GEM Voltage Wrong

Ask questions about the GEM here.
Post Reply
adamjacobmuller
Posts: 3
Joined: Sun Jul 10, 2022 6:19 pm

GEM Voltage Wrong

Post by adamjacobmuller » Mon Sep 19, 2022 10:29 pm

Hi,

I have two GEM and have some odd issues with both of them from the voltage monitor.

For Unit #1, I occasionally record improbably low or impossibly high values:

Code: Select all

> SELECT "volts" FROM "voltage" WHERE ("serial" =~ /^01120860$/) AND time >= 1663625927152ms and time <= 1663625951873ms fill(null);
name: voltage
time                volts
----                -----
1663625928000000000 121.8
1663625930000000000 121.8
1663625932000000000 121.6
1663625934000000000 121.7
1663625936000000000 121.7
1663625938000000000 121.7
1663625940000000000 121.7
1663625942000000000 121.7
1663625944000000000 121.8
1663625946000000000 11.8
1663625948000000000 121.7
1663625950000000000 121.7
>
This happens "a few" times a day and is plausible I suppose that there are some "micro" outages happening but I do have multiple UPS systems attached to the same mains feed and none of them show the issue, however, I also see spikes where the voltage increases to implausible values:

Code: Select all

> SELECT "volts" FROM "voltage" WHERE ("serial" =~ /^01120860$/) AND time >= 1662020921000000000 and time <= 1662020935000000000;
name: voltage
time                volts
----                -----
1662020921000000000 121.8
1662020923000000000 121.8
1662020925000000000 121.8
1662020927000000000 1218
1662020929000000000 121.8
1662020931000000000 121.8
1662020933000000000 121.8
1662020935000000000 121.8
>
As I'm writing this, I realize that this is very possible if that value is sometimes 1/10th the value (11.8 instead of 118v) and sometimes 10x the value (1218 instead of 121.8).


I have a 2nd GEM unit which also shows this issue, but, in addition to that all values are implausibly low:

Code: Select all

> SELECT "volts" FROM "voltage" WHERE ("serial" =~ /^01120861$/) AND time >= 1662020921000000000 and time <= 1662020935000000000;
name: voltage
time                volts
----                -----
1662020921000000000 68.9
1662020922000000000 68.9
1662020923000000000 68.8
1662020924000000000 68.8
1662020925000000000 68.8
1662020926000000000 68.8
1662020927000000000 68.8
1662020928000000000 68.8
1662020929000000000 68.8
1662020930000000000 68.8
1662020931000000000 68.8
1662020932000000000 68.8
1662020933000000000 68.8
1662020934000000000 68.8
1662020935000000000 68.8
>
Any ideas what could be wrong? Thanks in advance!
Teken
Posts: 2700
Joined: Wed Dec 15, 2010 4:09 pm
Location: The Bad Lands

Re: GEM Voltage Wrong

Post by Teken » Mon Sep 19, 2022 10:52 pm

Give this link a quick review as it relates to the power supply:

https://www.brultech.com/contact/
Teken . . .

My ongoing projects thread: http://www.brultech.com/community/viewt ... ?f=2&t=929
Buy me a cup of coffee: https://www.paypal.me/Teken https://gfinotify.com/ Discount Code: PC10
ben
Site Admin
Posts: 4266
Joined: Fri Jun 04, 2010 9:39 am

Re: GEM Voltage Wrong

Post by ben » Tue Sep 20, 2022 1:14 pm

How the unit #1 connected (to what, what type of connection, packet format, etc)? It looks to me like it's missing parts of the string every once in awhile. The "2" in the first example, the decimal in the 2nd example.

Any chance you're using a Prolific-based Serial-to-USB cable?

Check the PT settings on Unit #2 (under Channel Option in the GEM Setup webpage), it looks like they're off by close to half. Also check the phyisical PT connection to make sure it's plugged in all the way.
Ben
Brultech Research Inc.
E: ben(at)brultech.com
adamjacobmuller
Posts: 3
Joined: Sun Jul 10, 2022 6:19 pm

Re: GEM Voltage Wrong

Post by adamjacobmuller » Tue Sep 20, 2022 8:01 pm

Hi Ben,

> Also check the phyisical PT connection to make sure it's plugged in all the way.

This was (basically) it! It was plugged in fully, but, removing it and plugging it back in mostly fixed it (75v->110v), doing that again but cleaning up the connections on both sides with a paper towel and some compressed air fixed it completely.

> How the unit #1 connected (to what, what type of connection, packet format, etc)?

I am using the ethernet model, setup as a TCP server, so my client software connects to it and periodically (1/sec) sends an ^^^APISPK and reads the responses. Would lowering the baud rate between the GEM and the TCP server help? I'm on 115.2k now.
ben
Site Admin
Posts: 4266
Joined: Fri Jun 04, 2010 9:39 am

Re: GEM Voltage Wrong

Post by ben » Wed Sep 21, 2022 9:20 am

adamjacobmuller wrote:
Tue Sep 20, 2022 8:01 pm
Hi Ben,

> Also check the phyisical PT connection to make sure it's plugged in all the way.

This was (basically) it! It was plugged in fully, but, removing it and plugging it back in mostly fixed it (75v->110v), doing that again but cleaning up the connections on both sides with a paper towel and some compressed air fixed it completely.

> How the unit #1 connected (to what, what type of connection, packet format, etc)?

I am using the ethernet model, setup as a TCP server, so my client software connects to it and periodically (1/sec) sends an ^^^APISPK and reads the responses. Would lowering the baud rate between the GEM and the TCP server help? I'm on 115.2k now.
Which packet format are you using with APISPK?
Ben
Brultech Research Inc.
E: ben(at)brultech.com
adamjacobmuller
Posts: 3
Joined: Sun Jul 10, 2022 6:19 pm

Re: GEM Voltage Wrong

Post by adamjacobmuller » Wed Sep 21, 2022 11:55 am

Hi,

2 "ASCII with Wh"
ben
Site Admin
Posts: 4266
Joined: Fri Jun 04, 2010 9:39 am

Re: GEM Voltage Wrong

Post by ben » Wed Sep 21, 2022 1:37 pm

adamjacobmuller wrote:
Wed Sep 21, 2022 11:55 am
Hi,

2 "ASCII with Wh"
If you're doing 1s polls, I'd suggest moving over to one of the binary packets. Might be having some trouble keeping up.
Ben
Brultech Research Inc.
E: ben(at)brultech.com
Post Reply