QSI problems with LGB MTS III

[beavercreek]

After some embarrassing lack of function with my new C19/QSI Quntum combo at the local Essex meet. I managed to untangle why the MTSII (with LGB handset and Massoth Navigator) was having problems as it was set to serial, not parallel. So the MTS III and the loco then was fine and Dandy when the MTS III was changed to parallel..
BUT..
Although the loco was demonstrating very fine low speed control when run from the laptop with either the QSI Programmer or 'Sprog' with Decoder Pro3, it would only jump to a trot when then controlled by the 55016/MTS III or Navigator/MTS III systems.
Speed steps were set to 28 on the decoder and on the Navigator. Even at setting '1' speed on both controllers the speed would jump to a brisk pace. control after that is fine. It is just the initial start that is so 'crude'.
As I say it is fine with the laptop control systems. you set the speed to '1' and after a nice pause while the sounds of steam etc it gradually moves the wheels very slowly then to '2' a little faster etc. Very realistic starts can be achieved. The MTS ssytem configurations seem to miss all the slow start and jump to the equivalent of speed '4' on the laptop control configurations even though the MTS or Navigator handset is only at '1'
The decoder was reset to factory default settings and the same behaviour was displayed.Also another QSI decoder was tried with the same effect.
Am I missing something...can the MTS III only do 14 speed steps even if the Navigator handset is set to 28?
Or is there a compatibility problem between the QSI decoder (with all of its possible extensive CV c ontrol of how the motion can be configured) and MTS III?
The MTS III unit is fine with a Climax fitted with a Tsunami DCC board or a factory fitted DCC LGB loco.

 

[Cliff George]

How much voltage does the SPROG or QSI programmer put out?

Is the problem possibly that both MTS and Massoth put out far more voltage?

I think you can adjust the voltage on the Massoth, can you experiment and see if doing so makes any difference?

 

[beavercreek]

Hi Von Kliff
The Navigator is being used attached to the MTSIII unit so voltage is what that unit pumps out 22v probably. I think that the Sprog or QSI programmer put out a similar voltage but at lower amperage.
The loco will also not come to a perfect halt when using the MTS system..it very very slowly still creeps forward even when another loco address is selected on the controller. It is as if the QSI decoder has not been told to 'stand still'. This does not happen with the laptop controllers.

 

[Cliff George]

OK another idea, remembering I'm no QSI expert.

Does the QSI have an option to turn off the possibility of analogue operation? If so does switching it to digital only (as an experiment) make any difference?

Von Kliff

P.S I presume that you have to have an external power supply to power the QSI programmer and Sprog, you are using one that puts out an approproate voltage?

 

[beavercreek]

OK another idea, remembering I'm no QSI expert.

Does the QSI have an option to turn off the possibility of analogue operation? If so does switching it to digital only (as an experiment) make any difference?

Von Kliff

P.S I presume that you have to have an external power supply to power the QSI programmer and Sprog, you are using one that puts out an approproate voltage?

Ah Von Kliff, it is affirmative for ze option to turn off DC operation...I vill try that. Although I do wish to have the option to run ze trains on DC or DCC. It will show if there is ze problem.

As to power supply. Ze QSI programmer and the Sprog have suitable power supplies and both can run as 'command station' mode

Your input is always welcome Herr Von Kliff

 

[ntpntpntp]

I had the same thought as Cliff: The SPROG and the QSI programer don't put out anything near 22V, at a guess I think more like 15V. Remember they're intended for use with any scale, not just G. I'll go and put my RRAMP dcc meter across mine in a bit (I've never actually done that before!)

 

[beavercreek]

I had the same thought as Cliff: The SPROG and the QSI programer don't put out anything near 22V, at a guess I think more like 15V. Remember they're intended for use with any scale, not just G. I'll go and put my RRAMP dcc meter across mine in a bit (I've never actually done that before!)

Hi Nick
In my ignorance of the finer points of DCC.....Even if the Sprog and QSI programmer were at let's say 12 to 15V, why would this affect how the loco speed ramps up. If anything they would not be able to achieve higher speeds. Does track voltage affect how a decoder operates slow speed running? Even if The MTS III is pumping out 20v and the programmers only 15v the decoder should demonstrate exactly the same functionality, it is just the load that might be too much for the small voltage....Of course this is an ignoramous talking here! :wave:

 

[ntpntpntp]

Well, my understanding is that the speed you get for a given speed step is simply a proportion of the total voltage available. (Not strictly accurate because we're using a "pulsed" constant voltage and varying the mark/space ratio, but it will do).
Therefore if you give the decoder only 15V instead of 20V then each speed step will only give you 3/4 of the speed you see at 20V. Hence if you set up the decoder with the programmer so that speed step 1 gives you a nice slow starting speed, when you then give the decoder 20V to play with you get a faster speed for the same speed step.

 

[ntpntpntp]

Here's my QSI programmer under no load, showing 15V as I guessed

... and under load (full speed with my old Aristo U25B) it drops to less than 11V . (Ignore the amps, I don't think this is accurate here)

Powered by a 12V unregulated supply, no wonder the voltage dropped under load

 

[ntpntpntp]

Here's my SPROG under no load

(Damn, the "under load" photo didn't come out) but the voltage didn't drop very much
... but the SPROG is powered by an old laptop/printer power supply which I think is rather better regulated than the one I use for the QSI

 

[beavercreek]

Thanks to you Nick and Von Kliff.
I have fiddled again (no hairs on the palms yet), with CVs, and turning the DC analogue option off had no effect but was worth a punt.
I decided to bring in the 1200Z from its 'special' place and hooked it up and tried with voltages ranging from 14 to 20 and you are perfectly right Nick (you knew you were!) the voltage does affect the starting speeds.
So..I investigated the speed/load options (there are soooo many parameters with a QSI board). Settled on the speed curves and found one that gave good slow starting speeds at 20V with no 'hop' up to a trot.
Also there was no 'creep' at stop with the Massoth.
I have then played with the chuffs, the chuff trim and other chuff related options (wheel diameter, scale etc etc) and got a pretty close match to motion movement and chuff at all speeds.

The last one to play with is the back emf settings. I will need to get those spot on so that my inclines, both up and down,are taken in the loco's stride

Many thanks again lads, much appreciated

 

[ntpntpntp]

The last one to play with is the back emf settings. I will need to get those spot on so that my inclines, both up and down,are taken in the loco's stride

Funnily enough the BEMF on the QSI decoders seems to be tricky to get right in some cases, especially locos with multiple motors. I noticed my3-truck Bachy Shay wasn't too happy and used to judder at slow speed, in the end I turned off the BEMF. It's another thing on my list to go back and have another go at tuning up. Pretty sure I've read on Greg Elmassian's site that he has problems with the BEMF sometimes.

 

[beavercreek]

Yes, Greg's site is a great source of wisdom on many things train related. I have heard that QSI had drawn up a solution to low speed running with back emf but not sure where to find it.
Although everything seems to be fine with the decoder and loco I have noticed a little bit of judder at slow speed acceleration whilst on the rollers. Will need to test it again with different BEMF settings on the track when the weather improves.