It can run on the 360mm radius, but doesn't look so good while doing it...

One thing to note is that the overhang in front & back is quite notable, especially in gradients where the couplers can get stuck on the studs if the start of the gradient is too steep.

Hi Manfred, all
Ah, I thought I were the only one with a 37593 with two motors running against each other. Replacing the brushes did not help, but I suppose this still may be a debris / overlubrication problem. I still have to find the time to disassemble and clean the motors. Did you?

BTW, I just obtained a used 33592 on delta which does fine.

Indeed, the couplers getting stuck at a gradient made me rebuild some slopes, now looking much better with a more natural gradual slope.

Hi Manfred, all
That's very interesting! Although I would not tend to spend Eur 200 for upgrading my 8/14.
It appears that the 37593 decoder is just a 90602 variation with an increased power output. (Haven't checked this myself).
One would have liked to see a sort of double load regulation system to compensate between the two motors.
We may want to start a new thread in the digital section discussing this. [:p]

Sorry Monique,
2 men spoiling your impressions of the great 8/14 by arguing about load regulation issues. To get back to the heart of the matter: Yes, it's great, it manages the R1 radius and gradual slopes, and I'm sure that very few of them will have these finetuning issues. A must-have!

Yep,
this was exactly what I wanted to discuss in a separate thead.
How about supplying the two motors with a separate Lokpilot each?
These are programmable for double traction, and technically, there's no difference between an 8/14 and two 4/7s in double traction.
Obtain a delta version, add Eur 50 for two Lokpilots, and have a smooth drive!

Once again, we're still guessing that the Mä 8/14 decoder has no separate load regulation for each of the motors. Does somebody know the actual scheme and function of that dedicated decoder?

Hi Bart, I was thinking a bit more about this point.
Assuming that both motors are not exactly equal,I believe that the solution could be, as you said, the use of 2 decoders, but maybe the best to configure them as DCC 128 steps, in order to reduce to a minimum the differences between both motors at each of the speed steps. Do you believe this could work?
Regards,

Hi Bart, assuming that Jacques is right, as he certainly should be, you can make a celanup in both motors,especially the collectors, install new brushes, cross your fingers and see what happens.
If you are not satisfied with the results, try the twin decoders approach. You need an IB to drive both decoders in DCC128 mode, you should assign them different numbers and make your adjustments using consisting mode and, once both motors are running as you want, give them both a unique number.
To do this, I belive you need to install a new pickup shoe in the half-loco that doen's have it and make your adjustements individually for both locos.
Another approach could be POM, (Programming On the Mains, or Hauptgleisprogrammierung), depending on the decoder you use. Uhlenbrocl 76520 does support it, LokPilot doesn't inform about this support, several Lenz decoders do support this option. The advantage of this method is the ability to change some decoder parameters directly on the main track for one specific decoders address. You give a different number two each of the decoders, install them, adujst their CVs and, when ready, give them the same address.
Good luck, Bart.
Regards,

Bart,
A second pick up shoe might even prove mandatory in order to synchronize the two halves to obtain equally well operating units.
Hope somebody will bring an add on auxiliary driving decoder, with
S.U.S.I. and/or decoder lock becoming more popular.

Regards,
Manfred

Jorge,
You're right. I certainly should try.
However, in theory, things may become worse, since the single decoder then would 'sense' both motors.
No idea how that would work out.

Jorge,
sure, no hurry; This is one of many projects for next year..

Now just as a matter of general interest for those who may read this forum in the futute:
What would you recommend to someone who is planning to purchase an 8/14? A Delta (33591/3) or digital (37591/3) version?
In other words: Would the synchronizing issue discussed here be due to the specific Mä digital decoder (an thus a Delta version would be a safer buy), or to a possibility of the 2 motors having slightly different characteristics (in which case the choice of Delta versus Digital may not matter)?

Hi,

I have followed this discussion with great interest. I do not have any double traction projects myself but I have been thinking a bit about the problem.

I do not think that separate decoders and separate regulation is the best solution. The regulation tries to keep the speed constant independent of the load. If both motors are regulated independently there will always be a difference in target speed between the two decoders. Big or small depending on how well you can adjust it, but a difference there is going to be. Then, the marginally faster decoder will feel more load, and increase the drive strength. The slower one will feel less load ad thus decrease the drive strength. If the speed regulation were absolutely independent of load, you would end up with just one motor driving. This is exactly the opposite to what you want to acheive. Now, there is some load dependence and it might work anyway. LokPilot may work better than Märklin here since Märklin has better regulation.

The approach Märklin have chosen, with sensing on one motor and just replicating to the other one, should work well if the motors are equal. Small differences in strength/speed would at least not be amplified by the decoders. The problem, I think, is that the 5-pole DC motor has too little load dependence already without regulation. The standard AC motor is more load dependent. Maybe a solution could be to have a C90 motor in one of the locos and a delta decoder (with the same address) and an AC motor in the other unit. (I do not know, just trying to make intelligent guesses).

You may also test with adding small resistors in series with the motors. The faster/stronger motor will draw more current than the other one. There will be a larger voltage drop than for the slower motor. This will give a self-regulating effect that might even out the differences between the two motors (if the difference isn't too big of course). the resistors should be in the range 10-50 Ohm. Make sure you have sufficient power rating to avoid burning them.

All,
As Jorge pointed out, this thread is applying to all twin engine
locos or trains, such as the Ae8/14, F7 combos, PA1 combos and the new TEE.
Seems the discussion is leading us in the direction of the speed step number topic, slightly less powerful/more load dependable motors and decoders with adjustable degressive BEMF effect towards higher speeds (ZIMO).
Since there are several offerings of bi-or even trilingual command stations like the Digitrax Chief (only old trinary format), the IB,
ZIMO, hke and, hopefully, the new märklin system to come, I think
an Ae8/14 Delta version, converted to DC operation with the appropriate HAMO magnets installed and, controlled by a ZIMO MX64 HS,
could do the trick.
(MX64s are undergoing a redesign at this time.)

BTW, I learnt, that there were at least 5 different versions of 6090
armatures. I was able to identify at least 3 different versions.
Don`t know how many are around at this time and whether they are
identified by märklin with specific spare parts numbers.
They can be identified by their differing coil diameters from eyesight. In addition they`re different in top speed/slow speed/
speed curve response. So may be, some experimenting with different
armature shapes could ease the pushy pusher syndrome.
BUT, the single most important issue in my opinion is the adjustment
of the brush holders.

Question to Jorge.
What is the tinkering required to make C-track turnouts two-rail
compatible?

TIA,
regards,
Manfred

Hi everybody, is all ok with 2004?
I took a look on a Roco catalog. They offer two LKAB locos, both of them with TWO motors. For the AC version, they offer specific LokSound decoders.
Maybe the problem of both motors in parallel is already resolved...
Does somebody want to do a try?
Anyway, if it doesn't work, the decoder can be installed in a different loco and it's sound changed to the appropriate one...
Regards,