Post Go back to editing

# Tuneable 1 to 10A switched current source to drive a diode

Hi,

In our project we need to build an efficient tuneable current source with a nominal current of 5A and tenable in range 1A to 10A. My idea is to use a synchronous switched buck converter and manipulate the feedback loop for current regulation and control. The ADP1853 seems to be a good candidate for that. However, I am not able to build a stable loop with that regulator. The source is a diode with quite a long linear range (at 1A/5A, the voltage across the diode is about 1.45V/1.8V). The allowed current ripple of the source cuurent is about 1mA max.

I tried two possible solutions with were both unstable:

1) A simple arithmetic with OpAmps: Measure the current via a shunt resistor and amplify / subtract a set voltage / add 0.6V / connect to FB pin. Compensation was difficult here as I don't know how to calculate it in this scenario.

2) Measure the current via a shunt that is place after the inductor and before the output capacitor. Add a comparator that compares it to a set voltage and feed back to the COMP pin.

Does anyone has experience in such a design and could give me a reference to papers or a suggestion to calculate or evaluate the circuit in more detail?

Parents
• Since a circuit can tell more than a lot of word, I attached an example LTSpice simulation with the ADP1853 in a constant current circuit set to 5A. Tracking and softstart are disabled. Frequency is set to 1MHz. The current sense is disabled (100k resistor) and ILIM is set to roughly 10A. Ramp compensation is disabled. Therefore the internal ramp is defined as in the datasheet from 0.7V to 2.2V. A simple resistor R16 replaces the diode for now.

It regulates until it reaches about 5A by switching as fast as it can as expected. But then it bursts with constant switching on-times instead of constantly regulating with small on-times of the HS switch: (blue: 5A reference, red: current through R16, dark green: SW pin voltage, light green: COMP pin voltage)

If we take a closer look into the burst periods, we can see the HS on-time is very constant, no matter what the COMP voltage is. I would have guessed the on-time is directly proportional to the COMP voltage (between 0.7V and 2.2V as stated in the datasheet).

I hope this helps to clarify my question. I don't fully understand the (internal) feedback behaviour.

• The ADP1853 has power saving mode feature when it detects the output current is low. It is to improve efficiency at light load condition.

You may want to check on DC2132A demo board that features few ADI parts to achieve constant voltage, constant current power supply.