HM7000-21TXS doesn't work with DCC decoder test board

[A User]

Hi all,

I've just spent two frustrating evenings trying to figure out why I couldn't get two brand new 21MTC HM7000 decoders working. Turns out that there appears to be a compatibility issue with at least some DCC decoder test boards that causes the MCU that deals with motor control and sound to refuse to work.

The symptoms I was getting with the decoder plugged into the test board was as follows:

• The Hornby HM DCC app could see and communicate with the decoder via Bluetooth (including updating Bluetooth firmware on the MCU that handles Bluetooth operations), but was unable to load profiles/firmware to the APROM/SIPROM.
• Motor control, function outputs and sound features wouldn't respond to control requests via Bluetooth or DCC.
• Profile ROM, Boot ROM, Profile Version and Hardware Version for the decoder were all showing as "v0" in the app.

In a frustrated last ditch attempt I soldered some leads for track power to a 21MTC breakout board and chucked one of the decoders on that with no other connections to rule out the decoder test board as causing problems and sure enough everything worked as it should. I was finally able to load a profile to the board and after wiring motor leads to the breakout and plugging in the speaker was able to verify the @#$% decoder worked.

The DCC decoder test board I was using is brand new and comes from what I think is probably a fairly well known if niche supplier. Visually it seems fairly similar in design to many other decoder test boards on the market, so consider yourself warned that these decoders might not play nicely with a few, or possibly more than a few of the decoder test boards that are out there. In trawling through the forums looking for a solution to the problem I was having I did see a couple of recommendations to get a decoder test board as part of troubleshooting steps for those that were having issues with decoders - perhaps be a little wary of this as it may hinder more than it helps.

I don't have the schematic for the test board so it's hard to guess at exactly what is occurring and how to go about modifying the test board to work with the HM7000-21TXS decoders I'm trying to work with.

It seems likely that one of the signals in black text in the table below is conflicting with the board as they are connected to the motor/sound MCU that is being disabled. The blue signals are connected to the controller on the decoder that deals with Bluetooth operations, so they are less likely to be causing the issue. Even though it’s connected to the Bluetooth MCU I’ve ruled out pin 6 as the Hornby functional description is a bit suss by wiring it from the 21MTC breakout to one of the AUX function channels on the DCC decoder board as it would have been when connected directly to the 21MTC connector on the board (along with track power and motor outputs) and the decoder still operated as expected.

My questions to the brains trust are:

- Has anyone run into this behaviour before?

- Does anyone know which pin on the decoder is likely to be causing the motor/sound MCU to stop working when connected to the DCC decoder test board?

- Has anyone tested HM7000-21TXS with a DCC decoder test board that it works reliably and doesn't lock up with and can recommend the brand and model of test board?

[HST Mainline]

Are you sure it is the test board? The test board is simply a locomotive without wheels. It should just work.

[sir john]

I use an esu decoder tester/programmer never had a problem setting my hm7000 decoders up on it either with Bluetooth or dcc for which I use my nce powercab 

[ntpntpntp]

I agree with HST, the decoder board is just standing in for a loco's circuitry, motor and speaker. Doesn't make any sense to me, it's just a bunch of LEDs, resistors, a motor, and a speaker.    I have an older model ESU tester board and it's been fine with everything I've connected to it. 

[SteveM6]

Why keep the make of test board a secret?

[Rog RJ]

Lais???

[LTSR_NSE]

Have you given the decoder tester board a visual inspection in case of poor soldering causing a short/bridge between pins or PCB tracks?

[ColinB]

Nop can't be LaisDCC as I use that one and it works ok. Sound is a bit "tinny" but then when you look at the loudspeaker you can see why.  I can't see why the motor won't work or the speaker, it is pretty standard stuff. The other thing you could is plug in another 21 pin sound decoder if you have one and see if it woirks. Come to that any old 21 pin decoder will be good enough to test the motor function.

Edited August 6 by ColinB

[96RAF]

I have used a genuine ESU decoder test rig for the whole development phase for all plug orientations without any problems. I have 3 off these boards and use them as test locos to add load to the HM7K system.

My initial thoughts were Liais (no guarantee it meets any spec reverse engineered knock off) but Colin doesn't seem to have any problems with his.

Tell us the make and Hornby will test it.

[ColinB]

I suppose the motor could drawing too much current at start up, we do know from other posts this can be an issue. Then there is the speaker, it might be a 4 ohm, that would cause issues.

[ntpntpntp]

@ColinBI wondered about the speaker impedance having any effect?    My ESU's got a speaker switch (I forget the choice of values 8 ohm / 16 ohm I think) and also a speaker OFF position.  Lais' tester board doesn't appear to have a switch but the board says 8 ohms 0.5 watt speaker in all the images I've checked. 

Certainly whatever tester board it is could do with a detailed close-up inspection for faults as @LTSR_NSE  suggests.

Edited August 6 by ntpntpntp

[LTSR_NSE]

In addition to visual inspection (if nothing looks obviously wrong)…

I would recommend acquiring a set of male-female DuPont (jumper) leads.  They would enable you to connect decoder to tester board pin by pin to isolate the faulty connection/component.  Obviously you could achieve the same result by soldering up your breakout board pin by pin - however the leads are both quicker & simpler for multi-use!

[A User]

Checked and rechecked - it's definitely the test board causing the decoder to fail. Speaker on the board should be disconnected (switch is off). Board is from Tramfabriek - I've sent them an email and wanted to give them a chance to respond before I pointed any fingers their way, but it appears they are away for a few days.

Looks like I might have to swap the board for one from Lais or ESU, but I'll connect up the remaining pins with Dupont leads and see if I can suss out what's causing the issue before I send it back.

[Pendragon Sailing]

I use an ESU Board too, with no issues encountered.

[96RAF]

The Tramfabreick board is fairly simple device. The only noticeable difference from other boards is a coreless motor. I suspect a board fault such as a short trace to trace or similar. I had an ESU one where the 6-pin socket wasn't soldered to the board.

 

[Dukedog.]

I use a ZIMO test board not had any issues yet.

[ntpntpntp]

I must admit the Tramfabriek / DrM tester is a new one on me - I know of the ESU and Lais testers.      Should just to the same job though. 

How are you powering the tester board to run the HM7000 decoder - DCC or DC? 
 

[ColinB]

It is probably the coreless motor that is causing the issue, has anyone checked it works with HM7000?

[sir john]

@ColinB colin my esu tester has a coreless motor as standard, it's never failed to set a decoder yet

Edited August 15 by sir john
Misspelt

[ntpntpntp]

1 hour ago, ColinB said:

It is probably the coreless motor that is causing the issue, has anyone checked it works with HM7000?

Why would a coreless motor make any difference?  Most decoders can be tuned to run a coreless, and even if not tuned I can't see why the motor would cause the decoder to misbehave as described by the OP. 

 

[ColinB]

1 hour ago, ntpntpntp said:

Why would a coreless motor make any difference?  Most decoders can be tuned to run a coreless, and even if not tuned I can't see why the motor would cause the decoder to misbehave as described by the OP. 

 

Well you tell me, there isn't a great deal on that board that would stop it working. Many decoder manufacturers make a point about them working with coreless motors, there must be a reason for that. Putting on my electrical engineering hat, it is well known that they have the ability to draw more current, so I assumed that their current would peak at start up. That is the only thing that I can attribute that to. I don't have one of those tester so  I can't test the theory, I have a LaisDCC one which has a normal 3 pole motor.

[ColinB]

2 hours ago, sir john said:

@ColinB colin my esu tester has a coreless motor as standard, it's never failed to set a decoder yet

As I said on the other response, what else is there on that board to cause a failure, perhaps your coreless motor is a different rating. If I remember rightly Zimo actually state in their specification that their decoders are happy with coreless motors, if there wasn't an issue with them why would they state that. Doesn't one of the analogue controls have an issue with coreless motors and feedback control, I am sure I have heard Sam mention it. Bachmann use a lot of coreless motors but generally I use a Zimo decoder to control them so I don't know how a HM7000 will perform.

[ntpntpntp]

7 minutes ago, ColinB said:

it is well known that [coreless motors] have the ability to draw more current

Quite the opposite!  They are more efficient and draw less current than traditional iron cored motors.  The problem decoders can have with them is the difference in Back-EMF and their need for higher frequency PWM drive to prevent the heat build-up which they cannot dissipate.

[ColinB]

19 minutes ago, ntpntpntp said:

Quite the opposite!  They are more efficient and draw less current than traditional iron cored motors.  The problem decoders can have with them is the difference in Back-EMF and their need for higher frequency PWM drive to prevent the heat build-up which they cannot dissipate.

Ok generally the amount of current a motor draws is offset by the generator effect which gives you back emf. So if you get a lot of back emf the current drawn goes down. With a coreless motor as you rightly say you don't get any back emf, so therefore they must use more current. You further endorse that by saying that they get heat build up. What causes heat build up volts multiplied by amps, so I assume you avoid overheating by adjusting the mark/space ratio on the PWM. I won't pull rank but lets say I know a little bit about electronics especially PWM (we used it to control lots of things on a vehicle engine). I will be honest my motor knowledge is a bit limited I just did the labs and wrote up the reports but I did read the journal on how they worked and the issues with them. Either way as I said look at that board and assuming it hasn't got a fault, tell me what you think is causing the issue.

[ntpntpntp]

42 minutes ago, ColinB said:

With a coreless motor as you rightly say you don't get any back emf, so therefore they must use more current.

They draw noticeably less in my experience. That's easily demonstrable just run a coreless motored loco on a bench supply with voltage and current metering, and compare with a similar loco or having an iron cored motor.  When I'm servicing my N gauge locos I expect a typical motor current draw of cored motors to be somewhere around 150mA (quite a bit more when it comes to old 1970s/80s motors), but with locos I know to have coreless motors I see more like 50-75mA.   People upgrading locos with one of the many coreless motor kits out there are noticing the reduction in current draw as well as the improvement in smoothness of running.


Anyway, I'm with @96RAF and this could be a board fault.  It would be interesting to swap the motor for a cored type - or for someone else to connect a coreless motor to their ESU or Lais test board and try a similar test with a HM decoder.