HM6010 issues

[Steve e boy]

Regarding the HM6010.

First issue.

There are two options BLE/MESH to connect to the device. To update the firmware you need to be connected to mesh connecting via bluetooth you cannot update the firmware but no matter what device I use to connect I can only connect via Bluetooth.

Second issue.

It gives you two options to power the unit. Via the transformer that you have to purchase separately or by a DCC supply as clearly indicated in the instruction manual. But when you connect it via the DCC supply the unit powers up indicated by the orange light coming on but the Bluetooth will not work in any shape or form.

Does anyone have any resolve or insight on these issues because I am at a complete loss why you would have these options and can only use one.

[96RAF]

I assume you have followed the trouble shooting guide, if not then there is a link on the HM | DC forum main page to get you there. If so and you are still having problems let us look at your questions in reverse order.

For clarity of this post I will use the word module(s) to describe the 6000 or 6010 module and the word device to describe the hand-held set you have installed the app on (tablet or phone). The app uses the term device in the settings pages to describe the modules which I think is confusing (e.g. Device Provisioning or Device Edit, etc).

The power options are as you say using a 15v DC P9100 wall wart or connecting a DCC Track feed. You need to be aware that the DCC option is simply a handy way of providing power to the 6010 module and not a method of control. This has been provided in case a DCC layout user wishes to control their accessories using the app and it saves them the cost of a power supply when they already have a handy power source. At present there is a bug whereby the DCC power method prevents Mesh or BLE connection. This is being investigated and when corrected there will be a firmware update over the air to the module.

Going back to the BLE vs Mesh connection.

BLE is a simple local pairing of your chosen device with the module(s). The app talks direct to each module. This is useful if you only have one or two modules.

Mesh connection is a tad more complex as it allows the modules to set up their own wider group connection with the app under control of a host module, generally the one with the best signal to the app, however you can specify which module is the host if desired. This is a bit like cell phone towers where the signal bounces from tower to tower. Here the host (with the best signal) bounces the app commands to the other modules. Of course if you wish to contact the app provider for over the air updates then this is the only connection method that can do that.

Now to try to resolve your problem. Start from a level playing field and delete all modules from every hand-held device you have tried to pair (provision) with. This is essential as if a module is paired with one device then it cannot be seen by or pair with another device. To do this open the app on each device and disconnect each module listed, then un-provision it. If it says it can’t do this then you have to force delete it. Once you have done this on each device you have tried, then reset each module by holding the reset button (using a probe) until the red led flashes, then release it. The red and blue will flash then the blue will go out and the red will go steady. The modules are ready and waiting for pairing.

Now you can scan for modules and provision (pair) them to the app. This will save them for you. You may have to rescan to list the other modules again if you provision them one by one. At this stage the modules are paired with your device but not active. Once all are provisioned you can connect all to activate them and accept the auto-allocate for host. Your module(s) should work, but if not find and press the resume commands/operation button (can’t remember the exact label) for each module. Be aware that the 6010 module defaults to Point (pulse output) operation and if you are using it to operate lamps, signals, etc then you need to select these options in set-up (module edit).

If all else fails then you can call Hornby help desk and they will walk you through the same procedure.

I hope this helps and you get it working.

Once you are up and running then you can check for firmware updates but of late there are none. The app will tell you what revision your module is at and what the latest revision is. If both are the same then obviously there is no need to update, but if you do then you may need to set things up again. Normally the app will flag up if an app update is outstanding at start-up and ask you to press hot-update.

[NC1]

I am new to the forum, and have a background in electronic design.

I have started to use the 6010 to power point motors and was curious about the energy available to pulse them, particularly as I read elsewhere that the 6010's outputs do not have as much energy as some of the dedicated CDUs.

Having sorted out a tool to unscrew the "tamper-proof" screws, it was obvious that each channel had a CDU capacitor of 2200uF. Channels 1 to 3 have a fully-charged voltage of 14.25V, but channel 4 is only 11.94V. The quick-start instructions mention the nominal 12V for channel 4, but only in the "Street Lighting" section. It is not obvious in the user manual or the quick-start guide, that, when used for switching points, this channel will have less "ooomf" to drive a points solenoid.

The energy stored in a capacitor is proportional to its capacitance times the square of the voltage so this channel will, when driving similar loads, only have about 70% of the energy of the other three. This is likely to be a problem if the points have a stiff operation.

I wonder if any other members have found this to be a problem.

[Chrissaf]

The HM6010 has a very similar design to the Hornby R8247 DCC Accessory Decoder. This IMO also struggles to pull the skin off a rice pudding.

As you have identified, the charge voltage in these devices compared to commercial Analogue CDU products is very low. Thus you need a very powerful high uF capacitor value to compensate, but in my opinion using a higher charging voltage gives a far more reliable and robust operation. Some other brands of Accessory Decoder support this by using the DCC power just to provide the control function and have a separate DC power input of higher voltage to provide the point operating power. Some brands just provide a significantly higher uF value capacitor. Hornby have decided in their wisdom to concentrate on keeping the products as cheap as possible to manufacture rather than focusing on reliable operation.

[96RAF]

As you both say the drive voltage is more important than the capacitor value, as it gives more oomph to the solenoid. For the layman compare the height of a water tank versus the size of the tank. A large tank will not provide any more pressure than a smaller tank, but the higher the tank the more pressure.

Channel four as you will have seen at component level (see photo) is a heavier amp output to drive a turntable motor rather than the standard 150 mA output of the other channels.

The channel 4 component is a simple 12v 500mA output voltage regulator.

Your assumption is correct in that channel 4 is less suited to driving solenoids than the other three ports.

Even if the DCC port could be used as a power source, which at present it will not do (corrective work in progress) it would not provide any where near the optimal 19-24 volts that any external CDU uses.

[NC1]

Because the 6010 has to provide an output that meets 3 requirements, (1) maximum pulsed energy for point solenoids, (2) constant current to drive LEDs, and (3) a 12V higher current capability on one output; and all from the same 15V plug-top supply, needed for the 6000 module; the Hornby designers have done the best they could, to minimise size and cost.

Continuing 96RAF's excellent water tank analogy, the 4 water tanks are at maximum height (16V devices running at up to 15V - the results of an exploding electrolytic capacitor are not a pretty sight, or smell), and are taking up maximum space in the loft (the next preferred value being 3300uF, and roughly 50% more volume). The 6010 case is already slightly larger than that of the 6000.

96RAF's photo, above, shows that the capacitors occupy a significant part of the PCB area, but it is interesting to see that the white ink has additional circles for an alternative vertical-mounting option; however, as the circles overlap, not a practical option for the 4 capacitors.

[NC1]

I measured the coil resistance of several different point motors. The Hornby R8243 was 1.7 ohms, Peco PL11 (similar size) was 3.9 ohms, and Peco PL10 (double bobbin to go under point) was 4.3 ohms. So, it looks like the R8243 is going to be the most demanding from a CDU.

Thinking "Outside the box", or in this case, inside:-

Channels 1 to 3 have identical circuitry, fed from the same power source. It is unlikely that a user would switch a series of points without a short delay between each one. On my Samsung tablet, I am lucky to be able to get reliable finger movement in under 2 seconds between channels. So, following the water tank analogy, we have 3 identical (2200uF) water tanks, each fed via the same-bore (140mA) pipes, but, at any one time, we only take water from one. If we linked the outlet pipes together, we would have 3 times the available water. With the HM6010, this can be achieved by using a link wire between the 3 "C" sockets. I tried this, and it made a big difference to an R8243 point motor. This will not work for channel 4, as it uses a different capacitor charging circuit.

Note:- a "3C" supply will increase the power dissipation in the switching MOSFETs in the "A" and "B" channels, and these are likely to be damaged if the point motor wires short together - guess how I know? However, with care, they should be fine with a normal point solenoid load.

The waveform, below, shows the 3-channel switching waveforms (400ms ON, 400ms OFF, 400ms ON, OFF)

The equivalent for channel 4 is:-

I am sure that the above would NOT be endorsed by Hornby, but I found that linking just 2 channels made all the difference when holding an R8243 point motor adjacent to an unmodified R8073 point, and 3 more so.

[96RAF]

@nc1

If you check an R8247 DCC acc decoder, upon which the HM6010 is largely based you will find all four of the C ports are already commoned, so you have essentially reinvented that particular ‘wheel’ for the HM6010 thus proving the earlier design theory.

I have no idea why Hornby didn’t common connect the collectors on the three mosfet ports in the HM6010. Obviously they can not be commoned with the voltage regulator output.

[NC1]

The HM6010 outputs can also be used for signals and street lighting, so by keeping the common outputs separate, you avoid blinking signals and layout lighting if using one HM6010 for a mixture of point and constant loads.

I found that there was enough space in the case to add 4-off 2200uF capacitors piggy-back on the existing ones, and this certainly gives me more reliable point switching.

[96RAF]

@NC1

At risk of repeating what I have already posted...

Unlike the R8247 upon which the HM6010 is based the C ports are all separate to each channel. The R8247 has all the C ports common linked.

The performance of a CDU (internal or external) is ruled by the square of the applied voltage x the capacitance, so whilst adding extra capacitors gives you a bigger volume of energy it does not increase the intensity of that energy. e.g. you may now be able to throw 2 points if the original could throw one point. Think of a CDU as a water tank where the height of it above the tap is the voltage and the size of the tank is the capacitance. The high of the tank dictates the water pressure and the size of the tank how long the tap will run before empty.

The HM6010 performance is set at 15v DC input voltage so your energy level is dictated by that not the volume of capacitance. As port 4 is a set 12V output at higher amps by an internal voltage regulator for use by a turntable it is advised not to use port 4 for solenoids if at all possible.

In attempt to improve the reliability of solenoid throw the HM6010 output pulse time was amended from the original 100mS default to a new 500mS.

[Richard-1234084]

Is there such a thing as a circuit diagram for the 6010, especially the output section?

many thanks.

[Richard-1234084]

Is there such a thing as a circuit diagram for the 6010, especially the output section?

many thanks.

[96RAF]

Is there such a thing as a circuit diagram for the 6010, especially the output section?

 

 

There is but not in the public domain. If you expand on exactly what you need then I should be able to provide an answer.

But briefly ports 1, 2, 3 output 14vDC pulsed or steady at 150mA to drive solenoids, signals or lights and port 4 only outputs regulated 12vDC steady at up to 1A to drive a TT rotation motor or is also configurable to pulsed for solenoids, or steady for signals or lamps.

[Richard-1234084]

Thanks for the response. I’ll try and articulate the confusion. It relates only to ports 1-3 of the 6010. I am of course a novice to model railway electronics.

My background is, or was some time ago, electronics. So assumptions and underlying queries are.

Assumption 1. Is that the three wires on a points solenoid are a common and two live wires and that a single (?) pulse of suitable voltage and duration on either of the two live wires will pull the solenoid one way or another. That sounds simple.

Assumption 2. Is that each capacitor referenced in the above, charges and is then discharged into one or other of the two live wires to the points solenoid with a suitable semiconductor. If this is true so far, so good.

So question 1. Is Why is the waveform shown in the above showing more than 1 pulse?

The second question, or confusion, is around commoning outputs. I assume that the suggestion is that the capacitors are all commoned and then that when a pulse is required there is more charge available, although not more voltage. Power being proportional to the square of the voltage as discussed. It might mean that you couldn’t switch multiple points simultaneously but it might give you the max voltage for longer on one.

I guess as you say it’s a compromise.

So thanks for answering. Much appreciated. If there was a circuit diagram I could work this out for myself but as you say, it’s understandably proprietary.

And of course a compromise based on providing flexibility of usage is also understandable.

I’ll stop worrying about this and try it, maybe with some piggy back 2200 microfarad capacitors on each circuit if I have problems.

Still like to see the circuit diagram of course.

Rgds

[96RAF]

I can add to my previous based in your further assumptions and questions:

1. A solenoid motor is not polarity sensitive and most have a common and two directional wires. A solenoid will happily work on AC or DC. The output from an external CDU being DC, as is the output from the internal capacitors in the 6010. Note you cannot use an external CDU in parallel with the 6010 without some isolating device such as a pair of relays per channel. The 6010 would pulse into the relay coil which would fire the CDU pulse to the solenoid.
2. The centre wire of the 6010 ports is firmly positive with the outer wires being floating until switched negative in either direction by mosfets. The firmware has set these ports to double pulse, initially 100mS but now I believe 500mS in attempt to improve the reliability of throw. The port output can be set by the app to continuous for use with signals or lights, but the pulse characteristics are set in firmware.
3. Unlike the similar DCC R8247 which does have commoned C terminals across all four channels, the 6010 has completely separate channels. Three use VDC for their positives and the fourth has a regulated 12v positive.
4. I tried additional capacitors during beta testing to improve solenoid throw reliability, especially for paired points, with no noticeable effect, hence the adopted fix was a double pulse of longer duration as above.
5. If it helps your task I can quote component codes on the PCB e.g. Q or U numbers, but my NDA with Hornby prevents me posting the full or partial circuit schematics in public view.
6. Note that the DCC terminals on the 6010 are there to provide an alternative power input to that of the standard PSU, but at present the use of this affects blue-tooth connectivity and is not yet practical. Note that it does not parse DCC data in any way, merely dumbs it down into a power source.

[Richard-1234084]

Thanks for the response. Much appreciated. I am due a Birthday Present of a 6010 next Monday and will have a look at the innards, or as much of the circuit as I can deduce from the pcb. As an alternative, I have considered using an Arduino micro processor with suitable pulsing for points. There lies complexity of course. I am familiar with that approach but also that programming is fickle and you really need to ensure the solenoids are not overloaded. Some form of protection is required for Arduino hangs whilst driving a solenoid or more sensibly, just follow the 6010 route, listening to all the advice. Many thanks again. Quality comments and great response.

Rgds

[96RAF]

@richard

6010 with its top off...

Port 4 bottom left with heftier output. Bluetooth module top centre.

[Richard-1234084]

Many thanks for the most thoughtful of responses. Monday soon. I’ll start experimenting.

Rgds

[Richard-1234084]

@nc1

If you check an R8247 DCC acc decoder, upon which the HM6010 is largely based you will find all four of the C ports are already commoned, so you have essentially reinvented that particular ‘wheel’ for the HM6010 thus proving the earlier design theory.

I have no idea why Hornby didn’t common connect the collectors on the three mosfet ports in the HM6010. Obviously they can not be commoned with the voltage regulator output.

 

 

[Richard-1234084]

A bit late to go back to first principles but why not. The question is why have capacitors involved in the first place? Now I think I understand. The resistive impedance of the point relay is quite low. A sustained voltage across its input could lead to overheating and failure. So a good way of preventing coil failure is to switch using the charge stored in a capacitor. Worst case there, the capacitor fully discharges into the coil. An alternative would be to just switch the correct voltage onto the relay from a solid power supply, through a suitable switching transistor for the correct amount of time. But what if that timer failed. Meltdown.

So am I right? Because I am considering making my own controller through an Arduino microprocessor and the failure possibilities are high.

So that’s the question. Why the capacitive approach at all?

Rgds

[Richard-1234084]

OK people. I’ve done the sensible thing and read up about CDU systems.

No need to answer.

Rgds

[Richard-1234084]

OK people. I’ve done the sensible thing and read up about CDU systems.

No need to answer.

Rgds

[96RAF]

Simple answer is you need the oomph to throw solenoid point motors; without the capacitors the module would not be able to generate enough amperage direct from the available supply voltage.

Edit - I was already typing when your last post came in.

[NC1]

A few observations from my experiments ...

With a total nominal available input current of 1A, 3 channels at 140mA and 1 at 500mA, uses up most of that 1A. If the design was beefed up, it is likely that some of the components would overheat, particularly given the limited case ventilation.

The input is protected with a series diode for polarity reversal protection, so this drops the 15V to about 14.3V; and this is the final maximum capacitor voltage.

With some combinations of point and point motor, I found that a strong switching action would cause the moving section to bounce off the target fixed rail, but the second pulse ensured that any small gap was closed.

Where I had added another 2200uF capacior across the existing one, the "off" period, between the 2 pulses, was just about enough to recharge the capacitor to most of its target voltage for the second pulse.