openinverter forum

Yes buzzing is normal. When CSDN is NOT connected to 12V the lower IGBT will switch on and short out your battery.

I didn't have much luck charging from 240V, next time I'd put a heating element in series, so should you

Thanks again what about the burning smell and we're would you put the heater inseries with what.

In series with 240V.
Burning smell: probably wire insulation getting hot.

Just a quick update got charging up and running today. I found out if I put more than 190vac into MG1 it tripped the gen3 relays out. I also could only get my 24v battery to charge at 36v if I changed settings started to trip the relays.

Alright cool
How do you charge a 24V battery at 36V?

I had a voltmeter on the positive and negative posts of the battery . Before charging process began it read 25v and while charging it read 36v . So next step is to keep playing with the settings to get a 27v reading when charging the lead acid batteries and check the ampage . The charging process is started as previous post. Need to keep altering the charge settings to find out what they actually do as at the moment I don't totally understand them.

HA!

Charging works! My set-up for testing is a little crude, but much has been learned in the past day and a half. I tied CSDN to inverter logic power (12v) via a 470 ohm and grounded the CSDN side of that resistor as would happen when the DCSW is closed (by ground signal to the relay coil) to initiate charging (per Johu's suggestion).

Short story: As long as: CSDN, F&R are tied to 12v @ VCU on, nothing terrible happens if there is current limiting on the input to MG1.

"chargepwmin" is approximately the actual % of the HVDC bus you will see at the HV battery connections while charging

I had a time getting the voltage limits high enough to allow the system to function. My advice is to set the UDC limits to at least 1.5x the input on MG1.

While charging:
- If CSDN is "open" (ungrounded, or tied to 12v) WHILE charging, nothing really interesting happened.
- If the AC input was disconnected while charging, nothing interesting happened
- If the HV pack becomes disconnected, the HVDC bus voltage spikes, but nothing exciting occurs
- If "F" or "R" become disconnected from 12v WHILE CHARGING, things get truly exciting on the battery charging side (lots of I & U fast )
- Disconnecting the VCU & Inverter logic power (by pulling the ground), got a slight current spike to the battery, but not "excitingly" so

More to come after further testing

Good work zookeeper what voltage did you put into mg1 as I could not go above 190vac

Forgot to ask but voltage is your high voltage battery

Pcstreet wrote: ↑Fri Dec 18, 2020 8:22 pm Good work zookeeper what voltage did you put into mg1 as I could not go above 190vac

220 in and saw about 320 on the High HV bus

What voltage is the battery

Pcstreet wrote: ↑Fri Dec 18, 2020 9:21 pm What voltage is the battery

Initial testing, 12, planned for 240ish.

OK, more testing. This time it was "off script" and likely off the reservation! :O

What did I do?
- Applied AC power to V & W of MG1 (via my autotransformer so I have control)
- Powered up the Inverter only (VCU was NOT powered up)
- Connected a 60v battery pack to DC In
- Successfully CHARGED the 60v pack to 65v by varying the transformer output

So, if all else fails, I have a practical method of charging my EV battery w/o having to purchase a stand-alone charger.

Now to get the VCU involved in charging the 60v pack, but am having trouble configuring the PI settings

chargecur =2A - I *think* this is on the AC side "input for Boost" based upon the Parameter Wiki
chargekp - I messed about at 12v and did not see any significant changes
chargeki - I messed about at 12v and did not see any significant changes
chargeflt - I messed about at 12v and did not see any significant changes
chargepwmin - this seemed to work fine @12v, but first stab at 48v did not get the expected results (ie: 50% of the HV bus of 120v did not get me close). With 120v on the bus, I will set to 55% and see if I get 65v.

Something is not right, either with my understanding, process or settings. The ONLY way I get charging is with both CPWM and CSDN "unpowered", as in floating or grounded. If either is tied high (via a 470 ohm), the system boosts "pack" voltage into the inverter and I get no charging.

Process followed:

I pulled high CSDN and CPWM, the latter via a 470R resistor
I connected a 60V battery to the orange DC input and 115Vac to V & W phases of MG1.
I started BOOST mode (always use boost mode with this converter)
<5s later I disconnected CSDN

chargemode = boost
chargecur 2A
chargekp 10 dig
chargeki 10 dig
chargeflt 8 dig
chargepwmin 5%
chargepwmax 80%

I then disconnected CPWM and began to charge @ 55v but have no charge control

I also noticed that "udc" = the "high-high" voltage rail and not HV battery voltage, so I learned something new


edit:added video

Also did some 230V experiments of my own.

Hooked up 230V to two phases of MG1 with a precharge resistor (always on) and a relay. Relay controlled by the DCSW signal. The DCSW signal also controls the enable signal of the converter. Basically it will first enable the charging PWM at chargepwmin and then close the relay and enable the converter. Thus no unwanted short circuit or boosting.

I found last time I had the controller gains too high, thats what caused the fuse trips. Have chargekp=40 and chargeki=10 now. Also there might have been a glitch in software where PWM was shortly disabled when closing the contactor.

Those things are now fixed. Mind you, the power factor is probably not great, most of the sine wave remains unused as is usual with buck chargers. But it charges

Video coming soon!

EDIT: I'm wondering if for some reason the plot only shows the positive wave... Current passes through sensors BEFORE rectification so should swing both ways. Hmm.

Looking forward to the video! Eventually I will catch up to this stage, one day, I hope.

I'll give it a spin tonight

Ok, lets do an early forum release

Thanks Johu. Been trying to get back on my feet. End of year complicated, so project had to take 3rd place... I have prius contacters waiting to get wired up..

Just watched Johu's video again and it looked like the charger settings says he had it buck mode not boost. My question his is this right or an optical illusion as all the previous posts say it as to be in boost mode

I actually altered the software, so on the latest you indeed need to set it to buck. When using MG1 as rectifier and have battery connected to stock battery input

Thanks for your reply so now does this mean that you never use in boost mode . Also which revision software did you introduce this.

The latest is 5.0.R and I believe that is correct, not Boost Mode as 240VAC rectified will get you ~320DCV to the batteries.

Just tried charging again with revision 5 software in buck mode and nothing happened at all. It did not even say it had gone into buck mode. Switched back into boost mode and started charging again and got 17amps roughly going into battery measured with a clamp meter on the positive cable fed by a 60vac supply. The software was set up to supply 1amp so I should have seen about 3amps at the clamp meter. I'm wondering if I have a faulty invertor any advice or suggestions will be much appreciated.

No it has to work in buck mode, otherwise the controller polarity is reversed, i.e. when seeing too much current it will push even more!
Not sure what you did there. Also you're setting input current, not output. Output isn't sensed.