I am finding it hard to find a suitable buck

[king.oslo]

Hello there,

I am finding it hard to locate a buck controller which is suitable for my specs. I want to drive the V-out setting from MCU. I would like to avoid digital pot. Driving it with pwm from MCU through a buffer would be excellent. But then I expect it will be easier to get it stable with a buck with a ref-out (this seems to be very hard to find).  These are the specs:

- External N-FETs
- Vout max of minimum 15V
- Iout max of minimum 6.5A
- Surface mount package with exposed pins (not QFN and similar)
- Available@ farnell

Preferable:
- Less than 500KHz
- Synchronous

Thank you for your time.

Kind regards,
Marius

[amspire]

Need a few more details, like input voltage range, output voltage range, how well regulated do you want the output and what space limitations do you have?

[king.oslo]

Thanks Richard,

Off topic: I have worked loads on the voltage reference you helped with, but I have some problems with it which i have failed to fix. I will find time to do it today

Input voltage range: 18-17V
Output range: ~1-15V
Space: board size about the same as a credit card.

I am hoping for 5mv Vpp.M

[deephaven]

How about an LM25085?

[free_electron]

- Available@ farnell

well there's your problem.... why do you limit yourself like that...
go with digikey or mouser. they have 100.000 times more parts than farnell..

Exar has a couple of really cool ones. with external fets. programmable over i2c ... with controlled startup slope and shutdown slope.

[king.oslo]

- Available@ farnell

well there's your problem.... why do you limit yourself like that...

Cuz it is going to kill my wallet ship and import parts to Norway from DigiKey. Last time I did it, I paid $300 in import fees on an order of $100. And DigiKey gave me the middle finger when I asked for their help.

Thanks, I will check out Exar M

[ejeffrey]

I don't know about norway, but in the netherlands mouser is a lot better than digikey.  Shippping is cheap, and mouser will handle the import duties so that the shipping company won't charge you an extra 100% service fee to do them for you.  They also don't make you fill out an export declaration.

[amspire]

As far as varying the voltage, most of the IC's have an internal voltage reference of 1.25V or similar connected to one input of the comparator internally. Lets say your device is 1.25V. You then have a voltage divider on the the other input to set the output voltage. If instead of taking the lower resistor of the divider to 0V, you take it to a variable 0 - 1.25V source, then as the voltage of this source goes from 0 - 1.25V, then the output should in theory go from maximum to zero.

I have not tried this, but it should work. Make sure the source has a capacitor decoupling across its output. If the source opamp does not like a straight 0.1uF cap on its output (starts oscillating) add a 10 ohm or bigger resistor between the source output and the capacitor.

Richard.

[IanB]

Last time I did it, I paid $300 in import fees on an order of $100.

Wow, is it really that bad that Norway is not a member of the EU? I had sort of assumed that the EEA agreement between Norway and the rest of the EU would limit those kinds of import fee problems. (I'm assuming here that you would buy components from European suppliers and not US suppliers.)

There are a few people here from the UK. What do you guys do when buying stuff? Do you often feel compelled to buy from the US, or do you manage to find the parts you need in the UK and Europe?

[amspire]

As far as the choice of an IC, this post from Winfred Hill shows how to use the venerable UC3842 as a buck converter. Everyone supplies the UC3842 or compatible - including even Farnell. A few extra parts are needed as the output has to be inverted to be useful in a buck converter.

http://cr4.globalspec.com/comment/653625/Re-DC-DC-Converter-UC3842-buck-voltage-mode

Richard.

Edit: Except leave out the resistor and 12v zener on Vcc as shown in the circuit. The UC3842 only starts up at 15V and then shuts down if the supply gets below 10V.

[king.oslo]

Thanks guys. I got enough datasheets to read up on for a while now M

[king.oslo]

As far as varying the voltage, most of the IC's have an internal voltage reference of 1.25V or similar connected to one input of the comparator internally. Lets say your device is 1.25V. You then have a voltage divider on the the other input to set the output voltage. If instead of taking the lower resistor of the divider to 0V, you take it to a variable 0 - 1.25V source, then as the voltage of this source goes from 0 - 1.25V, then the output should in theory go from maximum to zero.

Richard.

Just wanted to ask (and learn) how you arrived at the conclusion that the error amplifiers does not compensate the voltage relative to the 1.25V reference. Say at V_FB = 1.25 = V_REF the output is uncorrected, and if V_FB increases, V_OUT is lowered to again put F_FB = V_REF? Like an inverting amplifier?

Thanks M

[amspire]

As far as varying the voltage, most of the IC's have an internal voltage reference of 1.25V or similar connected to one input of the comparator internally. Lets say your device is 1.25V. You then have a voltage divider on the the other input to set the output voltage. If instead of taking the lower resistor of the divider to 0V, you take it to a variable 0 - 1.25V source, then as the voltage of this source goes from 0 - 1.25V, then the output should in theory go from maximum to zero.

Richard.

Just wanted to ask (and learn) how you arrived at the conclusion that the error amplifiers does not compensate the voltage relative to the 1.25V reference. Say at V_FB = 1.25 = V_REF the output is uncorrected, and if V_FB increases, V_OUT is lowered to again put F_FB = V_REF? Like an inverting amplifier?

Thanks M

I am not completely sure what you are asking, but if you are asking if there is negative feedback to ensure that the feed back voltage ends up equaling the internal reference voltage, then the answer is yes. So if the voltage on the feedback input is too high, the regulator will reduce the output voltage till it the feedback input is matching the internal reference.

Richard

[IanB]

Just wanted to ask (and learn) how you arrived at the conclusion that the error amplifiers does not compensate the voltage relative to the 1.25V reference. Say at V_FB = 1.25 = V_REF the output is uncorrected, and if V_FB increases, V_OUT is lowered to again put F_FB = V_REF? Like an inverting amplifier?

Like Richard, I am also not quite sure what you are asking?

But another key point to bear in mind is that voltages are always defined as a difference between two points. There is no such thing as an absolute voltage of "1.25 V". For practical purposes you cannot have such a reference. So the regulator is controlling the voltage it sees to be "1.25 + X volts" where X is some external voltage offset under your control as the circuit designer. In most cases X is zero, but if you adjust X to be a negative voltage like -1.25 V then the regulator can regulate the output down to zero.

[king.oslo]

I understand that Richard says 0v at FB-pin = lowest duty cycle, 1.25v = highest duty cycle.

If I feed in 1.15V from a DAC to VFB, how do I know that the regulator does not drop the duty cycle to minimum, in an attempt to get VREF = VFB?

Does that make my question clearer? M