Help with changing loco address - Updated not working?

[PJ_model_trains]

Graskie said:

Byting bits is no problem for me, Fishy. After all, I produced a Machine Code game for the old Atari 32K games computer, which sold 20,000 copies in Europe many years ago. However, I'm sure I read somewhere that Hornby had drastically

decreased the time for RM to read all 256 CVs (0 to 255, of course). I can't seem to find it but, nevertheless, I understand the principle of "peek" and "poke" in Basic, and I really can't see why it still seems to take far too long reading CV's. Perhaps someone

could tell us why it does. Mind you, 11 seconds pause between each reading looks like 11 seconds x 255 = 46.75 minutes alone!!!!!! Why any pause at all? On my old Atari I could peek that number of bytes within seconds. What causes the delay with RM? Wiring,

eLink or Elite, the loco's decoder?

Atari, wow that brings back memories.

[Fishmanoz]

RDS said:

@fishmanoz
Very good posting but you may wish to recalculate, 2 to the power 7.

Fishmanoz said:
... Bit 7 is 2 to the power 7, which is 124.

Hmmm, but what's 4 here or there, apart from bit 2?

[Fishmanoz]

And I also agree that 11 seconds to read an 8 bit word is extraordinarily long. But then the Atari didn't have any of its peripherals attached to the data bus via a length of OO track either. And updating the RM software may not get you anywhere if the

 

delays are coming from the read speed inherent in the decoder either.

 

Reading 255 CVs is going to be needed only rarely. Most people won't be wanting to go beyond 29, with I think only speed tables of any significance above that. So the number to be

 

read can be reduced for a start without having to rely on crashing out to speed up the process.

[Fishmanoz]

And don't forget that once you go into the loco details window, there is a specific set and write loco address function via the Sigma button at the bottom.

[HornbyRailMasterSupport]

The time taken to read a loco's CVs is down to the chip in the loco, which is why RailMaster allows you to change the interval. Some chips can allow reading in as little as 8 seconds, however others can take upto 15 seconds.

 

When you ask RailMaster

 

to read all CVs from 1 to 255 although it could take upto an hour, in reality it takes a much shorter time. When RailMaster has identified the chip (CVs 7 and 8) it will ignore reading many of the CVs that the chip does not support, thus taking a shorter time

 

to read all applicable.

 

The "Sigma" (Sum in maths) button at the bottom of the CVs programming window allows you to set long (4-digit) loco addresses. The way NMRA have chosen to implement loco addressing is a little messy, with addreseses below 128

 

being written into CV1 and CV29 having to be changed to reflect this and addresses above 127 having to be written into CVs 17 and 18 and also 29.

 

RailMaster shields you from all of this by just entering the desired address. Essentially:-

 

1. Addresses

 

under 128 are written into CV1

2. For addresses above 127 use the SUM button

 

[PJ_model_trains]

HornbyRailMasterSupport said:

The time taken to read a loco's CVs is down to the chip in the loco, which is why RailMaster allows you to change the interval. Some chips can allow reading in as little as 8 seconds, however others can take

upto 15 seconds.

When you ask RailMaster to read all CVs from 1 to 255 although it could take upto an hour, in reality it takes a much shorter time. When RailMaster has identified the chip (CVs 7 and 8) it will ignore reading many of the CVs that the

chip does not support, thus taking a shorter time to read all applicable.

The "Sigma" (Sum in maths) button at the bottom of the CVs programming window allows you to set long (4-digit) loco addresses. The way NMRA have chosen to implement loco addressing

is a little messy, with addreseses below 128 being written into CV1 and CV29 having to be changed to reflect this and addresses above 127 having to be written into CVs 17 and 18 and also 29.

RailMaster shields you from all of this by just entering the

desired address. Essentially:-

1. Addresses under 128 are written into CV1
2. For addresses above 127 use the SUM button

Thanks for your detailed reply, this is explained well. Your reply also gives users confidence Hornby are

there in the back ground. Thanks again.

[gilbo2]

PJ_model_trains said:

Has anyone used the TawCraft decoder or other makes. I will be looking for 5x 4 decoders. Comments welcome. TawCraft decoder is on eBay title 'DCC Point Decoder Hornby, Peco, etc R8247 Type Decoder'

Like

everyone else, I'm now having to wait until the summer before Hornby can replace my 3 faulty R8247 decoders, so I wanted a 'plan B'! I have now tried the Tawcrafts decoders whilst I'm waiting and I have to say they operate my points brilliantly. To setup I

just wrote the base address to CV1 using eLink / RailMaster as if it was a loco. (You cannot read the CV's or use any of the Accessory Decoder programming options but writing to CV1 is straight forward and reliable). The only real downside is you do need a

separate DC power supply.

To test performance I wrote a short RailMaster programme and it would change all 4 points on the same decoder with as low as 0.1 second interval, and repeat forever!! Superb performance with strong point switching even under

this tough "test" condition. Almost created a buzzer :-)

It is such a shame that Hornby are having these supply problems of their fundamental DCC kit.