I have only dealt with about a half dozen decoders so far, but I am a TCS fan. My experience with Digitrax was not a bad one, but the TCS were easier setups and the motor control for me was better. I have an ESU 6 pin decoder on the way and can't wait to get it installed in my N scale switcher. I have a Kato FEF with the Loksound and it runs the best of all my stuff.
Tf you are looking for Motor Control Zimo and ESU are both better than Digitrax or TCS. As for having to choose from Digitrax or TCS I will choose TCS every time. I have never had a problem with a TCS decoder. However I have bought one to many Digitrax decoders New from various dealers that were Dead On Arrival. Digitraxx will warranty them but you have to send it back to digitrax at your expense and wait for a replacement to arrive. Then top that with having to modify the solder pads to make what's supposed to be a "Drop in Decoder" work. No Thankyou i may pay a little more for the TCS decoder upfront, but i don't have to pay extra for shipping and wait weeks to months for the replacement to arrive.
Post by davestubbs on Feb 12, 2017 13:15:19 GMT -5
Have to agree with tjdreams I've had to send back 3 Digitrax decoders and the solder problem on the pads is a pain. I also use my tcs decoded engines on DC layout and like the way the TCS works under DC. I had some trouble with digitrx but must admit I didn't dig into that problem too far.
Greg, Being relatively new to DCC and not having tried speed tables yet, I've found your discussions very informative and have sent me back to the "big books" from both manufacturers. But I must ask why?
TCS recommends using the motor trim and speed curve CVs and use the speed table as the last resort. When I look at the data from my locos, it seems from my data that motor trim and adjusting Vmax would cover what I've measured:
Yeah their would still be some variation but it would be within 20% and wouldn't you see that level of variation with load and grade?
Am I missing something important?
Also with Digi v. TCS their documentation describes the speed table processing as interpolation and averaging respectively. So if the description is mathematically accurate, I'd expect the same speed table to perform differently between the two decoders. I'll look forward to your results.
BTW: at my level of DCC, I haven't seen any difference between the two decoders.
Greg, I appreciate the time you've put into your reply. I'm at the point with DCC that I know just enough to be dangerous and I really wanted to understand your issues. I haven't seen your issues but then again I didn't know what to look for. I am wondering what the value of CV7 (manufacturer's id) is, just to see if all AZL4s are equal. I have both the DZ123Z0 and AZL4. The owner of the train store I do business with recommends TCS. He seems to feel that they put more effort into the basic throttle functions. He's apparently known for DCC into N scale so I tend to trust him. I was going to ask him about your issues this week but he was out sick when I visited. I've been doing DC for decades and always felt that DCC was a bit of cheating: there is a certain skill associated with a DC throttle controlling the locomotive in a prototypical fashion. But with AZL's drop-in decoder option, I thought it was really worth a try. I'll have to convert my test bench back to DCC and give it a try next weekend, although I suspect you'll have results before then. The speed data I posted is from an old photographic timer I had. Push to start, push again to stop. I used the Radio Shack IR emitter/Darlington pair to trigger it. At the time (around 2010-12) my main interest was just basic locomotive performance around a 220mm oval with varying load. Since then I've built the Arduino throttle I described in another thread in hopes of automating the process. I'll add a second IR sensor to my tunnel.
Greg, I'm so tired of listening to the Orville dam here in NorCal, I thought I'd look into it a bit as you suggested. It seems that what you describe is not uncommon and that TCS acknowledged the problem in the Z2 last year: www.therailwire.net/forum/index.php?topic=40182.0 (mid page)
Since it seems that you've installed more TCS decoders than I'll ever see, I was wondering if the problem is CPU specific? From my small sample of decoders here, it seems that they use both the PIC16 and PIC18.
Greg, If you contact TCS it would really good to know if they update CV7 with firmware changes and if so what the "fixed" value will be.
I agree it's good that TCS acknowledged the problem. Better if they reported it to the customer rather than the other way around. But at this time I'm not quite ready to directly fault TCS.
A common technology mistake is to assume the silicon is perfect. From my professional experience, this is far from the truth and even Intel made processors that couldn't multiply. I downloaded the errata for the PIC18 used on the AZL4 and the PIC16 used on the DZ123Z0 and MZA4. Because of errors in the silicon, it's not clear that PIC16 code loaded into the PIC18 would run correctly. Also it is possible that hardware errors in the PIC18 could cause the PWM to be loaded with the wrong data under the right circumstances. Since I don't know their code, this is just conjecture, There are published workarounds...if you go looking for them.
Don't forget to set CV3 & CV4 to 0 on Digitrax. This is the Momentum and the default is NOT 0. This means, every time the decoder looses power, it will re-ramp the motor speed. So you think the loco is NOT running and then is slowly (may) get going again. This is most apparent with dirty track. Even with the Digitrax decoder CV's set to 0, the TCS decoder will restart better on dirty track, quickly getting back up to speed.
Also on the Digitrax decoders, The NCE system Programming track, you need to go through the normal English Standard 'STD' setup, after you need to go through the Config. Just Enter, Enter, Enter, Enter and then answer Yes to Activate the Long Address. The decoder will run fine without doing this extra step but the headlight (F0F) won't turn on when the loco is put on the track or after dirty track recovery. If you do the extra Config section and Activate the Long Address, the headlight will always turn on when the loco is placed on the track (you know you have power). I don't see this on Digitrax system programming or the TCS decoders.
The default momentum settings are likewise not 0 on the TCS, it is set to 1 for both.
You know manufacturers' "comprehensive" manuals claim the factory setting for these CVs is 0 (page 32 and 21 respectively). Do I trust the manufacturers or the guys who have actually programmed and 100s the these things? It seems that the more I learn about DCC, the less I know. My vote above reflects that.