diy solar

diy solar

Solar Assistant (Possibly) Voids Warranty on EG4 Inverters?

OK, so if wall worts are supposed to be isolated due to certification as mentioned here (click links to go to posts):
maybe that is best for power...if you can handle the power outage issue. Unless I'm mixing up floating vs isolation?

Cable, possibly use RS232, an RS232 isolator and the RS232 to USB-A adapter as mentioned here:
and
liked/responded by SS/Eg4

Orange vs Rasp? Who knows.
 
Well there's another method if your uber paranoid. Plug a ups into an outlet that is being fed off the inverter. Plug a wall wort into the ups. Plug the pi into the wall wort. UBER PROTECTED.

This strangely enough is how I have a cell phone being powered off my 12v stuff. I have an inverter on the 12v stuff and with a ups plugged into it which powers various stuff and a timer plugged into the ups with a wall wort plugged into the timer which powers the an old cell phone I use to read bluetooth stuff remotely.

The reason for the timer is to keep from killing the battery in the phone since cell phone batteries bloat and burst if left charging constantly.
 
SA just mentioned to me in an emial:

"We communicate with the device using standardized USB protocol RS232/485, so if SolarAssistant can damage the device then so will plugging the device into anything else. Voltage spikes are an issue with the Voltronic inverters because they use a cheaper CH340 chip that fails and causes a voltage spike from its end, but there is never any damage done since these spikes are in the millivolt range and a reset usually gets the system up and running. "

I'm not faulting Voltronic, if they engineered the inverter like this, and there wasn't any problem with it the way it was designed, so be it. I'm just trying to figure out what freaking cable to use specifically for the 6500ex so I don't cause a prob and have to try to deal with SS support.

Anyone know where the CH340 chip might be? Is it in the RJ port or micro usb? Others have mentinoed the micro usb works for SA to the 6500ex. Maybe the problem is using the RJ port designed only for PC/Laptop firmware updates, and everything else is "not supported."
 
The reason for the timer is to keep from killing the battery in the phone since cell phone batteries bloat and burst if left charging constantly.
I hacked one of my Moto phones by wiring a barrel plug to the bms strip soldered to the battery pins on the phone for connecting a temp batt to it during boot up, then leaving it plugged in all the time...after boot, I disconnect the batt via the barrel so it will not bloat and stays on via the wall wart.
 
Yep. It boots FAST. About 2 or 3 times faster than with the flash drive.

How long does it take to boot a pi from solid state memory?

You young whipper snappers have no concept of "fast".
I used to have a computer with 8" DSDD (double-side, double-density) floppy drives. It was a CompuPro S-100 system with 8085 CPU, running CP/M.
When I turned it on, there were two "clicks" as the head were pressed to the media, and in less than one second it had booted.

For mission critical applications, when it absolutely, positively, must work every time, we use floppy disks.


And we most certainly do not use USB drives. Who ever thought it was a good idea to execute software automatically without the operator typing "run"?
 
Floating AC to DC converter for your RaspPi will not have 32v popping out. But as I said, double up the safety measures to hedge against me not knowing what I'm talking about / you not understanding / either of us not translating properly from theory to practice.
The wall wart is going to be floating since that's required for safety certification. Otherwise you may have 115 to 120V from ground with some fault combinations which is not a good day for someone touching the USB shield plugged into the wall wart.

Inside the wall wart the DC side is isolated from the AC side etc via a small power transformer. Which means the DC output will end up being referenced against the other side of that RS232 chain, if that chain is not isolated.

Older wall-warts had a 60 Hz transformer. Most these days are SMPS, with high-frequency transformer. Either way they are galvanically isolated. However, there is a non-infinite impedance at higher frequencies.

We were using a Meanwell switching power supply to feed our circuit boards, with a switched power strip for AC. When the switch was turned on (random phase to 60 Hz sine wave), there was a common-mode transient on the DC output, relative to ground. Testing repeatedly, I measured varying spikes, as much as -8V.

This worked fine powering an isolated DUT, but with a data cable (USB to I2C dongle) connecting it to a PC that had a USB cable to an AC powered scope, the voltage transient damaged the DUT, about 10% of our boards coming out of test, having. A quality bench power supply eliminated the problem.

In the case of this inverter with SA, you're stuck with whatever the inverter has inside, so isolator in USB adapter may be what is needed.
 
How long does it take to boot a pi from solid state memory?

You young whipper snappers have no concept of "fast".
I used to have a computer with 8" DSDD (double-side, double-density) floppy drives. It was a CompuPro S-100 system with 8085 CPU, running CP/M.
When I turned it on, there were two "clicks" as the head were pressed to the media, and in less than one second it had booted.

For mission critical applications, when it absolutely, positively, must work every time, we use floppy disks.


And we most certainly do not use USB drives. Who ever thought it was a good idea to execute software automatically without the operator typing "run"?
How ironic. Ive got one of those :)

But its much newer than the oldest in my collection. The trs80 model II's I have are older than it and have the 8" drives.

Even have one old vacuum tape stand up drives like they used in the scifi movies. Most of my oldest stuff got to hard to power reliably so I boxed it.

Thinking of putting it all on ebay since no one seems to care about the stuff anymore. I collected over the years but its just junk now I guess. Wanted to do a museum at one time. Oldest machines I have that still work are from around 1972.
 
But its much newer than the oldest in my collection. The trs80 model II's I have are older than it and have the 8" drives.

Even have one old vacuum tape stand up drives like they used in the scifi movies. Most of my oldest stuff got to hard to power reliably so I boxed it.

I was organizing my garage and found a rack of wire-wrapped cards from my childhood.
Stuffed in among them was a bit of paper punch tape, some program I had written.

1701102202369.png

Thinking of putting it all on ebay since no one seems to care about the stuff anymore. I collected over the years but its just junk now I guess. Wanted to do a museum at one time. Oldest machines I have that still work are from around 1972.

Try contacting Computer History Museum on Shoreline in Mountain View
(took over SGI building.)
 
Also, now I'm wondering if I should just use the micro usb port and avoid the "custom" rj port on the 6500ex...even with the SA approved voltronic usb-rj cable.
Definitely not. I think this was one of the problems with power feeding between th Pi and the inverter.
 
I just received this from EG4/Signature Solar support:

"I'm glad to see the system is running again, although I would personally not advise using Solar-Assistant as it will burn up the communication card of EG4 inverters. It sends an additional 32V back into the port it's connected to. Also it voids the warranty of EG4 inverters since it is an unsupported configuration."

So I have to say, WHAT? How does that work? The Raspberry Pi is only 5 volts. Has anyone else been told this? I'm using the same cable that comes with the inverter, used by a Windows computer to update firmware on the inverter. Does that mean that a firmware update can damage the inverter (6000EX)? Anyone from Signature want to chime in? @BenFromSignatureSolar ?
We apologize as this is not accurate information. It should not have been voiced to a customer in this way, and the source of the information has already been corrected. It does NOT void the warranty, and the time it "burnt up" was an isolated incident and cannot be proven that there weren't other factors involved.
 
We apologize as this is not accurate information. It should not have been voiced to a customer in this way, and the source of the information has already been corrected. It does NOT void the warranty, and the time it "burnt up" was an isolated incident and cannot be proven that there weren't other factors involved.
Thanks for finally posting this. I really don't understand why this took so long, nor why such confusion on the part of support. Many of us have been concerned about using Solar Assistant with our EG4 inverters because of this misinformation.
 
Definitely not. I think this was one of the problems with power feeding between th Pi and the inverter.
I agree I have an LVX6048WP and if I use the micro USB port my SA Raspi continues to receive power after I unplug it from the power supply. If I use the serial cable it shuts off immediately, so some cheap design of the control board is pushing power back out the usb port on the inverter.
 
Regarding the possibility of stray 32v being sent to the EG4 6000ex/6500ex’s display through the USB to RS232 voltronic cable, would it be possible to just disable the 5v pin on the cable? I am not sure if SA/inverter comms require the 5v to work or not. I read that some 3d printers have problems with Orange pi/Octoprint and users are placing a small piece of tape to cover pin 1 of the standard A USB plug. Just a thought.
 
SA just mentioned to me in an emial:

"We communicate with the device using standardized USB protocol RS232/485, so if SolarAssistant can damage the device then so will plugging the device into anything else. Voltage spikes are an issue with the Voltronic inverters because they use a cheaper CH340 chip that fails and causes a voltage spike from its end, but there is never any damage done since these spikes are in the millivolt range and a reset usually gets the system up and running. "

I'm not faulting Voltronic, if they engineered the inverter like this, and there wasn't any problem with it the way it was designed, so be it. I'm just trying to figure out what freaking cable to use specifically for the 6500ex so I don't cause a prob and have to try to deal with SS support.

Anyone know where the CH340 chip might be? Is it in the RJ port or micro usb? Others have mentinoed the micro usb works for SA to the 6500ex. Maybe the problem is using the RJ port designed only for PC/Laptop firmware updates, and everything else is "not supported."
This still seems to be missing some details.
I'm not super familiar with this chip, but from my brief reading, this is my understanding:
The CH340 is a USB to UART (and RS232?) series of chips.
It seems like in order to use the CH340 with RS232 levels, you still need to use an RS232 driver like the MAX232 series.
Those driver chips use capacitors to generate enough boost voltage to drive the lines to true RS232 levels.
-32V is large, even for true RS232. That said, if there was a problem with large negative voltage, I'd be looking at the driver chip, not the CH340.

It is possible to hook some versions of the CH340 directly up to the RS232 lines, but it lacks the charge pump of the MAX232 chips.... so I don't see how it would generate true RS232 voltages.

I'm not familiar with the Votronic part(s?) that failed. What was the actual spec the comms port was designed for? Did they use RS232? RS485? TTL serial?
 
Back
Top