Large Scale Central

Wireless DCC

Electronics Power and Sound
#8

Kevin Strong said:

Theoretical technical question…

If I recall, the output of the G-wire receiver wasn’t exactly easy to feed into non-QSI decoders. It could be done in certain cases, but it required tapping the output of the receiver into the decoder at points other than the track power inputs. Could you take that output and feed it into the DCC input of the Tam Valley Depot booster? Would the booster accept the unipolar signal and convert it to “proper” bi-polar through the normal “boosting” process?

Later,

K

Kevin

I have looked at the output of both QSI and Airwire. Both are pure DCC with the following exceptions. The bit timing is at 9600 which is slightly below the minimum for 1 bits. Most decoders have a 1 bit range that exceeds the spec so this is not a major problem in most cases.

The problem with the QSI output is that it is at a logic level and does not have the voltage swing required by the standard which is necessary for many decoders in the market.

Most decoders get their power from the DCC signal which is the main problem. Some decoders such as those manufactured by Lenz only require a logic level input at the track input so work fine with no special requremens for hooking up.

The last issue is that the QSI receivers require a 5 volt input which much be provided to power them.

Hope that helps.

Stan

#9

If you have a logic level, single-ended DCC bitstream, you could just drive an H bridge to get the higher-voltage (and current) bipolar DCC output to drive a standard decoder. But that’s a bit of work to support an obsolete product.

#10

I’d be happy to trade one of these for some constructive feedback. It’s an amplifier that takes a logic level signal and boosts it up to standard DCC levels.

It is a kit, but it’s a very simple circuit, takes maybe 15 minutes to solder up. I have tested QSI, TCS and Soundtraxx decoders with it but only with my own control system. It requires 5v/gnd and the logic level input, the power can be from any DC source, I use a 14.8v Lipo.

#11

Eric Reuter said:

If you have a logic level, single-ended DCC bitstream, you could just drive an H bridge to get the higher-voltage (and current) bipolar DCC output to drive a standard decoder. But that’s a bit of work to support an obsolete product.

Yes, it’s that simple… that’s also why AirWire embraced this early on. QSI just made something compatible.

I have enough QSI to last my life now, and the sound is way better than basically anything out there. Also while the G scale decoder is out of production, the custom ASIC developed by QSI is alive and well in other scales.

But I digress on the QSI… the LINX radio modem was chosen to be simple, cheap, and transparent (to the DCC protocol)

Greg

#12

Martin, what you have, combined with a “radio modem” looks perfect for creating a deadrail system from a normal DCC system.

Looks like an H bridge “ic” and a PIC processor.

What kind of amp rating, and what about output short circuit protection?

Greg

#13

Where would I find decoders for sale and what would a reasonable price be?

#14

Red, RLDHobbies.com and Reindeerpass,com are sources for TCS and Soundtraxx 4-5Amp decoders. QSI is very tough to find these days, I have one to my name. There are others but I have not used them, Greg and other folks here would know more about those. The TCS and Sountraxx sound pretty good to me and for just running trains and honking horns and stuff, they work well. More esoteric things I have not tried. These are motor/sound/light units, they do it all. Figure about $150 or so.

Greg, I don’t worry about short circuit as this is designed to be directly coupled to the DCC decoder, there should be no shorts as long as you wire it correctly (two wires is hard to mess up and it’s a cheap IC anyhow if you do blow it ha) Also, that’s just the amp, I have a modular system with interchangeable network boards that produces the logic level DCC. It’s driven by an Atmel Atmega328p at 16Mhz. As far as specs, 5 amps is on the data sheet but I have not instrumented it to take exact measurements like stall current and such.

I think I’ve posted this before, can’t remember, anyhow here is the complete install in a U25B if you want to take a look. Control board is on the left, Soundtraxx TSU 4400 in the middle, amp to the right. 2200mah 14.8v Lipo is in the fuel tank.

https://youtu.be/VFONdFwhyP0

#15

Stanley Ames said:

…The last issue is that the QSI receivers require a 5 volt input which much be provided to power them…

Ooh, yeah. Forgot about that. So much for that idea. Likely best (absent a source of compatible QSI decoders) for Red to sell off his existing G-wire receivers to those who have the QSI decoders and use the funds to buy a new receiver for other brands.

Don, is the Revolution system on the market yet? Last I checked with Reindeer Pass, it was not yet available.

Later,

K

#16

Kevin, the Revolution DCC is now available. I have my arriving next week. JK told me he tested with Soundtraxx and Phoenix. It seems this version has been selling in the UK with good results.

Don

#17

Kevin, the Revolution DCC is now available. I have my arriving next week. JK told me he tested with Soundtraxx and Phoenix. It seems this version has been selling in the UK with good results.

Don

#18

Kevin

I have been using the Revolution DCC equipment since the end of March.

So far I like it a lot.

Tom

#19

For small locos you can use the Zimo HO decoders as they are rated at 30 volts and are small. Current of the MX645 wired version is 1.2 amps and has 3 watt audio output plus has 2 servo outputs. I use these on all my single motor trains.

#20

I’ve had feedback from NCE and CVP. Both advise of compatibility issues between the NCE GWire cab and CVP receivers. So I’m now in the market to sell one never used throttle and 6 QSI Gwire receivers!

I’m finding it difficult to select a new system, its a minefield of poorly explained info out there. It would be great if there was a uniform standard as in the smaller scales.

#21

Red said:

… its a minefield of poorly explained info out there.

Yes, and that is shame. How is someone new to this technology supposed to get started, without wasting good money?

It would be great if there was a uniform standard as in the smaller scales.

Yes. yes it would be…

#22

Red said:

I’ve had feedback from NCE and CVP. Both advise of compatibility issues between the NCE GWire cab and CVP receivers. So I’m now in the market to sell one never used throttle and 6 QSI Gwire receivers!

I’m finding it difficult to select a new system, its a minefield of poorly explained info out there. It would be great if there was a uniform standard as in the smaller scales.

Well, you are looking at something that is relatively new, trying to move DCC, which is defined as a system with certain components, a certain architecture, and track power; to wireless, and eliminating certain important architectural details.

This is no different what the scale, there IS no standard.

DCC as defined by the NMRA is more than just the protocol, but certain functions, where a central command station is the only unit transmitting commands. This has been bypassed in most implementations. This creates unique problems that different people solve in different manners.

Since there is very little battery power used in all the DCC systems that exist, it would seem that this is the reason no one has determined a standard. Only in the last few years have people tried to “inter operate” between hardware from different manufactures in WIRELESS DCC implementations.

The fact that Airwire has violated the most fundamental issue, the actual timing of the protocol, shows that these manufacturers are really not interested in a universal standard.

So you are really seeing this in it’s infancy, in ALL scales.

Greg

#23

The Airwire T5000 throttle is in spec. Here is a plot of an RF capture (top) and frequency threshold plot (bottom), showing the symbols. The overlay in white averages the symbol rate, and you can see that the 1s are very consistent, and the timing of each half of the symbol is 55.75 µs. The command station spec is 55 to 61 µs, and decoders must accept 52 to 64 µs.

Perhaps Stan measured older equipment, but the current transmitter is dead on. Any DCC decoder should easily understand what the T5000 is sending.

#24

Eric Reuter said:

The Airwire T5000 throttle is in spec. Here is a plot of an RF capture (top) and frequency threshold plot (bottom), showing the symbols. The overlay in white averages the symbol rate, and you can see that the 1s are very consistent, and the timing of each half of the symbol is 55.75 µs. The command station spec is 55 to 61 µs, and decoders must accept 52 to 64 µs.

Perhaps Stan measured older equipment, but the current transmitter is dead on. Any DCC decoder should easily understand what the T5000 is sending.

Eric

Great news! I stand corrected.

Indeed I did the test many years ago when we were working on a standard for DCC direct. At that time they were transmitting the 1 bits at 9600 which resulted in one pits slightly under the standard. I had not realized that they redid the bit transmission in later versions. (you can actually transmit DCC using standard ACSII using a serial port at 9600 but that is another topic)

We came close to completing the standard for DCC direct but alas the DCC working group ran into other problems about that time and lost the wide manufacturer participation.

Stan

#25

Hi Eric, by being within spec I assume you mean within spec to send signals to their receivers which work with any decoder made in line with NMRA DCC protocol?

So now I need to find out about 900Mhz versus 2.6Ghz - (range, interference with other electronics, , interference by other electronics), which frequency equipment is approved for use where I live, which systems have built in sound and which do not etc. Hardest of all, finding out which systems will be around for a long while - I have already burnt my fingers with QSI before getting even one loco to run.

It seems that the locomotives are the cheap part, the control, sound and battery systems are the expensive part! Right now I run a loco with a $4 LED light driver with forward only; the motor makes a ringing sound like a DC loco on a DCC system set to the short loco ID reserved for non DCC(https://www.largescalecentral.com/externals/tinymce/plugins/emoticons/img/smiley-smile.gif) - at least it goes!

#26

Hi Red - Yes, that’s what I mean. CVP makes their own decoders (G3, Drop-In, etc.) that combine a radio receiver and DCC decoder on one board. They also makes a product called Convrtr that is just the receiver portion, and generates a DCC output that you can then hook to any standard DCC decoder. The DCC signal originates at the throttle, so it’s important that the throttle generate an NMRA-compliant DCC bitstream. The current throttle that they sell is compliant, but Stan found that the ones they made several years ago were slightly out of spec.

As far as the two frequency ranges go, lower frequencies are generally able to propagate more easily through and around things, so 900 MHz is theoretically better that 2.4 GHz in this regard. In reality, though, the difference is probably negligible compared to other factors like transmitter power, antenna gain, etc.

I chose Airwire because I like to hack and build my own devices. As it uses DCC, this is much easier than trying to reverse-engineer a proprietary protocol. Another benefit of DCC is that you can piggyback receivers to add function outputs. One of the serious shortfalls of the new RailPro system is that it only has 6 outputs. Once you hookup front and rear headlights, classification lights, and number boards, you’re done. No cab light, servo couplers, ditch lights, mars light, etc. If you’re ok with that limitation, it’s probably the most user-friendly interface. I have never used Revolution, so I can’t comment on that system.

#27

Thanks Eric. Your comments are most helpful! I’m compiling a spreadsheet of the different systems to make sense of them all, at the end of the day everything is a compromise of one sort or another.