Freerails Home 

Freerails IS ACCEPTING new Members ... To join Freerails ... See how to Register as a Member in the 'Joining Freerails' Forum

 Moderated by: .  
AuthorPost
Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi.  

Batteries are ok and give a good run time of a few hours, 
however require an on/off switch and a specialised charger.


In comparison, super-capacitors hold less charge,
but can be recharged in seconds, and no on/off switch required.

A short section of track powered with a 12V plug adapter,
re-charges the super-capacitor via the loco pickups.


I have a test track with a 1 metre section of powered track,
the rest of the track is un-powered. 

The loco charges the super-capacitor in seconds, on the powered track.
The loco will run up to 25 metres on an un-powered track.   

There is no waiting time for a charge.


The super-capacitor supplies a boost circuit, which outputs a constant 12V,
until the super-capacitor is discharged to approximately 3V.

The circuit is simpler than the battery circuits.

1. No switch
2. Fast charging, less than 5 seconds
3. Cheap charger, 12V plug adapter
4. Only 1 metre of track to keep clean

I have used 4 of 15 Farad 2.5V super-capacitorss in series,
to give 10V to the boost circuit.

Power from the rails is bridge-rectified to charge the super-capacitorss.


Photo attached shows my first prototype.  
It will be much tidier on a PCB.





Chris Morris
Registered
 

Joined: Mon Feb 15th, 2021
Location: United Kingdom
Posts: 13
Status: 
Offline
 
Not sure on capacitors.

Now, I only have experience with mains voltage capacitors,
& they can really go with a bang if given a feed of over voltage.
I have also had a few shocks off mains voltage versions.

But the more informed will give you advise on the pros & cons.


Michael M
Registered


Joined: Thu Jan 26th, 2017
Location: San Bernardino, California USA
Posts: 1851
Status: 
Online
 
Steve,

Looks interesting. 


Would like to see more.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi,  

Thanks for the feedback.

Supercaps are not too bad,
providing they are not overcharged too much.
  
Like all things electrical,
they need to be treated with a bit of respect.

I have not had any problems to date.

I will post more of my findings,
as I experiment with the circuit.

Regards

Steve


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi Michael

Thanks for your interest, 
early days.  

I will post updates as I do further testing. 
The next step is to get a prototype PCB.
 
Regards

Steve


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
I have been using Supercaps to overcome dead spots "for ages"
(that is since 2009, and using 6S@13V for a garden railroad)
and have several TT scale locos that run off a 5.5V Supercap.

The voltage is stepped up to 9V with a 'Pololu' in case of contact loss.
No capacitor problems encountered for years!


Addendum:
Just one of the circuits I employ:





Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi.

Excellent,
all we need now are some super supercaps.

Regards

Steve


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
A short video of a test I made in 2009.

The loco is using an analog supercap-based buffer.

The difficulty is to avoid lags in reaction time under normal operation conditions,
making the circuit much more complicated than the one for RC/digital.


Rod Hutchinson
Registered


Joined: Fri May 8th, 2009
Location: Mooroolbark, Australia
Posts: 544
Status: 
Offline
Helmut wrote: 
The voltage is stepped up to 9V with a 'Pololu' in case of contact loss.
No capacitor problems encountered for years!


Helmut

Is the input voltage to the Supercap from Track or Battery?


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi,

Thanks for the video, 2009 ... early days.

Supercaps are great.

If some of the new graphine types,
under development become available,
then it will be a game changer.
 
Interesting times we live in.

Regards

Steve


davecttr
Registered
 

Joined: Sat Nov 9th, 2013
Location: United Kingdom
Posts: 540
Status: 
Offline
Rod Hutchinson wrote:
Helmut,

Is the input voltage to the Supercap from Track or Battery?


I imagine the power is from the track,
as if it was from a battery, there would be no 'dead spots' to worry about!

The idea is very interesting though, well worth watching.

There is also a topic on RMWeb,
where someone is experimenting with using supercaps in DC locos.

The only things on that diagram I understand are the RX6 and the 'Pololu'.





Rod Hutchinson
Registered


Joined: Fri May 8th, 2009
Location: Mooroolbark, Australia
Posts: 544
Status: 
Offline
 
davecttr wrote:


" I imagine the power is from the track,
as if it was from a battery there would be no 'dead spots' to worry about! "


5v seems an unusual voltage. 

Is that normal for TT scale?

Rod


fallen
Registered
 

Joined: Wed Apr 3rd, 2013
Location: United Kingdom
Posts: 204
Status: 
Offline
 
It looks to me as if the power for the RX6 receiver and motor comes,
either from the track if it is powered, via the 1N4001 diode at the top,
or from the supercap via the 'Polulu' if there is no track power.

There is no battery, just the supercap.


The 5v is because the max voltage the supercap will take is 5.5v
The 7805 regulator reduces the track voltage to 5v, to safely charge the supercap.

The 'Polulu' steps the 5v from the 7805 regulator and supercap, back up to 9v,
to feed the RX6 receiver and the motor.


An interesting idea, worth watching.
Must get some supercaps!


Frank


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
The track voltage is a constant 12 to 14V DC from a switch mode power supply,
or should you be so minded, 15V DCC.

In my case, I operate the RC locos together with the standard DCC on module meetings,
in order to be independent of tethered walkaround controllers and DCC address management.

The supercap's only function is to keep the loco alive at contact loss.

Battery operation is covered here


Rod Hutchinson
Registered


Joined: Fri May 8th, 2009
Location: Mooroolbark, Australia
Posts: 544
Status: 
Offline
Helmut wrote:  
The track voltage is a constant 12 to 14V DC from a switch mode power supply,
or should you be so minded, 15V DCC.

The supercap's only function is to keep the loco alive at contact loss.


I misread the circuit. 

On closer inspection it now makes more sense to me.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi

I have done some further testing on the Poover PP3 Li-Ion battery,
the inbuilt protection circuit works well, and protects from over discharging. 

I have now simplified my circuit and omitted the ON/OFF switch.

So, a bridge-rectifier off the rails charges the battery (or supercaps),
the boost circuit is connected to the battery, and the 12v from the boost,
goes to the receiver, and then its output to the motor.

Charging is still the same on a powered section of track (I use 1 metre).  

The supercap version charges in a few seconds, 
the battery needs a few minutes depending on how big the circuit is,
then it gets a top up charge every time it runs on the charge track section.

I have now ordered some prototype PCBs,  
they measure 61 x 26 mm, and include the receiver.

Best wishes to all

Steve

:2t:


davecttr
Registered
 

Joined: Sat Nov 9th, 2013
Location: United Kingdom
Posts: 540
Status: 
Offline
 
Supercaps would be better as they can recharge very quickly?

For example, one of my passenger locos at 60mph scale speed hauling a passenger train,
would only be in the 1 metre charge zone for 3 seconds per circuit.

The locos power requirement demands a 2-cell Li-Po battery and voltage booster.
Would this be difficult to achieve without 3 wires to balance charge?

Still utterly confused over this, but it appears four 2.5V 15F supercaps in series,
would have the  equivalent of about 10mAh and give 10V ?

The loco described above would need about 3mA per circuit?


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
A set of 4x15F supercaps in series, stores a charge of 37.5 Ampsec. @10V

Without a step-up regulator, the usable charge is much less than that,
because the voltage drops continuously with discharge.

Now let's assume you have one and its cutout voltage is 3V,
the usable charge would then be ~26 Ampsec,
which gives you roughly a running time of 100sec @ 200mA motor current,
and a COP of 0.8 for the step-up.

In order to recharge, you have to put these 26 Ampsec back in.

A charge current of 1A maximum is possible only at very good contact conditions.
Intermittent contact may cause some detrimental arcing on wheel treads.

So if you limit the charging current to 500mA,
you need ~ 1minute to top it up after a complete use of available charge.


davecttr
Registered
 

Joined: Sat Nov 9th, 2013
Location: United Kingdom
Posts: 540
Status: 
Offline
 
Thanks for the clarification Helmut.

It looks like the idea won't work unless you pause on the powered track,
every circuit to give the caps time to recharge.

So no trains running around the continuous run,
while you are happily moving wagons in the freight yard.

It might be doable if you have say 4 trains that do one circuit in sequence.
For example

- long through freight
- express passenger
- local pickup freight
- local passenger

Each of the storage loops would need its own powered track?

Might be just as easy to fully charge the batteries before the running session,
which for me is up to 30 minutes run time per loco in a 2 to 3 hour period.


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
A supercap of ~1F suffices for the purpose it is ideally suited for:
Keep the loco running for up to say 1sec over a stretch without contact.

There you can keep recharge time short,
and space requirements for necessary components and circuitry are manageable.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
I agree with much of the above.

I have been looking for a value of maximum current on the pickups.
 
I had decided that 1A is a good starting point,
and as my loco has 3 pairs of pickups,
then 3A is a good starting point.  


I now have approx 30hrs run time,
and have stripped down the loco and pickups are fine.
However time will tell.

I have no evidence of arcing, but again time will tell,  
the voltage is low and there is no inductance in the circuit.


Beginning to think that the battery option has big advantages.

Cheaper than supercaps.
Much longer run time.
Smaller.


Regards, Steve

 

Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
I have arrived at powering my RC (Deltang) radio controlled TT-scale locos from the rails.

The occasional dead spots that hamper operation (much more than with DCC)
are overcome by using flat 0.47F@5V supercaps that were once made by Murata,
nowadays they are hard to come by as only cylindrical ones are offered.

The only firm offering them is CAP-XX  at prices that'll make you cry.
I was lucky enough to secure my batch when prices were affordable some years ago.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Several suppliers on ebay.  

Search for : " supercaps 0.47F 5V " 
    

Regards Steve


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Also listed at 'Farnell' UK   

Part number 2696625 - £2.09 + handing + VAT   

Quantity pricing is not too bad.

:thumb:

Steve


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
As I told you, cylindrical ones abound,

only prismatic ones are hard to come by at decent prices.


And I do not mean the standard 6.3x14mm² twin-packs,

but 2mm or less  thickness.


davecttr
Registered
 

Joined: Sat Nov 9th, 2013
Location: United Kingdom
Posts: 540
Status: 
Offline
 
Idiots question.
Will supercaps recharge while the loco is moving and powered via the rails,
or only if the loco is stationary?


I think I will try the supercap idea.
Set up a test rig first, with a loco on unpowered rolling road,
with motor fed from a Rx60-rectifier and fixed voltage power supply.


Which is best, DC or DCC ?

How do I construct a rectifier, which components needed?


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
The supercap will charge when there is voltage across the rails.

A rectifier is best bought as a ready-made item.
You need filtered DC.

Only when you operate your loco on DCC,
all that has been said about DCC applies.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi,   

I agree prismatic is best, but, the cost is a bit prohibitive.
 
I have been using 'Farnell' order-code 2696625
Less than £2.50 each, with an ESR of only 0.3 Ohms, they are a good buy.


I have used these in a stack of three connected in series (see photo),
this loco has been running for three years with no problem. 

It's preferable to take precautions when connecting in series, to balance the charge.
I suggest a 4V7 Zener-diode in series with a 220R resistor, across each capacitor.


With 2.3 Amp-seconds it will run on unpowered rail for almost 1 meter, 
this is more than sufficient to carry the loco over a reverse loop gap.

No voltage-regulator or boost circuit required, 
it's just bridge-rectifier, capacitor assembly, receiver.  
 

I can post a diagram if requested.

Regards Steve





Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
@Steve

I'd prefer to use a 5V1 Zener-diode,
as you can use the capacitor to its fullest then.


When operating on DCC, as I assume you do,
one must be sure the DCC voltage doesn't exceed 15V
(which is NMRA S 9.1. for H0/00).

However I found that almost all manufacturers,
deliver DCC output voltage ranges of more than that,
up to 22V, and still intend it for H0/00 operation.


The decoders will not fry, they withstand that voltage anyway.
But your capacitors will suffer.

The current surge when switching on after long standstill,
has to be accounted for, too.


The smaller scales don't have the space for such a capacitor bank,
and must turn to step-up converters, or internal low-voltage operation.

Of course life is much easier when you operate rail-powered R.C.
Even the surge is not much of a concern then.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi

Point taken, 5V1 Zener-diode in future.   

I no longer use DCC, generally 12V or 15V DC. 
Much prefer R.C.


You obviously have a good understanding of electronics, 
I am retired now, but much of my working life was electronics based. 

I do not have a model railway layout, just a test track, 
I like to look at control systems more than modelling. 


I have made my own PIC based decoder/motor driver,
and use a very cheap 433MHz 4 channel receiver. 

Push-button control on a handheld transmitter,
controls acceleration, deceleration, forwards and reverse. 
 
Not perfect, but it is cheap!!


Thanks for your input.

Regards Steve


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
@Steve

Do you use narrow-band FM to control different locos,
or do you use short, pulse-coded telegrams to address them?


I made such a system, using an IR-RC5-receiver chip as a decoder,
to control 32 locos, and I implemented 4 ancillary functions per address.
Each telegram was some 20msec long therefore.

But that was way back when in 1990 and intended for G scale.


In case you ever heard of Uhlenbrock IR-control
(then marketed by Chalk Rail in the UK)  that was my system.

It started as an IR system,
and I just replaced the IR-Diodes by 433MHz-modules.

I still have one working sample in my 'museum'.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi

Wow that's something, very impressive.
 
My own history with loco control only goes back to 2000. 
Made a few PWM track controllers with remote hand held controller. 
My neighbor who is an avid modeler has a couple of these on his analogue layout.


My latest controller is based around EV1572 433MHz RX and TX 4 channel, 
the outputs from the receiver are ON/OFF signals,
which the PIC sorts out into PWM for accelerate or decelerate, or forwards/reverse.
 
The PIC controls a full bridge motor IC.
I know this is very basic, but it controls quite well and is very cheap.
 
The RX is visible on the photo in my last posting,
the decoder/motor driver is underneath the RX and is the same size).


Have you looked at PP3 Ni-Mh batteries, they claim 900mAh!! charge at 1C,
discharge at 0.3C, which gives 270mA. max.

I have ordered one so will try it when my new PCB arrives.  
Cheaper than Li-Ion or supercaps.


Best wishes

Steve


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
Well, I can only find Li-Ion PP3's that claim to have 600mAh,
and only these may be discharged at a rather high rate.

Ni-MH PP3's I've found only in the 200mAh range so far.

What make have you found?


OK - 'Durable cell GTL' - found it :-)

Interesting enough, all what I've seen comes from China,
so be aware of a fake.

All the 9V reviews do not list them at all, the great manufacturers,
like Duracell or Ansmann, have only 270mAh to offer,
not without reason.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
I agree, must be fake,

it won't be the first China mislabelling I've seen.  


Well will give it a test, and it will be returned for refund if less than 900

Will let you know.


Cheers Steve


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
It all depends on what you want to achieve.

Supercaps are best if you employ command control via rail-pickup.
There you can overcome short dead spots and be safe of runaways,
as the limited charge allows buffering for rather short, but sufficient, times.
High capacities are even counterproductive.

When using R.C. with rail power supply,
you can use supercaps up to any capacity your vehicle's space allows for.
The key point is that you don't lose control when contact is lost,
as compared to Command Control.

The energy densitiy is lower than that of batteries,
which makes them the primary choice for any 'dirty rail' application.
As I've demonstrated, even the tiniest vehicles allow for R.C. battery operation.
 
A 'drawback' is the need for an on-off switch, however.
One can easily forget to switch off at the end of an operating session.
There are methods to avoid inadvertent battery draining by automatic shutoff.

Correction: I'm pretty sure I wrote 'Dead Rail' above, only then the sentence makes sense.

Last edited on Mon Mar 15th, 2021 10:17 pm by Helmut

Si.
Moderator


Joined: Thu Feb 23rd, 2012
Location: London
Posts: 5826
Status: 
Offline
 
It doesn't make sense to me, to say 'dead rail' in that context.

'dirty rail' does, as I read it applying to the option, prior to 'batteries'.


But then the 'term' dead rail, makes little sense to use in R.C. anyway,

the very next sentence, someone is talking about powered rails !


The term Battery Powered Radio Control, or 'B.P.R.C.', is descriptive,

and which has evolved & been almost universally adopted on Freerails.


Also Dead Rail is yet more confusing, as it has become a company name.

It's best to avoid the words DEAD RAIL, as some folk start popping rivets !


It is of course possible to interpret sentences, in several different ways.


Sometimes, also the vast amount of 'typos' Posted, benefit from proofing,

some Postings using all lower-case, acronyms Posted in lower-case etc.


It's just nice to have stuff up on the Board, that is easy to look at & read.

We do our best to make that happen, in whatever small ways we can.


For the most part those tireless efforts, result in an overall improvement.


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
To make it clear,
as soon as you opt for non-powered rails, or BPRC, the name says it all -

batteries are much to be preferred for their higher energy density,
and less space requirement for the intended purpose.


Si.
Moderator


Joined: Thu Feb 23rd, 2012
Location: London
Posts: 5826
Status: 
Offline
 
Hi Helmut  :brill:


I can only congratulate you, on your top notch & excellent Posts.  :bow:

Always well informed, entertaining & a pleasure to read !  :)


If I were to attempt to write, even a simple, Posting in German ...

... I wouldn't even know the word for "battery", let alone anything else.  :us:


This is despite owning & listening very carefully, to ALL my 'Kraftwerk' LPs !

My Dad preferred Schlager & ABBA, but of course I'm much too mooodern !!  :P

( yes, I know ABBA are Swedish ! )


I am amazed that anyone can even speak a 2nd language ...  :shocked:

... let alone write it down & 'technically' as 1st class, as you do so well !

I tried learning some French once, but couldn't get Past "gimme a beer" ...  :dope:

... & I'm pretty sure the barmaid was having a giggle about my attempt at that !
( could have been my plaid shirt or haircut though )  :old dude:


[toast]


Si.


That ^^ sentence COULD be read BOTH ways ...

... but 'dead' would be 49% correct & 'dirty' 51% correct.  ;)


I do like the way "... from 'station to station', back to Dusseldorf City ..." ...

... can refer to BOTH train stations & the David Bowie tour of the same name !


Trans Europe Express ... Trans Europe Express ... Trans Europe Express ... Tran Express ...

( my LP jumps at that point, predating Afrika Bambaataa's sampling, by at least 10 years ...

... but of course he DIDN'T 'sample' that record anyway, but re-programmed the beat ! )

:pimp:


Si.
Moderator


Joined: Thu Feb 23rd, 2012
Location: London
Posts: 5826
Status: 
Offline
 
Just to add to the 'experimental junk' on the work bench...  :brill:  ;)

...I thought I'd get some SUPER CAPACITORS  :shocked:  to play with ! 


I found an excellent & cheap place on eBay to get them from.  :P

The sellers Link is now out of date, but plenty on eBay to find.





They had TONS of different values & sizes & Voltages to choose from.  L:

I bought five  1.5 Farad  2.7 Volt  ones to check out.


They seemed to be the most suitable for space saving & power capacity.  :)

They were pretty cheap I think, at only £1.49p inc. P&P for five.


The kind of thing I had in mind, is similar to this schematic below.  :dt:





Also five of them wired in 'series', gives me a nice  13.5 Volt  rating.  :thumb:


So I'm following any  SUPER CAPACITOR  developments closely !  :cool:


:java::moose: :dt:


Si.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi, 

Ideally you need some way of making the super-capacitors share the charge,
otherwise the super-capacitor with the lowest capacitance, will over voltage.

I would suggest you might try a 2v7 Zener-diode with a series resistor,
across each super-capacitor.

Regards

Steve


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Hi again,

On second thoughts,
just a 2v7 Zener-diode across each super-capacitor,
no resistor.

Steve

:s:


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
When you look up the specs. of these super-capacitors,
almost every manufacturer recommends, if balancing is used,
a parallel resistor, that draws ~10x the leakage-current, of that specific capacitor.

For example, a 3F/2.7V green-cap. specified for 8µA of leakage-current.
For balancing, a 33K resistor in parallel to it will suffice.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
  
I agree that resistors do give a degree of voltage sharing.

However if one super-capacitor in the string, has 10% lower value,
then it will charge to a 10% higher voltage, this is not good,
and could cause a further reduction in capacitance.
 
When the string is discharged,
the lower value super-capacitor will then become reverse biased,
this too is detrimental.

I would suggest that if resistors are used,
then look at the voltages on each super-capacitor when charged,
and check all are 0V when discharged.

Steve


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
That's why one should use capacitors from one batch, ordered all at once.
Probability is very high that they will be taken from the same string.

I have made stacks of 10 in series,
and their charge voltages differed by ~1% which is negligible.


The drawback of common Zener-diodes is their undefined transient region.
Sharp-knee Zener-diodes are the only ones really making sense.

They are available only for 2.4V max. from 'On-Semiconductor',
which means you can charge an arrangement of 2 only at 4.5V


I think the actual risks in our applications are exaggerated.


Addition:

Here's a result of a 'torture' I applied to a Panasonic super-capacitor.
I used a 3.3F/2.3V Panasonic-HW that I charged @5V via a 10-Ohm resistor.

As soon as the voltage across the cap exceeded 3.3V, the leakage-current increased rapidly,
and the voltage rose to 3.6V to increase further with a rate of 10mV/5 seconds.

That is, the leakage-current changed from a few µA to ~130mA.

Taking the power away resulted in a quick drop back to 3.3V,
and an asymptotic decrease to 2.5V within 10 minutes.


My conclusion is that it might be possible to apply 140% of the rated voltage,
without any long-term problems under our circumstances.

The usable charge will NOT change, however.
The capacitor voltage still read 2.3V after 15 minutes and 1.1V after 12 (!) hours.


Still, if you have $5.38 to spare,
you can buy an ALD810026 to balance four caps.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
I agree.

If the super-capacitors are from the same batch,
then ok, resistors are good.

I had a bad capacitor in a string once, and things went a bit wrong,
so perhaps I am being a bit over cautious now.

Thanks for your input.

Regards Steve


Si.
Moderator


Joined: Thu Feb 23rd, 2012
Location: London
Posts: 5826
Status: 
Offline
I can see 2 of Freerails top technical brains ^^ operating at FULL POWER !!  :java::brill: :java::brill:
Cognitive cog wheels grinding away in the ol' grey matter.  :dope: :slow:


So I thought I'd throw a spanner into the works !  :P


And propose a  'Bi-Polar Super-Capacitor'  :shocked:  for regular 12 Volt D.C. trains.  L:

An electrical 'flywheel' that can operate whilst reversing polarity & train direction.  :mex:


Just in case of a dead rail ... WHOOPS ! ... I mean dirty rail (or railS) !  :old dude:

( Is 'dead rail' just 1 un-powered rail ? ... Or should that really be 'dead railS' ? ...  ;)

... boB need not reply, regarding the singular rail question ! )  :pimp:


:java::moose: :dt:


Si.


Si.
Moderator


Joined: Thu Feb 23rd, 2012
Location: London
Posts: 5826
Status: 
Offline








:old dude:


Si.


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
Well, for that purpose may use these:

https://www.monacor.com/products/components/speaker-technology/crossover-networks-and-components-/lsc-2200np/

Needs less space.





John Durbetaki
Registered


Joined: Sun Jan 28th, 2018
Location: Oregon USA
Posts: 18
Status: 
Offline
 
The capacitors you are suggesting are pretty small,
compared to the power requirements.


A super-capacitor of 2.7F can also be referred to as 2,700,000uF
The 220uF capacitor is 1/12,000 the size of the super-capacitor.

If a 220uF capacitor is charged to 15v,
that is about 0.025 joules, which is just 1/40th Watt/second.

So if the motor is using 0.2A at 15v (3.0 Watts),
the motor will run for less than 8 milliseconds (0.008 seconds).


A 220uF capacitor powering a 3.1v LED at 5mA (15.5 milli-Watts),
means the LED will stay lit for 1.6 seconds.


Also, putting polarized-capacitors 'back to back',
almost always ends really badly.

Want to try it?  Go ahead.

Just make sure you do it in a safe place (outdoors),
with a shield between you and the capacitors.

Hearing protection is advised...


bobquincy
Registered


Joined: Sun Jan 27th, 2013
Location: Florida USA
Posts: 324
Status: 
Offline
 
My single plastic rail, with any metal on both sides,

becomes a capacitor, non-super type.


boB


bobquincy
Registered


Joined: Sun Jan 27th, 2013
Location: Florida USA
Posts: 324
Status: 
Offline
 
And the 2.7F super-capacitor powering the 200 mA motor,

will drop 66% of it's voltage in about 5 milli-seconds.

boB



Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
@bobquincy

NO, Sir!

Let's assume you use a 2.7F capacitor @2.3V 
The charge is C*U=6.2Cb
If you discharge that at 0.2A, the time to zero is 31seconds.

For convenience, let's say the motor stops at 1.5V,
then you still have 11seconds run-time.


Two electrolytic capacitors back-to-back, will NOT explode or whatever.

To extend their useful life, you can bypass each of them by an anti-parallel diode,
to make sure that they do not see any reverse-charging current.


bobquincy
Registered


Joined: Sun Jan 27th, 2013
Location: Florida USA
Posts: 324
Status: 
Offline
 
I stand corrected:

T (seconds) = R*C

In that time the voltage of the capacitor,
will drop to ~ 37% of the initial charge voltage. 


If the capacitor is at 2.3 V and the motor draws 0.2 A (about 11 Ohms),
the T is 11 * 2.7 or about 30 seconds to 0.85 V.


Your calculation of 11 seconds run time appears to be good.  :)

boB


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
Okay, now that we've settled that,
there remain two drawbacks of direct buffering in analog operation:


1. Such a huge capacity in parallel to the motor,
will make it virtually uncontrollable with a standard controller,
because of the stored charge.

Your controller has to be of the four-quadrant-type,
i.e. actively discharging the capacitor to the set voltage.


2. Supercaps AFAIK do not respond well to the simple circuit Si. featured.
You have to use diodes, too.
The simple circuit needs double the number of capacitors.


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
The use of capacitors to store energy and "keep things running" is well proven,
but not very easy to implement on analogue.

Has anyone any experience of using an inductor between controller and rails ? 

I have been considering this,
and as it is not on board the loco, there is no problem with size and weight.

I am thinking about an iron cored inductor with air gap which will store energy,
according to how much current is flowing, and the value of inductance.

If the current to the loco is suddenly reduced, due to dirty track,
the inductor tries to maintain current, by increasing the voltage at the dirty track,
this voltage can be quite large.

Does anyone have any experience of this technique ?  

I intend to do some trials,
but if it is a non-starter I don't want to waste my time.

Regards to all

Steve


Helmut
Registered


Joined: Sun Feb 17th, 2013
Location: Friedberg, Germany
Posts: 1242
Status: 
Offline
 
An effective method, and taking up less space,
was the RF-ionizer once sold by 'Relco' in the UK.

I still have one of these,
and it is rather effective in bridging dirty spots.


Excerpt from their website:

In the late 1970s the first consumer product utilizing RF technology,
was developed and marketed – The 'Relco Track Cleaner' for model railways.

This product was demonstrated on the BBC TV programme “Tomorrows World”,
consequently it became hugely popular between model railway enthusiasts,
within Europe, North America and Australia.


BUT there are caveats:

They bake the dirt onto rails and wheels if you don't do regular cleaning maintenance.
They will damage DCC decoders and coreless motors, due to the 50V RF voltage.


Now imagine what your proposed ignition-coil circuit will do to your locos!


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Very pleased for the info,  
yes I can now see the problems.   

Ah well, back to the drawing board.

Thanks for your input.

Regards

Steve


Steve Soar
Registered
 

Joined: Sun Feb 24th, 2019
Location:  
Posts: 47
Status: 
Offline
 
Back to super-capacitors.  

There seems to be two basic types, one of which is the EDLC,
this has symmetrical electrodes and construction, so although marked + & -
I wondered if they can be used in a bipolar circuit.  

So I have connected 3 off 0.47F 5V capacitors in series,
and mounted them on a running loco chassis (see photo). 





Polarity is correct  when running forwards, but is reversed when running in reverse.  
This appears to work well on my dirty tracks, and will run down to a much lower speed,
with the capacitors connected than without capacitors.  

Forwards or reverse is ok,
but I don't know if running in reverse will eventually damage the capacitors.  
I have run this in reverse for over 1 hour with no problems.

When running at higher speed (10V) and supply is switched off,
then loco continues to run for approximately 1 metre forwards or reverse.
Running on my uncleaned tracks is much improved especially at creep speed.

The super-capacitors were bought off ebay  0.47F  5V

I will continue to run for longer periods.

Regards

Steve



Addition :

Now completed 4 hours running in reverse, all working well,
still giving 1 metre run when supply switched off (from a 10V rail voltage).


I have also tested capacitors on a power supply,
and they appear to work well with reversed voltage.  

Not seen any degradation .... yet!


Testing to date has been with a linear DC controller.  

I have just tested with a PWM controller,
and as expected this needed a small ferrite-inductor in one of the rail supplies.

All looking very promising, I have not cleaned this track in months,
but loco is now running very well, even at very low speed (a big improvement).


There is a problem with emergency stops from high speed,
the super-capacitors run the loco for over 1.5 metres before stopping.  
 
So if a quick stop is required, then it could be done by shorting out the rails... !! 
(after switching off the supply) 



Regards
Steve



UltraBB 1.172 Copyright © 2007-2016 Data 1 Systems