. . .

Your QSI decoder is NOT responding to both 03 and 4538. That is impossible.  What you are seeing is power applied  to the track and the decoder comes alive.  The ADDRESS in the DCC packet is what the decoder will  respond to once it is powered up. This long  versus short address is determined by CV29 of the decoder.

No decoder will respond to a READ while in OPS mode, so that is certainly a typo on your part.

If your DCS50 is reporting a short condition, then the engine is causing a short condition. Find the source of the short. Could be in tender. Could be in engine. Could be the decoder.  Could be a tender truck spun around.

Marc

Clarify

I know that OPS can not read. Yes, that was a typo. What I should have said is: I put the loco on the layout. With OPS, I changed CV8 to 08 which was "supposed" to change the ID back to 03. But with the DCS50, when I input 4538, the loco ran - lights, bell, whistle and so on. When I input 03, the loco started making  noises and the light came on. That's all I did. I do not get a short on the layout with the DCS50. The short is when I put the loco on the pt - JMRI shows a short. And I tried various ways to read the decoder. I created a roster from the decoder list on JMRI and tried to read it to the decoder. That did not work. I obviously, had tried to "read the decoder" and JMRI simply fails. I can not do anything with the loco on the pt with JMRI. Since I can't even get it to recognize a roster/xml, I can not try anything further with JMRI. Another thing I failed to mention is: On the layout with DCS50, I tried pressing f6, once,twice & multi times. I was told to try that. What it is supposed to do is "over my head". In fact, this whole QSI thing is beyond my brain power. I'm simply stymied. As for the loco. I had gone over it carefully messing with the tender and loco. All seems to be in order. Even the darn plug that can be accidentally put in upside down. But I made sure it was right side up. (And that bothered me too. A USB hub can't go in upside down, but the BLI loco to tender plug can. Phooey!) So that's why I'm on here asking for help.

. . .

"  But with the DCS50, when I input 4538, the loco ran - lights, bell, whistle and so on. "

This tells me the engine/decoder is on 4538 as an address, since it is operational.

"When I input 03, the loco started making  noises and the light came on. "

03 is not the address that is active in this decoder. noise and light are simply a reaction to a Track power 'on'.

" The short is when I put the loco on the pt - JMRI shows a short.  "

This is an indication that the decoder is overwhelming the DCS50 program track output and the DCS50 is seeing it as a short and passing on the same info to JMRI.  Could be solved by using a program track booster such as the Soundtraxx PTB-100. My DCS50 has problems programming Tsunami decoders unless the PTB-100 is in the loop.

If the engine runs fine on MAIN on address 4538, there is no problem with it.

Marc

If it truly IS a QSI decoder,

If it truly IS a QSI decoder, CV08 set to a value of "08" won't reset the decoder.  If you look in the manual for QSI variants, there are three separate CV changes to effecting a QSI factory defaults reset, and there's actually one more step, often forgotten...remove track power after pressing the final 'enter', and then restore track power.  Acquire Add "03" on your throttle, and you'll have your decoder back.

This may be nothing more than sticky/dried lubes with no volatiles left.  The decoder is otherwise fine....?

IF it is not a QSI, but a P2 decoder, they don't do anything until you activate the throttle with a speed step...at least that's how they work in my recollection....perhaps their literature says something else.  So, if you have performed a correct reset for what is actually a P2, with CV08, you don't have anything happen until the throttle gets a command to move the locomotive.

. . .

" If you look in the manual for QSI variants, there are three separate CV changes to effecting a QSI factory defaults reset, and there's actually one more step, often forgotten...remove track power after pressing the final 'enter', and then restore track power.  "

This is only valid for a reset done on MAIN,  which is not a good idea. You must have the RIGHT address for this to work. You could inadvertently reset the WRONG engine doing this or not  reset it at all if you are uncertain of the current address. 

Doing the reset on the Program Track is best and does not require this "remove track power"  since this is automatic during Service Programming.

Marc

. . .

... And since you are encountering an "Overcurrent situation " on your program track with this engine/decoder, you are in a catch 22. 

If it is a QSI decoder the reset to factory involves these CV's and values. But check your manual.

write CV: 49 to: 128 mode: Paged [AWT-EventQueue-0]
write CV: 50 to: 255 mode: Paged [AWT-EventQueue-0]
write CV: 56 to: 113 mode: Paged [AWT-EventQueue-0]

If it is a Paragon2, then CV8 = 8 should reset the decoder.

If doing this on MAIN, use your 4538 address as it seems to be the valid address as per your comments. And cycle track power as  ??? states.

Marc

OK. Now you've confused me. Let me clarify once more.

With the DCS50, I only can use OPS on the main/layout/running track. That is all. I can not use pt because it isn't wired up. All programming with DIRECT or PAGE is on a pt with a SPROG2. As y'all should know, SPROG2 is a command station and a programming interface with JMRI. So, when I say I'm on the layout, it is only with DCS50. And I can not read CV's. But I can use OPS and input a change. I did that with CV8 to 8 and it apparently did not change 4538 to 03. (We now know that changing CV8 does not reset a QSI decoder.) As stated in the QSI instructions. Then, on the pt, all I get from JMRI is that there's a short on the pt. Error #309. Now is this all clear? I did write to BLI and they replied that they do not support QSI anymore in any way. They did give me several YouTube videos. I have already tried all of them. I failed to get the decoder to default with the way they described. I may have miss understood the video. And I may try again. However, the mystery still exists. The loco runs correctly on the DCS50/layout with #4538. That indicates to me that the decoder is working. And as explained above, if it shows power on 03, it's only that. Not that the loco has 03 in it. So, to me the problem lies strictly in JMRI. Also, in BLI's written response, they claim that QSI can not be programmed with JMRI. So - if that is so, how is it that some of you can program a QSI on JMRI? The correct decoder is listed but only gives me a short circuit? And that defies the fact that the loco does run. So as of now, I can not do the CV changes in JMRI. Only on the layout. i.e. Input CV's, but can not read them. So, I can try that. Also, as I mentioned, I can diss connect the SPROG, and reconnect the pt wires from the DCS50. But that's where I need a booster? Or will the 12v auto bulb connected between the DCS50 pt wires to the pt track? (That's how I got a Bachmann loco to work-input the light bulb.) Is this clear?  I've done enough locos now with JMRI to know how Decoder Pro works. It's just the quirk of QSI and JMRI not communicating.

...

You have a lack of understanding of how JMRI works. JMRI does NOT communicate with decoders, that is what your command station does. JMRI just sends and receives information from the command station like a simple throttle. Connect your DCS to a programming track and see what you can do without using JMRI. If you can't do it with a regular throttle and DCS you can't do it using JMRI.

https://www.broadway-limited.com/support/manuals/Generic%20Steam%20Q1a%20Operators%20Manual.pdf

Refer to the reset instructions in the above pdf.

BLI I-1

While modeling the 56-57 era on my PRR Buffalo line I had a number of BLI I-1's , None had a QSI decoder in them. They all had BLI Paragon decoders in them. I don't believe that BLI ever offered a I-1 with a QSI. I know the M's did have them and they were horrible. --------  Ken 

First run of I1s were Precision Craft?

I think Ken is right.  If memory serves the first batch of "Broadway" I1s came in the Precision Craft box, and had ESU decoders.  Did they then do a run in the Broadway "Blue Line" series (sound decoder but no motor decoder)?

Scott Chatfield

Wow!

Let's not go off on a tangent. The loco is a BLI PCM PRR I1sa with a QSI decoder. That is a fact. Verified. And is not an ESU. It's the decoder that BLI started using as Precision Craft Models, and when QSI went downhill, BLI bought them out. And Pagagon 2 & 3 are "supposed improvements" on it. This loco does now run on the layout with a DCS50 with #7538. It lights, makes noises, has a bell and whistle/correct Banshee. BUT: I can not get JMRI to recognize it. So, Peter, whether I am mistaken with JMRI or not, does not answer my problem. I don't care about semantics. I am not a programmer. I am just a MR with a loco that I would like to change CV's via JMRI. And I'm not angry or trying to insult anyone. I'm just trying to find if someone has some experience with JMRI and QSI. The groups.io is dormant. The BLI groups.io is almost dormant. And this MRH forum is about the best there is. Has the most knowledgeable people I know. And I sincerely doubt if Digitrax groups.io can help. This is not a Digitrax problem. And the JMRI groups.io isn't helping yet. I'm trying to reach a large an audience as I can to get help. It's what we call "frustration".

Marc, not correct.  If you

Marc, not correct.  If you use Add "00" and do the reset in Broadcast Mode or "Paged Mode", you will get the reset done that way.  It is only in "programming on the main", or Ops Mode, that you must use the current address programmed into the decoder.  In Paged Mode, you'll get all addresses currently polling track signal.  So, remove all locomotives if there are others, leave the problem child on the main, and use Paged Mode addressing "00".

. . .

I fail to see how a Digitrax DCS50 user could combine PAGE mode and OPS mode  as they are totally different modes of programming.  Page and Direct are  reserved to Service mode programming  as per the Digitrax Loconet  Personal Edition manual. OPS is a mode on to itself and reserved to MAIN.

Using 'broadcast' is a dangerous and in the 'right hands' can cause great grief.  And that is why I never mention it.  MIght as well put the Zephyr in it's BLAST mode, does the same as broadcast out of both PROG and RAIL outputs. Just disconnect the layout from RAILA/B and you are safe.  But this means Morgan must wire up the PROG outputs.

- - - - -

Morgan.

Andrew Crosland (the guy who created and sells the SPROGs) says the Sprog should work with QSI decoders without a problem.  Since your issue is with the SPROG, I would write him and request support. 

There are a few hits regarding SPROG and QSI  and one fix was upgrading the power supply because of the "in rush current" issue.  The in rush current clips the power pack and the SPROG reports a "SHORT". But I am confidant Andrew Crosland will set you up.

Marc

Should be solved!

Thanks Marc. But it was Bill Chown who came to my rescue. He and Andrew are the SPROG guys to go to!  The answer given is to put a 33 ohm resistor in the one wire from SPROG to the pt. As of right now, I haven't done that yet. This morning I have a swollen foot and am in pain. So, it will be a day or more before I actually try this solution. He also said it might take a 25 ohm. So I do have them and can try them as soon as I am able. So for MRH, this problem is solved. And thanks to all who are on here helping.

. . .

Good for you.

Marc