DJM class 71 Slow & Jerky. CV settings?

[ianthetrog]

Just got the DJ models class 71. It's very noisy and jerky at low speeds and flat out only goes at about half speed. 

Anybody know if cvs can be altered to aid this? Using a Bachmann decoder with Hornby elite controller. Have tried another decoder but no different

Thanks for any advice before it goes back.

Just got the DJ models class 71. It's very noisy and jerky at low speeds and flat out only goes at about half speed. 
Anybody know if cvs can be altered to aid this? Using a Bachmann decoder with Hornby elite controller. Have tried another decoder but no different

[dpgibbons]

First off I suggest you remove the decoder and test on DC to ensure it's not an issue with the loco. 

[RFS]

A Bachmann decoder is not a good choice for this loco, especially if it's the 36-554 version. You may need to adjust the Back EMF settings in CV 54/55 - see http://www.Bachmann.co.uk/pdfs/decoder_settings.pdf

 

Top speed is quite low - mine is 55 mph with a Lenz Silver decoder.

 

Bear in mind also that if you want the head code lights to work on DCC you'll need a decoder with a minimum of 5 functions. 

[NIK]

Just got the DJ models class 71. It's very noisy and jerky at low speeds and flat out only goes at about half speed. 

Anybody know if cvs can be altered to aid this? Using a Bachmann decoder with Hornby elite controller. Have tried another decoder but no different

Thanks for any advice before it goes back.

Just got the DJ models class 71. It's very noisy and jerky at low speeds and flat out only goes at about half speed. 

Anybody know if cvs can be altered to aid this? Using a Bachmann decoder with Hornby elite controller. Have tried another decoder but no different

 

Hi,

 

I haven't received my DJ Models Class 71 yet so I haven't got a lot of info but have you removed the capacitor that is normally in the wiring circuit/circuit board that leads to the motor?.

 

I'm only guessing about the capacitor being the cause as the Class 71 has a coreless motor - with higher resistance but much less inductance than standard DC permanent magnet motors so maybe the surge of current to the capacitor is throwing out the back-emf pulse width control of the DCC decoder.

 

On the Dapol OO Class 73 the capacitor is a smallish rectangle soldered directly to the PCB and I carefully removed it with a small tipped soldering iron and some solder wick.

 

Alternatively to try something that wont modify the loco anymore - see if any of you decoders have the facility to turn the back-emf feature on and off, preferably via a function button.

 

Then you could see if the back-emf feature is causing the slowdown and what its top speed on DCC might be.

 

I assume you've set the DCC decoder Vmin, Vmid and Vmax and any other CVs appropriately to give maximum top speed before you tested the class 71?.

 

Regards

 

Nick

Edited May 20, 2017 by NIK

[chris p bacon]

First off I suggest you remove the decoder and test on DC to ensure it's not an issue with the loco. 

I'm with dpg.

 

Before doing anything see how it runs with straight DC.

[ianthetrog]

Thanks for the advice chaps. Nothing works so must have got a Dud.

[chris p bacon]

Thanks for the advice chaps. Nothing works so must have got a Dud.

When you say "Nothing works" what do you mean ? In your first post you say it's "slow and jerky" so it did work (of a sorts) 

 

What was it like when powered on straight DC (no chip)

[dpgibbons]

Just in case this isn't obvious - if you remove the decoder you need to substitute a blanking plug to get the loco to run on DC

[DJM Dave]

Thanks for the advice chaps. Nothing works so must have got a Dud.

Just read this thread, i'd also like to know what does and doesn't work please and maybe i and others here can help?

 

cheers

Dave

[csvt2004]

Well just fitted mine with a loksound decoder. Jerky at low speed and poor top speed. I disabled the bemf settings on the lok programmer as advised earlier on this list. Now runs beautifully.

[ianthetrog]

Thanks again. 

I cant try it without DCC chip, as my layout is 100% digital control and I dont possess an analogue controller anyway. 

I have tried it with a Bachmann 557 chip, a Sapphire and a Laisdcc chip. All give the same result.

I appreciate all your suggestions but as a non technical person, all the jargon goes right over my head (Vmin, Vmid and Vmax bemf etc?)

To reiterate the problem. At low speed it jerks badly whilst picking up speed. The motor is noisy, like an old diesel tractor ticking over, Plus it derails at points, even going straight forward.

As stated, I am not very technical and have never had to deal with anything like this on all the Hornby, Bachmann & ViTrains locos I have. Just very disappointed that such an eagerly anticipated addition to my fleet requires such a faff.

[7013]

Ianthetrog looking at the cl71 thread on the DJModels section coreless motors have to be tuned to the decoder to get the best performance. I agree that there is a lot to learn about DCC, I have been using it for years and still struggle to understand some of it. Get yourself a good book and start the learning process. When you do start to master some of the concepts you will be surprised how much you enjoy it.

For now I suggest you find someone who can help you tune the decoder for optimum performance with a coreless motor.

[chris p bacon]

Thanks again. 

I cant try it without DCC chip, as my layout is 100% digital control and I dont possess an analogue controller anyway. 

Is there a club nearby or do you know anyone else with straight DC, you have to make sure the motor runs well first, then you'll know if it is the decoder settings.

[SRman]

My green E5004 arrived today. Ran it in on DC as usual, no problems there. I fitted what I thought was a TCS decoder but when I read it on the programming track, it was from manufacturer 99 - Lenz!

Anyway, I set it up with the Lenz decoder, with my usual settings of 0 in CV2, 25 in CV3 and 18 in CV4 (later changed down to 15). Like the earlier reports, it ran fine at faster speeds but was a bit chattery at very low speeds. From the slight surging at those slow speeds I would say it's a BEMF problem. The top speed seemed fine to me.

I'll wait for the weekend before connecting JMRI Decoder Pro to the programming track and see what I can do with the settings for BEMF - it may even need to be disabled completely.

My other 71, E5003 has an ESU LokSOund v4 and that ran beautifully straight away after fitting the decoder. While I have Decoder Pro connected, I'm going to see if I can activate aux 3 and attach it to a function to get the headcode lights working - someone suggested the PCB may already have the transistors in line to allow this to work, but we'll have to wait and see.

I'll report back if I have anything positive to share.

[Shroomy]

Ianthetrog looking at the cl71 thread on the DJModels section coreless motors have to be tuned to the decoder to get the best performance. I agree that there is a lot to learn about DCC, I have been using it for years and still struggle to understand some of it. Get yourself a good book and start the learning process. When you do start to master some of the concepts you will be surprised how much you enjoy it.

For now I suggest you find someone who can help you tune the decoder for optimum performance with a coreless motor.

Interested in what book you would recommend as I would like to know more

[Budgie]

My green E5004 arrived today. Ran it in on DC as usual, no problems there. I fitted what I thought was a TCS decoder but when I read it on the programming track, it was from manufacturer 99 - Lenz!

 

Anyway, I set it up with the Lenz decoder, with my usual settings of 0 in CV2, 25 in CV3 and 18 in CV4 (later changed down to 15). Like the earlier reports, it ran fine at faster speeds but was a bit chattery at very low speeds. From the slight surging at those slow speeds I would say it's a BEMF problem. The top speed seemed fine to me.

Try setting CV50 to 32 and CV9 to 63, and see what happens.

 

I'll report back if I have anything positive to share.

Please report negative results too, so that we can avoid the same pitfalls.

[Shroomy]

Try setting CV50 to 32 and CV9 to 63, and see what happens. Please report negative results too, so that we can avoid the same pitfalls.

Budgie how will this affect bemf and consequently the loco please

[Budgie]

Budgie how will this affect bemf and consequently the loco please

From the decoder information booklet, section 6.2

Six different motor types (not to be confused with motor designs, such as a bell-type armature motor) can be selected in CV50 to adapt to the respective locomotive model. These motor types include parameter sets which have been specially adapted to the respective models. Moreover, it is possible to carry out fine-tuning via CV113 or CV114 when selecting motor types 4 or 5. Of course you can switch off both the high-frequency drive as well as the control system itself. You can still use CV9 to adjust the repeat rate.

...

The decoder still has a so-called EMF switch which makes it possible to adjust the decoder to different motor types. Depending on the motor type used, it is possible that a digitally controlled locomotive cannot reach an adequate maximum speed compared to a locomotive in conventional operation. If this is the case, activate your EMF switch by setting Bit 6 in CV 50. The locomotive will then reach a higher maximum speed while the minimum speed is also slightly increased.

 

Setting CV50 to 32 makes the EMF switch active and says the motor type is 0.

 

When I made the change to CV50, the voltage I measured across the coreless motor on my ESU decoder tester immediately went up from 9.6 to 14.7.

Edited June 1, 2017 by Budgie

[Shroomy]

From the decoder information booklet, section 6.2Six different motor types (not to be confused with motor designs, such as a bell-type armature motor) can be selected in CV50 to adapt to the respective locomotive model. These motor types include parameter sets which have been specially adapted to the respective models. Moreover, it is possible to carry out fine-tuning via CV113 or CV114 when selecting motor types 4 or 5. Of course you can switch off both the high-frequency drive as well as the control system itself. You can still use CV9 to adjust the repeat rate....The decoder still has a so-called EMF switch which makes it possible to adjust the decoder to different motor types. Depending on the motor type used, it is possible that a digitally controlled locomotive cannot reach an adequate maximum speed compared to a locomotive in conventional operation. If this is the case, activate your EMF switch by setting Bit 6 in CV 50. The locomotive will then reach a higher maximum speed while the minimum speed is also slightly increased.Setting CV50 to 32 makes the EMF switch active and says the motor type is 0.When I made the change to CV50, the voltage I measured across the coreless motor on my ESU decoder tester immediately went up from 9.6 to 14.7.

Brilliant thank you Budgie

[RFS]

From the decoder information booklet, section 6.2

Six different motor types (not to be confused with motor designs, such as a bell-type armature motor) can be selected in CV50 to adapt to the respective locomotive model. These motor types include parameter sets which have been specially adapted to the respective models. Moreover, it is possible to carry out fine-tuning via CV113 or CV114 when selecting motor types 4 or 5. Of course you can switch off both the high-frequency drive as well as the control system itself. You can still use CV9 to adjust the repeat rate.

...

The decoder still has a so-called EMF switch which makes it possible to adjust the decoder to different motor types. Depending on the motor type used, it is possible that a digitally controlled locomotive cannot reach an adequate maximum speed compared to a locomotive in conventional operation. If this is the case, activate your EMF switch by setting Bit 6 in CV 50. The locomotive will then reach a higher maximum speed while the minimum speed is also slightly increased.

 

Setting CV50 to 32 makes the EMF switch active and says the motor type is 0.

 

When I made the change to CV50, the voltage I measured across the coreless motor on my ESU decoder tester immediately went up from 9.6 to 14.7.

 

Unfortunately it doesn't seem to make the loco go any faster.  In fact top speed was reached on mine on speed step 100, and increasing to speed step 126 didn't cause the loco to go any faster.  

[SRman]

I did have a play with CV50 settings last night, trying different motor types (does anyone have a list of what each type actually represents - I couldn't find one anywhere).

I also tried combinations of the bits in CV50, first of all trying values of 1 to 5 for the motor types, then 32 to 37 (additives of 32 plus motor types. I got the most silent running with a value of 37, but it still did a bit of the slow (crawling) speed chattering. 

I have not had any problem with smoothness from slowish medium speeds and above, nor with the top speed.

I hooked up Decoder Pro (JMRI) and played with some of the BEMF settings as well but didn't entirely achieve what i wanted. Turning off BEMF altogether resulted in a smoother and quieter crawl but slightly unpredictable results for initial movements. I cranked up the CV2 setting to compensate but even that didn't give me predictable starting notches (I expect a locomotive to just creep on speed step 1).

I will try Budgie's settings when I get back home from work this evening (thanks for those, Budgie). 

Incidentally, I also tried plugging in a TCS EU621 (6 function) decoder but that produced all sorts of lighting anomalies. It buzzed initially but settled down to produce very smooth running at all speeds, but I couldn't get the lighting working correctly at all, so plugged the Lenz decoder back in.

While I had the JMRI window open, I also tweaked the cab light operations so that F1 operated the number 1 end light and F2 operated no. 2 end (reversing them from the default and making that a little more intuitive as to which function works what!). I also dimmed all of the lights while there.

So, overall, I am getting closer to what I want from the locomotive, but it's not quite all the way yet. I am still happy wit hit though, it is a nice bit of kit.

[SRman]

Budgie: I tried CV50 =32 and CV9 = 63 but it made absolutely no discernible difference to the slow running chatter, although it made the loco slightly more responsive to speed step 1.

I then tried some other values for comparison, but all seemed to me to be close to the same, with the low speed chatter: for the record, I tried CV50 =0, and CV9 = variously 0, 20, 40 and 63 (which seems to be the highest value). I was not looking at top speed for any of the settings. CV9 = 20 seemed to reduce the chatter slightly.

While on the subject of CV changes (although this may be a bit fraudulent since I used Decoder Pro and don't know what the actual CVs or their values were!) I tweaked a couple of the function mappings on E5003 with its ESU LokSound decoder. My intention was to get the headcode lights working on F20, but what I have currently achieved is the headcode lights work as soon as the loco has power, but switch off if I activate F19 for the cab lights (which are set to work according to the direction selected). Switching off the cab lights lights up the headcodes again. Strange but I can live with this until I work out exactly where I went wrong.

For the record, using Decoder Pro, I ticked a box under F20 on line 28 and also a box under aux 3 for the same line.

At least it proves what someone suggested elsewhere, that the DJM model has the necessary amplifiers built-in to its circuits to allow the logic output of the LokSound decoder to actually work the headcode lights.

I'll keep you all posted as and when I experiment further.

p.s. WHile fiddling with the light settings in Decoder Pro (currently using v 3.10), I dimmed all of them a bit. The headcodes and cab lights could still be dimmed further, but that can wait until my next session.

Edited June 2, 2017 by SRman

[Al51]

 

Incidentally, I also tried plugging in a TCS EU621 (6 function) decoder but that produced all sorts of lighting anomalies. It buzzed initially but settled down to produce very smooth running at all speeds, but I couldn't get the lighting working correctly at all, so plugged the Lenz decoder back in.

 

 

 

I have a TCS EU621 decoder fitted to my E5004, the only lighting anomaly I had was that F1 worked the reverse direction cab light and F2 the forward one, easily sorted by reversing the decoder direction in CV29. Both headcode lights come on with F3, but I think that is how it's wired. 

 

Movement is very smooth although I've only tried it on the rolling road and a very short piece of track.

 

Al

[SRman]

Yes, I'm not quite sure what was going on with the TCS one in mine. The white marker lights at one end remained on at all times, whether selected or not, and the cab lights at one end wouldn't respond. 

Anyway, I have played around more with the settings on the Lenz decoder, again through Decoder Pro on the computer. I turned off the BEMF completely, then tweaked the start voltage to a value of 14 (after several experimental values ranging from 1 to 15). There is a bit of a delay on moving off when the throttle is opened but the slow running is absolutely dead silent now, as it should be.

After all of this, though, I have ordered a Zimo MX634D, and when that arrives I'll pop that in to see whether I can improve things further.

Now, to the other problem I reported with E5003's lighting; I solved that by rereading the function mapping sheet in JMRI, and worked out that I must have gone a bit cross-eyed when I ticked the box for F20 on line 28 ... what I had actually ticked was the 'off' part of F19. Having now corrected that, the headcode lights work as they should when F20 is selected for the ESU LokSound decoder.








I have also dimmed the lights further since taking the photos.

Edited June 3, 2017 by SRman

[Nigelcliffe]

...............

 

Now, to the other problem I reported with E5003's lighting; I solved that by rereading the function mapping sheet in JMRI, and worked out that I must have gone a bit cross-eyed when I ticked the box for F20 on line 28 ... what I had actually ticked was the 'off' part of F19. Having now corrected that, the headcode lights work as they should when F20 is selected for the ESU LokSound decoder.

 

 

 

 

You are using a geriatric version of JMRI to setup an ESU decoder.   Given the rate of code-changes in JMRI to support ESU sound decoders, I would recommend you upgrade to a current version.   As well as improved code, the new interface is a lot better than the hundreds of checkboxes in your screen shot.

 

 

- Nigel