Duel mains supply for srvo project

[sid]

I want to build around 15 small pcb circuits each with an 08m2 on board which will control an individual servo.
I was thinking that it might make sense to have one power supply for all the picaxe circuits that is regulated and smoothed and another power supply feeding the servo's. The supply for the picaxe's probably won't draw much current but I'm assuming that the one for the servo's will need to be able to supply several amps as there might be more than one servo operating at a time.
I was thinking along the lines of using a couple of transformers to achieve this, but I've never built a power supply circuit before and I'm guessing its probably not as simple as 240v into a transformer and 5v out of the transformer.
I realise that protection in the form of fuses either side is a must but I would be grateful if anyone could advise or share a circuit diagram.
Rgds,
Sid

 

[MartinM57]

The days of building your own mains power supplies are over really - just search out a good quality 5v wall-wart/laptop charger type supply and you will get there quicker, probably better, and much more safely

 

[hippy]

The safest and recommended course is a pre-built and certified power supply from a reputable manufacturer. 'Power bricks' for consumer products, games consoles, USB hubs, routers, even computers can often be found for 5V with quite high amperage. A PC PSU is probably the most ubiquitous source of high current 5V supply.

You should still take care with any high current power supplies.

 

[sid]

The safest and recommended course is a pre-built and certified power supply from a reputable manufacturer. 'Power bricks' for consumer products, games consoles, USB hubs, routers, even computers can often be found for 5V with quite high amperage. A PC PSU is probably the most ubiquitous source of high current 5V supply.

You should still take care with any high current power supplies.

Thanks for the advice

 

[MFB]

With PICAXE decoupling capacitors and good ground wiring it should be possible to power everything from a single regulated supply. After all, model aircraft power the servos and very sensitive radio receivers from the same battery.

 

[srnet]

After all, model aircraft power the servos and very sensitive radio receivers from the same battery.

Indeed they do.

All the servo connector header pins are normally in one block, with all the grounds and +5V connected together. So the layout is in effect single point ground and +5v.

 

[bluejets]

Indeed they do.

All the servo connector header pins are normally in one block, with all the grounds and +5V connected together. So the layout is in effect single point ground and +5v.

Yes, but good also to point out these receivers also have evolved with sophisticated noise rejection circuits (JR ABC&W system for example)

 

[westaust55]

While the use of a single supply for the PICAXE chip(s) and servomotors may well work, there have been a number of recent posts/threads where there were problems with the PICAXE resetting, etc which were resolved once the servomotors were provided with a separate supply.
Where separate supplies are used, do not forget to conenct to Gnd/0 volt lines together.

 

[MPep]

While the use of a single supply for the PICAXE chip(s) and servomotors may well work, there have been a number of recent posts/threads where there were problems with the PICAXE resetting, etc which were resolved once the servomotors were provided with a separate supply.
Where separate supplies are used, do not forget to connect to Gnd/0 volt lines together.

At the end of the day, it still comes down to good decoupling, and providing a good (read: large) capacitor as a localised reservoir for current for the PICAXE. I would use a diode to suply a cap that is wired across the PICAXE. When the voltage line sags a bit because of the servo, the PICAXE has enough of a current reservoir that it will kep going without reset.
My 2c worth anyway

 

[Dippy]

Absolutely. Though be careful with 'large' capacitors. 'Big' can be good, but 'monster' could result in problems.
Throwing a hundredweight of fat caps randomly at a circuit is NOT the best way of doing something.

This kind of thing is crying out for the long overdue 'Tutorial' section.
Maybe someone could post an article as a 'blog' (ghastly phrase)?

Where you have to share a supply, the things to consider are:-
1. (As said a million times) Good decoupling/bypassing for microC section.
2. Local regulation, decouple and reservoir for sensitive uC and other bits.
3. Careful (star style) wiring or track placement for power.
4. Ditto for ground.
i.e. minimising the tracks/wires having to 'share' power supply/return currents with noisy devices.
And this should good practice when ADCing too - so dump Autorouters right now
5. Transient/noise attenuation inductors/voltage suppressors/filters etc. on sensitive power lines
6. Physical separation or screening of tracks and/or wires.
e.g. physical spacing or metal shielding or use of planes on multi-layer boards.
7. Similar consideration for high impedance signal and analogue lines.


I'm sure there are more things and sometimes it's a bit of a 'black art'
But, something that is NOT a 'black art' is component choice.
For example, a great fat electrolytic will NOT decouple/bypass HF as effectively as a good ceramic cap.


It would be nice if one or two of the big brains would do some nice basic schematic sketches to show a couple of examples of a good way to attack this (all too common) problem.

 

[Buzby]

Maybe someone could post an article as a 'blog' (ghastly phrase)?

Please NO !.

Not a blog, or that other pile or random meanderings, the Wiki.

If there is to be a tutorial section, which I fully agree with, keep it fixed structure and well indexed.
Siomething like the BASIC commands by categories at the top of this page.

 

[Dippy]

It was the only section I could think of that could be up-and-running today in a PICAXE-specific place.
I know what you mean Buzby, so if you can think a better place, where it doesn't require difficult searching actions in 2 months time, then please suggest it. Maybe a normal post - with a clear (and correctly spelled/spelt) title.
Someone (sorry Technical) can organise it later.

The subject of a Tutorial section and it's format (and reviews,checking, listing etc.) has been discussed at length before and I didn't really want to go off at a tangent in this thread...

 

[jedynakiewicz]

Whatever the format, I agree with Dippy that such a tutorial or set of guidelines would be invaluable. Since I started working with PICAXE for a range of projects involving motors, servos etc., I think that the No.1 issue that I had was with the power supplies. Unexpected resets due to power drops with motors cutting in; interference from poorly-supressed motors, I have had it all. It required many hours ploughing through the "knowledge-base" that is this forum in order to understand where I was making my mistakes. One by one my problems were solved. For me, the best place to have this information would be as a section in a revised Picaxe Manual 3. Of all the manuals, this is one that most seems to need updating anyway. The existing power supply information is very basic and could be so helpfully updated for anyone using Picaxe for robotics, modelling etc. by a paragraph on each of the aspects highlighted by Dippy.

With regard to the original question in this thread; yes, Sid, I very much favour a separate power supply for the Picaxe chips; I have found it to be a very reliable way of designing the power supply. I use mains transformers with separate windings and run a dual 5v supply to the boards; one for the Picaxe and one for the servos. With regard to the actual design, it very much depends upon the current needed to drive fifteen servos; if they all initialise at once upon power-up then there will be a correspondingly huge current surge. Thereafter, it will depend upon how many servos are actively moving (or actively resisting movement) simultaneously. Rapid sell a range of transformers with separate windings in a range of VA outputs - you need to calculate your current requirements. I use dual output 7.5 v windings which, when rectified, give me around 8.5vDC that can then be smoothed with big caps and then stabilised down to 5 volts with 7805 regulators or their higher current derivatives. I use separate fuses in the 7.5v AC lines to keep things sensible. I have found this to be a really reliable way of powering things.

As Hippy and others have mentioned, safety is the order of the day and building your own power supplies is not without potential hazards if you don't know the issues. Nobody wants to see a forum member electrocuted or transformers smoking with shorts in unprotected circuits! Particular attention needs to be given to the casing of the unit so that it not only provides appropriate insulation but also is robust against damage if it falls off a work-surface (drop test!). Transformers are heavy and need to be firmly attached to the casing to avoid displacement in event of a fall. Given these aspects, one can see why the recommendation is buy things-ready made.

 

[Eclectica]

Arguably the power supply topic is bottomless, but not a bottomless pit!

A successful approach with some of my recent designs have incorporated a single 24V DC switch mode PSU main feed, which also feeds Recom 24-05 DC/DC converters. [fed through a small inductor with filter capacitor to common 0v]. The 5V out of the DC/DC is rock steady and feeds the PICaxe. The 24V is used for switching relays and fans etc. via MOSfets.

There are many approaches!

Regards
E.

 

[John West]

The days of building your own mains power supplies are over really -...

They are? Dang, I didn't know that.

 

[Dippy]

..and is a "duel" power supply where two PSUs fight each other?