TM44 analog logarithmic speed steps (not only linear)

Hello

I am using Dinamo with Z scale
Even i discussed this with Leon sometime ago, i didnt see any adjustments yet.
The linear speed steps for analog locos are not convinient.
The problem is, that with the min Voltage of 10 Volt, many Z scale locos are running 300+ km/h due to a missing step in the gear.
That causes, that i have only few speed steps left from zero to realistic speed.
If i could choose from linear to logarithmic speed steps, i could use much more steps which would cause a realistic acceleration and deceleration.
I see that i can choose it for motors and servos, but not for locos.

Isnt it possible to have this choice?
This is the only thing i have to complain with Dinamo but this causes as well that i cant convince people to use Dinamo at the moment.
Leon, couldnt you put this in the "To-Do" basket for the next release to help Z scale users?

Regards
Klaus

I have no idea if this is possible, but I would also be interested in this enhancement to allow better speed control of analogue Z gauge locomotives. It is not always an easy option to convert such models to DCC, although I have seen that DCC would also offer great improvement of speed control.

Thanks,
Les.

Dear Les, i think it can be done relatively easy but Leon doesnt pay attention to Z scale at all.
It's more than a year ago where i asked him to implement that, but there seems to be zero interest in getting z scale people as a customer.
Just today i did a speed measure with a new loco (88225 - the new BR 194 with brushless motor). I can use only 15 speeds from 63.
DCC is much easier but yes, for some locos its expensive. like steamer.
I am a bit disappointed about the attention on z scale, even i would have good access to the community (and therefore customers).
But look at the website and the efforts in Dinamo. Its more or less zero for many years. Unfortunately.

Hello Klaus,
Yes - I understand and share your concerns. I have bought into the Dinamo system and especially the TM44 for use with Z gauge due to the flexibility of running analogue and digital locomotives on the same layout, and also allowing a limited conversion of some locomotives to DCC in the future where that is possible. It's up to Leon, of course, but my feeling is that this should be a relatively straight forward adaptation in the TM44 firmware with a simple switch selection in the configuration tool. Of course, I cannot be certain, not knowing how everything works.
I hope that Leon will consider adding this possibility in the future. I would be happy to help testing if it ever became a reality.
Kind regards,
Les.

Hi Les,

I also have no idea yet if it is possible. Is your post based on your own findings with Dinamo and Z-scale loco's?
If so, which loco's have you tested and with what results? And what is the track-voltage (power supply) you use?

Best regards, Leon

Just some facts:

To be able to run sound locos, the power supply to the TM44 has to be 12,0 Volt
Any less will cause the sound is not as good as it should/can be.
Older 3-/5-pole locos are running at 300+ km/h at 12 V power supply.
This fact is reducing the available steps dramatically, if the real speed is about 80 - 100 km/h
Sometimes there are just 5 steps left
Newer locos, especially with one more gear step, are much better.

If the speed steps could be logarithmic and not only linear, there would be much more steps left for the speed between zero and 80-100 km/h

Even if i use 10 Volt, its not much better
With less than 10 Volt, some Decoder are not running anymore.

I just know from DCC (D&H and Velmo - which are mainly used in z gauge) that they use CV48 for speed characteristics.
0 = linear
7 = strongly curved

This would be fantastic to have for analog locos in z gauge

Klaus,

The first email from me to you and thus the first contact between you and me was February 23, 2022. That was 8 months ago. So if you write "It's more than a year ago where i asked him to implement that, but there seems to be zero interest in getting z scale people as a customer", this is demonstrably incorrect. Is that detail important? I think so. It sets the tone for the discussion.

I have written several times that, IF what you demand is possible, I expect that it will require a lot of effort, which cannot be seen in isolation from a number of other major changes. I also have my doubts whether the adjustments you demand are the solution. But apparently you know exactly how it works and how easy it is to adjust. I don't think you can have a clue, because you didn't design the hardware or write the software. Maybe you are right and it should be very easy, but in this case it is not. Maybe it's just me, I'm happy to acknowledge my limitations.

FYI: Fact is that there has been a serious chip shortage for about 2 years now and it is expected that this will remain so for a while. If systems can no longer be produced, the functionality no longer matters. This caused me to set priorities and they haven't worked out in your favor. Sorry for that, but I stand by the choices I made.

Kind regards, Leon

Klaus,

A BR 194 does not have a "brushless motor". It is impossible to drive a brushless motor with DC. It always takes electronics to do that. The BR 194 has (probably and as far as I see in the documentation) a "coreless motor".

Best regards

Question:

I have noticed that the cruising behavior in forward/reverse, especially at low speeds and especially with Z scale, differs considerably. Initially I was under the assumption that this was due to the motor/gear. But I noticed that it partly depends on the direction on the track. So if the locomotive moves faster reverse than forward and you physically flip it around, it will move forward faster than reverse. I noticed that at least with one locomotive quite strongly.
Can anyone confirm this is the case with them too?

Best regards, Leon

Vraag:

Ik heb gemerkt dat het rijgedrag vooruit/achteruit vooral bij lage snelheden en vooral bij Z-materieel aanzienlijk verschilt. Aanvankelijk was ik in de veronderstelling dat dit lag aan de motor/aandrijving. Maar ik heb gemerkt dat het mede ligt aan de richting op het spoor. Dus als de loc achteruit sneller rijdt dan vooruit en je keert hem fysiek om, dan rijdt hij hij vooruit sneller dan achteruit. Ik constateer dat in elk geval bij één loc vrij sterk.
Zijn er mensen die dit kunnen bevestigen dat dit bij hen ook zo is?

Met hartelijke groeten, Leon

Hi Leon,

I've never seen that phenomenon and I have 140 locos.

Differences between forward and reverse are usually caused by the gear.
Especially after many years of running in the same direction all the time.

Regards
Klaus

You are correct, sorry

Leon,

to make a long story short:
Is the requested adjustment on your To-Do list and if yes, any idea when it will be implemented (or tried to be implemented)?

Regards
Klaus

Congratulations, but I think the amount is irrelevant in this respect.
I noticed this before, however never this strong and until now I also assumed it was motor/gear. Until I did a real test with a brand new loco I did not have before.
See attached diagram. Positive direction is to the right, negative direction is to the left. And it does not matter how the loco is on the track, so there is no difference forward/reverse and therefore it cannot be motor/gear.
This is a brand new DB220 with coreles motor. Power supply to TM44 = 11.5V.
The characteristic in negative direction is quite nice, 30 usable speed steps and a minimum sustainable speed of 15km/h. It can go slower, but not very well reproducable. The characteristic in positive direction sucks.

DB220.png

Leon,

what is causing this in your opinion?

Did you try an older loco (running 300+ km/h with step 63?) as well?

An old and rather worn DB361 has a similar behaviour, although not as strong as the DB220.
The DB361 does not move before step 18 (even with kickstart on) and it reaches 90 km/h at step 29.
A logarithmic speed curve does not solve the problem that you have only 11 usable steps, since the delta's will just move up for the center region. It could even turn out to become worse.

Thats a bad candidate ;-)
The original max speed is 60 km/h
usually i add 10 to 20% which still looks ok
That makes it even worse

I'll try to do the same test you did with some old, medium and young locos
Lets see what i get

i can use one of my own analog controller on my test layout.
Maybe there is a difference between using this compare to the TM44

Will post the results as soon as available

You don't have to test 140 loco's. I'm just interested if you (or anyone) can reproduce the phenomenon I see: strong difference in behaviour by direction of travel, not (just) caused by forward/reverse of the loco.
I suspect your new BR194 has a similar motor/gear as my DB220.
And yes, if there is a difference, it is caused by Dinamo. I have a prime suspect in mind, and if I can prove it's the culprit, I may be able to come up with a solution.

Leon,

Did a trial with 5 locos

I sent the file by email because its too large to attach here

Thats the test layout

IMG_6573 (1024x323) (800x252).jpg

Leon, en anderen,

ik wil me graag als Z locomotief tester aanmelden alleen kan ik deze test pas vanaf 1 december doen. Is dat dan nog zinvol. Want ik denk dat je gelijk hebt Leon dat de locomotief analoog een verschillend rijgedrag vertoont voorwaarts in de andere richting. Dat wil graag nader voor je onderzoeken, maar dat lukt me nu dus even niet.

Met vriendelijke groet,

Geert

Dear Leon and others,

If needed I will test for you several Z analog locs, accept that I can't start before the first of December. Is that ok and helpfull.

Regards,

Geert

Hallo Geert, Altijd nuttig hoor.
Mvg, Leon

Hi,
Are there any news available?

Cheers
Klaus

Leon,

zoals afgesproken een test gedaan vanuit DinamousersCC waarin ik een betrekkelijk lang blok ( 1m) gebruikt heb om te testen, dit blok gekoppeld aan naast liggend blok als overloop. Meeting verricht waarbij Rechter railstaaf ( A ) en linker railstaaf ( B ) bij vooruit rijden. Beide blokken worden gevoed uit dezelfde TMH en hebben ieder een 3tal bezetmelders. (extra CD16 bij geplaatst ). Rails voor aanvang test gereinigd. Test uitgevoerd met analoge locomotieven die weinig tot geen uren gereden hebben.

BR94 met klokanker : Snelheid ingesteld op 7 in programma vertraging 1. Rijdt richting vooruit aanmerkelijk sneller dan richting achteruit ook bij omgedraaide loc.

BR260 met 5pool : heeft hogere snelheidwaarde nodig om te rijden. Snelheid ingesteld op 11 in programma vertraging 1. Opvallend is dat de loc met korte - (achter) zijde beide kanten uit harder rijdt dan met lange- (voor) zijde.

BR41 met 3pool Zonder tender: heeft nog hogere snelheidwaarde nodig om te rijden. Snelheid ingesteld op 21 in programma vertraging 1. Door hogere start waarde geen noemenswaardig verschil richting vooruit / achteruit ook niet na omdraaien loc.

Binnen iTrain snelheidsmeting locomotief op ijkspoor bemerk ik duidelijk verschil tussen vooruit tov achteruit.
Bijgevoegd iTrain meting Dinamo 64 stappen analoog BR94 klokanker:

Hopende iets te kunnen bijdragen. Mochten er onduidelijkheden zijn of andere metingen laat 't me weten.
Met vriendelijke groet,
Geert

iTrain snelheidsmeting BR94 klokanker analoog Dinamo.jpg

Hallo Geert,
Ik maak op uit je verhaal dat je een soortgelijk effect ziet als ik.
Het verschil naar links/rechts (ik vermijd termen vooruit/achteruit omdat dit te verwarren is met de richting van de loc) is het sterkst bij klokankers minder bij de 5-polige motors en blijkbaar verwaarloosbaar bij de 3p. Dat laatse kan ik niet testen, want ik heb geen 3 polige, dus dank daarvoor. Je kunt ook zeggen: hoe gemakkelijker een motor opstart, hoe groter het verschil.
Mvg, Leon

Leon,

jij verwoordt het (effect) correct .

Groet en dank,
Geert

Klaus, and possibly others running scale Z with TM44,

Can you try the attached TM44 firmware?
On my test track it seems a tremendous improvement on low speed.
Warning: BETA version!
Occupation detection may not be fine yet. It needs to be modified. I applied a quick fix, but needs some more coding.
This version will not interoperate with TM-H's!

Best regards, Leon

Leon,

could you please send me a link to the AVrootloader and any other config files i need?
I forgot where i can download those.
As soon as i have installed all of them, i will trial your version and report.

Thanks
Klaus

Leon,

i still get this message when i lock in at dinamousers.net

You have not been granted a customer status but have already applied for it on February 5, 2022

Could you please change that?

Can i find the necessary software there?

Regards
Klaus

This forum: Dateiarchive > Software > VPEB Bootloader

To update the TM44's you have to set your RM-C in Bootloader Transparent Mode.
If not done already: Download and install DinamoConfig1.40
Start DinamoConfig, Choose COM-port, click "Bootloader Transp. Mode" in tab "RM-x", Close DinamoConfig.
Update TM44's using AVRootloader. Use TM44 manual paragraph 8.2 from point 6 onward (if needed).
Reboot RM-C when done to get your system back running.

Best Regards, Leon

Leon, if i click on your link for DinamoConfig1.40, nothing happens.
Is it because i am not registered as a customer?
Where can i find the AVRootloader?
Regards
Klaus

All 3 links in my last post work on my PC, my MacBook and my smartphone. Checked, doublechecked.
You are registered as customer. Checked, updated, checked.

Hi Leon,

got it done now.
Thanks for your help.
As soon as i got the firmware update done, i will test the new version.

Regards
Klaus

Leon, you are my hero.
Huge difference at low speed now with any kind of motor.
Thank you so much.
Regards
Klaus

Leon,

indeed for me too. The "Slow motion " is much Much Better.

Dank Leon, ook ik was in de gelegenheid om te testen. De benoemde BR94 in testbericht rijdt nu zelfs met stand 3 in DinamoUsersCC en in beide richtingen gelijkwaardig en langzaam.

Nogmaals dank,

Geert

Yes, both directions in i-train with turned loco work the same good.
Thank you again Leon

Dear All,

did some speed measuring drives and one of them was the BR 03 1001 (Märklin 88850)
This loco has a max speed of 130 km/h
With the old firmware i had 17 steps with a lowest speed of 59 km/h
With the new firmware i have 37 steps with a lowest speed of 28 km/h
Massiv positive change
Thanks again Leon

Regards
Klaus

88850_MESS_NEU (500x493).jpg

To complete my previous post, here the old speed measuring curve

88850_MESS_ALT (500x495).jpg

Klaus, Geert (and others reading along),

Good to read that you are happy with the improvement.
FYI: These modifications are in the core of the TM44 code, which is written in Assembler for precise timing and speed. It's a lot of byte swapping, inverting, shifting, or'íng and and'íng to squeeze pwm, dcc, hfi, etc in. So I needed a period of high concentration to change this properly. Thank you for your patience.

NOTE that the code you have has some bugs. Attached a version 1.2.2.A. that has 2 fixes:

• Loco is not running in the wrong direction anymore when HFI is switched on
• Occupation detection is changed to match the modified pwm waveforms

The attached version still is a beta-version. I will not release the new version before also a number of other changes are implemented, but I guess it will be ok for you for the time being.

Geert, I thought you use TM-H and CD16. Beware that this modification will not (ever) be available for TM-H. The TM-H hardware cannot support the way the H-bridges are driven by the TM44 now. Sorry for that.
And ... DinamoUsersCC is quite outdated. I advise you to install DinamoConfig 1.40. You can testdrive with that too and it offers quite some additional configuration and testing options.

Best regards, Leon

Dank Leon,

voor je heldere uitleg.

"Geert, I thought you use TM-H and CD16. Beware that this modification will not (ever) be available for TM-H. The TM-H hardware cannot support the way the H-bridges are driven by the TM44 now. Sorry for that."

Leon, naast de TM-H's heb ik ook een 2 tal TM44 wel ik geupdate heb en daar de laatste ( deze ) test op gedaan.
Bewust tijdens de opbouw van de baan de TM44 geplaatst in het zichtbare stations-gebied ivm het aantal bezetmelders per blok, wat ik in het schaduwstation en rangeerterrein heb opgelost met TM-H en extra CD16.
Dus gelukkig kunnen testen met heel positief resultaat. Misschien in de toekomst het rangeerterrein ook uitrusten met TM44's.

"And ... DinamoUsersCC is quite outdated. I advise you to install DinamoConfig 1.40. You can testdrive with that too and it offers quite some additional configuration and testing options. "

Bedankt voor de tip. Natuurlijk de DinamoConfig 1.40 al lang geïnstalleerd en zelfs gebruikt tijdens de update om de RMx in Transparante mode te zetten. Maar daarna onbewust teruggegrepen naar DuCC.

Just to let all followers of this thread know that Leon has announced a new firmware update for the TM44
https://www.dinamousers.net/forumthread14117-TM44-firmware-1-23-Experimental?topics_offset=0

Thank you, Klaus.

Dear Leon,
Thank you for your efforts in making these changes and improvements. I must apologise for not having been an active participant in this discussion. At the end of last year "real-life" threw some unexpected distractions in my direction. Fortunately I am getting past them now, and I have been setting up a pair of TM44 units on my test layout which will make my life easier. If I have any particular feedback to add, I shall make it in the thread that you started concerning the 1.23 experimental release. I note the comments there concerning the use of the updated Dynamo Config program (version 1.41).
Thanks again and kind regards,
Les.