Advice Needed V-Box blew up :(

[Archived-1]

Looks a great job Evan well done

 

[Vipersan]

Last photos ..
There it is complete ..
Hopefully I'll get to move Christina again soon..
;)

btw ..have changes been made to photo upload ??
Can only highlight and upload one at a time now ?

 

[Captain Jack]

Very nice! You should stick some labels on

 

[Vipersan]

Very nice! You should stick some labels on

Yep ..
Been tidying the shop for an hour or so ..but so far not uncovered my dymo ..so it will have to wait..
lol

 

[Captain Jack]

VS, how's the heat generation from the power supply?

 

[Vipersan]

VS, how's the heat generation from the power supply?

Negligible in standby/idle mode ..
No doubt it draws a bit more running a motor But my only test thus far is with an actuator connected but static ..and left switched on overnight ..
Barely warm to the touch on the psu ..
The most significant heat was from the 5v regulators ..which is why I moved them off pcb and mounted on a seperate heatsink.
rgds

 

[Captain Jack]

Hi Evan, need your opinion.

So, in order to do away with manually switching each board on and off, I've come up with this idea.

On the little used declination V-Box, I am planning on keeping it turned off by default when the receiver is powered on. For this, I need some sort of a circuit to "push" the power button as soon as the receiver is switched on (or a second later) to turn the box off. I already have designs for the circuit but I am wondering what's best.

The first design is that the circuit is powered by the 5V line from the V-Box and triggers a relay for half a second to turn the V-Box off when the power appears on the receiver/LNB line. This then stays dormant until the next time receiver is switched off and back on.

The second design the whole circuit is powered by the recever/LNB line and, again, triggers a relay to turn the V-Box off.

My personal preference is the former - have the circuit powered up from the V-Box power supply rather than the LNB line itself. Hopefully this would introduce less noise into the system etc.

What do you think?

 

[Vipersan]

Sounds like a plan Alex ..
Look forward to seeing what you end up with ..
Like you I prefer the former..

Now ..
My build has undergone a few changes ...
I've added extra fuses ..
..an 8amp fuse in the mains line ..
But this will have to be upped to 10 amp in light of what I intend to do next..
Space may be an issue..
I also have 4 amp fuses in each 24volt power line to the relays on each Gbox pcb ...
Just in case of motor failure or indeed excess loading in the event I accidentally drive to the actuators limits.

So far the dual positioner is working ok after a fashion..
The 24v psu now uses a later version of the one I originally used but is physically bigger..
Again ..cut to fit in the space (same ratings 8.3Amp)
BUT ...noise IS an issue ...
When I say noise I mean electrical noise..
Spikes etc as the relays switch ..
and this has caused me a few days of head scratching ..and no I didn't find an easy solution.
As you know I simply drive the 24v line to a 5v regultor on each Gbox pcb ..and mostly this works ...but occasionally during a switch of direction ...or start up/shut down of motor movement ...the Gbox electronics goes dark ...
Sometimes simply to standby..
Sometimes requiring full power down then up again..
Sometimes erasing ALL sat position data..
I suspect this is why there is a seperate winding on the transformer and a thus a seperate 12v ...stepped down and regulated to 5v for the electronics..
So ..I hope to emulate this by seperating completely the electronics from the motor drive ..
to this end I've ordered a small 5v regulated load protected switch mode power supply from Hong Kong..for £4 ..
Small footprint and small in price ..
I just hope it works ok ..
It will mean there will be no common ground ....no shared supplies ...and hopefully NO spikes and intermittent shut downs ...and the electronics will then be assured of a constant uniterrupted 5v .
rgds
VS
This is very much ..still a work in progress..

 

[Vipersan]

This is the space to the left of the Gbox pcbs ..in which I hope to squeeze the 5v psu ...
Edge on..
we shall see....

 

[Llew]

BUT ...noise IS an issue ...
When I say noise I mean electrical noise..
Spikes etc as the relays switch ..
and this has caused me a few days of head scratching ..and no I dsn't find an easy solution.

Protection diodes across the relays VS?

 

[Captain Jack]

I am grateful that it's you being the guinea pig here.......

 

[Vipersan]

Protection diodes across the relays VS?

Already fitted Llew ...and along with some 10uH inductors in the supply line and ground line to the 5v regulators ...it improved the frequency of the resetting ..but not completely.
So a decision was made to fully seperate and isolate the motor supplies ...and the 5v supplies to somewhat sensitive control circuitry.
If I get a reliable fix ..the extra cost and additional build time will be well worth it.

 

[Waveguide]

I just found my copy of National Semiconductor's "Voltage Regulator Handbook".
(why do things you need disappear and return again when not needed)
A chapter deals with automotive power supplies and mentions the real specification
of normal voltages in a car, +80 V to - 50 V and not +12-14 V as you might expect.
Anyway, the cure is quite simple with a few well chosen extra components to
limit the voltage spikes. A resistor, a diode, a zener-diode and a capacitor makes
life easier and a normal voltage regulator won't see the horrible stuff on the input
voltage. Examples show 12 to 5 V regulator applications and higher input voltages get the
same treatment.

 

[Vipersan]

I just found my copy of National Semiconductor's "Voltage Regulator Handbook".
(why do things you need disappear and return again when not needed)
A chapter deals with automotive power supplies and mentions the real specification
of normal voltages in a car, +80 V to - 50 V and not +12-14 V as you might expect.
Anyway, the cure is quite simple with a few well chosen extra components to
limit the voltage spikes. A resistor, a diode, a zener-diode and a capacitor makes
life easier and a normal voltage regulator won't see the horrible stuff on the input
voltage. Examples show 12 to 5 V regulator applications and higher input voltages get the
same treatment.

Any chance you could scan the relevant page(s)...with perhaps a sample schematic which is often given..?
rgds
VS

 

[Waveguide]

No problem at all, I just have to haul four carloads of furniture (2 m3 each) from
the hallway so I can get upstairs to the scanner. (IKEA sale)
It's three pages in all, schematics and explanations, I'll try jpg-images here first.
A pdf-document to some mail-adress if the jpegs don't work?

(The hauling of boxes will be fixed later today, if not I will starve, blocked the kitchen too.)

 

[Llew]

To save Waveguide the hassle (and backbreak) VS, here's a pointer to the (very large) pdf -

https://ia800909.us.archive.org/13/...SemiconductorVoltageRegulatorHandbook1980.pdf

 

[Waveguide]

... that's the one!
Pages 7-32 to 7-35 and figure 4 to be exact.

Now I will have a beer to celebrate that I created my own indoor Stonehenge.
The boxes are upright now in a row like the monoliths.

 

[Vipersan]

So ..further experimentation this morning..
Inductors having already been tried along with increasingly larger electrolytic capacitors..
..and eventually transient suppresion along the lines of those outlined in the SVRH 1980 pdf ...
All failed to produce trouble free results.
After much head scratching earlier in the week...I could not understand why the designers of the G/V box use transformers with seperate windings for Motor and control supplies ..and why they simply didn't use one DC supply stepped down for 5v.
..and why I kept getting resets and unpredictable results.
I finally had the idea that transients were causing the problem and that these affected the sensitive control circuitry.
In the end I tried leaving the ground line common to both halves of the circuit but used a seperate 12v bench psu to feed the 5v regulators..
much as the transformer design did,..
Still I had issues ??
But as I suspected these transients were being generated on the ground line as well ...
Isolating and seperating both the 24V motor ...and 12v > 5v ....lines both DC source and crucially GND lines has so far produced results which are consistant and NO re-sets ..
fingers crossed this isn't going to suddenly change.
Conclusion..
I will go ahead and fit a totally seperate psu and supply for the control circuit as planned.

There may well be another way to approach this issue ...but this is by far the easiest.
rgds
VS

 

[Vipersan]

...What is really needed is a switch mode psu with seperated secondary windings providing 2 regulated outputs at 24V at 6 amp and 5v at 2 amp with no common ground ...
Not an off the shelf item sadly ...so a compromise has to be reached....or design and build the whole dual AZ/EL unit from scratch.
Not a prospect I relish..
and why re-invent the wheel unless it was a commercial enerprise ?
;)

 

[Waveguide]

A bouncing ground in the supply to logic circuits is just as bad as spikes in the +5 V line.
Drawing a schematic of all the ground connections with all wires replaced by small resistors
can sometimes help to understand what parts are common between the 24 and 5 V
parts of the circuit. V-shaped lines are nice if they start from the actual power source.
The large filter capacitor is such a source. Y or X-shaped lines will make voltage drops
visible on the 5 V part too, they share parts of the wiring.

Another common solution is to add an isolated DC/DC-converter to get lower voltages
from a 24 V supply. There are many that have an input voltage range that includes 24 V
and has a 5 V output. You can connect the grounds or leave it floating. Depending on
what components are connected to both 24 and 5 V the ground can be left floating or
connected with a resistor to keep the voltage difference within limits. Relays and optocouplers
don't mind the voltage difference. (in Sweden they cost around 30£ (for 10 W output)
from a supplier that is considered to be expensive)