Hi everybody. I recently bought a BLI mikado with the factory QSI decoder. I finally figured out how to program the long address via CV's 17 & 18, then CV29 (It wouldn't let my NCE system do it via it's normal method on the programming track. I had to figure out the individual values for 17, 18, & 29 and manually program them in on the main). Problem is, when I shut off the layout, the decoder jumps back to address 3 when I power back on. Am I missing something, or could it just be a bum decoder? Any ideas?
Mike
Water Level RouteAny ideas?
Just one.
It may be possible that there is a reset jumper still in place? Or if the decoder has a reed switch, the tiny electrodes are "shorted" (closed) causing the decoder to stay in reset mode OR if it has a pushbutton (probably only found on later decoders) the button is stuck or otherwise closed also calling for a reset.
QSI reset jumper by Edmund, on Flickr
Just maybe.
Good Luck, Ed
Water Level RouteProblem is, when I shut off the layout, the decoder jumps back to address 3 when I power back on. Am I missing something, or could it just be a bum decoder? Any ideas?
That seems to be a fairly common issue with QSI decoders. They often lose programming on one or more CVs, sometimes including the address, when you turn the DCC system off. If you have one where this happens repeatedly, it's probably going to continue until you replace the decoder.
One of the layouts where I operate regularly had this same problem with multiple QSI decoders, including those supplied with Broadway locomotives. All the QSI decoders were eventually replaced because of it.
Rob Spangler
That leaves me uneasy. I must have eight or nine QSI decoders, now at least 13 years old. So far, I haven't had a problem with any of them. I guess, though, that nothing lasts forever.
selector That leaves me uneasy. I must have eight or nine QSI decoders, now at least 13 years old. So far, I haven't had a problem with any of them. I guess, though, that nothing lasts forever.
Rich
Alton Junction
Water Level Route Hi everybody. I recently bought a BLI mikado with the factory QSI decoder. I finally figured out how to program the long address via CV's 17 & 18, then CV29 (It wouldn't let my NCE system do it via it's normal method on the programming track. I had to figure out the individual values for 17, 18, & 29 and manually program them in on the main). Problem is, when I shut off the layout, the decoder jumps back to address 3 when I power back on. Am I missing something, or could it just be a bum decoder? Any ideas?
Which NCE system? Ph-Pro or Power Cab?
Is the loco the original Paragon?
The biggest problem that I have encountered with factory installed QSI decoders in BLI locos is Advanced Consisting. The QSI is quirky, losing the consist on power down, not clearing CV19 when a consist is killed, dropping or adding locos from or to the consist.
I have never experienced your particular problem with programming long addresses. However, others have had all kinds of weird things happening when programming QSI decoders, particularly on the Programming Track.
One other comment. I have the NCE PH-Pro 5 amp wireless system. I find it useful to tether the ProCab to a UTP panel when programming, either on the Programming Track or POM (Programming on the Main).
Lately, when I program a decoder (any decoder) on the Programming Track, the LCD screen on my ProCab tends to overwrite instructions and values on the LCD screen. Don't know what that is all about, but it does not adversely affect the programming of the decoder.
So, your issue could be the decoder or the throttle.
richhotrain Crandell, I would say that you are good. In my experience with QSI decoders, all of mine factory installed in BLI locos, if they are problem free out of the box, they stay that way. If they are problematic out of the box, they stay that way. Rich
Crandell, I would say that you are good. In my experience with QSI decoders, all of mine factory installed in BLI locos, if they are problem free out of the box, they stay that way. If they are problematic out of the box, they stay that way.
Thanks, Rich. I thought I had a late model QSI Titan, or a Revolution, both post-2008, go toast on me a couple of years back, but I somehow convinced it to begin to play nice again. I forget what it was all about...maybe it felt neglected because I rotate my locos in about a two year cycle to keep them fresh.
Guys, thank you so much for all of your input. Sorry about the delayed response. I've been sick the last couple days. As soon as I'm feeling 100%, I'll start by opening up the tender to see if there is a reed switch/reset button issue. Yes, its the original paragon series BLI.
Rich, I have the very same NCE system but haven't had the troubles with the throttle display when programming wirelessly, but I'll keep an eye out for it. Thanks for the tip!
Worst case, this loco jumps to the front of the line for a decoder replacement. Would be a bummer, but there are worse things in life.
Thanks fellas.
So I got the tender apart last night and there is no reset button or reed switch that may be sticking. Absolutely nothing that could be disconnected that would indicate a reset jumper. Guess I better start looking for a tsunami/econami for it.
Thanks again guys.
Some are easy to spot:
IMG_7572_fix by Edmund, on Flickr
Others are tucked in closer to the board (alongside the chip socket):
IMG_7824_fix by Edmund, on Flickr
On this one the two pins this side of the big cap is where a jumper has to be placed for a "hard" reset:
BLI_UP-tender_board by Edmund, on Flickr
and here, a little rubber button does the job:
BLI_P2-reset by Edmund, on Flickr
Maybe yours was made without a method for resetting.
Are you sure it isn't a Blue Line decoder?
Yeah, it does seem to be a faulty, or at least quirky, decoder.
I have a lot of QSI decoder-equipped locos, and I have developed a love-hate (mostly hate) relationship with them. So, I decided to research this issue further.
I found several websites that discuss various QSI upgrade chips, each a different version, that deal with faulty QSI decoders. One such upgrade was the Q1A Version 7.0 that apparently would solve my problem associated with Advanced Consists. That is not your problem, but I wonder if one these upgrade chips would solve your problem of programming long addresses.
If you are interested in investigating the upgrade chip solution, let me know and I can provide some links for you to contact before incurring the expense of a brand new sound decoder.
Ed, it came to me in a Paragon box labeled for an unlettered light 2-8-2. This would seem to match as the loco is clearly a light 2-8-2 and was decaled by the previous owner. The decoder has voice reply when changing CV's, kind of like a TCS but not exactly, so I'm pretty certain it is the QSI decoder the box suggests. There is clearly no reed switch or button. There are no pins like what you showed either, just a bunch of empty sockets (yes wires in some here or there from wheels, etc). The decoder a two piece deal where an upper board plugs down onto a lower board, kind of like a 21 pin style, but the pins are about evenly split half at one end, half at the other. I didn't think to grab a pic, but I can tomorrow if it's helpful.
Rich, I didn't know the chips could be replaced. I assume this requires desoldering the old and soldering the new, or do they come with other components mounted on one of the two layers to the boards like I described above?
Well, if I were you, I would try using CV1 and set a 2 digit address (two first numbers of the address) and see if it keeps that. I'm a firm believer in delaying purchases!
Simon
Water Level Route Rich, I didn't know the chips could be replaced. I assume this requires desoldering the old and soldering the new, or do they come with other components mounted on one of the two layers to the boards like I described above?
Here is what Mark has to say about QSI upgrade chips.
This is a OEM hardware upgrade that involves replacing the Q1 chip to a Q1A version that allows downloadable software. Q1A Version 7.x Firmware requires the Q1A chip to be installed. Key advantage of Q1A chip upgrade is it allows all future software upgrades to use the new QSI programmer. No more chip replacement required. Chip swapping is not a user friendly upgrade and one can easily damage the socket if one does not have the correct tool to take the chip out of the Q1 chip out of the socket. If you are using a small screwdriver....you are doing it wrong and may crack the corner of the brittle socket. QSI solutions does sells an inexpensive chip extractor tool and save you time and grief. The tool will really pay off if you have a lot of Q1A upgrades to do.
I have not found a specific reference to your issue, so I cannot say that there is a chip upgrade for it, but it is a possibility. You might contact Litchfield Station for advice.
HO DCC decoder upgrade chip by QSI – #601-HO999 – Litchfield Station
Lastspikemike QSI decoders are quirky, no doubt. F6 is perhaps a necessary function to wake up the decoder in some circumstances. Also, the QSI can seem dormant or unresponsive until you apply some throttle. Don't assume it isn't at the ready until you actually direct it to move and it doesn't.
QSI decoders are quirky, no doubt. F6 is perhaps a necessary function to wake up the decoder in some circumstances. Also, the QSI can seem dormant or unresponsive until you apply some throttle. Don't assume it isn't at the ready until you actually direct it to move and it doesn't.
By the way, POM may yield better results than using the Programming Track.
Those QSI upgrade chips go back some fifteen years. They can be changed out without soldering since there's a big IC socket on the board. I did about a dozen of my early PCM/BLI engines back in 2006.
I really doubt you'd be able to find any. Maybe some dealer has some tucked away in a drawer somewhere, or eBay. You have to be careful about which chip you get as the sounds and driving characteristics are specific.
https://sites.google.com/site/markgurries/home/dcc-tips-and-tricks/qsi-decoder-versions
I was somewhat disappointed that when I put the upgrade chips in my NYC Niagaras I lost the neat Leslie A200 Tyfon air horn sound that was on the original chips.
Regards, Ed
gmpullman Those QSI upgrade chips go back some fifteen years. They can be changed out without soldering since there's a big IC socket on the board. I did about a dozen of my early PCM/BLI engines back in 2006. I really doubt you'd be able to find any. Maybe some dealer has some tucked away in a drawer somewhere, or eBay. You have to be careful about which chip you get as the sounds and driving characteristics are specific. https://sites.google.com/site/markgurries/home/dcc-tips-and-tricks/qsi-decoder-versions I was somewhat disappointed that when I put the upgrade chips in my NYC Niagaras I lost the neat Leslie A200 Tyfon air horn sound that was on the original chips. Regards, Ed
Yeah, QSI Solutions met its demise some years ago, but Litchfield Station could still have some since they do advertise at least one upgrade chip on their website. That is, if there is even an upgrade chip to solve the OP's problem. This is clearly a a longshot.
If QSI is defunct, maybe it would be best to stick with a "live" brand like Soundtrax or ESU or TCS.
Rio Grande. The Action Road - Focus 1977-1983
Lastspikemike http://cs.trains.com/mrr/f/744/p/242811/2703567.aspx QSI decoders are quirky, no doubt. F6 is perhaps a necessary function to wake up the decoder in some circumstances. Also, the QSI can seem dormant or unresponsive until you apply some throttle. Don't assume it isn't at the ready until you actually direct it to move and it doesn't. My Intermountain F7P ABB set will not illuminate the headlight until the throttle is cracked open, just for example. If you power down using F8 it requires two presses. The first press has a volume mute effect. I recall if you shut down using F8 then maybe you need to press F6 to wake it up?
http://cs.trains.com/mrr/f/744/p/242811/2703567.aspx
My Intermountain F7P ABB set will not illuminate the headlight until the throttle is cracked open, just for example. If you power down using F8 it requires two presses. The first press has a volume mute effect. I recall if you shut down using F8 then maybe you need to press F6 to wake it up?
I had forgotten this, and thanks for mentioning it. I have one locomotive that behaves that way, a Lionel HO Challenger. It, for some reason, goes to sleep on its own if I fail to engage it on my throttle over two or more sessions. As soon as I acquire it, it will pop back into sounds mode and act as if nothing had happened. It's an earlier QSI, about the same vintage as the Paragon Hudsons of 2003/4.
And, some of us do actually use, and like, that coma-mode that QSI introduced. It may seem weird to some users, but I quite like it, along with that super muting customization where you can partially mute the decoder with F8, and not have it binary sound/off. But, we forget if we have put the decoder to sleep with successive presses of F9.
riogrande5761 If QSI is defunct, maybe it would be best to stick with a "live" brand like Soundtrax or ESU or TCS.
First, use the three step QSI reset and do it on the mainline, not the Programming Track.
(1) Set CV49 to 128(2) Set CV50 to 255(3) Set CV56 to 113
As engine resets you will hear the 3 toots.
Test out the locomotive using address 3.
Second, try to program the long address on with POM (Programming on the Main), not on the Programming Track.
richhotrainFirst, use the three step QSI reset and do it on the mainline, not the Programming Track. (1) Set CV49 to 128 (2) Set CV50 to 255 (3) Set CV56 to 113 As engine resets you will hear the 3 toots. Test out the locomotive using address 3. Second, try to program the long address on with POM (Programming on the Main), not on the Programming Track.
Is it normal on power up of these QSI decoders that they will always give the 3 toots, or only when resetting? Mine does this every time the power turns on.
Water Level Route richhotrain First, use the three step QSI reset and do it on the mainline, not the Programming Track. (1) Set CV49 to 128 (2) Set CV50 to 255 (3) Set CV56 to 113 As engine resets you will hear the 3 toots. Test out the locomotive using address 3. Second, try to program the long address on with POM (Programming on the Main), not on the Programming Track. I'll try the reset process you detailed tonight. For what it's worth, I tried programming the long address on both the programming track and the main previously with no luck. I did not have the reset instructions you provided though, so I'll give it another shot. Is it normal on power up of these QSI decoders that they will always give the 3 toots, or only when resetting? Mine does this every time the power turns on.
richhotrain First, use the three step QSI reset and do it on the mainline, not the Programming Track. (1) Set CV49 to 128 (2) Set CV50 to 255 (3) Set CV56 to 113 As engine resets you will hear the 3 toots. Test out the locomotive using address 3. Second, try to program the long address on with POM (Programming on the Main), not on the Programming Track.
I'll try the reset process you detailed tonight. For what it's worth, I tried programming the long address on both the programming track and the main previously with no luck. I did not have the reset instructions you provided though, so I'll give it another shot.
The three toots never occur upon power up. Or, should I say, the three toots should never occur upon power up. That could well be your problem. I now wonder if the decoder is resetting to factory default upon power up which would delete the long address. Whenever you reset a QSI Quantum decoder to factory default, the long address is cleared out and deleted as part of the reset.
richhotrain Water Level Route richhotrain First, use the three step QSI reset and do it on the mainline, not the Programming Track. (1) Set CV49 to 128 (2) Set CV50 to 255 (3) Set CV56 to 113 As engine resets you will hear the 3 toots. Test out the locomotive using address 3. Second, try to program the long address on with POM (Programming on the Main), not on the Programming Track. I'll try the reset process you detailed tonight. For what it's worth, I tried programming the long address on both the programming track and the main previously with no luck. I did not have the reset instructions you provided though, so I'll give it another shot. Is it normal on power up of these QSI decoders that they will always give the 3 toots, or only when resetting? Mine does this every time the power turns on. Those three programming steps are the standard QSI Quantum decoder reset to factory default. The three toots are confirmation that the reset to factory default has occurred. The three toots never occur upon power up. Or, should I say, the three toots should never occur upon power up. That could well be your problem. I now wonder if the decoder is resetting to factory default upon power up which would delete the long address. Whenever you reset a QSI Quantum decoder to factory default, the long address is cleared out and deleted as part of the reset. Rich
Those three programming steps are the standard QSI Quantum decoder reset to factory default. The three toots are confirmation that the reset to factory default has occurred.
Even though OP says there is no reed switch, it sounds like the switch could be stuck to the reset position. If it can't be unstuck, I think its possible to remove a reed switch and it would not effect anything, except the ability to reset the decoder manually if needed.
- Douglas
Let me add a few more thoughts to this most recent discussion.
There may or may not be a reed to use in the reset. There may or may not be a wand to activate the reed switch switch. There should be a jumper on the decoder to reset to factory default. Various versions of the QSI Quantum decoder used different methods.
Personally, I like the three step CV reset process. Once the three step reset is completed, if POM is used to reprogram the long address, it is advisable to disable voice read back that normally occurs with QSI Quantum decoders. So, CV62=0, to disable voice read back.
If the OP wishes to pursue the reed/wand/jumper approach, the NCE website provides some good guidance, as follows:
The Broadway, Atlas, and Lifelike equipped with QSI (ID 113) DCC sound engines use a different reset procedure. Check the manual that came with the engine. Some use a magnetic “wand” for the reset. If there is no "wand" then open the engine or access hatch and find the reset jumper. Remove the jumper and restore the track power. There will be an audible acknowledgment from the engine at the end of the reset sequence (usually 3 toots of the whistle/horn). Replace the jumper and test out using address 3.
If a "wand" is present, locate the internal reset reed switch and place the "wand" over it for a moment. When track power is restored the engine will go through its reset procedure and let you know when done. To find the internal reset (reed) switch, place the engine on active track. Slowly move the "wand" over the top of the engine or tender until the engine alerts you, audibly responds, by a message.
Lastspikemike In theory you could remove the reed switch and leave the contact points with enough wire to connect a jumper manually if need be. Same idea as the rest button which you could also wire in place of the reed switch. From the three toots problem it seems likely the decoder is constantly in reset mode when it powers up. That would explain the symptoms.
In theory you could remove the reed switch and leave the contact points with enough wire to connect a jumper manually if need be. Same idea as the rest button which you could also wire in place of the reed switch.
From the three toots problem it seems likely the decoder is constantly in reset mode when it powers up. That would explain the symptoms.
In lieu of removing it, I have read in the past where others were successful in unsticking the reed by flicking the glass tube with their finger. Being careful of course.
By the way, the reason for disabling the voice read back is to avoid confusing the decoder if the POM instructions are keyed to quickly while the voice red back is still responding to a prior instruction.
In fact, it may be best to disable the voice read back even before the 3-step reset to factory default. Key in each command slowly and then pause before keying in the next command. If you do it slow enough and pause between commands, you don't necessarily have to disable the voice read back.
But, prudence dictates that you disable the voice read back in any event.
During the 2004-2005 time frame, I bought a number of BLI Paragon steam locomotives, most of which I have now sold to trim my steam roster.
However, I still have two of them, a 2-10-4 and a 4-8-4. The relevant instructions differ a bit, so I can summarize both sets of instructions for resets. Incidentally, both boxes have a magnetic wand, for what that is worth.
In the 2-10-4 instructions, the reference to reset is as follows: A misbehaving engine can be reset by toggling the switch to the RESET Position as shown above and cycling the power to the locomotive. The sound system will say "Reset". This indicates that the electronics are back to factory settings. Toggle the switch back to the Standard Position and operate as normal.
There is no mention of the wand or reed switch. When the instructions say to "toggle", the switch is actually a slide bar.
In the 4-8-4 instructions, there is a reference to the Quantum Reset Jumper to return the locomotive to factory default. The instructions read: Some newly released products may utilize a reset switch instead of a jumper. See separate instructions.
The instructions then continue: Turn off power. Remove the jumper by pulling up. Reapply power, the whistle will respond with three toots. Turn power off and reinstall the jumper. The decoder is now reset to factory default.
The separate instructions read as follows: To reset the electronics to default settings, off track, toggle the switch to the left, place on track and listen for 3 whistle toots. Toggle switch back to the right. The decoder is now set to factory default.
From all of this, my guess is that the OP is correct. There is probably no reed switch to wave the wand over. There may or may not be a jumper. I do recall resetting to default through the use of the jumper on some of my BLI steamers, so the jumper is there on some BLI steam locos.
Okay guys. Just got a look again at the problem engine. My apologies, there are reset pins, but there is nothing attached to them, so that isn't the problem. I did the reset procedure Rich provided, and reprogrammed the long address using POM, cycled power, and the decoder reset itself (3 toots!) and it's back to address 3. Time to address the decoder. I appreciate everyone's input on this. I just need to decide if I chase the chip replacement or swap it out entirely. Thanks again everyone.
Water Level Route Okay guys. Just got a look again at the problem engine. My apologies, there are reset pins, but there is nothing attached to them, so that isn't the problem.
Okay guys. Just got a look again at the problem engine. My apologies, there are reset pins, but there is nothing attached to them, so that isn't the problem.
In my instructions, there is a reference to the Quantum Reset Jumper to return the locomotive to factory default.
The instructions do say if there is a jumper to turn off power. Remove the jumper by pulling up. Reapply power, the whistle will respond with three toots. Turn power off and reinstall the jumper. The decoder is now reset to factory default. So, the instructions do say to reinstall the jumper.
Do you see a slide switch under one of those covers on the tender?
If there is one, that is another way to do a hard reset.