Hi guys.
I'm trying to control a 500k ohm potentiometer using some sort of digital means in order to make a programmable buck/boost converter...

How can this be achieved? The biggest digital potentiometer that I could find was a 10k one....

 

Without a schematic it is impossible to say. Do you want to add a small motor to the 500K pot you have, or alter the circuit to use the 10K digital pot you can buy, or what?

ak

 

Why does the buck/boost converter need such a high value pot?

 

Why does the buck/boost converter need such a high value pot?

Its a ltc3780 buck boost converter. It has a 500k pot for voltage control and 250k pot for current limiting.... So any ideas how can I digitally achieve this. Thanks so much :')

 

Hello,

Do you have the intended schematic you want to use?
On the webpage I do not see the 250K and 500K you mention:
http://www.linear.com/product/LTC3780

Bertus

 

Hello,

Do you have the intended schematic you want to use?
On the webpage I do not see the 250K and 500K you mention:
http://www.linear.com/product/LTC3780

Bertus

Thanks for the reply

In this video the guy changes the potentiometers to external ones with a value of 500k and 250k. You can also see in this schematic that there is a 487k potentiometer..

 

I basically want to copy his variable power supply but replacing the potentiometer s with a keypad, arduino and LCD display.. I already figured how to measure voltage and current but I have no clue how to digitally control 500k,250k potentiometers...

 

I see a 487K fixed resistor in the voltage output sense feedback loop, but not a pot. Also, most digital pots cannot run between +48V and GND. What is the max output voltage you are trying for?

ak

 

I see a 487K fixed resistor in the voltage output sense feedback loop, but not a pot. Also, most digital pots cannot run between +48V and GND. What is the max output voltage you are trying for?

ak

Thanks for the reply
If you watched the video the buck/ boost converter can output a max voltage of 30 volts. The module is rated at 120w so it technically can output 4 amps at 30 volts. I honestly cant find the 500k, 250k potentiometers in the schematic but I'm 100% sure that the module uses 250k/ 500k potentiometers....

 

That's because that is not the actual schematic for that board design. Linear Technology makes the control chip at the heart of the circuit, and the schematic is from their applications group, not the board vendor. It's probably very close, with pots substituted for fixed resistors. R5 is a 220K resistor that sets the current limit threshold, so that's where the 250K pot goes.

Because the current sense adjustment resistor is so close to GND, it probably can be replaced with a digital pot. The problem for voltage adjust is that it touches the output. You might be able to leave R7 alone and replace R8 with the digital pot.

HOWEVER - both the voltage and current set circuits have capacitors in them that are part of the control loop frequency compensation, and varying the resistors shifts the corner frequencies. Whether or not this will send the regulator into oscillation can be calculated by someone much better at this than me. I'd start by replacing R8 with a 10K pot and adjusting it all over the place to see what happens.

ak

 

Between some unknown dudes video off the internet and the manufacturers data sheet , I would go with the data sheet every single time.

 

This website shows the 500k pot on the feedback pin. http://wiki.beyondlogic.org/index.p...ost_Converter_Ebay_Automatic_lifting_pressure
The digital pots I have used run off 5vdc. Most of the pots are 8 bit which is 256 taps, this may reduce your resolution a bit, so I would stick with a old fashion pot.

 

both the voltage and current set circuits have capacitors in them that are part of the control loop frequency compensation, and varying the resistors shifts the corner frequencies. Whether or not this will send the regulator into oscillation can be calculated by someone much better at this than me. I'd start by replacing R8 with a 10K pot and adjusting it all over the place to see what happens.

The VOS+ feed resistor R4 (10Ω) and C47 (0.1u) (as in the datasheet) would certainly affect the feedback loop frequency response, but I'd be surprised if reducing R8 and R7 by a factor of 100 even would make much difference. The reactance of C47 at the operating frequency is only a few Ohms, so if R7 were, say, 5k that shouldn't greatly affect the phase shift which R4 and C47 produce. That's theory, of course .

 

So im all confused by all of this. So is it possible or not? From what i understand it is too complicated?

 

So is it possible or not?

My money says not, unless (a) you can source digital pots which can withstand 30V (or whatever your max output will be) and (b) the regulation will work with pot resistances ~10k or less.

 

O

My money says not, unless (a) you can source digital pots which can withstand 30V (or whatever your max output will be) and (b) the regulation will work with pot resistances ~10k or less.

Ok thanks :') is there a good guide to make a switching programmable variable psu with a max voltage of around 30v?

 

The pots values make zero sense to me.

Having said that, if you still want to go down the path, attenuate the high voltage and then amplify it with an opamp whose gain is controlled by the pot.

 

The pots values make zero sense to me.

Having said that, if you still want to go down the path, attenuate the high voltage and then amplify it with an opamp whose gain is controlled by the pot.

Do you have a real schematic?

John

 

I found this high voltage digital pot, it can handle 36 VDC.

http://www.mouser.com/new/microchip/microchip-mcp41hvx1-potentiometers/

Microchip's MCP41HVX1 36V Digital Potentiometers are devices that have dual power rails (analog and digital). The analog power rail allows high voltage on the resistor network terminal pins. The analog voltage range is determined by the V+ and V– voltages. The maximum analog voltage is +36V, while the operating analog output minimum specifications are specified from either 10V or 20V. As the analog supply voltage becomes smaller, the analog switch resistances increase, which effect certain performance specifications. The system can be implemented as dual rail (±18V) relative to the digital logic ground (DGND).

I don't know if that part is a good solution or not, but thought I would mention it

 

is there a good guide to make a switching programmable variable psu with a max voltage of around 30v?

Haven't checked, but people here usually advise that it's cheaper/easier to buy one than build one.
Ready-built ones normally have a fixed potential divider or trimmer feeding back from the output to set the ouput voltage. You might be able to hack that, using a digi-pot (30V rated?).