Crosland

15 hours ago, WIMorrison said:

They range from 16v to 23v and 2.8A to 3.6A - strange values but there you go.

If they have a two pin power cord, then they are designed to be below 65 W. Anything above that almost always has an earth. Otherwise the filtering gets very difficult.

14 hours ago, peach james said:

remember that the decoder eats about 4v as a minimum due to H bridge ect.  

It's 2 V, tops, in the bridge rectifier in the decoder. The H-bridges in the booster/command station and decoder will both be MOSFET based these days and very low resistance. If the booster/command station can be powered by AC then there will be another bridge rectifier and another 2V max drop. Some systems allow the track voltage to be adjusted down.

The loss in the decoder will be a lot better than 4 V.

Picoscopes have DCC decode software, but they can be pricey (mine is an MSO and was a few £hundred). You may need to check if it works on the cheaper models https://www.picotech.com/library/oscilloscopes/digital-command-control-dcc-protocol-decoding

Decoders would normally go bang and maybe let out a puff of the "magic smoke". Anything more than that is probably internal wiring, maybe the varnish on the motor windings.

6 hours ago, Gareth Collier said:

Of course on occasion the "evidence" photo means a trip around the neighbours to see if you can find a front door the same colour as in the photo.

We get regular posts on our local facebook page of the "that's my parcel, but it's not my door" kind. Often they are in the next town/village.

Evri are a total s**tshow around here. They seem to take the satnav "you have reached your destination" as having reached the exact address, without realising a postcode can cover quite a number of properties.

7 hours ago, Daddyman said:

Well, RM collected today, but I had to book a new collection and will have to spend 30+ mins on the phone to get the original £3.99 back. 

Why did you need to book a new one if it was collected?

If the collection fails you can simply reschedule it at no extra cost (I've done this). It's only "no extra cost" whilst collection is free. If/when they start charging for collections you will lose the collection fee for failed collections.

If you longer want to post the item you can request a refund online. It takes a few days (again, done this).

I recommend creating an account on the RM website and logging when sending. The system remembers who you are (to populate the sender address) and you can easily see your previous transactions.

17 hours ago, Fordbank said:

 

Apparently the early model ZTC 511s do not like capacitors. ( Early version 511s programme at very low voltage – typically around 1v - to protect the decoder and this voltage is exceeded by that stored in the stay alive.

If that's really what you were told, then they are not knowledgeable. No DCC decoder can be programmed at 1 V.

What they probably meant was that there is only a very low current available for programming and the very high inrush current to charge the stay-alive triggers an overload detection.

This can also be an issue when trying to turn the power on for a layout with many stay-alive fitted locos. Better decoder/stay-alive combos will limit the charging current.

I have found the collection service to be utterly reliable. It's brilliant, in fact, and a much better experience than the rude counter staff in the village PO.

The only times I have had issues, including chasing the postman down the street to ask why he didn't collect, turned out to be my errors.

I've never seen it. I would use individual wires and twist them if you just want to keep them together.

2 hours ago, DCB said:

See Pic. I think that must be a red  comes alive around 1.9 volt 

LEDs don't suddenly "come alive". All that example shows is a Vf of 1.9 V at 5 mA. You need to look at the luminous intensity v. forward current graph in the data sheet, which you will see is pretty much a straight line, no sudden turning on on coming alive. A modern high efficiency LED will light at less than 5 mA but, as you can see, you can see you are well down the curve and the Vf will not decrease very much in your example. 

In the other direction, if you try to control the voltage, you can see a small change in voltage (0.1 V) leads to a near tripling of the LED current. Voltage control of LEDs is too imprecise, which is why they should always be current controlled.

A LED operating normally at 1.5 V is a physical impossibility, unless someone has discovered new materials and rushed them into production without telling anyone. They would probably be in line for a Nobel prize in physics.

7 hours ago, DCB said:

The Diode tester is for testing Diodes.  LEDs work differently as they don't actually pass any current until a set voltage is reached , it used to be 1.9 volts for a red and over 3 volts for Clear LED but for modern ones used in Christmas lights etc its a lot less as they work from a single AA battery  ( I bought a load after Christmas last year and operate them  by AA battery with switching on /off  from the layout power by using a relay .

Most common LEDs go POP when tested with a 9 Volt battery without a resistor.

Sorry, but you are showing your lack of understanding of LEDs again.

LEDs are diodes, they behave much like diodes but are somewhat more fragile, e.g., very low reverse breakdown voltage. There is no "turn-on" voltage at which they suddenly start working. LEDs are, in any case, current driven. That's why you use a resistor to control the current.

The technology has not changed and typical Vfs have not changed. If your LEDs work on a single AA battery then there is some other electronics involved, to boost the voltage.

ALL common LEDs will go pop on 9V, and always have done. Any that don't have built in current limiting. You will often see these sold as, e.g., "12 V LEDs".

The only reason you get away with running a LED from a coin cell with no resistor (as you have stated in the past) is the high internal resistance of the cell acting as the current limit.

13 hours ago, spamcan61 said:

Well I was going to measure the output of my SPROG on my Noddy oscilloscope versus my Fluke 77 DMM but weirdly the SPROG has developed a short circuit input so experiment cancelled.

If it persists, contact us directly to discuss, but I will be away for a few days so you may not get a response straight away.

Andrew Crosland

sprog-dcc.co.uk

What readings you get will depend on how you connect the meter.

If you measure on DC from one track to the booster "ground" or 0V then you will probably see half the voltage as the DMM will average the square wave that is switching between 0 and +V. See https://www.digitrax.com/tsd/KB909/track-voltage-measurement-on-dcc-layouts-with-dire/ for more details.

If you measure across the rails then you need to set the meter to AC and, as someone said, it can be a bit of a crapshoot. There is, somewhere, a web page by someone who tested a lot of DMMs for measuring DCC, but I can't find it at the moment.

That's interesting, thank you! 

If they program fine then there should be nothing to worry about. Does "programming" include both reading and writing?

Traditionally the ACK pulse was generated by pulsing the motor, often resulting in audible feedback and the loco creeping along the programming track. The effect depends upon the motor, as well as the decoder. It's also possible to simply switch a resistor or other load across the tracks to generate the ACK. This would not result in any noise or movement.

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.

7 minutes ago, spamcan61 said:

I thought to POM the loco had to be in a 'programming' state rather than underway, but as a DCC dabbler I could well be wrong.

You've got them the wrong wat around. Service mode needs a programming track and the loco is put into service mode to tread or write. POM is Program on the Main, also known as ops mode, works anywhere on the layout, but is write only unless you have something like Railcom.

Yes, but you might need to change a setting in the 'phone (at least if it's Android) to connect to a network with no internet.

You need a wireless access point or router. Be sure to change the network name (SSID) and admin password away from the defaults.

The individual items should have instructions on how to connect them. These days it can be as easy as pressing a button or two and everything is setup automagically.

16 hours ago, spamcan61 said:

I would say for DCC, as essentially a bipolar square wave, RMS and peak AC voltages should be the same.

And the prize goes to...😀

A very easy way to measure the track voltage is to simply rectify it and add a bit for the diode losses. Maybe add a small smoothing capacitor. The DC voltage you measure will be the RMS, which is the same as the peak.

16 hours ago, spamcan61 said:

I suspect this may also confuse DMMs measuring AC voltage as they may well assume a sine wave input.

True RMS meters will work, but they need to have a high enough frequency response. Most seem to be 1 kHz or below.

3 hours ago, martin_wynne said:

But my position with Templot is that I'm 75 years old.

Apologies for being gloomy, but is there a succession plan for Templot?

I'm not a user yet, but may be after the next house move in a few years when I can have a permanent layout room.

2 hours ago, WIMorrison said:

any multimeter will work, it may be a slightly wring in the absolute value displayed, but it will slightly wrong everywhere therefore good enough for testing and fault tracing.

It may be more than slightly wrong. It really depends on the multimeter. It's counter-intuitive in that cheaper ones can be better than dearer ones.

1 hour ago, RAF96 said:

RRAMP meter but they are not as cheap as a Lidl or Screwfix multimeter. The accuracy depends upon how close the sampling rate is to DCC frequency. 

Many do not sample the waveform. They have simple circuit that is only good for low frequency sine waves (e.g. 50/60 Hz derived form the mains). Even many true RMS meters do not have a frequency response adequate for DCC.

1 hour ago, spamcan61 said:

It would be interesting to know what a cheap 'scope reads. (not a hint !)

It should be at least as good as any cheap multimeter.

23 minutes ago, spamcan61 said:

Sorry me again ;-) - I can't read the x axis there, what frequency range are we looking at roughly for those square waves please?

DCC uses cycles of 112 us and 200us, so appx 9 kHz and 5 kHz.

23 minutes ago, spamcan61 said:

 

I'm wondering if my very cheap handheld 'scope has enough bandwidth to measure DCC voltage reasonably accurately.

 

One of these to be specific:-

 

https://www.ebay.co.uk/itm/204530362829?mkcid=16&mkevt=1&mkrid=711-127632-2357-0&ssspo=tk_m-8fksgm&sssrc=4429486&ssuid=CUw97j2zRRy&var=&widget_ver=artemis&media=COPY

It claims a sample rate of 1 Msps so it will be good for DCC.

Those don't look like  the original pins, they look like they have been soldered on on place of a wiring harness.

Whoever cut them short, needs a new pair of wire cutters :)

photo?

How thin will the thinnest parts be when scaled? Will it even print?

19 hours ago, jamesed said:

Unlike CVs 2, 5 & 6 which are specified by the NMRA there is no standardisation for CVs 3 & 4. 

Actually there is a recommended standard in NMRA standard S-9.2.2 https://www.nmra.org/sites/default/files/standards/sandrp/pdf/s-9.2.2_decoder_cvs_2012.07.pdf I believe (from posts by someone who was on the standards committee) that this was based on one manufacturers practice at the time, but the details are lost.

The problems are that (a) it was standardised after decoders were already available with whatever the manufacturer did and, (b), manufacturers continue to ignore the DCC standards in all sorts of ways.