Should I isolate grounds from the minus (-)DC power supply?

Discussion in 'Off Topic' started by Joan Genebriera, Apr 20, 2021.

  1. Joan Genebriera

    Joan Genebriera Cyanogen Customer

    Joined:
    Jul 10, 2016
    Messages:
    60
    Location:
    Catalonia, Spain
    First of all excuse this off topic note but with a frequency of once a week, working 4 nights a week, my Maxim DL 6.26 goes into hang situation without any apparent reason.

    When this happens there are multiple applications through the Observatory Control function, such as: image camera, tracking camera, MaxDome, focuser and telescope control by TheSkyX. As I am sure that Maxim DL is a solid piece of software, I have verified and updated, all the device drivers and upgrade ASCOM to v 6.5.1.0 but I still have a doubt that concerns if the ground connection has to be isolated from the (-) minus connection of 12VDC PS ?

    At the observatory there is a 24VDC PS for the telescope mount and a 12VDC PS for all other instruments. The observatory is controlled from a room 80m away via an IP line and a USB server and is precisely on the USB server where there is an internal direct connection between ground and the (-) 12VDC, hence this is distributed to all devices connected by the USB server. Could this be the cause of the Maxim hang?

    I suppose the right thing to do would be to install a ground cable to the chassis of all the observatory devices and connect this common ground to two long copper bars buried next to the observatory? Any other suggested action? Thank you.

    Regards
    Joan
     
  2. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,369
    Location:
    Earth
    Bom dia Joan, (my Catalan is not so good!),

    1. Check the operating system settings:
    https://forum.diffractionlimited.co...connects-turn-off-usb-selective-suspend.7848/

    2. Consult a professional licenced electrician.
    Usually these people are experts and are trained to deal with issues of grounding, bonding, earthing, proper ground rods, lightning protection, etc.

    3. DC power and common ground point
    I run a solar-powered observatory. I've learned a few things.
    a. There should be a ground rod hammered into the earth (depending on local soil conditions), perhaps 3m in length.

    b. All AC power grounds should go to a single panel and single ground point connection, like that ground rod.

    c. DC power should be handled with care. Fuses (like Automotive or Marine) for DC should be used on all power sources and to all devices.
    That's why I like the RigRunner.
    Here's an example - the RigRunner by West Mountain Radio. MFJ makes a similar device:
    http://www.westmountainradio.com/rigrunner.php

    I like DC distribution devices - amateur radio people invented a standard way to run +12V power that is very reliable, and RF filtered.

    d. If there are different "ground" potentials and the devices aren't isolate, this can cause major problems for USB and RS232 communications.

    e. Some computers are not grounded and float - common with laptops. This can cause issues.

    f. Usually you want all +12V and 0V DC connections to connect a single common point, and feed from there.

    g. In the case of different voltages, the grounds usually need to be connected, if the devices are not isolated.

    h. SBIG cameras like the original ST-10 with the "extra cooling" option had a second power supply, and it has to be isolated from the camera's main power supply, but the +12V from the second power supply connects to the case (backwards design) but since it was an isolated transformer power supply, it didn't matter.

    i. Scopes like the original Meade LX200 classics had the case floating above DC ground, so the red bar graph current display would work. This was a dumb design, but a nice feature for monitoring the battery. The problem was hooking a grounded accessory to the scope would short out the power monitoring circuit, and could put a 0.5A to 1A through that mechanical connection. Big problem for cameras.

    j. Newer SBIG cameras have a resistor/capacitor network between the case and the electronics ground. Don't connect them together.

    k. If you run solar charge controllers or inverter-based switching power supplies, there can be a fair bit of noise, both high frequency on the DC and radio frequency interference.
    Noise suppression and ferrites might be needed.

    l. Solar charging - if you have solar panels, usually the aluminum frames have to be connected to a common ground - that ground rod, to protect against lightning.
    The +/- from the panel should NOT be connected to the DC output ground - a lot of bigger charge controllers need the input and outputs separated.
    Don't underestimate the problems you can have with these and with battery power as well.

    m. Dew controllers - some dew controllers use RCA round jacks, but they wired them backwards, with +12V on the outside, and ground in the middle. The cables can short if the plug outside hits the scope. Very hard to detect when this happens.
    I use heat-shrink tubing to fix this, or buy a better dew controller.

    n. DC power packs - be careful, some have a centre negative pin, outside positive. Some have a DIN connector, and in many cases the outside metal of the DIN is not connected to DC ground, and might be connected to AC ground. DC ground is on a pin or a few pins. They need to be kept separate.

    That's a lot. Start with the USB issue - it's the most common thing we see.
     
  3. Joan Genebriera

    Joan Genebriera Cyanogen Customer

    Joined:
    Jul 10, 2016
    Messages:
    60
    Location:
    Catalonia, Spain
    Hi Colin,

    Nice and long list for verify my electric installation. I will start at once.

    Thank you for you kinds words!

    Joan
     
  4. Greg Crawford

    Greg Crawford Cyanogen Customer

    Joined:
    Oct 11, 2014
    Messages:
    101
    Hi Joan,

    A few years ago I had a nearby lightning strike which destroyed some of my observatory equipment. Fortunately I was insured for most of the destruction, but knew I had to learn something about lightning. Since my second hobby is amateur radio, I turned to my radio club for help. I think what I learned can address one of your statement/questions, namely:

    "I suppose the right thing to do would be to install a ground cable to the chassis of all the observatory devices and connect this common ground to two long copper bars buried next to the observatory? Any other suggested action? Thank you."

    Please do not do this. I will explain why. If you have a nearby lightning strike, the ground (meaning the soil) at point zero will have a massive over-abundance of electrons causing a huge increase in electrical potential. Since electrons do not like each other (like charges repel), they will spread out into the surrounding ground until the potential difference is dissipated. In other words, the excess electrons will flow from the location of the highest potential (ground zero of the strike) to the areas of lowest potential. This is a flow of electric current, with the voltage dropping in steps the further you go from ground zero. Most buildings have a single ground rod for their electrical mains. If you add a second ground rod some distance away and you have a nearby lightning strike, the over-abundance of electrons near one ground rod will have two paths to areas of lower potential. One path is through the earth itself. The other path is up one of your ground rods through your earthing wires to the other ground rod. This second pathway has much less resistance to the flow of electrical current than the path through the ground (soil). When seen as an electrical circuit, you have two resistors in parallel: the resistance of the soil is one circuit and the resistance of your earth wiring is the other. Ohms Law of electrical behaviour means most of the electrical current will flow through your earth wiring and this is a very dangerous thing. A typical lightning strike might have a potential of 50,000 volts pushing a massive amount of amps through your household wiring system. The earth wiring system could heat up like a toaster and burn your buildings down. Now for the disclaimer: I am not an electrician or electrical engineer, so follow Colin's advice to consult with someone who is.

    With global warming, electrical storms will increase in frequency and intensity, and thus the risk will increase. Even without a lightning strike, overhead clouds with high electrical charge can induce current to flow in long leads. So reduce the number of long leads and at the first sign of rumbling thunder disconnect those long leads from observatory equipment. I admit I have become a bit neurotic about this. But what can you do when the air force keeps flying their Joint Strike Fighters over your house and the young lady across the road keeps revving up her motorbike?

    As for long leads, a USB lead that covers 80 metres seems to be operating way past its viable length. Why not put a PC (even a mini-computer) in the observatory and connect to it by either Ethernet or Wi-Fi?
     
    Last edited: Apr 21, 2021
  5. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    9,921
    Greg is correct - what you want is a single point ground.

    In my observatory, the pier is grounded. Everything else in the observatory is tied to the pier ground. That way if there is a lightning strike in the vicinity, the ground might suddenly get a very negative potential, but everything in the observatory moves with it.
     
  6. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,369
    Location:
    Earth
    Greg and Doug make excellent recommendations.
     
  7. Tim Povlick

    Tim Povlick Cyanogen Customer

    Joined:
    May 8, 2020
    Messages:
    316
    Location:
    Kayenta Utah, USA
    One of the DC devices used here also has the 12 DC(-) terminal tied to the chassis ground forcing all the other 12V gadgets to also be at ground on the DC(-). The large metal bits (pier, mount scope) are bonded to a ground wire that is tied to the AC powers ground. According to "Grounding and Binding for the Radio Amateur", if a single circuit is run to a remote building, then a grounding electrode is not required (NEC 250.32)

    At the observatory here, a local PC (on the pier), Maxim DL works like a champ over wired ethernet; no glitches or crashes are observed (operator error notwithstanding).

    Suggest using Wireshark (WS) with the USB add on to check if something about the USB transfers is causing the failure. WS is not to difficult to use but requires a review of the docs. Also Windows Event logger.

    Good Luck....

    Tim
     
  8. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,369
    Location:
    Earth
    Tim - good comments. Windows Event Viewer's System log can often give useful insights on what is happening on a machine, including OS errors and hardware issues.

    I use WireShark from time to time. I don't recommend it for most people as it's rather complex. Plus, the USBcapture driver can add extra load to the machine when running.
     

Share This Page