Mount Power Fuse Box

[Gina]

As a result of problems with my EQ8 mount, testing showed the problem to be due to the Maplin 13.8v PSU that I used to power it.  The PSU is in the warm room with a cable to the pier and mount.  The problem was not volt drop in the cable and wasn't cured by smoothing and decoupling with capacttors at the pier end but was cured by using a sealed lead acid car battery at the pier end. 

In view of the destructive power in a battery of this sort, fuses are essential to prevent a fire in the event of a short circuit or component fault.  Hence this fuse box for mounting on the pier, which also provides a switch for the mount (maybe overkill since there is a power switch on the mount) and volt/ammeters to show battery status and charging current plus current drawn by the mount.

[hughgilhespie]

Hi Gina,

I am also going down the same route. Although an absolute beginner, everything I have read suggests that 99% of problems in imaging are down to power supply / connector issues. So, I want to at least try to eliminate as many of these as possible from the start.

I have built my UPS 14 volt power supply that will power everything. It has a 12 volt 22 AH battery back-up that is always on when the main PSU is running so in the event of a mains failure, there is enough power for a smooth shut down. When the main 14 volt supply is switched off, a 'smart' battery charger keeps the back-up battery fully charged. The battery charger is isolated when the 14 volt supply is active though as it has a very 'noisy' output - the main 14 volt switch mode unit is a good 'un with very low output noise. I have also put 100 nF filter capacitors on the outputs. The outputs from this box are fused with a 10 amp fuse.

I am now making the box that will be attached to my pier. This will take the 14 volt input from box 1 and distribute it to:

Mount -EQ8

Focuser - Arduino design by Robert Brown

Dew Heaters - Arduino design by Robert Brown

Powered USB 4-port hub

Camera - eventually.

All outputs are individually fused and I have allowed for a separate power supply to the camera socket if needed - I have read that they can be fussy about where the supply comes from.

The focuser and dew heater bits will be mounted in small diecast boxes inside the main box as will the 4-port hub. Each module has an LCD display that is mounted on thetop panel. The focuser also has a manual control box on a 1.5 meter cable, just a couple of push switches and LED's basically. Hopefully this will be redundant when I achieve automation nirvana!

I spent a lot of time looking for quality locking dc power connectors and settled on Neutrik SpeakOn connectors. These are professional audio connectors, rated for up to 40 amps and tough enough to be dragged around stages, tripped over,  thrown in the band's minibus, etc. They are available with different numbers of poles so it's possible to have  scheme that prevents damage if the wrong plug and socket combo is used.

The small signal connectors - two temperature sensors and a humidity sensor for the dew heaters and a temperature sensor for teh focuser -are Neutrik lockable mini XLR connectors. Also supposedly very reliable.

The USB 2.0 inputs and outputs are also from Neutrik. These are lockable and reversable - i.e you can mount them so they present an A or B socket. You can buy ready made leads that have the matching locking connectors.

It is all taking FAR longer than I expected but hey - the joys of being retired! I have now finished the metalwork for box 2 and will start on the electronics next. Then final wiring up, testing - finding out why it doesn't work.....

Regards, Hugh

[Gina]

Here is the fuse box attached to the pier and connected to car battery and mount with power coming from the PSU in the warm room through the cable pipe.  The ammeter has yet to be connected (I've mislaid the connecting leads )  The dual A/V meter at the top shows the battery voltage and current coming from the PSU.

[Gina]

The power feed to the EQ8 mount with battery and fuse box does NOT provide reliable operation of the EQ8 mount   The control drops out after a short period and all control of the mount is lost.  The only differences from the test version which seemd to cure the problem were that :-

1. Crocodile clips replaced with proper battery terminals with supposedly good connections
2. Fuses added.
3. Digital A/V meter inserted between PSU and battery.
4. Instead of feeding power to the battery terminal then separate feed to mount, a pair of short, thick cables connect the battery to the fuses and V/A meters

I find it hard to comprehend that the test system worked and that the properly engineered version doesn't.  Or is it a coincidence and something else is causing the problem? 

If the problem is with the fuse box, the only thing I can think or is the extra resistance of fuses and ammeter.  Seems unlikely but I have now added capacitors across the feed to the mount.  A 470 microfarad electrolytic and a 220nF ceramic capacitor.  I'm hoping to test this out tomorrow.  Could it be that the 4 wire connection to the battery made a significant difference?

Here is the new circuit.

[Gina]

Just had a look outside and it IS a clear night as forecast - so that's yet another imaging opportunity missed due to gear trouble

[Gina]

Still doesn't work!   13.3v showing on the voltmeter should be fine.  I think the problem may be elsewhere, but where? I guess it's a matter of elimination... Aarrgghh

[Davey-T]

Could you put a dummy load on it instead of the mount just in case it is the mount playing silly beggars

Dave

[Gina]

When I say it doesn't work I mean that the mount works for a while after connecting with CdC but then stops responding even though the display shows it as still connected.  The power/charging/battery looks fine.  The battery is charging and the voltage is steadily increasing and the current decreasing.  When the mount is slewed the current goes up and the voltage down a little just as expected.  When the PSU is connected the voltage always reads 13v or above - without, the the voltage is 12.5 without the mount slewing and goes down to 12.4v when is.  I am now trying with the PSU disconnected and the mount running off the battery.  This is a standard way of running the mount and eliminates all question of problems with the PSU or wiring etc.

[Gina]

Stil mount problem but battery voltage down to 12.2 so now maybe running a bit low so I've swapped batteries for my imaging system backup battery - same capacity but fully chaarged.  Running at 12.6v with mount off and 12.5v with mount running and set for lunar tracking with set pointing at the moon in CdC.

It has been running for half an hour now and it's still connected.  Here's a screenshot from TeamViewer from about 20m ago and 10m after setting the alignment to the moon.  Even with lunar tracking enabled it is drifting off target which is not right really.  However, I have been messing abount with the mount and might have upset the PA a bit.

I was just about to take another screenshot showing that the pointing had drifted off by about a third of the moon's diameter when the rectangle suddenly disappeared, indicating the the connection to the mount had been lost.  I'll just pop out and check the battery voltage - back shortly...

A bit later...  Battery reading 12.3v.  I've connected the PSU again and battery charging at over 3A - that went flat rather quickly for a load of 1.25A   Maybe my batteries are past their best

[Gina]

And we're off again - PSU connected, battery at 13.3v.

[Gina]

Gone again   I suppose I could try separate charging and load wires to the battery.  Or maybe I need a new battery...

OR a new hobby!!

Or maybe more sensible, take the EQ8 off and put the NEQ6 on the pier instead.  See if that works...

[Gina]

This 12V 7.2Ah Universal Sealed Rechargeable Lead Acid Battery looks like a reasonable bet for a new battery.  Nothing like the 45AH capacity of the batteries I'm using ATM but I think adequate as a backup/regulating battery for the mount.  With the rate at which the car batteries are being discharged by the mount I think they've pretty much had it

[Gina]

I did eliminate one other possible cause of the problem and that's the USB port - swapped from USB port on laptop to powered USB hub at the pier head - no difference.  Had to specify the com port as it didn't detect that automatically - no problem, just checked Device Mangler and set the port in ASCOM.

I have a spare USB to serial adapter so could try swapping that.

If anyone has any suggestions please do tell

[Oddsocks]

I am inclined to think it is a comms problem rather than mount power, I suspect if you had the EQ8 handset connected and were using that to control the mount you would not have any issues.

Are you able to run the laptop just using internal laptop battery, it might be there is an unwanted earth leakage loop to earth via the usb cable - laptop - laptop switch mode PSU, and it is this that is causing the USB comms to drop out sporadically.

I know from experience with remote (medical) data printers that we had to ensure they were running on the same hospital supply phase as the source camera in a different part of the hospital otherwise it was impossible to maintain a consistent data flow, you could see it quite easily when scoping the individual RX and TX lines that there would be huge superimposed 50Hz hum signals playing havoc with the data timing pulses even though the leakage current was just a few mA.

Do you have a common earth grounding point in the observatory that your EQ8 power, pier metal work, hub power, camera power and the laptop PSU can use?

William.

[Gina]

Thanks for your reply William

I am inclined to think it is a comms problem rather than mount power, I suspect if you had the EQ8 handset connected and were using that to control the mount you would not have any issues.

Yes, I think you're right.

Are you able to run the laptop just using internal laptop battery, it might be there is an unwanted earth leakage loop to earth via the usb cable - laptop - laptop switch mode PSU, and it is this that is causing the USB comms to drop out sporadically.

Yes, I could try that.  Good thought

I know from experience with remote (medical) data printers that we had to ensure they were running on the same hospital supply phase as the source camera in a different part of the hospital otherwise it was impossible to maintain a consistent data flow, you could see it quite easily when scoping the individual RX and TX lines that there would be huge superimposed 50Hz hum signals playing havoc with the data timing pulses even though the leakage current was just a few mA.

Ah yes, I see.  I have a small hand held oscilloscope (if I can find it ) so could check for hum or other interference.

Do you have a common earth grounding point in the observatory that your EQ8 power, pier metal work, hub power, camera power and the laptop PSU can use?

I haven't got one ATM but I could arrange it.  Just as long as I don't make things worse with earth loops   I'll give this some thought  

Thank you

[Gina]

I think I shall draw up a complete wiring/circuit diagram showing all power lines and all earth lines and connections.  I think things like cameras and filter wheels that have both power and USB connections probably have a common earth/ground/0v.  This could lead to circulating earth/ground currents and interference to the USB data.  Unfortunately I've never seen an optically isolated USB hub.  With data lines of up to around 5 metres between computer in warm room and devices on the pier, there is plenty of opportunity for interference   This includes the mount, of course, though the power to that can be separate from the other devices.

Although I can see all sorts of possibilities for interference, I have been running the mount and all the other equipment quite successfully in the past.  OK not all the time it has to be said - there have been the odd times when the mount data connection has failed, sometimes after several hours of perfect operation.  So I guess you can say that the problem has always been there, just that it's many times worse now.  I have used different mounts, different cameras and even different USB hubs and different computers.

I have done all sorts of things to eliminate problems but all I seem to have achieved is a system that is now totally useless

I will NOT be defeated though!  I shall look into where problems might occur, and the possibility of reducing data and power cable lengths etc.  Maybe I should try having the laptop beside the pier - I could have another in the warm room connected by WiFi and TeamViewer if I want to be near the equipment (as opposed to indoors).

This is now top priority as far as astro projects are concerned - nothing can be done without the mount working properly

[Oddsocks]

As a quick first step, connect and power everything up as if you were working normally but do not plug any usb cables into the laptop, then measure with a voltmeter between the laptop chassis and each usb cable's plug housing, one at a time, and then between each cable. If all the various components are on the same ground reference there should be no measurable voltage difference between laptop chassis and usb cable shield.

I have found so many problems over the years with leaky laptop switch mode PSU's, I'm quite certain half the eq serial interface cable failures I keep reading about on this forum are caused where the laptop is on one switch mode PSU and the mount is on another. Even on laptops specifically supplied for medical use are not immune, I once came across a hospital laptop running an ultrasound package with a firewire ultrasound probe for portable prenatal scanning that had a leaky PSU, with 110v at 20ma leaking down the firewire cable to earth, it provided quite an unpleasant shock for the operator and the patient!

In all the major scanning suites that I worked in we had a standard method of power installation to avoid ground loops and the problems they cause.

All three phases earth and neutral for all power and services in the suite arrived at a common set of bus rails in the main equipment cabinet. Anything with a metal frame or grounded cover was fixed in place with isolating components so that it was not in contact with any part of the fabric of the building, even the medical gas pipes and domestic water pipes were isolated with plastic inserts and the pipework was earthed suite-side to the common earth bus.

We took up to 200A from each supply phase bus for our kit and within our kit all power earths from motors, transformer cores, stab inverter chassis etc were connected to a common earth bus. All 0v lines from the 60 or so different individual power supplies, signal grounds and signal cable shields were connected to a separate ground bus.

The main earth bus, signal ground bus and incoming neutral were linked together by a single copper braid strap at the entrance to our processing racks and all the racks were full EMI/RFI enclosures.

This meant we never had earth loops or external ground paths since any earthed component in the suite was connected to a single ground point. When things did go wrong it was often because the isolation from fabric ground earth of some individual piece of equipment in the room had been compromised in some way, a common failure was water seepage under the floor caused by the scrubbing out of the suites following particularly messy operations and soaking into the concrete, shorting to the steel rebar!

As part of the anual certification tests we had to remove the braid between the earth, signal earth and neutral bus bars and measure the isolation resistance, anything less than 10 Meg Ohm was a failure and the suite would be shut down until the cause found and rectified.

If possible for the observatory to avoid unwanted ground loops or ground paths all chassis earths from all powered equipment should have its own single earth wire back to a single earth terminal and from there a single connection to mains earth, then all 0v lines should be treated the same with a common 0v terminal and a single connection should be made between the 0v and earth connection.

But, as you will know from your years in design, the ideal is difficult to achieve, particlarly where these nasty switch mode PSU,s are used since the 0v of the one for the camera may be floating 50v or more above incoming neutral and the 0v of the one for the laptop may only be 10v, so everything has to be measured carefully before commoning together to ensure any floating voltages are able to be pulled down safely without damage.

I guess that being rural you don't get a local sub station earth and rely on a buried rod for earthing so I wonder if your mount is acting as a possible ground rod, it would vary it's ground path depending on water table levels and dampness of the pier in which case a ground path may exist between your house supply, via the laptop switch mode PSU - laptop - USB cable - EQ8 to real earth!

You could try 3d printing some shouldered isolation washers to fit between the EQ8 mounting bolts and the pier head to make sure there is no possible earth route beween EQ8 and real ground (we used to use 4mm thick GRP ones in medical installations but sadly I threw all that stuff away when I retired).

Another possibilty, I wonder if you know any local electricians that can loan you a site-safety isolating 240 - 240 transformer to supply the observatory as a test, this would isolate the neutral and live of the house supply from the observatory supply interrupting any possible supply leak route via the switch mode PSU's.

Lot to think about.......

William

[Gina]

Thank you for all that William   Certainly food for thought

You're right about the earth leakage path through the mount, metalwork and concrete pier - I hadn't thought of that

There is currently only one USB connection to the laptop - everything is fed from a powered USB hub next to the scope, but I can check that.

I know that laptop PSUs can be leaky - my MacBook Pro is a prime example, I can feel the leakage by just touching the boby of the laptop.  I would have expected better from such a top-end machine - they cost enough!!

Yes, the mains supply is just two wire with a copper earth rod outside the kitchen with a heavy(ish) cable to the power distribution earth connection.

I don't know of a local electrician (though I could enquire) but I could run everything off batteries with no mains connection as a test.

I thought of power line problems but didn't really thing of earthing - tut tut - not good enough   I'm sure I would have thought of this many years ago - I guess my age is showing

My word, there's certainly much food for thought here    Thank you again for your help William - much appreciated

[Gina]

Here's a block diagram showing the mount power and data connections.  The ground of the 5v regulator chip is actually physically connected to the mount via the dovetail - this is probably NOT a good idea    I have also shown earth leakage through the pier.

[Gina]

I can disconnect the power supplies and run everything off batteries, and also isolate the 5v regulator tab from the dovetail.

[fondofchips]

One of the many Obsy things that makes you scratch your head but when sorted is part of the pleasure of the project.  I asked a experienced workmate at work explaining what I wanted to do & the best way to do it.

My spec - Decent power supply on top of the job with battery back up & capable of working for long sessions during a power cut.  (Absolute bliss, scope working & all the LP is switched off).

I could get a used 180ah diesel engine battery from work so that was sorted.   I went for a battery conditioner like this one: - http://www.thegreenreaper.co.uk/battery-chargers-optimisers/tecmate-battery-chargers/accumate-pro-12-24v-battery-charger-maintainer

The battery is outside, the cold does not affect it when connected to the conditioner.  I have just switched on all my loads in the obsy one by one & it didn't drop below 12.6 volts, settles at 13.8 volts with no load on it, highest 14,5 volts.  The conditioner kicks in even if a only switch on one of my 12 volt Red LED strips.  The important thing here is not to match a battery conditioner with a battery that is too large for it.  An underpowered conditioner simply won't keep up with the load, a drop in voltage can damage mount motors, as the current tends to increase when the voltage decreases.  Your current meter is a good idea.

Cheers,

Fondofchips.

[Gina]

Yes, a nice bit battery would be good but I can't afford it ATM.  If I can get my gear working from mains power I shall be a happy bunny   A regulating/smoothing battery would be possible though.

[Gina]

Been thinking and made my mount testing arrangement even simpler.  I don't need any other USB device than the EqDir USB to serial adapter cable which I can plug directly into the USB cable from the warm room.

[Gina]

Testing...

GONE!!

[Gina]

Battery voltage 12.3.

Connected PSU - 12.4 and rising.  Now I'm just going to get a late lunch - been out this morning...

Gone again