SCN5 / SCN6 Hardware Thread

Re: SCN5 Hardware Thread

I think they are too expensive, and cheaper ones work as well. Those converters mentioned from Ferrari_man and Fahim use FTDI chips, which are known to work well.

But it is recommeded to use automatic ones, as they are easier to set up.

regards

Re: SCN5 Hardware Thread

They are FTDI chips, but aren't they a different FTDI chip implementation? I can see that the ebay links posted above are definitely not the 232R chip.

I have asked several times for people to post any performance numbers, or comment on the stability of their adapters. Noone has, so as far as I'm concerned, they are likely to work, but remain untested and we have no performance data on them.

To me, this makes the safe bet the more expensive FTDI based chips from Trossen until someone posts information stating that the other FTDI chips work well with Vista, dont crash, dont have IRQ confilcts, and achieve a reasonable performance result.

@Ferrariman - Did you get your adapters and get them up and running yet? Sure would be nice to have a confirmed low cost alternative. I am building a second sim, and would buy those adapters if they achieved similiar performance.

Re: SCN5 Hardware Thread

Dear sirs,

bvillersjr writes:

I'm not saying that there isn't a better solution out there, but I am saying that this solution is the only one known to perform at this time.
End quote:

egoexpress writes:

But it is recommeded to use automatic ones, as they are easier to set up.
End qoute:

There is differential of opinion. 2ms, auto, non auto.
But from bvillersjr advice I have learnt 150mm SCN5 are the best actuators for this application.
I expect X-sim software to work aswell for this actuator as it would for 100mm.

To conclude I need to now find the best RS485/USB adaptor x 2.

The one bvillersjr suggested is $37 approx. I take it this would be non auto.

Kind Regards.

Re: SCN5 Hardware Thread

Both my SCN5s and the adaptor are with customs right now... they need some papers on them. I think I will get them next week.(customs in Denmark are really slow)

Maybe Thanos has got his.

I have two exsams left before chrismas and two after new years and most of my time are going that way.

Ferrari_man

Re: SCN5 Hardware Thread

Dear sir,

Can I ask which scn5 size and adaptors you have purchased?

Kind Regards.

Re: SCN5 Hardware Thread

I believe that my adapter is technically an auto-adapter per an evaluation of it by SirNoName a couple of weeks ago, but I switch it in X-Sim using DTS. I use a switching time of 1ms, a receive delay of 0ms and a switching timeout of 30ms. These settings achieve 140'ish SCN5 updates/sec per the SCN5 Plugin dialog box. I say 140'ish because it seems to vary between 140 and 154.

There may well be a faster adapter out there. I'm not sure how important this is though. There must be some maximum number of updates per second that an SCN5 can handle via it's half-duplex interface anyway. Unfortunately, I don't know what that limitation is, so I don't know if pursuing a faster adapter than the one I have would be of any benefit.

: Does anyone know what the max inputs that an SCN5 can handle is? More importantly, what would it do if more packets were sent than it can handle. Does it discard information? I'm not sure if this is even possible to send more information than it can handle since the baud rate is predetermined at 115200. Maybe the better question is how many updates per secnd can be sent at 115200 baud?

EDIT: Note switching this adapter with DTS improves the updates/sec to 176, but the switching timeoput has to be upped to 55ms in order to achieve this or the actuators won't be detected at startup.

Re: SCN5 Hardware Thread

First, I have to thank bvillersjr for his help with what to do.

After changing numbers and trying all kinds of combinations this is the best I can get with the $7 ebay adapters.

Any changes lower the numbers. Both adapters can do 200 outputs/second. Sometimes one is 166 and the other 200, most other times they are both 200. If I change any setting (for example receive delay to 1) both always get 166. Which way is better? Does anybody know what the physical limit/second of the scn5 is?

Re: SCN5 Hardware Thread

Looks like a winner. I'm going to buy a set for my next sim and give em a try as well. I don't think I would be to concerned about seeing a different number on on actuator than the other. It seems that these numbers update on some kind of timed interval. Mine fluctuate as well, however, I havent seen one actuator different from the other. You may have just been switching to view the other actuator at the time of the update? Is it a common occurance that they are both report different numbers?

Re: SCN5 Hardware Thread

They are not different if at lower outputs (166 and lower), but over 166 once it's 200/200, then 166/200, 166/200, 200/200, 200/200. It's always the same adapter that is lower, so it might not be the same quality.

Re: SCN5 Hardware Thread

Hmm, that is a little corcerning. I've never seen that, although the best I can get is 176 updates/ sec and you say it happens more often at higher speeds. Are you using the latest SCN5 Plugin? If so, we will need some advice from EgoExpress or SirNoName on this one. I'm not sure how the profiler would deal wuith the variances if they occur frequesntly. Knowing SirNoName, it's already handled, but better check to be sure.

Re: SCN5 Hardware Thread

Can I ask you both a question about the FTDI RS485 adapters you use?

If you create a hyperterminal connection with them (115200,8,n,1,flow control=none), can you see what you are typing? I just wonder if these adapters include the extra ECHO back remover circuit, that I added to my DIY USB_to_RS485 adapter.

@Sirnoname: One of my DIY USB_to_RS485 adapter is on the way to you already! I hope it will be delivered to you soon!



Regards, Thanos

Re: SCN5 Hardware Thread

Which corresponds 5ms! That is enough for sure. Even 100 outputs per second would be enough! Sirnoname gave an explanation on the very beginning of this thread.

Regards
Christian

Re: SCN5 Hardware Thread

What aboput those of us that do not use VSync, and therefore have a higher framerate than the monitor refresh rate? Mine varies between 100-200.

Re: SCN5 Hardware Thread

Ok, lets say the converter should have at least an output speed that is equal or higher than the framerate, that your pc is able to deliver constantly.

@Mr.Burns
Did you assigned your comport as serial printer already, as Thanos suggested recently?
Would be great if you could tweak your converter to output stable 200outs per sec. Cause I really would like to recommend this converter.
Do you have a possibility to check them on another computer to see if those irregularities in output speed arent caused by your pc or serial PCI-card?

Regards
Christian

Re: SCN5 Hardware Thread

Didn't try anything else yet. I don't use a serial card, it's all through usb. I will try on another computer and the serial printer to see if it will make a difference in a little while.

Re: SCN5 Hardware Thread

You have to set the serial printer option for USB->serial converters as well.

Re: SCN5 Hardware Thread

What is the max speed of the SCN5 controller?

Re: SCN5 Hardware Thread

Tried another computer and the same results.
I think 166 is good enough for $7.

Don't have any serial printer option in the com port advanced section. No Miscellaneous Options at all.

Re: SCN5 Hardware Thread

Most of us run games probably with framerates alot under 100fps. I think a converter with at least 150 outputs/per second and a single pc with dual- or multicore-processor is sufficient enough. Frex, for example, drives them with a single pc as well.

I cannot see why we should try to get 2000 outputs/sec, if we are not able to deliver gamedata in that speed.

I think we should recommend those 7$ converters.

Regards
Christian

Re: SCN5 Hardware Thread

If sirnoname can try the ones I sent, he could confirm if they also go as high, and it's not just a luck thing. Also, they are serial only, you still need rs232-usb adapter (which I found for $2, so the total cost was $9) and it was all free shipping.