Jump to content
 

Fabrication of Small Capacity DIY Locomotive Stay-Alive


Recommended Posts

9 hours ago, pygmalion said:

OK folks, I have finally reached stay-alive experiments.

I made 4400uF stay-alive (2x2200uF) for my 860014 LaisDCC decoder with black and blue stay-alive wires.

.........


Now I'm am clueless. Either I damaged LaisDCC somehow by providing stay-alive with DC turned on, or the decoder is just a piece of junk.

Does anyone have any ideas on this?

 

 

My starting point is "piece of junk".  However, it ought to work.  There may be a setting in the Lais for how long it runs without a valid DCC packet ?  No idea what Lais did when they cloned an old TCS decoder.  

 

Such a circuit will work on a Zimo decoder, or anything else which is half-decent.   4400uF (if you mean 4400uF, not 440uF ??) should give a clearly visible run time.  440uF probably won't be visible.     

My most recent build is about 2300uF in a small 4mm scale 0-4-0 fitted with a Mashima motor and a 100:1 gearbox.  That gives 6 to 30mm of running distance without power, depending on the speed when power is removed.  

 

 

- Nigel

Link to post
Share on other sites

7 hours ago, Crosland said:

Why do you expect the LED to stay on for longer?

 

Maybe the motor is running for a fraction of a second. After that the stay alive is drained and nothing will work.

 

For a test, I turned off motor (but leaving LED on) and then disconnected power. Since motor was not working, I expected to see LED on for long enough to see that stay alive is working.

 

Is there any other way to test stay alive function?  I think 4400uF should be enough to get something visible.

 

Maybe it is not exactly 4400uF but it is definitely more than 440uF. My multimeter measures 3.9mF. There are two 2200uF 25V electrolytic capacitors in parallel, so ERS should be small.

 

Edited by pygmalion
Link to post
Share on other sites

34 minutes ago, pygmalion said:

 

For a test, I turned off motor (but leaving LED on) and then disconnected power. Since motor was not working, I expected to see LED on for long enough to see that stay alive is working.

 

Is there any other way to test stay alive function?  I think 4400uF should be enough to get something visible.

 

Maybe it is not exactly 4400uF but it is definitely more than 440uF. My multimeter measures 3.9mF. There are two 2200uF 25V electrolytic capacitors in parallel, so ERS should be small.

 

 

Ways to test things  ?   

Decoder not in loco, instead fitted to a test rig, so you know where any power is going.   But, 4400uF should give something visible on a motor. 

Or, measure the analogue DC current consumption of the loco, then you can do some very rough estimates of capacitor run time.  
Or say what motor/model is being powered here. 

 

Decent decoders, such as a £20 Zimo, will work.  

 

 

- Nigel

 

Link to post
Share on other sites

On 12/06/2021 at 18:54, Nigelcliffe said:

 

Ways to test things  ?   

Decoder not in loco, instead fitted to a test rig, so you know where any power is going.   But, 4400uF should give something visible on a motor. 

Or, measure the analogue DC current consumption of the loco, then you can do some very rough estimates of capacitor run time.  
Or say what motor/model is being powered here. 

 

Decent decoders, such as a £20 Zimo, will work.  

 

 

- Nigel

 

 

My motor is the Mitsumi M24N-1 and I even measured its electrical characteristics (you can find them published here).  So without load, the motor consumes something on the order of 15mA, while LED consumes something on the order of 10mA. So theoretically LED should glow for at least 0.4s and the motor should work for at least 0.2s ( giving 20mA for the motor with load, I have my loco on my DIY test bench).  I think this should both be visible.

Test 1: To see what's going on, I also measured the voltage at the stay-alive using a multimeter. It takes about a second for the voltage to drop below 10V (it seems that only the decoder itself is drawing current).

Test 2: I also charged the stay-alive and connected it to the same resistor-LED combination I use in my locomotives, and LED glowed for a few seconds...

Test 3: I also opened the decoder and saw that the stay-alive wires are mounted directly behind the diode rectifier - exactly where you should expect them.

I would assume that LaisDCC decoders would not support stay-alive unless LaisDCC sold their own stay-alives. And as @Art Dent pointed found out, their stay-alives are just capacitors plus a charging resistor. I guess @Art Dent forgot to mention the diode, which I also properly used.

 

So maybe LaisDCC does use some trick to force folks to buy their stay-alives?  It would be very useful if @Art Dent could post the picture of the LaisDCC stay-alives, this might help solve the mystery.

 

Edited by pygmalion
Link to post
Share on other sites

I have made two more tests:

Test 4: I reduced the input voltage of the DCC system to 10V and found that LaisDCC decoder works perfectly at these low voltages.

Test 5: Using five 1F 2.7V capacitors with low ESR, I created another stay-alive (effective capacitance 0.2F, but too large to measure with my multimeter). After turning off the power, it took a few dozen seconds for the voltage to drop below 10V. However, the motor and LED stopped immediately.

Conclusion: either some setting in my LaisDCC decoder is wrong or LaisDCC has some trick to prevent the use of DIY stay-alives. This would actually make them smarter than other decoder manufacturers :/

Edited by pygmalion
Link to post
Share on other sites

On 12/06/2021 at 08:54, Nigelcliffe said:

 

Such a circuit will work on a Zimo decoder, or anything else which is half-decent.   4400uF (if you mean 4400uF, not 440uF ??) should give a clearly visible run time.  440uF probably won't be visible.

 

Well... I've never heard of Zimo, and they're practically my neighbours (250km/150miles from where I live). Yes, they are about twice as expensive, but now I'm sorry. I've already f***** up myself, as  my entire locomotive fleet is already equipped with LaisDCCs.

 

Edited by pygmalion
Link to post
Share on other sites

Upon further research on @Nigelcliffe's suggestion, I found that TCS decoders require CV182 = 2 when stay-alives are used. This is mentioned in the document by Bruce Petrarca, which curiously was shared as document http://laisdcc.com/Stayin_Alive_laisdcc.pdf (no longer available on the internet, but could be found again by  Internet Archive).  So there is a teensy bit of hope that I can still get this to work.  The whole issue is also mentioned in another discussion on this very forum, but remains unresolved.

But the problem is that I can't access CV182 with JMRI Decoder Pro! Any suggestions on this alley?

 

Edited by pygmalion
Link to post
Share on other sites

40 minutes ago, pygmalion said:

 

Well... I've never heard of Zimo, and they're practically my neighbours (250km/150miles from where I live). Yes, they are about twice as expensive, but now I'm sorry. I've already f***** up myself, as  my entire locomotive fleet is already equipped with LaisDCCs.

 

 

A bit of history:   If you do some digging of internet stuff of about 5-8 years ago, you'll find there were complaints from TCS (long established DCC decoder maker from the USA) about illegal copies of their decoders being made in China and badged under a number of brands.  Lais is one of those who TCS named in various places.    The CV structure of a Lais decoder is identical to old TCS designs.  And there's a quirk in the speed performance curve which is unlikely to be deliberate, yet is present in both.     

I doubt there is anything clever binding a Lais stay-alive to a Lais decoder, there are not enough components in the stay-alives to make it possible.  Instead it will be a CV setting you have to find and change, possibly the one you mentioned.   JMRI isn't blocking you, its just whoever made the JMRI file didn't include that CV, you can set it manually with the single-CV-programmer, or edit the JMRI file to add it.    

 

 

If you need help sorting out a Zimo, there's loads of people who will help, their manuals are thorough, and their helpdesk will reply (in English as well as German).   If there's a feature missing in JMRI, at least four people I know will help update the Zimo files (not counting me).  Their decoders are amongst the best you can buy.   

 

 

- Nigel

Link to post
Share on other sites

54 minutes ago, Nigelcliffe said:

 

A bit of history:   If you do some digging of internet stuff of about 5-8 years ago, you'll find there were complaints from TCS (long established DCC decoder maker from the USA) about illegal copies of their decoders being made in China and badged under a number of brands.  Lais is one of those who TCS named in various places.    The CV structure of a Lais decoder is identical to old TCS designs.  And there's a quirk in the speed performance curve which is unlikely to be deliberate, yet is present in both.     

I doubt there is anything clever binding a Lais stay-alive to a Lais decoder, there are not enough components in the stay-alives to make it possible.  Instead it will be a CV setting you have to find and change, possibly the one you mentioned.   JMRI isn't blocking you, its just whoever made the JMRI file didn't include that CV, you can set it manually with the single-CV-programmer, or edit the JMRI file to add it.    

 

 

If you need help sorting out a Zimo, there's loads of people who will help, their manuals are thorough, and their helpdesk will reply (in English as well as German).   If there's a feature missing in JMRI, at least four people I know will help update the Zimo files (not counting me).  Their decoders are amongst the best you can buy.   

 

 

- Nigel

 

I see, Nigel. I'm a total newbie, two years ago started playing with old Mehanotehnika locomotives my dad gave me over 30 years ago. Since the locomotives themselves are not a great value (you can usually get them on EBay for $10 - $20), I decided to go with cheap decoders. LaisDCC were included in JMRI Pro Decoder and even have their manufacturer's number, so I thought it was all legit.

Now, after digitizing more than 10 locomotives, I'm stuck - it doesn't make sense in terms of time and cost to switch to Zimo, no matter how much I'd like to. So, yeah, I'm SORRY, but I can't help.


For anyone concerned, I edited the JMRI file and changed CV182 to 2 (default was 0), but that didn't help. So I'm back to square one.

 

Marko

Link to post
Share on other sites

On 08/03/2019 at 17:18, pygmalion said:

 

Thank you for your information about marketing gimmick.  I suspected so, but it is easier when I have first-hand confirmation.  I will try using five (not four) pieces of this capacitor in series (I have realized that for the previous one minimum order is 12000 pieces...):

 

https://www.digikey.si/product-detail/en/illinois-capacitor/DGH105Q2R7/1572-1758-ND/7387508

 

I will let you know the result.

 

I think 5 x 2.7 = 13.5 is enough for my use, after all I am powering my DCC controller with 12V.  But even if I powered my controller with 14V, I could still shield my capacitors with 12V Zener in parallel, wouldn't I?

 

Trying to recall my physics lessons,  capacitors in series / parallel,  capacitors in parallell add,  1F+1F = 2F, capacitors in series divide , 1F in series with 1F equals 1/2 a Farad. If so then 5 of 2.7F capcitors in series equals 1/ 13.5, ie about 0.07 farads

Link to post
Share on other sites

  • 1 month later...
On 12/06/2021 at 07:31, pygmalion said:

OK folks, I have finally reached stay-alive experiments.

I made 4400uF stay-alive (2x2200uF) for my 860014 LaisDCC decoder with black and blue stay-alive wires.

I also forgot to turn off DC (CV29 = 38) so the decoder was just acting all crazy. After I turned off DC (CV29 = 34), the decoder works normally, but the effect of stay-alive is ZERO. After the line loses power, the motor and/or LED stop immediately.

I know that the motor should only work for a fraction of a second, but I expected LED to work for at least a second.

I checked my stay-alive against LED and LED is kept on for a few seconds. I measured the voltage on the stay-alive while it was mounted on the decoder, and the voltage slowly drops after I turn off the power, but LED stops immediately.

Now I'm am clueless. Either I damaged LaisDCC somehow by providing stay-alive with DC turned on, or the decoder is just a piece of junk.

Does anyone have any ideas on this?

 

Hi,  Did you get any answers to this? I'm trying work out how to disable CV29 and I don't really understand how.

Link to post
Share on other sites

45 minutes ago, posneg6 said:

Hi,  Did you get any answers to this? I'm trying work out how to disable CV29 and I don't really understand how.

 

Article on CV29, written years ago.  I really should have put a penny-per-view on it to fund buying live-steam locos....

http://www.2mm.org.uk/articles/cv29 calculator.htm

 

You don't "disable CV29".   You might want to use CV29 to disable DC operation of the locomotive. 

 

  • Like 1
Link to post
Share on other sites

15 minutes ago, Nigelcliffe said:

 

Article on CV29, written years ago.  I really should have put a penny-per-view on it to fund buying live-steam locos....

http://www.2mm.org.uk/articles/cv29 calculator.htm

 

You don't "disable CV29".   You might want to use CV29 to disable DC operation of the locomotive. 

 

Hi Hi I've been told to Disable CV 29 to enable stay alives to work.

 

Link to post
Share on other sites

7 minutes ago, posneg6 said:

Hi Hi I've been told to Disable CV 29 to enable stay alives to work.

Whether they said it or not, what they meant was "Disable the DC operation bit in CV29".

 

Either they misunderstood themselves, or were just talking shorthand, assuming you were familiar with the various options that can be set in CV29.

 

Link to post
Share on other sites

On 06/08/2021 at 09:30, posneg6 said:

Hi,  Did you get any answers to this? I'm trying work out how to disable CV29 and I don't really understand how.

 

Stay Alives do not work with 860014 (and probably the entire PanGu series). On the other hand, Stay Alives work with 870014 (and probably the entire KungFu series). Both are true for the LaisDCC brand as well as for the homemade Stay Alives.

I contacted the guys at LaisDCC, and while they did not explicitly admit that 860014 does not work with Stay Alives, they could not tell me how to fix the problem. I was able to negotiate some sort of decoder swap (it was marketed as Stay Alive compliant + I am a regular customer + I did several web publications using their decoders).

Stay Alive issues aside, both the 860014 and 870014 work fine for me, even with old, really lousy Mehanotehnika motors (all of which I will be replacing with nice Mitsumi motors now). Too bad the 860014s have this problem, because now I have a lot of work to do to replace them in all locomotives...

From a brighter perspective, maybe this means that LaisDCC decoders are not copies. After all, the supposed originals all work with Stay Alives.  They also sell small Stay Alives, the size of which is impressively small. So in this situation I will continue to use them.

Edited by pygmalion
  • Like 1
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...