openinverter forum

Hello, I have an error measuring the DC bus voltage with the "get udc" command. The command reports a DC bus voltage of 22.71 when my meter shows 0v. Is this normal or perhaps I have not correctly applied the voltage sense fix?

Are there any obvious issues with the way I applied the voltage fix?

Yes it's correct and it only affects the temperature reading. The voltage sense needs to be calibrated with udcofs and udcgain.
With no voltage connected set udcgain=1. Then read udc and set udcofs=udc . Then apply a known voltage and set udcgain=udc/u_known

GrapeNuts wrote: ↑Mon Mar 11, 2019 1:56 pm Hello, I have an error measuring the DC bus voltage with the "get udc" command. The command reports a DC bus voltage of 22.71 when my meter shows 0v. Is this normal or perhaps I have not correctly applied the voltage sense fix?

Did you follow calibration instructions? You have to set the voltage offset and gain. So your offset at 0V would be 22V then. If you connect your main bus how much does it show? What is your system voltage? Mine is 380Vdc and if i connect 50Vdc battery i get 60Vdc or more whereas at 360Vdc system is spot on. I like the ATtiny sensing but there are also other options.

EDIT: Sorry i see Johannes has answered already .

Thanks! The calibration procedure corrected the issue. My voltage is now reported +/- 3v of measured voltage at several intervals between 0 - 350v. I saw no mention of that procedure in the documentation I have been following at https://openinverter.org/invkit.html. Is there another document I should follow once I have my kit assembled?

I just added it to the wiki which will supersede the traditional pages soon. https://openinverter.org/wiki/index.php ... sor_boards

johu wrote: ↑Thu Mar 28, 2019 7:20 am I just added it to the wiki which will supersede the traditional pages soon. https://openinverter.org/wiki/index.php ... sor_boards

Hi

I am looking at your voltage comparator schematic and i am wondering what is the result of 680k resistors. Could i use 360k to 4k7 resistor divider with the same circuit? What would i get out of the battier?
My goal is to get a linear signal that would tell another chip HV voltage in frame from 2.5V to 5V signal. Could that be done with your circuit?
EDIT: could i transfer 400Vdc sensing directly across Si8920 isolation barrier with MCP6001 to get 2.5V to 5V linear?

Yes with resistors R8 and R7 you can set the offset voltage. I tried with ltspice but I guess you can find the formula somewhere on the web.
The input voltage into the Si8920 chip must be between -250mV to 250mV so goes negative also. The voltage between AOP and AON varies from -1.6 to 1.6V (8.1x gain)

johu wrote: ↑Thu Apr 18, 2019 7:23 pm Yes with resistors R8 and R7 you can set the offset voltage. I tried with ltspice but I guess you can find the formula somewhere on the web.
The input voltage into the Si8920 chip must be between -250mV to 250mV so goes negative also. The voltage between AOP and AON varies from -1.6 to 1.6V (8.1x gain)

Hm... i will just use your schematic then since i still have to use another decision making chip to trigger precharge, reset and enable for controler. Also i could use it to input other events like trigger A/C inverter, throttle for a genset or a chargepump . It could also handle overtemp protection for inverter. I will test it with another board this weekend and measure output from 360Vdc across isolated barrier and opamp.

tnx J

After reading this discussion I also took a look at the isolated voltage sense schematic of Johannes and noticed a point, not sure if I'm correct though. The pin number 2 of OPA344 is negative supply (V-) and pin 5 is positive (V+) according to the datasheet (http://www.ti.com/lit/ds/sbos107a/sbos107a.pdf). But on the schematic, pin 5 is connected to the ground and pin 2 is connected to VCC. Am I missing something?

There is two pinout versions, I selected the weird one (with full intention of course ) It's the 6001R part.

johu wrote: ↑Fri Apr 19, 2019 8:59 pm There is two pinout versions, I selected the weird one (with full intention of course ) It's the 6001R part.

Understood, thanks.

johu wrote: ↑Fri Apr 19, 2019 8:59 pm There is two pinout versions, I selected the weird one (with full intention of course ) It's the 6001R part.

Hah, it got me scratching my head also. And then I saw the negative R amp on digikey. That figures, programmers like to use negative logic all the time .

TNX Johannes