Interesting PSU behaviour

[powerlord]

Posting in case it's of use for anyone else.

I use a decent quality PC PSU to control both my mounts. I have this setup with a few power resistors to load the PSU, and run 12v/5v to both of my EQ6s which have asiair, cooled camera, eaf, 5v heated dew shields, etc.

It's all been working flawlessly for over a year.

I've now moved it all into the new observatory, and part of the move meant extending the 12v/5v run from the PSU to the mounts by a few meters. I knew I might have to contend with a voltage drop, but as things all powered up I thought it was ok.

However, in the seconds between clouds and rain over the last week or so I've been trying it out and was getting weird behaviour on the furthest away mount - disconnects from it (its on ethernet), and non guided  images showing star trails even with 10 sec exposures.. then last night I noticed when doing a goto the EQ6 was acting weird - it moved just the DEC, then the RA rather than both at the same time.

So today I did some testing - and sure enough - I was getting maybe .1- .2 of a volt drop over the 2m run, but still just above 12v at the mount (12.15v)... BUT.. put some load on it - crank the temp down on the camera, etc.. and that drops to 11.8v .. crank cooler up more, and then try a goto - and it starts the same behavioour - only moving one axis at a time - and the voltage at the mount (i.e. into asiair) has dropped to just above 11v!

So there's yer problem - I mean.. I've got a degree in electronic engineering, so this is hardly a surprise, except sometimes you forget the simply things... sigh.

Anyhoo, clearly my PC PSU is totally unsuitable now to drive my mounts at a distance and will need to be replaced. I've got 12v 5v metal enclosure PSUs with pots on them, so can wire up these and adjust the voltage to allow for the drop fine, but I do worry that the varying current load is still going to play havok with the voltage over the cable. e.g. if I set the voltage at the PSU to 13v, this might get me a nice 12v still at 5a.. but when that load goes down to 1a that voltage is going to be up to 12.8v or so...

And from what I've seen, these bits of kit are not exactly well designed - no stabilizing or boosting voltage.. so I'm unsure just how high I can push that no load voltage safely...

The alternative, is that I fit voltage stabilizers (aka voltage boosters) at the mounts. this sort of thing:

 

https://www.ebay.co.uk/itm/203591993751

This should be able to keep the voltage a steady 12v in dependent of the drop in incoming voltage, and independent of current draw - really making it the same as a 12v psu more closely connected to the mount. that's the theory.

I'll let you know if it works.

[Stuart1971]

Not sure why the voltage should drop, just because you are drawing more load, assuming your not going over the recommended load on the PSU, I have a 15 amp PSU and no matter if my kit is drawing 3 Amps or (exaggerated) 13 amps, it still produces 13.8v at the Pegasus power hub….

You will deffo get a voltage drop over longer cable distances, but that drop should be constant….but you will know all this if you have an electrical engineering degree, so I am not trying to undermine that….👍🏻

[Dr_Ju_ju]

Check that your connections are solid & have no tarnishing.... 

[powerlord]

voltage drop across a length of cable is current dependent. more current - more drop. not constant. V = IxR at the end of the day. R is yer cable resistance here, I is the current, and V is the drop across it. more current, more drop.

All my connections are soldered or solid plugs, BUT I did use single core 1.5mm house 3 core for the 5m run to the far mount as I needed something flat to fit under the laminate - I suspected this would not be ideal, though it is pure copper.

this calculator can show you roughly:

https://www.calculator.net/voltage-drop-calculator.html?necmaterial=copper&necwiresize=0&necconduit=steel&necpf=0.85&material=copper&wiresize=0.4066&resistance=1.2&resistanceunit=okm&voltage=12&phase=dc&noofconductor=1&distance=5&distanceunit=meters&amperes=6&x=53&y=27&ctype=nec

with 6a across 5m of 14awg it's showing 0.5v drop (vs .04v drop at 1a.

At 10a, it can be a full 1v, or more if some of the wire is aluminium (don't think it is).

However what I'm seeing is worse than that - I suspect either my wiring e2e IS higher resistance than I thought, or the PSU just isn't handling load well (but it used to do it fine), or it's drawing more current at least in a spike - that's what I suspect as I could see the drop in voltage as the mount starteed to move, and then it recovered a bit.

At 10a, it would have gone down to 11v.

So my 2 options are:

1. replace pc psu with a 12v and 5v psu, with the voltage pushed up a bit.. BUT the voltage reaching the mounts will vary with load due to the long cable run.

2. add voltage stabilizers close to each mount (in base where cables comes in).

It's interesting that you say @Stuart1971 that you are using 13.8v though - thats into mounts, cameras, etc ? That suggests I can push voltage on option 1 up quite a bit. For your setup, if you only have a small run between your PSU and hub, and it's good quality thick cabling the drop can be minimal though.

I've got two aircraft type 3 pin sockets between the psu and the mount, and if there's any extra resistance, it's probably there - though the ones I'm using are rated for 20a, I noticed they looked like they were made of quite thin mystery metal. possibly replacing these with XT50s or something would be useful.

Before I do 1 or 2 though, I'll measure the resistance of the wire in case there's a quick win as you point out, and it is just a rubbish connector somewhere. 👍

Certainly, I tried replacing the supply to the mount with a regular 12v power supply I had lying about, and that does a good job of staying at 12.10v or so, dropping to 12v under lots of current initially (camera running 60% cooling, and then move mount), then recovering a bit to 12.05 or so).

The main issue comes down to me using a PC PSU I think - I've got away with it for now, but the output of those is 12v on the nose - monitoring and controlled to be exactly 12v by the onboard controller. Whereas really, for our uses, as you have pointed out, it's better to have something a bit higher - 13v or so.

stu

 

[Carbon Brush]

In a lot of PC type power supplies, the 5V (high power) output is well controlled. The 12V output is not well regulated and relies on a certain load on the 5V provide a sensible output.
This is probably where your mention of resistors comes into play.

I don't think the problems are voltage drop as such, but the PSU not responding well to the reactive load of long cables. The designer was thinking in terms of a few cm within the PC case.

In my observatory I use a decent (not PC) quality supply generating about 13.5V, mounted on the wall. This connects to the AWR motor drive on the pillar.
However, at the pillar I have electrolytic bulk + suppression capacitors. These were included on day 1 as a precaution. Never had any issues.

HTH, David.

[powerlord]

Good idea. Yes, I have a 10 ohm resistor on the 5v supply and another on the 12v supply. I did replace the 5v one during the move... I'll need to check what th old one is.. Maybe I've made it too low a load.

Back in the beginning, I made up a wee 5v supercapacitor box for USB dew heaters connected to USB sockets, as they pulse on and off even on low, and it was tripping the USB. The supercaps smoothed that off. Not really a solution here unfortunately.

But yeh.. I think as the runs are longer the pc cpu just isn't a very good fit.. Oh well. 😢

The main reason for posting though was to describe that eq6 behaviour as I'd not seen it mentioned before. Its a good indicator that voltage is low.

What value caps have u used?

Edited March 22, 2023 by powerlord

[Carbon Brush]

By all means try a few thousand uF for the bulk storage. It is easy to get carried away trying to overdesign when you don't know the reactive load, or the PSU characteristics.
But actually having a PSU with a main output 12V (or a bit more) is a much better idea.

I use a TDK DIN rail mounted supply, screwed to the wall with a cover over the mains terminals.
Other brands and mounting methods are available😄

[Stuart1971]

52 minutes ago, powerlord said:

voltage drop across a length of cable is current dependent. more current - more drop. not constant. V = IxR at the end of the day. R is yer cable resistance here, I is the current, and V is the drop across it. more current, more drop.

All my connections are soldered or solid plugs, BUT I did use single core 1.5mm house 3 core for the 5m run to the far mount as I needed something flat to fit under the laminate - I suspected this would not be ideal, though it is pure copper.

this calculator can show you roughly:

https://www.calculator.net/voltage-drop-calculator.html?necmaterial=copper&necwiresize=0&necconduit=steel&necpf=0.85&material=copper&wiresize=0.4066&resistance=1.2&resistanceunit=okm&voltage=12&phase=dc&noofconductor=1&distance=5&distanceunit=meters&amperes=6&x=53&y=27&ctype=nec

with 6a across 5m of 14awg it's showing 0.5v drop (vs .04v drop at 1a.

At 10a, it can be a full 1v, or more if some of the wire is aluminium (don't think it is).

However what I'm seeing is worse than that - I suspect either my wiring e2e IS higher resistance than I thought, or the PSU just isn't handling load well (but it used to do it fine), or it's drawing more current at least in a spike - that's what I suspect as I could see the drop in voltage as the mount starteed to move, and then it recovered a bit.

At 10a, it would have gone down to 11v.

So my 2 options are:

1. replace pc psu with a 12v and 5v psu, with the voltage pushed up a bit.. BUT the voltage reaching the mounts will vary with load due to the long cable run.

2. add voltage stabilizers close to each mount (in base where cables comes in).

It's interesting that you say @Stuart1971 that you are using 13.8v though - thats into mounts, cameras, etc ? That suggests I can push voltage on option 1 up quite a bit. For your setup, if you only have a small run between your PSU and hub, and it's good quality thick cabling the drop can be minimal though.

I've got two aircraft type 3 pin sockets between the psu and the mount, and if there's any extra resistance, it's probably there - though the ones I'm using are rated for 20a, I noticed they looked like they were made of quite thin mystery metal. possibly replacing these with XT50s or something would be useful.

Before I do 1 or 2 though, I'll measure the resistance of the wire in case there's a quick win as you point out, and it is just a rubbish connector somewhere. 👍

Certainly, I tried replacing the supply to the mount with a regular 12v power supply I had lying about, and that does a good job of staying at 12.10v or so, dropping to 12v under lots of current initially (camera running 60% cooling, and then move mount), then recovering a bit to 12.05 or so).

The main issue comes down to me using a PC PSU I think - I've got away with it for now, but the output of those is 12v on the nose - monitoring and controlled to be exactly 12v by the onboard controller. Whereas really, for our uses, as you have pointed out, it's better to have something a bit higher - 13v or so.

stu

 

Oh, yes I have never used 12v, many mounts require at the very minimum 12v, my EQ8 works from 11v up to 15v, and the higher the better, most EQ6 mounts will run perfectly fine on 13.8v, in fact many people say it’s a must…the Pegasus hub requires a minimum 13.8v too

All my QHY268 cameras also use the 13.8v from the Pegasus power hub too, my understanding is that the cameras will take up to 15v too…

i have approx 2.5m between the PSU and my mount, with the Pegasus hub at 2m and then 0.5m from that to the mount….

My PSU puts out approx 14.1v and it’s approx 13.8v at the Pegasus hub…

Edited March 22, 2023 by Stuart1971

[EarthLife]

Yes the 12V outputs from the PC PSU's I've seen and used are not regulated, it's the main 5V or 3V output that's regulated, the other voltage outputs will bop up and down as they please.

And yes adding a high capacitance (low ESR) at the far end (the load) can certainly help as well with noisy loads like motors. The electronics in these consumer type scope mounts are very cheaply made, they are most definitely not over engineered etc, they can be very picky and have little to no protection against certain conditions.

Get yourself a proper 12V to 14V PSU that is regulated and can handle at least twice the peak current (it will last longer if you do).

[EarthLife]

One way is to have a 12V lead acid battery as your supply, keep a charger (constant voltage PSU etc) attached to it (and powered) that will maintain a constant 13.5V at the battery terminals (lead acid maintenance voltage).

 

[powerlord]

40 minutes ago, EarthLife said:

Yes the 12V outputs from the PC PSU's I've seen and used are not regulated, it's the main 5V or 3V output that's regulated, the other voltage outputs will bop up and down as they please.

And yes adding a high capacitance (low ESR) at the far end (the load) can certainly help as well with noisy loads like motors. The electronics in these consumer type scope mounts are very cheaply made, they are most definitely not over engineered etc, they can be very picky and have little to no protection against certain conditions.

Get yourself a proper 12V to 14V PSU that is regulated and can handle at least twice the peak current (it will last longer if you do).

Yeh I've got plenty of them, just need it fit one.

[DaveL59]

Other thing to consider with a PC PSU, the 12V isn't really geared to serve a high current load, it only spins a couple fans and HDD's maybe an optical after all. A large server type with capacity for several fast HDD's on the other hand may suit tho they'd be expensive and even then a dozen HDD's isn't that high a current load.

[powerlord]

So, made up a new 12v/5v supply last night. Will test today.

@DaveL59 - maybe in the distant past.. these days several 100 watts are used by the GPU alone on the 12v line. Out of the 480W my old PC PSU is rated for, the vast majority of that is on the 12v. Modern mobos pull it off that and regulate it themselves.

[powerlord]

tested new PSUs - working a treat now. My cables all tested at nice low resistance (0.1-0.2 ohms). I set the PSU to 14v, and the 5v to 5.5v.

Even under full load its now looking nice and healthy 🙂