YaMoRC-Digikeijs Forum

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

ray.jorgenson@... (Ray Jorgenson) — Thu, 03 Apr 2025 09:43:57 -0700

It sounds like it's time to drag out an oscilloscope and find out for sure what's happening.

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

bocabobgamble@... (Bob Gamble) — Thu, 03 Apr 2025 04:37:44 -0700

Jay,

I am using PSX and PSXX breakers from DCC Specialties.

https://dccspecialties.wpengine.com/wp-content/uploads/2023/09/PSXX-Manual-Rev-C-Rev-G.pdf

I believe that these are set up with the defaults, 2 amp breaker trip setting.

I have the PSXX (newer model that has support for bringing up capacitor charging keep alives) on my main yard track.

Regards,

Bob Gamble

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

jay.lennox@... (Jay Lennox) — Wed, 02 Apr 2025 14:03:39 -0700

Karst and Bob

Thank you for your feedback on my question.

First, Bob could you share what brand circuit breakers you are using and how you have them setup (jumpers, CV settings, etc.) We have been trying to use the NCE EB1 circuit breakers and the only way that they will trip before the DR5033 trips is if I set the DR5033 short circuit delay to 300ms - a setting of 225ms does not work.

Karst, you are correct, the need to circuit breakers is to minimize the effect of human error 😉. We do appreciate the fact that the DR5033 boosters, and the YD7001 and YD7403 handle what we call 'inrush current' - in one yard we can have 12 to 14 sound equipped locos sitting when the layout power goes on.

Karst, you said: Am I right, that the sole supply to that breaker is the Track-Out power from the booster ? If so, then there probably lies the problem: The YD7001/YD7403/DR5033 cut the power immediately ( within a single DCC packet's duration ) as soon as a short and/or overcurrent is detected. They will then retry pulsing the power back on in about a 10ms time interval until the short-delay expires.

I am now wondering if I am getting into a race condition between the DR5033 and the NCE EB1. By default the EB1 has a short circuit response time (delay time) of 16ms (it can be increased to 512ms). I did a quick search on the NMRA site and it appears that a DCC packet's duration can be between 95 and 9900 microseconds, less than10 milliseconds. Is it possible that the DR5033 is detecting the short before the EB1, and then it starts pulsing at 10ms intervals, which the EB1 cannot see as a short with its 16ms detection time? Further, since the DR5033 still sees a short when its circuit delay is reached, it signals a short and cuts the power. If this is true, why does a setting of 300ms for the DR5033's short circuit delay allow the EB1 to detect the short before the DR5033?

I am testing on the workbench as follows: Track output of DR5033 is set to 3 amps, and is directly attached to the DCC input of the EB1. EB1 is set to short at 2.2 amps. On the track output side of the EB1 I have an on/off toggle switch - turning the toggle to on creates the short.

Jay Lennox - Toronto

Background on what we are now trying to do at York Railway Modellers:

We have a large 40 foot x 40 foot (12 meter x 12 meter) layout room. There are two main cities, Toronto and London, Ontario, Canada), with large yards that are connected by about 900 feet (275 meters) of mainline track plus passing sidings. Each main city has a turntable and round house (we model 1955 during the steam to diesel transition period).
There are 12 town and industrial yards along the main line - some have complex yard configurations that reflect what was in each town in 1955.
The layout is powered by 7 DR5033's, and we have a YD7001 command station (it is not used to power the track). Breaking the layout into 7 power districts, was our start point when converting from DC to DCC in late 2019 and early 2020.
The longer term plan was to add circuit breakers to introduce sub-power districts, probably for yards, as we gained experience with how to break up a DCC layout for operations.
When we converted to DCC we had 8 members. We have almost doubled our membership in the last few years, and now on operations nights when someone runs a switch in a yard, others are impacted.
We are looking to move onto our longer term plan to add sub-power districts by using circuit breakers to isolate a yard from the mainline and other yards in the same power district.

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

bocabobgamble@... (Bob Gamble) — Wed, 02 Apr 2025 08:24:14 -0700

Am I right, that the sole supply to that breaker is the Track-Out power from the booster ?

Yes, this would be the correct way.

I have three DR5033 and seven breakers. They work fine as long as the breaker trip current is less than that of the booster. This may not be obvious to some... If the booster trips then there is too much of a load on the track between the sum of the breakers. I don't get this but it is possible. There are breakers that do the keep alive charge up cycling, so one would need these if there is a lot of this type of equipment in the block (like a yard or engine facility). Or just dedicate that block to a booster that does this.

Regards,

Bob Gamble

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

karst.drenth@... (karst.drenth) — Wed, 02 Apr 2025 07:52:37 -0700

By the way, that it is IS possible to get a correct "breaker-operation" is proven by our YD7432 reverse loop controller 😊

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

karst.drenth@... (karst.drenth) — Wed, 02 Apr 2025 07:50:29 -0700

Ok,

Thanks for the explanation Bob. So it IS symptom curing ( minimizing the effect of human error 😉)

Am I right, that the sole supply to that breaker is the Track-Out power from the booster ? If so, then there probably lies the problem: The YD7001/YD7403/DR5033 cut the power immediately ( within a single DCC packet's duration ) as soon as a short and/or overcurrent is detected.
They will then retry pulsing the power back on in about a 10ms time interval until the short-delay expires.

This method is needed to fill-up any buffer capacitors in all kinds of equipment, including loco(sound)decoders. Those capacitors often pose a "short" during initial charging.

Greets,
Karst

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

bocabobgamble@... (Bob Gamble) — Wed, 02 Apr 2025 07:35:50 -0700

Karst,

Note: I simply cannot understand why one would install such breakers. The YD and DR power devices really are fast enough to adequately turn off the power in case of a short.

Typically it comes from a desire to break up a layout into smaller power blocks than is strictly necessary based on booster power output. The breaker is less expensive than a booster and some can do auto reverse. The desire for finer granularity is usually driven by operating sessions where you don't want everyone on the power district impacted by one operator who tries to drive through a switch turned against them. But you could make the argument that it would be more elegant to use a booster for all of these smaller blocks, it would just cost you more money and you would have a lot more power supplies unless you cobbled together some sort of power distribution.

Regards,

Bob Gamble

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

karst.drenth@... (karst.drenth) — Wed, 02 Apr 2025 06:44:31 -0700

On Sun, Mar 30, 2025 at 02:35 PM, Jay Lennox wrote:

Further, what would be the risks of doing this ?

Well... first I think one should solve the cause, not the symptom. Obviously the cause is that that ominous circuitbreaker does not come after it's specs... ;)

It will take longer for the DR5003 to detect a short, and hence power will be applied for 3 times as long at the site of the short,

Almost... the 5033 will detect immediately the short, it will "just" retry applying power until the short-delay expired. This could lead to a kind of low duty-cycle of power, since the re-engaging delay is set to about 10ms. it "could" be that because of this the circuitbreaker does not trigger...

which leads to a longer heating danger to whatever is causing the short.

Correct, but not with the full power as described above. Mostly the power-pickups on the trains can take damage from this, since they are usually made of very thin material. I have seen them really glow yellow on certain Lima models...

Are there real risks to decoders installed in locos, etc ?

if the short is cause by the decoder, it is probably already too late. otherwise, shorts usually happen before the power enters the decoder.

Any other risks that anyone can think of ?

not that I can think of.

Greets,
Karst

Note: I simply cannot understand why one would install such breakers. The YD and DR power devices really are fast enough to adequately turn off the power in case of a short.

Re: Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

peter@... (Pete S.) — Mon, 31 Mar 2025 14:32:54 -0700

Hmmmmmm, I have never needed to vary this setting, but, I think the issues could be:-

Killing the item that caused the short - But it's already faulty so why worry.

Stressing the semiconductors in 1/ the block detection circuit; 2/ the output circuit of the 5033; 3/ the power supply.

The energy from the fault goes somewhere, and you'll get away with it a few times, but longer fault times mean more hidden damage, and eventually something will fail.

I think you should fix the fault, not the protection.

All the above IMHO of course........ Other opinions are available....

Pete S. - New Zealand

Increasing the DR5033 Short Circuit Delay to 300ms - what problems could happen ?

jay.lennox@... (Jay Lennox) — Sun, 30 Mar 2025 14:35:52 -0700

I am wondering if anyone has any experience with increasing the short circuit delay on the DR5033 from 96ms (default) to 300ms or more ? This is LocoNet CV 13 that is being increased on the DR5033.

Further, what would be the risks of doing this ? It will take longer for the DR5003 to detect a short, and hence power will be applied for 3 times as long at the site of the short, which leads to a longer heating danger to whatever is causing the short. Are there real risks to decoders installed in locos, etc ? Any other risks that anyone can think of ?

Thank you for sharing any experience you have.

Jay Lennox - Toronto

Re: NCE EB1 Circuit Breaker - is anyone using these with a YD7001 and DR5033 boosters ?

ray.jorgenson@... (Ray Jorgenson) — Fri, 28 Mar 2025 08:04:22 -0700

It could be you're experiencing 'switch bounce'; no switches close in one clean motion, but instead bounce open & closed multiple times in a multiple msec or tens of msec interval before staying closed. Slow logic does not see this, but fast logic can. It could be that the EB1 is being fooled. Try a different method of creating a short & see if anything changes.

Re: NCE EB1 Circuit Breaker - is anyone using these with a YD7001 and DR5033 boosters ?

jay.lennox@... (Jay Lennox) — Thu, 27 Mar 2025 11:03:49 -0700

Stuart

The trip time on the EB1 is controlled by CV130. I have set CV130=1, which should give a 16 ms trip time. In theory CV130 =1 when shipped from NCE - I purposely set it to make sure it was 1 after it did not work out of the box. To get the EB1 to trip before the DR5033, it had to set the DR5033 trip time to 300 ms - at 225 ms the DR5033 trips before the EB1. I know this makes no sense. Still investigating.

Iain

Good to know you have an EB1 working with a European brand command station.

I have set the EB1 trip current at 2.2A - it is the 1.1 version of the EB1 and the trip current options we have are 1.5A, 2.2A, 3.3A, 4.1 A, etc. We have the DR5000 set to supply the maximum 3000ma, or 3A in the configuration variables, so the maximum trip current I can set on the EB1 is 2.2A.

Testing methods;

I am only doing testing on the workbench, using a momentary switch on the track outputs to simulate a short. The EB1s need to work when I install them on the layout to avoid disappointed club members :-)
I have tested with the YD7001 directly and used two DR5033s in combination with the YD7001 - same results.
I have tested these combination with two EB1s to ensure one is not faulty - same results.

I have spent about 12 hours over the last 10 days working on this. I have a problem open with NCE support, but they have not been very helpful except to tell me that the EB1 always works as stated (which is exactly why the club decide to go with the EB1 so that we could isolate our yards from the mainline).

Things I plan to try

Change CV130 to say 4, which should give a trip time of 64 ms, and then see how low I can make the DR5033 trip time.
Add a Digitrax command station to the testing mix to see if the problem has something to do with the combination or YaMoRC plus Digikiejs and the EB1 - a long shot, but needs to be tested.

If you have any other thought please let me know. Thank You. Jay Lennox

Re: NCE EB1 Circuit Breaker - is anyone using these with a YD7001 and DR5033 boosters ?

w.iain.morrison@... (Iain Morrison) — Thu, 27 Mar 2025 10:05:55 -0700

I use a couple of NCE EB1 with my Z21 in the accessory bus and only changed the current trip to 2.5A to ensure the EB1 trips before the Z21 or the booster. I wonder if you have reduced the current trip on the EB1?

Don’t seem to have any issues.

Many Thanks

Iain Morrison
https://wimorrison.com/

Re: NCE EB1 Circuit Breaker - is anyone using these with a YD7001 and DR5033 boosters ?

stuart_w_baker@... (Stuart Baker) — Thu, 27 Mar 2025 09:51:51 -0700

Jay,

I have not used this combination before, but I have a couple of thoughts. What have you set the trip time on the NCE EB1 circuit breaker? I think the minimum setting is ~10ms. Can you try the minimum setting and see if it improves performance?

Thanks,

Stuart

NCE EB1 Circuit Breaker - is anyone using these with a YD7001 and DR5033 boosters ?

jay.lennox@... (Jay Lennox) — Thu, 27 Mar 2025 09:18:16 -0700

My club, York Railway Modellers, purchased several NCE EB1 circuit breakers and we cannot get them working with our YD7001s and the DR5033 boosters without increasing the circuit breaker trip time in the command station and boosters to 300 ms (default is 96 ms). I am concerned that setting the circuit breaker trip time in the command station and boosters to 300 ms could lead to damage to them or the decoders in the locomotives.

Has anyone else used the EB1 with the YaMoRC command stations and the Digikeijs boosters? If yes, could you let me know you set the CV in the command station and boosters.

Thank You. Jay Lennox.

Re: Resistor for block detection, best size?

w.iain.morrison@... (Iain Morrison) — Tue, 25 Mar 2025 03:16:46 -0700

You will see that the block detection (i.e. when current is drawn) appears slightly before the Railcom ID is displayed. This is because it is faster being almost the instant a loco enter the block. The Railcom ID is then obtained from the loco and sent via Loconet to the command station. It is a sequential process and will always be sequential.

The current is drawn by the decoder, even if the loco is stationary, and without a decoder you cannot get a Railcom ID.

Iain Morrison

Kaninchenbau

Re: Resistor for block detection, best size?

chad@... (Chadbag) — Mon, 24 Mar 2025 20:48:58 -0700

On Mar 20, 2025, at 06:19, Iain Morrison via groups.io <w.iain.morrison@...> wrote:

FYI Railcom has nothing to do with occupancy, it is simply reporting the details from the decoder. The actual occupation comes from the current draw by the locomotive (or wagon).

So let me ask a dumb question, as I’ve only set up sample blocks with DR5088RC and have seen the RC ids show up in my blocks but don’t actually detect any current.

The question:

Why do I care about he occupancy current? If the RC id is showing its obviously in the block. It won’t show an id for the block if it’s not in the block. I get the technical difference, but functionally, it is detecting occupancy. Because, as I said, the existence of the id shows its occupying the block.

My question is one of understanding as my current “layout�� is a simple oval to test my software as I get set back up in our new home over time and start a bigger and more complicated test layout and then a “real layout�� to keep me going until I die :)

Thanks

Chad

Re: upgraded DR5000 to YD7001 throttle issue

steve_donald_@... (Oldngrey) — Fri, 21 Mar 2025 14:54:29 -0700

Yes, YaMoRC's karst.drenth posted in the thread.

Re: Programmer Busy reading CVs with Ops mode over Railcom with YD9401

jscottwalton@... (Scott Walton) — Fri, 21 Mar 2025 13:35:39 -0700

I tried 2, 3 and 4. With 4 I get some (fewer) Programmer busy messages. With 2 and 3, I don't get Programmer busy, but I do see pauses., fewer with 2 than 3. It seems that 3 is the best setting for this. I'm doing all this with a mostly idle railroad, so the only messages are occaisonal "occupied" messages -- no throttles active. Overall, looks like a definint improvement.

Scott Walton

Thread exception attempting to run YD9401

jscottwalton@... (Scott Walton) — Fri, 21 Mar 2025 13:32:01 -0700

Shortly after I open the app, I get this exception:

This is on Windows 11. All attempts to r]use the app result in a "click" and no response. When I click on "OK", the app closes. The same result if I run the app with no connection to the command station (other than network). I have successfully run it on a Windows 10 system though.

Scott Walton