DSX9 Fault

[Steve]

Seems my DSX9 has developed a fault.

I've noticed recently (maybe over the last 3/4 weeks or so) that a couple of my helis weren't flying quite the same as they used to. So I went back to basics and started checking pitch curves and servo centres and stuff as you do.

What I found was that the pitch curve had risen by 3 deg, so I had -7, +3 and +13. That's a bit odd I thought, so I reset the servo horns and rejigged the heli back to -10,0,+10 with no probs. Out of curiosity, I then checked my other 3 helis, and guess what, they were all out by +3 deg as well, now the only common link between the 4 helis is the Tx.

More investigation revealed that the tail and throttle servos on all 4 helis were fine - they had not moved at all and it was only the swash servos that were affected.

Got talking to xygax in the chat about it the other night and he kindly lent me his HiTec servo analyser which can measure the signal output from your Rx in each channel. So I set up a fresh blank new model heli in my Tx with mechanical single servo mixing, 50-50-50 throttle curve and 50-50-50 pitch curve and measured the output to each channel from the receiver. What I got was 1500us (or very close - within +-5) on all the channels except ch6 which showed 1590us (i.e. miles out).

I then changed the heli to a 120 deg CCPM model and got a reading of 1560us to each of channels 2,3 and 6 (again miles out). This reading of 1560 (where it should be 1500) is equivalent to +3 degrees of pitch on the swash.

So it appears that I have developed a fault in channel 6, which is then slightly diluted and shared between itself and channels 2 & 3 in a CCPM machine due to the mixing.

Can't say I'm best pleased as it's less than a year old, but at least it hasn't led to a crash. A clubmate of mine has a spare DX7 which he has lent me to use for now while I send the DSX9 back to Macgregor, and guess what.... when I set up a heli the DX7 this morning, the swash went straight back to where it should have been all along. So I think the evidence is over-whelming and I'll be on the phone to Macky-G on Monday.

[Smoothound]

Not good

At least you found it before it caused any real problems though

[pchristy]

Be careful about reading too much into these figures! Its possible you have got a genuine fault, BUT....

Servos are designed to operate over a range of 1-2mS, with 1.5mS as the nominal neutral - or putting it in your terms, 1000-2000 uS with 1500uS neutral. Most sets by default only use 1100 to 1900 uS, with the remaining 10% being taken up by the "trim".

So, your figure of 1590uS represents an error of approximately 10%! The 1560 uS figure represents an error of 6%. That's certainly higher than I would expect to see - a figure closer to 1% would be more reasonable - but would a 10% error produce a 3 degree change in pitch? That implies that you have 30 degrees of pitch range, which sounds a bit on the high side! Are you sure of those readings?

Also, have you measured the outputs from the receiver outputs with all the swashplate mixing turned off? I'm wondering if one of the stick pots has worked slightly loose. Measuring each channel individually may enable you to isolate it.

Also bear in mind that having the "expo" function turned on can have a bearing on these measurements, so try taking them with the expo both on and off!

--
Pete

[Bigbellypopper]

the other thing thats worth checking is that the slider switches on the side of the transmitter have been inhibited, as from what I can remember, if you move these, the pitch will change

[Tim Procter]

the other thing thats worth checking is that the slider switches on the side of the transmitter have been inhibited, as from what I can remember, if you move these, the pitch will change

Good call Russ, Yes you need to inhibit the "Pitch trim" in the setup menu.

If this has not been done then catching the trim gimballs on the side of the TX will cause this very problem.

Good post Russ.

[xygax]

The thing that gave it away is that there were 4 helis set up the same in the first place so all were to the same place at 0 pitch. all are CCPM and all have the same issues at this point in time.

This would indicate that the transmitter has changed....! The throttle servo pulse was measured whilst establishing the cause of the problem and its center point hasnt moved. This should remove an ambiguity that is mechanical within the stick movement on the raido.

This leaves the digital section within the Tx. Somthing very odd has ocoured as the relationship between the throttle and pitch has become disconected or slipped. The reason is unclear but may be some inbuilt constant or factory setup has become couroupted and this is worrying. Has it finished moving or will it continue only the manufactures can answer this..... and its safer to return than have a heli in flight that has a pitch range of 20 degrees all above zero..... !

Just my 2p

Steve

[Steve]

Like xygax says - all 4 helis have been affected, all by the same amount and all in the same direction. The sliders are inhibited on all my helis - only use them for planes.

To reiterate - as a test I set up a brand new blank heli in normal mechanical mixing mode (i.e. one servo only for each of elevator, aileron and collective) and set the throttle an pitch curves to 50% flat line. This showed all channels giving about 1500 except channel 6 (the collective channel) giving 1590 despite being on a 50% flat line with all endpoints at 100%, no subtrim, exp off etc etc - i.e. a blank fresh model that's never been used in the Tx.

I then switched it to a CCPM heli with everything else remaining the same. Now channels 2,3 and 6 are all showing 1560. CCPM collective is two electronic mixes from channel 6 to ch2 and from ch6 to ch3, so now they're all affected since the problem source is with ch6. The reason it's not showing 1590 in CCPM is because of the Swash Mix function which is set at 60% by default - this is what is reducing the observed value from 1590 to 1560 or there abouts.

If as Pete says, the Tx uses about 400 points from the centre to the extreme (i.e. from 1500 to 1900 being the max travel of the servo one way), then this would be achieved by setting the servo travel limits to 150%. They're not at 150%, they're at 100%, so now we're only actually using 267 points from centre to full travel. The displacement of 60 points in one direction is 22.5% of the full travel in that direction. Now bearing in mind that the servo is a rotating mechanism, and a small rotation about it's centre has more effect on the blade pitch than a large rotation towards its extremes, then this 22.5% can easily be 3-4 deg of final observed pitch.

[pchristy]

If as Pete says, the Tx uses about 400 points from the centre to the extreme (i.e. from 1500 to 1900 being the max travel of the servo one way), then this would be achieved by setting the servo travel limits to 150%.

I don't think I can have explained that bit too clearly....

The 400 points from center to one end would be achieved with the travel adjust set to 100%. If you up it to 150% it will go beyond that. Most servos can be overdriven by about 50% before they jam up - which is what you want to avoid at all costs!

You have to remember that to some extent we are limited by standards that were set back in the early to mid 60's when "digital" proportional systems were invented. Ever since then, any new system has had to retain "backwards compatibility" to ensure that servos will work with newer systems and vice versa!

When Doug Spreng invented the original "digital" servo back in the early 60's, he used a pulse that varied from 1-2 mS at a 50-60 Hz frame rate, and everyone has copied that since!

Modern computer radios default to these standard settings, but you can overdrive them considerably. If you set all the travel adjusts to 150%, and set all the swashplate mixing to 100%, then you could (in theory) drive the servo to its end stops. Most computer transmitters will limit to stop you doing this, as you could destroy the servo if you tried!

But to get back to the original topic, Russ makes a good point about the sliders (and one I should have pointed out), but also remember things like the hovering pitch trim as well.

As I said, you may well have a genuine bug in there. But it will help the service engineer who has to fix it no end if you can give him a detailed analysis of the fault. And that means being certain you've eliminated all the possible "red herrings" and narrowed it down to the minimum number of scenarios possible!

Best of luck, and I'll be interested to hear how you get on!

--
Pete

[stevo_m]

I read on a site in the US of people re-calibrating their TX's, I think in this case it was the 9303, so same as the DSX9.

You need to search the net to find how to access the service menu. This will then give you the option for re-calibrating the TX.

Or alternatively send it back to Mcgreggor who will do it for you.

Strange how 4 of you have the same problems though? Are they all the same age?

Steve

[Tim Procter]

Accessing the X9303 and XP9303 Service Menu:

• Pull Throttle Hold switch into the forward position.
• Pull AILE D/R switch up.
• Power the set on while keeping ENT button on front panel pressed.
• Rotate roller activator until Trim Step is highlighted.
• Press roller activator on Trim Step, rotate to AILE.
• Switch AILE D/R down, press roller activator, switch AILE D/R up, press roller activator.
• Repeat Step 6 four more times and on the 5th (total) time access should be granted to the Service Menu.

Here is a list of things you can do - should be in order as you scroll through each item. This is not a complete list, and there are likely lots of other things that can be done.

• From here, SW test allows you to test each switch and microswitch.
• The next menu item is Hard. Con. By rotating the selector bar it becomes obvious what can be changed by soft selection. Where the highlighter does not move it indicates that this is a hard connection on the PCB, but in conjunction with the table this can be modified accordingly.
• Next item is Stick Neutral. Center all sticks and side trimmers, press CLR followed by STO. This then stores all the centers for the sticks.
• Next is Stick Ga. Moving all sticks and side trimmers to max & min positions finalizes the stick calibration.
• Moving the selector across to the right allows for setting front panel voltage, and at what voltage the alarm will activate for low battery.

There are various ways to exit, and each one has a different effect. Total reset of system and loss of models stored, reset of the config data only. For straight calibration purposes, rotate back to EXIT and press the button next to : Reset start (No Initialized)

[Steve]

Thanks Pete - nicely explained.

busterboy - hmmmm... do I dare go into the underworld???

Some good stuff here - gonna do some more research with the set.

[antsis]

I've just bought a DSX9 so will be interested in what McGregor find. With regards the pitch trim, I was also advised by Mat at TJD to inhibit this.

[Steve]

Ok, just done some more testing with my Test model setup - the one with flat line 50% pitch and throttle curves. I've now inhibited everything that I possibly can in the device select menu - including the pitch trim, both the sliders and whatever else is inhibitable.

I've now got 1500 on every channel, so maybe there isn't a fault after all.

Now need to reactivate things one at a time time till I reproduce the problem.

[xygax]

Uh oh....

Operator Error ????

[Steve]

Seems to be the hover pitch, as has been suggested that is throwing things out of sorts. According to the output from the analyser, the best solution seems to be to inhibit it in the device menu, and set the trim step for this function to zero (i've also set the trim step for the hov thro to zero as well as I never use these two functions anyway).

Right, now need to see if this sorts out a real heli - seems to work on the test model...