Model Railway Forum banner
1 - 1 of 1 Posts

262 Posts
Discussion Starter · #1 · (Edited)
Digital Four Function Locomotive Decoder
Hornby Railways R8215
Review by Doug Teggin

I purchased a few Hornby locomotive decoders and a couple of stationary accessory decoders.

I set up a test station on my desk (in from the cold garage) and connected up my Lenz Set 100 with the LZV100 command station and the LH100 hand controller.

I have a short test track set up as a main line. I also have the rolling road connected to a switched section of the main line for testing faster running of the locomotives (preventing them hurtling off the edge of the desk).

I have a programming track set up for dedicated programming purposes.

Train Rolling stock Railway Vehicle Rolling

Here we have the Hornby DCC Decoder (R8215) on the left and a Lenz JST Gold on the right together with a 1 Euro coin for scale purposes.

This Hornby decoder at 17 x 10 x 3.5mm is very small and will fit in many places that the Lenz will not.

For the moment, I find the short wires to the NEM 652 plug fine. Sometimes longer wires from other decoder get in the way and take up valuable space inside the loco.

It is slightly interesting that the layout of components on the decoders that I have are not exactly the same as the decoder in the marketing material Hornby have used (below), indicating that product development has been ongoing since the product was first announced.

Circuit component Networking cables Passive circuit component Hardware programmer Computer hardware

Passive circuit component Circuit component Green Hardware programmer Electronic engineering

Photo: Hornby

I'm testing the decoder in a few locos. As I don't have any of those new fangled Diesel or Electric locos with lots of functions, I'm using the American geared Climax loco from Bachmann that allows easy access to the DCC decoder. It has directional lighting so that can be tested too.

I'm using a couple of Hornby locos to test running characteristics. I am very used to the Class 08 which I use quite a bit so I'll be able to judge well any differences to how the loco handles with the Hornby decoder.

Passive circuit component Circuit component Hardware programmer Electrical wiring Capacitor

Normal operation
Maximum current carrying capacity of the decoder in sum1 A
Rated continuous current output in sum500mA
Continuous motor output current500mA
Function output current100mA each
Speed Steps14, 28, 128
Dimensions17 x 10 x
* according to Hornby leaflet

Toy Automotive lighting Motor vehicle Vehicle Wood

  • Control of the motor's rotational speed (load compensation)
  • Acceleration and deceleration separately adjustable
  • Selectable for operation with 14, 28, 128 speed steps
  • Programming on the main track
  • Four On/Off function outputs. Two of the function outputs are dedicated for the head / tail lights
  • Operation on standard DC systems (analogue operation) possible
  • Motor overload current protection
  • With NMRA RP-9.1.1 / NEM652 medium plug

List of supported CVs
CVMin-MaxCV DefinitionDefault Setting
11-127*Locomotive address3
30-255Acceleration momentum5
40-255Deceleration momentum5
29BitDecoder configuration1**
1Default direction0
2Speed steps 14, 28, 1281**
* according to Hornby leaflet

** according to Hornby leaflet - for speed steps in CV29, bit 2 is set on with a '1', but this value is actually equal to decimal '2'

First test: Default settings

After plugging the decoder into the loco and setting the loco on the track, it answers straight away to address #3. Forward and reverse movement is smooth without any visible jerking or problems. Function 0 turns on the directional lighting that works perfectly and as expected.


I have compared the Hornby decoder directly to a Lenz Gold - perhaps unfair as the Hornby decoder costs less than £10 and the Lenz costs about £30, but the Lenz was the only other decoder I had at hand so that was the reference.

As it turned out, the Hornby decoder ran very well and had a smoothness comparable to the Lenz and a slow running speed much slower than the Lenz.

After testing the decoder on the Bachmann Climax, I switched it over to the Gronk that had had a Lenz Gold fitted previously. On my Hornby Class 08, with both decoders set (and reset to their default settings):

At speed step 1

  • The Lenz Gold took 38 seconds to rotate the driving wheel 1 revolution
  • The Hornby Decoder took 128 seconds to rotate the driving wheel 1 revolution

At speed step 28

  • The Lenz Gold rotated the driving wheel 3.6 times per second
  • The Hornby decoder rotated the driving wheel 2.8 times per second

The slow running characteristics of the Hornby decoder are very good (better than the Lenz Gold) and the speed curve is very well mannered. I found that setting the Acceleration delay to a higher number - around 50 - started the loco going very slowly and built up speed in a very prototypical way. It was as if the decoder had a constant ratio curve.

Always be wary of setting the breaking delay too high. Model railways don't have prototypical lengths to stop and when you need your loco to stop, you don't want it to keep on chugging around the track. Keep the number below 10, preferably at about 5 or 6.

Acceleration & Deceleration

Playing around with the acceleration settings, the Hornby decoder takes 1 minutes and 12 seconds to reach top speed when the CV3 is set to 99. When CV3 is set to 255, it takes about 3 minuets to reach top speed. This decoder has 256 levels of acceleration and 256 levels of deceleration.


The decoder runs very hot indeed at higher locomotive speeds. The diodes on the decoder at where the heat is and it is enough to burn you fingers if you touch it at high speed - I was testing it on a rolling road so I was able to touch the decoder at high speed.


I wasn't sure how the 500mA would stack up against decoders that offer 1A of normal output current. I tested the decoder in the Class 08 expecting to burn it out when subjecting it to some stress tests. At full speed, I held the wheels and flywheel still so that full power was applied and a full stall was evident. The motor hummed on and off - as is the decoder was switching on and off. Perhaps this is what the auto cut-off is all about. No damage to the decoder at all.


The leaflet that comes with the decoder says that the decoder can be programmed in Operating mode, Address-Only mode, register CV mode, paged CV mode and Direct CV mode.

Programming modeRegisterInfo programmedInfo read
Operating mode (PoM - Programming on the
DIR (direct) modeADR (address)No (normal)No (normal)
ACC (acceleration) Yes (normal)No (normal)
DCC (deceleration)Yes (normal)No (normal)
CV modeCV1No (normal)No (normal)
CV3Yes (normal)No (normal)
CV4Yes (normal)No (normal)
CV29Yes (normal)No (normal)
Programming TrackDIR (direct) modeADR (address)No (decoder not responding)No (decoder not responding)
ACC (acceleration) Yes (normal) No (decoder not responding)
DCC (deceleration)Yes (normal) No (decoder not responding)
CV modeCV1 (ADR)Yes (normal) No (decoder not responding)
CV3 (ACC)Yes (normal)No (decoder not responding)
CV4 (ACC)Yes (normal)No (decoder not responding)
CV29Yes (normal)No (decoder not responding)
Register modeR1 (ADR)Yes (normal) No (decoder not responding)
R3 (ACC)Yes (normal)No (decoder not responding)
R4 (ACC)Yes (normal)No (decoder not responding)
Page modeP1 (ADR)Yes (normal) No (decoder not responding)
P3 (ACC)Yes (normal)No (decoder not responding)
P4 (ACC)Yes (normal)No (decoder not responding)

Note: The limitation of supported CVs make it impossible to use 4 digit addressing. So it only works in 2 digit mode. I can run a 2 digit locos alongside my 4 digit locos using my Lenz system.

Even though the leaflet says the address can be programmed in the range of 0-255, as it is a 2 digit system, it can only accept addresses in the range of 1-99.

Using Lenz CV-Edit and JMRI to access the decoder

I have tried using the Lenz CV-Edit PC based program and the JMRI software to access and program the decoder. As the loco address can't be read in any mode, the software fails to acknowledge the presence of the loco. Even though as the system goes through the CV table, it briefly nudged the loco on some CVs, an indication that there is communication of sorts.

JMRI works in the following modes: Paged, Direct Bit, Direct Byte & Register.


This is a small, cheap well mannered decoder. Good running characteristics.

I am a little disappointed at the lack of supported CVs. I don't know how chips on a decoder are programmed, but I have a feeling that some good juicy CVs have probably been deactivated rather than been left out. Perhaps to make things easier for the first time user, but who knows...? I would have liked to see speed control in one form or another integrated into the device. I don't see how that could have been left out.

It has programming issues that perhaps need to be addressed by Hornby. Obviously we would like to be able to read the loco address and use software to program this decoder.

The leaflet has a few errors in it, address range and acceleration ranges etc, a couple of typos - nothing too serious, but perhaps enough to confuse some first time DCC users. A bit more info from Hornby would be welcome.

I am not an expert on NMRA compliance and can't say if it does or does not comply with the NMRA DCC standards. It probably passes 'the test', but as there are issues with programming, it may not work with other DCC systems. We have to wait and see how the Hornby Elite works with this decoder and it will be interesting to see if the device can read the info programmed on it. Hopefully we'll be testing that when it comes out.

I say again that it is small, cheap well mannered decoder. Perfect for fitting into steam locos with limited internal space, perfect if you starting out with DCC on a limited budget and perfect if you need a smooth running simple decoder for a little engine. I'll be using these on my layout for sure.

- December 2006

All text, photos & graphics (except where otherwise mentioned) ©2006 Doug Teggin - All rights reserved.
1 - 1 of 1 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.