N may not be the best choice as its quiescent current consumption is pretty high (i.e. draws power from battery even when the capacitor is full). Take a look for example to TPS62125, which has very low Iq, low light-load current consumption and 100% duty-cycle mode (by-passes the switching at low input voltages). Unfortunately it comes only in WSON package, which is not easy to solder by hand.
https://www.digikey.com/en/products/detail/texas-instruments/TPS62125DSGR/3503783

Digi-key has also evaluation board for that part available, which is handy, if you need only a small number of boards.

Cheers,
Heke
AsamaLab

1 Like

Hello Heke,

Thank you for your reply, it has brought to my attention the low Iq and current limit! I went this route instead: LT1111CS8-12-PBF.

What do you think of this choice?

I really like your IC and would have gone with that, but I did forget to mention we are going to run the 6V valve with 12V for the 100ms pulse. I have tried it and spoke to the manufacturer and they do not see a problem with this either. We will charge the capacitor to 12V prior to turning on the transistor.

Thank you,

Entechbob

Hi Entechbob,

Thank you for reply. OK. Sounds solid. That explains your need for boost converter. The chip you are planning to use, the LT1111 looks like a perfect fit. You only need to be careful not to short the output, as there is no short circuit protection, but that is perhaps not so critical in this case.

As your application is battery operated, you could also extend the battery life by fiddling the power consumption of the pulse source. The epic 555 chip consumes around 10mA all the time. That may become the major consumer, if your solenoid firing interval is long. You could consider a low power drop-in version of the timer, such as TLC555, reducing the Iq to 250uA.
https://www.digikey.com/en/products/detail/texas-instruments/TLC555CDR/276979

Cheers,
heke
AsamaLab

1 Like

Heke,

That is great to hear! I will consider the short circuit issue as someone may attach the battery incorrectly. Do you know of a low power solution for the short circuit protection(or if its even necessary)?

Currently, we are using this 556 IC, ICM7556IPD±ND, and it says 120-240uA supply current for the ICM556, is that the all time consumption? We are using the dual 556 version because we need to send out a positive pulse for opening and a separate positive pulse for closing. We are only pulsing for 100ms and the 12V capacitor is powering all the IC in the design, including this 556 IC.

Thank you!

Entechbob

Hi Entechbob,

If the end user does not have an access to the output side of the boost, IMHO, then you won’t need a short circuit protection, as the short can be caused by component failure only and the short just drains the battery (may also blow the converter’s blocking diode).

However, as the user may install the battery wrong way around, input reverse polarity protection is perhaps needed. A forward diode will do the job although some power is lost to it. A less lossy solution is to use a PMOS. Take a look at:

ICM7556 is a good choice.

Cheers,
heke
AsamaLab

1 Like

Heke,

Ok, I am learning new things all the time!

I will definitely check out the design guide and most likely implement it as well!

Thank you for all your help, hopefully I can pay it back/forward!

Take care,

Entechbob

1 Like