Jump to content
 

Home brewed DCC keep-alive - enough capacitance?


chaz

Recommended Posts

Can some one in the know give some advice please. I am rebuilding a K3 mogul and building a new tender for it.

 

The loco has a V3.5 Loksound decoder (the 1A version) and I want to add a keep-alive. I do know where on the decoder to wire to. The motor is a Maxon which I believe is very efficient with a low current draw (?). There is plenty of room in the tender for four 4,700uf capacitors with a voltage rating of 35V. My plan is mount the components (capacitors, resistor and diode) on a piece of stripboard and to link loco and tender with a polarised plug and socket.

 

snap shows the capacitors and the tender....

 

P1030938-2700x508_zpsb4121bb3.jpg

 

4 x 4,700uf = 18,800uf

 

Can someone advise me please is this likely to be enough? Obviously I will be able to set up a test later with the loco' towing a wagon with the keep-alive perched on it but some idea of whether it will work or not would be useful.

 

I would be grateful for informed answers from those with experience of DCC keep-alives.

 

Chaz

Link to post
Share on other sites

Thanks for that, DM. You say "find out how much current the entire loco uses (incl. sound!)" - I suppose the easiest way to do that is to hook the model up to my DC bench PSU which has an ammeter. But that will have to wait until the decoder is in and wired - by the time I get that far I might as well just hook up the keep-alive temporarily and try it out.

 

I was hoping that someone who has done a similar effort could say whether it's likely to be worth proceeding or not?

 

Chaz

Link to post
Share on other sites

Yes, you're right Nigel. thanks for reminding me. The manual talks about this - the one that maintains the decoder, as opposed to the motor does not need to be so big. Maybe one more 4700uf for that function, which I think I have room for but the question still remains - is it going to be enough? I think I am just going to have to try it and see.

 

Chaz

Link to post
Share on other sites

My understanding of Loksound V3.5 was that one needed two stay alive circuits attached to different places on the decoder. But I've never tried building one for a V3.5.

 

Well, I can confirm that you are right, hooking up 4 X 4700uf to a Loksound V3.5 at the output side of the rectifier diodes didn't work. Using a meter showed that they charged up fine, but when the power was cut they just held their charge and both motor and sound stopped instantly. As there was no sign of the charge slowly dropping this suggests that the chip had shut down and so could not keep the motor or sound going.

 

I will add another capacitor for another test when I have the time - probably over the weekend.

Link to post
Share on other sites

My findings too. Once you have sorted how to keep the processor alive then I would have thought that even a couple of 4700uF caps would be enough. Just don't forget to post your solution once you find it ;)

 

Will do. Experiments are ongoing.

Link to post
Share on other sites

The only circuit published by ESU for use with their V3 decoders was for the LokPilot V3 21MTC version in the LokPilot V3 manual, and that clearly shows that there are two different circuits required.

 

The LokPilot V3 and the LokSound V3/V3.5 are different animals though, as the LokPilot V3 had Railcom which could be activated through a firmware update, but the LokSound V3/3.5 did not have the necessary components for Railcom. So there may be other differences in the internal circuitry apart from the obvious one of the LS3/3.5 supporting sound, and the Railcom issue.

Link to post
Share on other sites

OK, another experiment with two keep-alives connected as suggested by ESU.

 

The small KA has a single 2200uf 35V capacitor with a 100R charging resistor and bypass (discharge) diode - ESU suggest 470uf so 2200uf should be plenty

 

The large KA has a 4 x 4700uf 35V capacitors in parallel with a 100R resistor and bypass diode - ESU suggest 2200uf so ditto

 

Still no joy, the sound and the motor cutting out as soon as the track power is cut.

 

Testing with a meter shows the following

 

the small KA is charging to around 5V and when power is cut it discharges in a few seconds to about 0.6V - suggests it is powering the chip the way it should

 

the large KA is charging to approx 15V - when power is cut the charge is held - very slowly discharging suggesting leakage in the circuit rather than use

 

------------------------------------------------------------------

 

I have set CV29 to value 6 - so that bit 2 - analogue operation - is on - I also tried turning bit 2 off - value 2

 

I have tried setting CV50 in all possible settings in its range 0-3 with no effect.

 

is there another CV that needs setting?

 

I have checked all the connections and the diode polarities and can't see anything wrong

but something is obviously stopping the large KA from discharging

 

If anybody has any ideas I would be grateful to hear them.

 

EDIT from a complete cretin....

 

Just did a search on keep-alives and found a list of CV settings that GavinD gave me in this thread...

 

http://www.rmweb.co.uk/community/index.php?/topic/43324-stay-alive-capacitors/page-2

 

pathetic isn't it - I had forgotten this very helpful stuff.  I will give it a try tomorrow - WTS

Link to post
Share on other sites

A disappointing start to the day.

 

I fired up the laptop, plugged in the ESU interface with a Loksound V3.5 with the two KAs wired as per. I spent a few minutes setting to CVs to the values suggested by GavinD viz...

 

CV29 - 2

CV49 - 3

CV50 - 0

CV51 - 0

CV124 - 6

 

Still no effect from the KAs - when the track connection on the interface is disconnected both the sound and motor stop instantly. As before checking voltages across the capacitors shows that the small KA loses its charge in a second or two, voltage dropping from around 5V to about 0.6V. The large KA charges to 15V but holds its charge after disconnection.

 

It may be that my only solution remaining is the expensive one - to retire the V3.5 decoders and replace them.

 

Chaz

Link to post
Share on other sites

 

The large KA has a 4 x 4700uf 35V capacitors in parallel with a 100R resistor and bypass diode - ESU suggest 2200uf so ditto

You have checked the bypass diode is the right way round? Although you should still see slow discharge through the resistor if it was not.

Keith

Link to post
Share on other sites

You have checked the bypass diode is the right way round? Although you should still see slow discharge through the resistor if it was not.

Keith

 

Yes Keith, thanks, several times (you know how it is when something doesn't work), working both from first principles and from the circuit diag. from the ESU website.

Link to post
Share on other sites

Are you connecting the 15V KA unit to the right place in the circuit? The fact that it gets charged but not discharged means there has to be a diode on the decoder that prevents the stored current reversing back into the system. If there is, try short-circuiting it with some tweezers and if successful, disable that diode with a (thin) strand of copper wire. It's probably best to keep the diode in situ (warranty!).

 

I assume so. I have followed the wiring diag. from the ESU website - the only thing I have varied is the size of the capacitors, but as they are charging up to the supply voltage that can't be the problem. I may well have made a silly mistake but I can't see it.

 

As for short circuiting components on the board I won't be doing that - at the moment the decoder works fine even if my attempt at a KA doesn't. I don't care about the warranty which I have probably voided anyway by soldering wires to the PCB but I do care that a £90 decoder keeps working.

 

Chaz

Link to post
Share on other sites

I assume searches in German have been done, such as "esu loksound v3.5 Kondensator".   I've looked at that term, and of the pages revealed, the following two may be helpful

http://www.s1gf.de/index.php?page=Thread&threadID=27849     ( has a diagram in link at top)

http://www.s1gf.de/index.php?page=Thread&threadID=10234     ( has a diagram half way down the thread )

 

 

(  I repeat that I've never tried this on an LokSound V3.5 and am not likely to need to try it ).

 

- Nigel

Link to post
Share on other sites

I assume searches in German have been done, such as "esu loksound v3.5 Kondensator".   I've looked at that term, and of the pages revealed, the following two may be helpful

 

http://www.s1gf.de/index.php?page=Thread&threadID=27849     ( has a diagram in link at top)

 

http://www.s1gf.de/index.php?page=Thread&threadID=10234     ( has a diagram half way down the thread )

 

 

(  I repeat that I've never tried this on an LokSound V3.5 and am not likely to need to try it ).

 

- Nigel

Hi,it might be an idea to have a word with Richard Johnson from DCC Concepts Australia.He has a wealth of knowledge of this subject.He also owns the DCC Concepts firm.which make their own Stay alive units.,they coined the name Stay Alive others are known by Keep alive.

I am just getting to grips with it myself. DCC Concepts TCS have ready built units to fit onto Decoders

As far as im aware Zimo are the leaders when it comes to keep alive being fitted to their sound decoders.Their website is very good on this subject.Their sound decoders are made so as you only have to add capacitors to them as everything else is built into the decoder,so no need for resistors n the like

Ive emailed Richard myself as to problems ive had with this subject,as ive said hes very knowlegeable n a very friendly person

David

Link to post
Share on other sites

Hi,it might be an idea to have a word with Richard Johnson from DCC Concepts Australia.He has a wealth of knowledge of this subject.He also owns the DCC Concepts firm.which make their own Stay alive units.,they coined the name Stay Alive others are known by Keep alive.

I am just getting to grips with it myself. DCC Concepts TCS have ready built units to fit onto Decoders

As far as im aware Zimo are the leaders when it comes to keep alive being fitted to their sound decoders.Their website is very good on this subject.Their sound decoders are made so as you only have to add capacitors to them as everything else is built into the decoder,so no need for resistors n the like

Ive emailed Richard myself as to problems ive had with this subject,as ive said hes very knowlegeable n a very friendly person

David

 

Thanks for suggestion, David - when I take a rest from mangling bits of brass I will do so.

 

Chaz

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...