Managing start up power surges in DCC

[GoingUnderground]

Did you try asking the question on the Facebook group?

 

As far as I know, as I don't have a CabControl, only an ECoS, and have to go on what's in the CC manual, there is no user interface to adjust the short circuit sensitivity.

 

Have you tried the idea that's been posted on the ESU Forum of putting a 4.7 Ohm resistor in parallel with a simple on/off switch in series with one pole of the Main track out? The idea is that the resistor limits the initial current draw and is then switched out of circuit before starting a running session. Pretty cheap, and very easy to fit. If it doesn't work then all it will have cost you is a few pence and a bit of time.

[Crosland]

2 hours ago, Hamburger said:

Modern boosters have 2 types of protection:

 

1. If the current exceeds the nominal value, it switches off (overload protection).

 

2. If the current suddenly increases within a very short time, it also switches off (short-circuit protection, called dI / dt switch-off). This happens even if the current does not reach the maximum value!

 

Are you sure about that? Can you point to a manufacturer's data sheet/specification that shows it?

 

[Crosland]

2 minutes ago, GoingUnderground said:

Have you tried the idea that's been posted on the ESU Forum of putting a 4.7 Ohm resistor in parallel with a simple on/off switch in series with one pole of the Main track out? The idea is that the resistor limits the initial current draw and is then switched out of circuit before starting a running session. Pretty cheap, and very easy to fit. If it doesn't work then all it will have cost you is a few pence and a bit of time.

 

Just make sure it can withstand the 100W+ instantaneous power dissipation! 

 

You will not need a 100W resistor, since the startup surge is not sustained, but nor will a 1/8W carbon film jobby be suitable.

[Hamburger]

 

4 hours ago, Hamburger said:

In the simplest case, this is a resistor between booster and layout that is bridged shortly after the booster is switched on.

 

1 hour ago, GoingUnderground said:

The idea is that the resistor limits the initial current draw and is then switched out of circuit before starting a running session. Pretty cheap, and very easy to fit. If it doesn't work then all it will have cost you is a few pence and a bit of time.

Let's calculate: (on the safe side)

U booster = 20 V

R = 50 ohms (example)

U / R = I = 20/50 = 0.4 amps

P = U x I = 20 x 0.4 = 8 watts

short-term possible resistor overload factor: 5 to 8

8 watts / 5 = 1.6 watts

===>> resistor 47 ohms / 2 watts

 

Just as a try (do not try to operate your layout with that):

install a fat resistor between booster and layout, (let it hang in free air or rest on a surface that can neither melt nor burn)

switch on the booster

see if the booster goes off or not (if the resistor becomes hot, switch off)

 

When you found a suitable Resistor:

Take a push button switch that is normally closed.
Connect it in parallel with the resistor.
(The switch must be able to withstand the entire booster current at all times.)
Press the switch button, turn on the booster, release the switch button (somehow that reminds me of starting a very old truck).

 

Edited April 14, 2021 by Hamburger

[young37215]

2 hours ago, GoingUnderground said:

Did you try asking the question on the Facebook group?

 

As far as I know, as I don't have a CabControl, only an ECoS, and have to go on what's in the CC manual, there is no user interface to adjust the short circuit sensitivity.

 

Have you tried the idea that's been posted on the ESU Forum of putting a 4.7 Ohm resistor in parallel with a simple on/off switch in series with one pole of the Main track out? The idea is that the resistor limits the initial current draw and is then switched out of circuit before starting a running session. Pretty cheap, and very easy to fit. If it doesn't work then all it will have cost you is a few pence and a bit of time.

Hi Keith 

 

The CC manual is silent on any short circuit parameters, it is generally pretty lightweight on technical stuff fullstop. I posted the question asking what are CC short circuit parameters on Facebook 2 days ago; no responses as yet. I sent the same question to ESU Technical Support at the same time; again no response as yet. Same on the ESU forum as you have seen. I do'nt expect to be given a way to vary CC short circuit settings by ESU, however detailing what the short circuit parameters are does not seem an unreasonable request of them. If the parameters are known then designing a way to circumvent them should be possible.

 

I have read the comments regarding the use of a resistor and switch. I think I understand what is proposed and why it might work, I just want to digest the suggestion and draw it up on a diagram to make sure I have got the required wiring correct. I also want to take counsel from someone more knowledgeable about the PSX's to understand what adaptive load reset is actually doing. My main reason for installing the circuit breakers was to benefit from this, if it does not work for me then I see little value in having them especially as you suggest, they may also be drawing power on start up. 

 

Thanks again to everyone for your continued counsel and assistance. 

 

[RAF96]

A similar problem was discussed on the Hornby forum with regard to inrush from multiple DCC Concepts acc-decoders tripping the circuit.

The proposed and adopted answer was to install a bank of isolation switches and feed them in in batches.

[tender]

You could try a Tam Valley Startup Module. I had this on one of my layouts and it cured the problem.

 

Quote:

TurnOn Module - Soft Start Circuit For Large Inrush Currents of Sound Decoders and Stay Alive Decoders
If you have a lot of sound decoders and decoders with stay alive circuits on the track, they can confuse the booster* when it starts up in to thinking there is a short because of the sudden current draw of all the capacitors. This circuit prevents that problem by putting a small resistance in to the circuit to slow the current rush and once the voltage gets up a 10 Amp relay bypasses the resistor so that there is a straight connection to the booster*.  (If you have no idea what the previous explanation is about and you are not having any touble with starting your layout then you don't need the TurnOn.)

 

*or Command Station

 

https://www.tamvalleydepot.com/products/dccbooster.html

 

Available from:

https://www.coastaldcc.co.uk/products/tamvalley/power-management/

 

[young37215]

I spoke with Kevin at Coastal DCC this morning who advised that to the best of his knowledge the ALR in the PSX was to all intent and purposes a soft start. There is nothing in the PSX instructions to explain ALR so I have e-mailed DCC Specialities in the US to see if they can provide greater clarity. Having read the supporting literature about the Tam Valley start up module and for the modest £17.50 it cost, I have bought one to see if this solves my issue. I now have three potential solutions namely: 

1. Tam Valley start up module, simple wiring change to install between ESU ICU/booster and the PSX's required
2. Further division of power districts using DPST switches, more complicated and extensive rewiring to sub divide existing power districts
3. install resistors and switches, the best size of the resistor needs to be found through testing which unerves me hence my purchase of the Tam Valley module 

Watch this space, I will report my findings when I have installed the Tam Valley module  

[young37215]

The Tam Valley start module has arrived, been fitted and does nothing to resolve my problem! Option 1 of the list of potential solutions has not delivered a resolution to the problem. 

 

At the time of writing neither DCC Specialities or ESU have responded to my requests for information about their products. I had hoped that such basic and straightforward information was readily available but it seems I am wrong. Poor customer service in my opinion.

 

In recent days the start up issue has been less pronounced and I have a useable layout. On some occassions I start track power with the first button push, on others it can take several presses. As it stands today the situation is manageable, the long delays have fallen away which I can only attribute to the PSX's. If this continues then what I have is workable as it stands although I have additional DPST switches and am considering further dividing the fiddle yard into 2 sub districts. It will be a pain to re-wire the fiddle yard in situ so I plan to make the changes when I install point motors which will require the baseboards lifting on to the workbench.  

 

I do appreciate the counsel and guidance  that has been given; thank you again for your time and efforts. As ever if I find anything more or get a reply from either vendor I will update the thread.

[melmerby]

On 14/04/2021 at 16:02, Crosland said:

 

Just make sure it can withstand the 100W+ instantaneous power dissipation! 

 

You will not need a 100W resistor, since the startup surge is not sustained, but nor will a 1/8W carbon film jobby be suitable.

A Vishay 10W wirewound resistor will stand a 10x overload for 5 seconds (manufacturers datasheet)

[Harlequin]

Just an observation from years of bitter experience debugging software:

 

When a problem resists the application of numerous logical solutions it's increasingly likely one hasn't understood the problem correctly. Then it pays to revisit one's basic assumptions - one of them may be wrong.

 

[Hamburger]

Sometimes I feel ripped off on this forum.

[young37215]

On 20/04/2021 at 14:53, Harlequin said:

Just an observation from years of bitter experience debugging software:

 

When a problem resists the application of numerous logical solutions it's increasingly likely one hasn't understood the problem correctly. Then it pays to revisit one's basic assumptions - one of them may be wrong.

 

 

Cannot disagree with the sentiment but everything here points to an accurate understanding and quantification of the problem by more than 1 person. The only real missing bit is the outstanding information from ESU about the short circuit trip programming thresholds. Without knowing the thresholds it is trial and error to see what stops the spurious start up trips occuring, with the information it might be easier to produce a definitve solution.

 

Today was a classic case in point with the intermittency and extent of the issue. It took me 20/30 attempts to connect track power this morning when previously this week I have been connecting on first or second attempt. I know locos were in different positions on the layout but their numbers had not changed. I have no idea why it is so hit and miss, I could understand the problem being the same each day but the wide variation in how it prsesents itself is beyond my comprehension.  

 

I heard from DCC Specialities main man Larry Maier on Tuesday although I only discovered his reply in my Spam folder today. Larry sent a very clear explanation of what ADR is which made it obvious that it could not solve my problem. ADR is essentially a slowing down of the speed of trip that the PSX uses as its threshold to determine shorts so that the momentary surge from capacitors charging does not cause unnecessary trips on the PSX. I have my PSX on the shortest time settings so as to maximise the PSX's short circuit protection and there is no obvious value to reducing the PSX trip time because it is my booster that is tripping, not the PSX.  

 

It looks like further sub division of my power districts is the next step along this journey of discovery. 

 

[Crosland]

Sorry to resurrect an old thread, but I have some insight and wanted to ask if your problem was ever solved?

 

Whilst working on a new booster design I was investigating "soft start" options and realised that no one mentioned the effect of layout wiring, or I missed it if they did.

 

At one, impossible, extreme, zero resistance between the booster and the discharged capacitors in the decoders, and a limitless PSU would result in an infinite current spike but the capacitors would charge very quickly.

 

In the real world, the peak current is determined from the track voltage and the wiring resistance, by ohms law, at time 0. The current decays quickly, and exponentially, as the capacitors charge. The aim is for the current to decay to a "safe" value before the booster (or other device) trips. A higher resistance between the booster and the locos will limit the current and flatten the curve, possibly resulting in too high a current flowing at the trip time. This is just another reason to employ hefty (low resistance) bus wiring.

 

As an example, 15V track voltage, 0.2 ohm wiring into 20,000 uF will result in 6 A flowing after 10 ms. I used a very short delay as switch mode PSUs can cutout very quickly if overloaded. The reason some systems cut out very quickly is to prevent the PSU cutting out.

 

The Tam Valley TurnOn Module switches a 0.5 ohm resistor into circuit at startup. After the same 10 ms the current will be almost double, at 11 A. Possibly still enough to trip the booster. Indeed, the OP reports that this module did not solve his issue.

 

Examples calculated using https://calculator.academy/capacitor-charge-current-calculator/

 

An interesting test would be to power up with all of the locos at the farthest point from the booster (maximum effect from wiring), and again will all locos as close as possible to the booster.

 

Technical bit: A more robust booster design would be to employ a current limit, not a trip. That, however, requires a lot of heatsinking, and very careful thermal design, as the limiter will probably be a MOSFET operating in it's linear region.

 

[young37215]

The start issue has substantially reduced to the extent I rarely think about it since I extended the number of droppers in the main 13 road fiddle yard to one per road. Originally the fiddle yard was built with only 4 droppers, my assumption is that the original wiring was inadequate for what was being asked of it.  The BUS is divided into sub power districts where everything gets powered up at the same time on start up. Most days I start track power on the first or second attempt, occassionally a few more attempts are required but much less than was originally the case.  The BUS wire is 24/02 and the droppers 16/02, the link to my layout thread shows what I did in the fiddle yard in more detail and contains a wiring diagram. 

 

I bought a DCC Concepts Alpha meter a few months ago which shows that for most of the time power is at 15.5 volts drawing around 1 Amp. The voltage setting is controlled/set by ESU's Cab Control system and which has been on the same setting since I first bought it. The Alpha meter has been useful in it tells me I do not need a booster.

 

I hope this helps with your activity, please let me know if any more questions arise.

 

 

[Crosland]

That's interesting, thank you! 

[melmerby]

Since this topic started my EB-1 failed (taking ages to switch on) and I replaced it with a PSX breaker.

That was a bad move as I was getting exactly the same problems as the OP, even with the PSX set to it's most "lenient" settings. It always tripped at start up.

 

I had a LDT booster that I had not yet used so put that into circuit instead set to 4A trip.

It works perfectly. It's connected to the Roco Bus on the Z21