LM34914 in buck-boost configuration

Hi Fred,

Can you confirm that you require a traditional inverting buck-boost configuration? If this is the case, most conventional buck regulators, can be applied. Can you provide us with specific information about your input voltage range and desired negative output voltage and current. For some interesting reading on inverting buck boosts, please refer to SLVA059A or AN-2027.

Thanks,

Anston

Hi Anston,

What do you mean by "traditional"?

Vin=21-27V

Vout=-7.5V

Iout=0.2A

We also need the same in positive voltage (7.5V, same Vin and Iout), hence the plan to use the same controller and probably the same discrete components around.

We are looking for a low cost and low noise solution (some post filtering is considered).  The ripple given by the Type 3 config in LM34914QuickStartCalculator.xls speadsheet is satisfying.

I already had a look at SLVA059A .  I will read through AN-2027.

Thanks

Hi Fred,

Our expression for 'traditional (inverting) buck boost' is to distinguish it from the similarly named buck boost that is non inverting and generally has four switches in its configuration. 

Now that we have your operating specifications, the configuration is clear.

There are three subjects we need to discuss. For your first application, which is a non inverting buck, this should be very straight forward from the datasheet, that presents no surprises. The second application, is an inverting buck boost. In this topology the "GND" rail of the IC is re-referenced to the negative output voltage. There are many situations where the application load is placed between the +7.5V and -7.5V outputs. As a cautionary note, add a reverse biased schottky diode between -7.5V and GND. This will reduce the possibility of the negative regulator from refusing to start. Also there is a subtle point you should be aware of with inverting buck boost's, in your application, the output power level of the positive rail and negative rail are the same. Internal to the inverting application the peak switch current and average inductor currents are higher than in the non inverting buck application. This is because the inductor current in the inverter is the sum of the DC input and output currents of that stage.

A different topology you might consider given your relatively low output power level is to use a coupled inductor on the positive output buck regulator. This topology is called a buck with an over-winding. By adding a SEPIC cap to this topology the cross regulation between the two outputs is quite good. Though only the positive output is truly regulated. It isn't clear whether there is a cost advantage using a single regulator and the more complicated inductor configuration.

 2117.[E2E] Fred P.pdf

Thanks,

Anston

Hi Anston,

Many thanks for these good advices!

I especially like the coupled inductor topology.  Coilcraft makes some good and affordable OTS coupled inductors so it looks like the cost can be lower with this topology.

Do you have any documentation to suggested on that topic?  I'll go for some simulation to get the feeling of it.

However, one thing worries me: the "noise coupling" network from the inductor to FB pin will be hooked to the positive side only.  What kind of impact is to expect?

Thanks  again!