Receiver switches.

[Phil Green]

example ideal diode, picked at random from a huge range, 15mv forward voltage, £5

https://www.digikey.co.uk/en/products/detail/analog-devices-inc/LTC4451AV-TRMPBF/15848872

alternatively ideal diodes are widely available on pre-assembled boards, eg:
https://www.amazon.co.uk/Controller-Battery-Charging-Irrigation-Protection/dp/B07QGW5J1H

 

I use a 50A one instead of a split-charge relay on my campervan, lots of uses where the voltage drop of a conventional diode is unwanted.

 

Here's an introduction if you're not familiar with them: https://www.ti.com/lit/an/slvae57b/slvae57b.pdf

[sorry for todays post-fest, its been a quiet day...   🙂 ]

 

Edited August 5 by Phil Green

[Chris Walby]

But why use a diode, its just another 2 links in the failure chain you don't need.

 

The failure mode of batteries is insufficient voltage to power the RX or switch/wiring that does not provide sufficient voltage for the RX.

 

If you have enough prospective current to burn the servo out without losing the RX or if a battery/wiring fails then the other battery takes over, simples

 

Tin hat on 🙂 

[Phil Green]

49 minutes ago, Chris Walby said:

But why use a diode

Well, its two ideal diodes for a redundant battery setup, and its because the cell manufacturers advise against paralleling NiMh directly. 
Best practises are widely published so it makes me nervous when people post "I've been doing [something fundamentally wrong] for years and its been fine" because it often comes across as a recommendation that we all do it that way 🙂

Edited August 5 by Phil Green

[Nigel R]

Nice one Phil the pre rolled board with the controller and fets looks perfect. 

 

[Simon Clark]

It would appear that I am effectively being accused of promoting something that is 'fundamentally wrong'! I think that this deserves some kind of response:

The method that I have been successfully using for many years now was not just the result of some random thought but was actually the result of some very careful consideration. I had considered the possibility of using diodes but wanted to make sure that these would be absolutely necessary. I was fully aware at the time that the general recommendation was not to connect Nimh cells in parallel but after further reading and analysis, it was (and still is) pretty clear that the issues relate to charging in parallel. It is very easy to see why this would be problematic and I would never consider doing this, let alone recommend it. Having allayed my concerns re this, I then got all the charge and discharge data together and started to calculate what would happen under different scenarios. These calculations were then verified by some actual (although relatively crude) testing and finally by many years on actual in field testing.

Anyone may think this this is a bad solution and have other ideas but please do not state that this approach is 'fundamentally wrong'.
Of course it is up to the individual to do whatever they want to, but the question was asked if this system worked OK and I have answered this to the best of my ability and, I think, from a position of some knowledge on this.

 

Simon

[Ron Gray]

32 minutes ago, Simon Clark said:

finally by many years on actual in field testing.

That is the point that speaks volumes. If you had only used that setup once or twice then it would be wise to treat it with caution. You have shown that ‘best practice is not always correct. As an aside, who decides what is best practice and what tests were done in determining it?

[Nigel R]

2 hours ago, Simon Clark said:

it was (and still is) pretty clear that the issues relate to charging in parallel.

 

No, the scenario I mentioned occurs when connecting in parallel, after charging. Source for me is simply the well known charge curve of nicad/nimh.

 

I appreciate my previous post was not very clear on that point, apologies for that.

 

But - as Phil previously stated, cell manufacturers do not recommend parallel connection. You have even stated this yourself:

 

2 hours ago, Simon Clark said:

I was fully aware at the time that the general recommendation was not to connect Nimh cells in parallel

 

I don't think anyone is doubting you did reading and analysis.

 

Devil's advocate - what would it take here to change your mind?

 

 

 

16 hours ago, Chris Walby said:

But why use a diode, its just another 2 links in the failure chain you don't need.

 

It mitigates the possibility of a failure where the cell shorts out - without the diode this would cause you a total loss of power / discharge between the packs / other issue.

 

The diode can fail in several ways.

Fail open, you have a redundant power source which can take over.

Fail short, that power source shows higher voltage and is thus dominant.

 

i.e. both failures are already mitigated by the redundant nature of this setup.

 

Edited August 6 by Nigel R

[Nigel R]

1 hour ago, Ron Gray said:

As an aside, who decides what is best practice and what tests were done in determining it?

 

Usually the manufacturer, as a first port of call.

 

I mean, consider Laser and castor oil - always used to be a fun thread - you would often solidly back the manufacturer's recommendation IIRC - quite sensible I think.

[Ron Gray]

8 minutes ago, Nigel R said:

you would often solidly back the manufacturer's recommendation IIRC

Quite correct, and not just because that is what they recommended, my own findings, having used the dreaded stuff agreed with them. The same recommendation was also made to me by another engine manufacturer which backed it up, so it was more than just the manufacturer saying so. In this case Simon's own fairly extensive testing has shown him that 'best practice' is not necessarily correct.

 

A similar situation is the charging rate for Li-Pos, 'best practice' says 1C (unless the manufacturer says otherwise). Testing by others has shown that this is not necessary so I changed my charging methods based on their findings. Guess what, their findings are spot on.

Edited August 6 by Ron Gray

[Martin Harris - Moderator]

1 hour ago, Nigel R said:

 

No, the scenario I mentioned occurs when connecting in parallel, after charging. Source for me is simply the well known charge curve of nicad/nimh.

 

I appreciate my previous post was not very clear on that point, apologies for that.

 

But - as Phil previously stated, cell manufacturers do not recommend parallel connection.

 

 

I don't claim any specialist knowledge, but like Simon, some years ago I put some thought and practical testing, using various combinations of equal and unequal charge state packs, into this subject - admittedly using NiCd rather than NiMH and have since used simple paralleling with a variety of cell technologies in many models.  On my simplistic level of understanding, the small voltage depression on reaching full charge doesn't appear to be likely to cause any sort of runaway situation.  There appears to be only a few millivolts difference which would surely be quickly equalised by the source battery's voltage drop once any energy was being transferred - assuming there was a sufficient imbalance to actually initiate any meaningful chemical reaction.

 

I suspect it might be that the manufacturers envisaged "in situ" charging of permanent installations when making the recommendation not to make parallel connections.

[Paul De Tourtoulon]

2 hours ago, Nigel R said:

 

Usually the manufacturer, as a first port of call.

 

I mean, consider Laser and castor oil - always used to be a fun thread - you would often solidly back the manufacturer's recommendation IIRC - quite sensible I think.

Bla bla bla, here we go, you had to put my favourite lubricant up on the stand ! Even on a battery issue, From the horse's mouth, me still at least 2%,

OPERATING INSTRUCTIONS LASER ENGINES
Running In
Oils, fuel and general Lubrication
Synthetic Oil
Caster Oil
Lubricate
Glow Plugs
Valve Clearance
Carburettor
Starting the Engine
Fault finding & Dismantling
July 99
SYNTHETIC OIL will generally give better performance than castor oil as
it does not produce carbon or lacquer. Used correctly, synthetic oil is
preferred to castor but it will not withstand overheating and may not
protect as well against corrosion. 1 - 2% Castor oil added to synthetic
oil can give additional protection against corrosion and seizure. Make
sure the engine is LUBRICATED as described below.

 

 I was taught that electricity was like water, two different tanks with a tube connecting them, if there is any difference in height/pressure, resistance in this case, one will go into the other.

 

Bye off to another subject

Edited August 6 by Paul De Tourtoulon

[GaryW]

8 hours ago, Nigel R said:

 

Devil's advocate - what would it take here to change your mind?

 

 

With respect Nigel it's not for you to change anyones mind or for you to personally feel the need to how ever you may disagree or feel they shud be doing things different or more like you do ..... it's for eas idividual rc modeller to decide to do thjing and operate their models in a way that clearly works for them... as for whatever manufacturers do or don't recommend......  as the word is " Recommend " being what it is " A recommendation " doesn't mean it cant be done or shouldn't be done.. In the same way you was trying to convince me not to run 1 rx packs in parallel in my DB Centurion again that s my choice to keep it the way the previous owner set the plane up not for you to convince me otherwise ...how i have my models set up is for me to decide........and to be honest.......  how many rc modellers really 100% truthfully follow a manufacturers recommendations,, think you will find the number not as high as you might think....

Edited August 6 by GaryW

[Simon Clark]

Thankyou Gary - some well made points.

 

For those who may be thinking about using the double switch/battery setup advocated by myself (and some others) a concern has been raised about a phenomenon where if the two batteries are in very specific and different states of charge then one battery will discharge into the other and this will not stabilise. The reason that I think this will never happen is this: 

 

In order for current to pass from battery 1 into battery 2, then the voltage on battery 1 must be higher than that of battery 2. If you take the discharge curve of battery 1 and compare to the same charge curve for battery 2, there is no starting condition that has the voltage on battery 1 always higher than that of battery 2. In practice, if battery 2 starts from a point of mild overcharge with a slightly lower voltage than a fully charged battery 1 at peak voltage, then some current will initially flow from battery 1 to battery 2.  The voltage of battery 1 (being discharged) will very rapidly fall to the same level as that of battery 2 which may see a very slight drop in voltage and so current flow will stop and the whole system will stabilise. It is easy to see this from the charge and discharge curves of any Nimh battery you care to choose. 

 

Of course, the proviso is that both battery packs are nominally the same type and capacity.

 

Simon

[Martin Harris - Moderator]

The only references I can find to the “runaway” situation are when attempting parallel charging, where the voltage depression of the fully charged pack biases the charge current to this pack, resulting in gassing/overheating problems and insufficient charge to the less charged pack. 
 

I don’t believe that anyone is advocating charging in parallel - the packs will only be connected to each other while under discharge where the discharge will bias towards the higher voltage pack, self regulating the overall discharge of the two packs, even with dissimilar capacities. 

[Nigel R]

9 hours ago, Simon Clark said:

this will not stabilise.

 

No, not what I said:

 

On 05/08/2024 at 15:50, Nigel R said:

Will that last forever? Probably not

 

Anyway.

 

11 hours ago, GaryW said:

In the same way you was trying to convince me not to run 1 rx packs in parallel in my DB Centurion again that s my choice to keep it the way the previous owner set the plane up not for you to convince me otherwise ...

 

Do what you want, this is just a discussion, no need to get upset.

 

 

[Simon Clark]

52 minutes ago, Nigel R said:

No, not what I said:

Hello Nigel, I apologise if I have misunderstood your points. Just for clarity, could you explain exactly what you think would/could happen up to the point where it has become stable and what problems you think that would cause.

Thanks

Simon

[Albert Ace]

Thanks to all. Sorry to have aroused such passion.

Next question: what's the best fuel and how much oil content ?.ONLY JOKING.!!!!

Albert.

P.S. I've used standard switches for 45 years.......

[Andy Stephenson]

 60 Amp connectors on slide switch for only £3.35. What could possibly go wrong!

[Jonathan M]

Sorry to butt in on the entertainment, but why is this:

 

 

Better than this (fitted to my IC Gangster)?:

 

 

Or better than this (fitted to my Middle Phase)?:

 

 

 

[Andy Stephenson]

The RS switch has a manufacturer's datasheet available...

https://docs.rs-online.com/01c1/0900766b80020e3d.pdf

 

Can you find one for the other two?

[FlyinFlynn]

If you go through the specs of that C&K switch you will find it is only rated at the quite impressive 12 Amps at 12v AC, but that drops to ONE amp 125V DC !!..... and it is still one of the better switches available.  

 

I'd open up the Etronix heavy duty JR switch to ensure it uses both poles in parallel. I wouldn't entertain the lever switch for this application having seen inside so many examples.

[Cuban8]

11 minutes ago, FlyinFlynn said:

If you go through the specs of that C&K switch you will find it is only rated at the quite impressive 12 Amps at 12v AC, but that drops to ONE amp 125V DC !!..... and it is still one of the better switches available.  

 

I'd open up the Etronix heavy duty JR switch to ensure it uses both poles in parallel. I wouldn't entertain the lever switch for this application having seen inside so many examples.

My advice, for what it's worth after dismantling one of these switches and inspecting it, is to say there are better solutions available. Read that as you see fit.

[RottenRow]

19 minutes ago, FlyinFlynn said:

 

If you go through the specs of that C&K switch you will find it is only rated at the quite impressive 12 Amps at 12v AC, but that drops to ONE amp 125V DC !!..... and it is still one of the better switches available.

 

The reason for the low DC current rating is because it is more difficult for a DC current to be broken, due to arcing of the contacts. This would be more significant at 125V DC compared to the 6V or so that we are using in our models. Also a slide switch has no positive switching action, meaning that the contacts operate at the speed at which you operate the switch, though there is usually a detent that does provide some spring assistance to the switch action. This means the contact opening time can be longer than with, say, a toggle switch, where the contacts are opened and closed by a spring or springs (in a decent make anyway…).

 

That particular switch will have a thermal current rating of at least 12A, and will therefore carry a current of 12A quite happily at 6V DC.

 

Brian.

[Andy Stephenson]

I think Phil Green pointed out that you aren't switching into or out of a high current as the radio is usually in quiescent mode at the time the switch is operated.

[Nigel R]

12 minutes ago, FlyinFlynn said:

ONE amp 125V DC !!.

 

Quite normal - AC is always rated much higher - the voltage crosses zero so any arcing naturally disappears quickly. The direction of erosion, from the arcing, also varies.

DC arcs take longer to die out. The erosion is always in one direction. Hence the lower rating.

 

As noted by Andy and Brian:

 

We only use 5V or thereabouts - not 125V.

 

When you switch off an RX, normal situation, all servos at idle, any current through the switch will be very low, 100mA or thereabouts for "your average sport model" (well within the current rating even at 125V DC). If the current isn't very low, you almost certainly have bigger fish to fry (stalled servo? burning servo?) and any small arc that may result from a 5V (ish) contact break is a bit "down the priority list" as it were.

 

On 07/08/2024 at 10:54, Simon Clark said:

could you explain exactly what you think would/could happen up to the point where it has become stable and what problems you think that would cause.

 

Sure. Many assumptions; it is an abnormal scenario.

 

Nimh cell resistance is usually quoted as between 30 to 100mO - going to assume 30 for now.

 

Cell #1 is charged perfectly to peak. Voltage will be quite high. e.g. it will read (maybe) 1.8V. When it is disconnected it will almost immediately drop to around 1.5V

 

IF Cell #2 has been overcharged for whatever reason. If the peak during charge was (the same as cell #1) 1.8V and then charge continued until the cell was 1.6V.

(that would be quite the overcharge, but, chargers do go wrong) When Cell #2 is taken off charge it will drops a comparable amount to cell #1, say to 1.3V.

 

Now parallel them up.

 

Initial current, (assuming the usual simple internal model of a battery)

Vdrop = 0.2V, Rtot = 0.06Ohm ( 2 cells in series with 0.03Ohm each).

 

Gives I = 3A

 

The discharge curve at this current (1.5C ish) is sharp and of course, cell #1 terminal voltage will decay quickly. How quickly? Minutes, around 5 minutes or so.

 

Within cell #2, the terminal voltage will also fall, but slower.

 

The situation will self limit. There isn't enough energy in cell #1 to sustain the scenario.

 

Meantime, further overcharge + temp rise within cell #2.

 

Significant? Maybe. Depends.

 

 

Likely outcomes

 

1) some further damage to cell #2

2) your power source is now busy self discharging and doesn't have much spare capacity to run the RX + servos.