Author Topic: Benchtop PSU 3xDPS3005  (Read 3572 times)

0 Members and 1 Guest are viewing this topic.

Offline KalvinTopic starter

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Benchtop PSU 3xDPS3005
« on: May 02, 2018, 08:59:19 am »
Hi,

I have started designing a very simple, switched mode, benchtop PSU with three separate outputs, based on DPS3005 adjustable power modules. The mains transformer is 120VA with four separate secondary windings: 24VAC@1.3A, 16VAC@2.2A, 16VAC@2.2A and 9VAC@1.1A. The 9VAC winding will be used for any logic needed for relay operation etc. The 9VAC winding might also be used in later state to power a Raspberry Pi Zero W.

Here are the main specs:

- Three separate outputs with 2x12V@2A and 1x20V@1A output which can be connected in series or parallel
- Each output can be enabled/disabled individually (DPS3005 has enable/disable button)
- All the outputs can be enabled/disabled simultaneously by a switch (three relays with SPDT contacts disconnecting the load from the PSU)
- Each output can be programmed to CV or CC mode with voltage and current limit (built-in to DPS3005 by default)
- Each output will provide voltage, current and power display (built-in to DPS3005 by default)

The diode at the output is actually a transient suppressor diode (TVZ) which will clamp any high voltage spikes at the output from inductive loads. The noise suppression is done using simple PI-filters at the DPS3005s inputs and outputs (If I have understood correctly, the DSP3005 will generate some switching noise from its inputs, thus the filtering also at the inputs).

Unfortunately the DPS3005 modules I have at the moment do not contain the serial port control ability, which would have been a very nice addition to the power supply capability. With the serial port control capability one could add a Raspberry Zero W inside the box which would control all three modules and provide some long term measurement statistics etc. any any other scripting ability with WIFI connectivity The three channels and the Raspberry Pi could be isolated very easily with simple opto couplers.

Any comments or suggestions are welcome!

Br,
Kalvin
 

Offline KalvinTopic starter

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Benchtop PSU 3xDPS3005
« Reply #1 on: May 02, 2018, 09:44:23 am »
One idea that has been lingering in my mind is to use INA226 for each output and a simple MCU to measure the output voltage/current/power for logging purposes. This would require three cheap DC/DC-converter providing a 5V power supply for the INA226 and MCU for each channel. Also, two opto couplers per channel are required to provide an isolated communication channel between the INA226/MCU and the three Uart-to-USB adapter boards connected to the Raspebrry Pi Zero.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: Benchtop PSU 3xDPS3005
« Reply #2 on: May 02, 2018, 10:51:27 am »
"use INA226 for each output "

How would you keep all the INA226's within their -0.3 to +36V common mode range, especially if the DPS3005s have a mosfet switch in their GND before the sense resistor.
.  That took much longer than I thought it would.
 

Offline KalvinTopic starter

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Benchtop PSU 3xDPS3005
« Reply #3 on: May 02, 2018, 11:41:24 am »
"use INA226 for each output "

How would you keep all the INA226's within their -0.3 to +36V common mode range, especially if the DPS3005s have a mosfet switch in their GND before the sense resistor.

Yes, I am aware that the DPS3005 has a low-side current sense and a low-side MOSFET which will enable/disable the output.

As INA226 and the respective MCU will be powered from an isolated DC/DC module, I will connect the INA226 across the output diode (which will be a TVZ in the final design) and connect the 0V of the MCU/INA226 to the negative output (cathode side of the output diode), the INA226 should not see any excessive voltages. Also, each channel will be floating relative to each other due to the isolated DC/DC modules and opto-isolated serial connections, there shouldn't be any problems.

Edit: Added the updated diagram to include the INA226s.
Edit2: Moved the diode/TVZ to the output side of the INA226 blocks where they belong.
« Last Edit: May 02, 2018, 03:09:15 pm by Kalvin »
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: Benchtop PSU 3xDPS3005
« Reply #4 on: May 02, 2018, 01:09:58 pm »
I didn't know you were adding sense resistor, from there you could connect to the 226's 3005's own GND current sense resistor.
I think you'd have to provide some protection to Vbus from the reverse -1V or more when the protection diode conducts. Vbus's input impedance is 800k, so easy enough to protect.
« Last Edit: May 10, 2018, 06:28:59 pm by StillTrying »
.  That took much longer than I thought it would.
 

Offline glarsson

  • Frequent Contributor
  • **
  • Posts: 814
  • Country: se
Re: Benchtop PSU 3xDPS3005
« Reply #5 on: May 02, 2018, 02:09:06 pm »
Sense resistor for the ina266 after where the DPS3005 module senses the output voltage? This means that the output voltage will change with the current...
 

Offline KalvinTopic starter

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Benchtop PSU 3xDPS3005
« Reply #6 on: May 02, 2018, 02:32:17 pm »
I didn't know you where adding sense resistor, from there you could connect to the 226's own GND current sense resistor.
I think you'd have to provide some protection to Vbus from the reverse -1V or more when the protection diode conducts. Vbus's input impedance is 800k, so easy enough to protect.

Yes, the INA226 input protection should not be an issue. Probably a simple protection scheme with series resistors with 1N4148s will do the trick.

I have not yet checked how bad the output noise ripple is and what kind of filtering is required to keep the switching noise levels below, let's say10mV, but I am quite certain that there will be a considerably voltage drop at higher load currents due to the filter's DC resistance. Less ripple is better, and 10mV ripple is something that I would like to achieve. For lighter loads I would like to see less than 1mV ripple, but I guess this may be quite hard to achieve with DPS3005 module.

EEVblog user plazma has made some ripple measurements here reporting 25mV ripple at 5A with DPS5005 module:
https://www.eevblog.com/forum/projects/dps5005-portable-mini-lab-psu/msg1282741/#msg1282741

Now that I was writing this message it occurred to me that I could have two alternative output filters:

- One for higher loads up to 2A passing more switching noise with less voltage drop.
- One for lighter loads up to 200mA with improved switching noise attenuation.

Selecting the filter can be accomplished with the same MCU controlling the INA226. Need to check whether the selectable filter is worth trouble, as this power supply is going to be used as a general purpose power supply for moderate loads and circuits in which low power supply noise is not a definitive requirement. If the noise is an issue at light loads, one can always add some external filtering to the outputs for improved switching noise filtering.
 

Offline KalvinTopic starter

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Benchtop PSU 3xDPS3005
« Reply #7 on: May 02, 2018, 02:50:24 pm »
Sense resistor for the ina266 after where the DPS3005 module senses the output voltage? This means that the output voltage will change with the current...

That is going to happen as there won't be any means for compensating the voltage drop in the current sense resistor. The maximum useful voltage drop across the current sense resistor is 81.92mV which means that if we want to achieve 2.5A maximum output current, the sense resistor value will to be 32 milliohms. So, the expected voltage drop at full load will be around 0.1V, not including the voltage drop caused by external wires from the output bananas to the actual circuit.

If one wants to measure smaller currents accurately at lighter loads, for example in range 0 ... 250mA with improved resolution down to tens of microamperes, the sense resistor needs to be increased to 320 milliohms giving the same 0.1V voltage drop at 250mA. However, this would call for a range switch which can be implemented using the same MCU controlling the INA226. The range switching could be implemented using a simple relay or using a MOSFET as a switch.

Edit: The INA226 has 16-bit ADC with maximum 81.92mV shunt resistor voltage range and nominal 2.5 uV resolution. That means the current measurement resolution will be around 100uA at 2.5A output load range and approximately 10uA at 250mA output load range. With some oversampling one might be able to achieve somewhat better resolution if really needed.
« Last Edit: May 02, 2018, 02:59:27 pm by Kalvin »
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: Benchtop PSU 3xDPS3005
« Reply #8 on: May 02, 2018, 03:11:30 pm »
I thought you'd make 2 connections to the 3005's sense resistor, to make sure neither goes more than -0.3V the 226s and filters would have to swap places.
.  That took much longer than I thought it would.
 

Offline KalvinTopic starter

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Benchtop PSU 3xDPS3005
« Reply #9 on: May 02, 2018, 03:23:48 pm »
I thought you'd make 2 connections to the 3005's sense resistor, to make sure neither goes more than -0.3V the 226s and filters would have to swap places.

That is one good option for the current sense, but I decided not to modify the DPS3005 module hardware as one can now replace the modules as needed. Unfortunately I do not have the schematics for the modules which makes experimenting a bit difficult.
 

Offline KalvinTopic starter

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: Benchtop PSU 3xDPS3005
« Reply #10 on: April 18, 2019, 09:29:00 am »
Old thread, but here is the latest version of this design providing three separate isolated outputs using the inexpensive DPS3005 (or similar) power supply modules.

- Universal DC-input from 9V to 36V, which allows using the power supply from a car battery, an inexpensive 19V PC battery charger or a 24 AC-to-DC converter.
- Each channel has its own isolated DC-DC-module providing the isolated input voltage for the DSP3005 module.
- Maximum power per channel limited by the DC-DC-module and/or DSP3005 power ratings.
- Multiple channels can be connected in series or parallel without problems with the ground loops etc.
- No problems with the earth leakage current.
- Very small in size, easy to carry around in the lab or in the field.
- Optional support for the USB or Bluetooth connection for controlling the DSP3005 module.
- Optional support for the relay controlling the output enable/disable with an opto-coupler for a) an individual channel or b) multiple channels in parallel.
- The DC-DC-modules used are CUI Inc. 30W, Vin: 9V - 36V, Vout: 15V 2A and 24V 1.25A.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf