Search for Replacement 9x Potentiometers!

[ReSt]

The pot should have a value of around 5 kohm like the original ones or up to 20 kohm measured from one outer pin to the other outer pin.
The absolute value is not that important, as it is used as a voltage divider and not as a variable resistor.

There are linear pots and logarithmic pots. We need a linear pot. That is when you turn the axis by let's say 10 degree, the change of resistance between one outer leg and the sweeper will be the same, independent of the relative angle of the axis.

Most pots have the full resistance variation with an angle of 270 degree. We need a pot the has its full resitive variation with an angle of only 60 degree. This is important, because the resolution of the servo signal is directly affected by this.

As I said above, the pot is used as a voltage divider. One outer pin is connected to +5volt, the other outer pin is connected to ground. Depending on the position of the sweeper, you can measure a voltage at the sweeper that varies from 0 and 5 volt. If you have a normal 270 degree pot, you must turn the axis the full 270 degrees. If you have a 60 degree pot, you need to turn the axis only 60 degrees to get the full range.

Reinhard

[mpjf01]

Thank you, that helps a lot.

A further question. Some of the radios I have use a pot for the stick and a separate pot for the trim. The 9x seems to have just the one. If I was to replace the whole gymbal can the dual pot type be used? (assuming that the other specs are met if that's even possible).

And. If a pot with the correct physical characteristics could be found (such that it could be mounted in the gymbal) but with different electrical characteristics - say a different overall resistance or a different value over the 60 degrees, could such a change be accommodated in the firmware? If it could, I would not be suggesting that we have different firmware for every different pot that may be available, just agree to standardise on one.

As a matter of interest I have a now disassembled JR XF631 radio. I measured the pots (which are still attached to the gymbals) and it seems to me that they are 5k ohms. The voltage varies from 5v at one end of stick movement to around 1 volt at the other. I don't know what effect if any the small trim circuit board on the gymbal may have on the measurement.

[kaos]

Any one know if the Flysky CT6A or CT6B pot can be used on T9X? they are both made by Flysky. any chance they are the same. if they are, there are cheap CT6A/B out there people wants to throw away. Actually, I still have a CT6A sitting there. would be great it can be used for spare parts for T9X.

[Kilrah]

And. If a pot with the correct physical characteristics could be found (such that it could be mounted in the gymbal) but with different electrical characteristics - say a different overall resistance or a different value over the 60 degrees, could such a change be accommodated in the firmware?

As ReSt said, actual resistance matters little. The 60° are important though, while standard 270° pots would work you'd end up with low stick resolution unless a circuit is added between the pot and radio. This has been mentioned several times in this thread already.

[ReSt]

A further question. Some of the radios I have use a pot for the stick and a separate pot for the trim. The 9x seems to have just the one. If I was to replace the whole gymbal can the dual pot type be used? (assuming that the other specs are met if that's even possible).

Probably not, as trimming in the 9x is done via switches.

Reinhard

[jhsa]

Guys, what do you say to this? A friend of mine found this a while ago.. He even posted it on RCG..
These sensors as far as I understand can connect via serial, I2C, or even send out an analog signal.. All programmable, end points included. But I think there are different versions that do different things. we would just have to find out the one that would work with the 9x. It even comes with a PCB that could maybe be adapted under the gimbal. It's also possible to buy only the chip, I think. These are triaxis sensors. we would need only one per stick..

http://www.rcgroups.com/forums/showpost ... ount=10969

http://www.melexis.com/mlx90333

http://de.futureelectronics.com/de/tech ... .aspx?IM=0

João

[AlanGP]

Four pages later, here is my view. The actual pot resistance is not that critical, since the wiper is connected to an analogue input on the Atmel processor, which is high impedance and doesn't actually care, only being interested in the voltage (0 to 5 volts) at the input. In addition, the range of stick and by default, servo movement (max to min) can be calibrated using the stick calibration feature on the (er9x) TX.

So, within reason, any value of pot could work. The more important issue is the physical dimensions of the pot in order to fit the gimbal.

In the final analysis, if you have already spent considerable time and effort and a little money fully modding a 9X, for example, to Frsky ER9X telemetry standard as I have then one option would be to buy another 9X for spare parts - or wait until HK release the 9XR spare gimbals and use the pots from that.

Alan

[Clivew]

Yes, I agree any pot value within reason will work, but what about the resolution?
I don't think the 60 deg pots are manufactured without good reason

[AlanGP]

True - I hadn't really considered that.

However, if a linear pot is fitted and the gimbal calibration is carried out, how is the resolution relevant?

A specific angular movement of a gimbal will deliver an equivalent servo movement and an end stop move of the stick should deliver a maximum servo throw. I have tried a number of replacement pots, not fitted inside the gimbal, and it seems to work. The ER9X software seems to cope very well in the area of stick movement/servo calibration.

Alan

[MikeB]

If you have a 60 degree pot, you get nearly the whole 5 volts swing at the analog to digital input. On er9x, this corresponds to the digital value going from 0 to 1023, so you can have 1024 distinct positions of your servo. If you fit a 270 degree pot, you reduce the voltage swing to about 1.1 volts, giving a digital value that only changes by 225, so you get just 225 distinct positions of your servo.
If this is good enough for your models then fine you are OK.
The software copes, but your resolution is not so good.

Mike.

[LTMNO]

Guys, what do you say to this? A friend of mine found this a while ago.. He even posted it on RCG..
These sensors as far as I understand can connect via serial, I2C, or even send out an analog signal.. All programmable, end points included. But I think there are different versions that do different things. we would just have to find out the one that would work with the 9x. It even comes with a PCB that could maybe be adapted under the gimbal. It's also possible to buy only the chip, I think. These are triaxis sensors. we would need only one per stick..

http://www.rcgroups.com/forums/showpost ... ount=10969

http://www.melexis.com/mlx90333

http://de.futureelectronics.com/de/tech ... .aspx?IM=0

João

I like this... I would buy that if it were usable...

[martin555]

I am interested too what will be the best replacement.

recently I just purchased a 9x (mode 2 with module) from HK, and the center point of the stick could not be calibrated, it always moved away from center by itself.

I BEGGED HK to make possible for me to PURCHASE some replacement pot-s from "the manufacturer" but that "would not be possible".

[jhsa]

Maybe if Mike would be interested in have a look at these sensors we could donate some € or $ so he could buy a couple of them.. What do you say Mike?

João

[Flaps 30]

Seems like a good idea. Until you start considering the amount of work required mechanically to use the sensors shown with the 9X.

What is required is something like a drop in replacement or a simple modification to install a hall effect device + magnet + amplifer (if required) so that the existing gimbal can be used.

I have yet to experiment with the hall effect devices that arrived a couple of weeks back. Hopefully they will work well enough to give a reasonable resolution without the need for an amplifier. Shame that to do the test fully will require the ripping apart of a good pot. It would be nice to find a pot that has the same mechanical dimensions/fitting as what is there now.

[jhsa]

What is required is something like a drop in replacement or a simple modification to install a hall effect device + magnet + amplifer (if required) so that the existing gimbal can be used.

But it is what the sensor above is.. a Hall effect sensor.. Only triaxis.. Also that sensor can be programmed and calibrated to output the range we want.. There must be a easy(er) way to install it. it has mounting holes, maybe they could be used..

João

[MikeB]

First problem is actually getting any, Digikey don't seem to have any stock, only Future Electronics seem to have any.

Mike.

[jhsa]

I think Future Electronics is here in Europe (Germany) right?

They output PVM also. I guess they could be directly connected to the m64/128 analog inputs replacing the pots without the need for any other circuits/boards, right?

João

[MikeB]

They also seem to take 13.5mA each at 5 volts.

Mike.

[jhsa]

could that be too much for the 5V reg on the 9x board? Is the reg working on the limit and can't handle extra 30mA? I'm asking because I never really measured it..
Could we use a regulator for the sensor boards or would that mess with some reference voltage??


João

[MikeB]

All that will need to be checked.

Mike.

[jhsa]

So.. should we take that as a yes? would you order a couple and have a look?? And if it works of course you would be the first to own a 9x with some fancy gimbals.. followed by all the rest of us I would guess..

João

[MikeB]

As I have plenty to do, perhaps we could wait to see how Flaps30 gets on.

Mike.

[jhsa]

ok.. But I think that it might be more difficult to install a sensor inside every pot and be reliable than this solution..

[Flaps 30]

My idea is to put the magnet on the shaft assembloy inside the pot. The sensor itself will be on the outside of the pot, with a thin fibreglass sheet cover between the sensor and the pot which should make the setting of sensor fairly easy to do. I have some time on Monday. So I will start playing with the hall sensor and the magnets and report back.

[Flaps 30]

Had some time today to play around with the Hall Effect Device and it does look promising. The magnets I have are not really suitable for use with the system, so I have a few different types coming soon. They should be with me in a couple of days. The magnets may be a sticking point with the idea. We shall see.

The device that is being tested is the Allegro A1302. Supply voltage used was 5.08 Volts. Results as follows:

Output at rest/no magnetic force=2.52 Volts
Low=0.1 Volts
High=5.04 Volts

Note that the measurements were taken with no load (other than the meter) on the output. The spec sheet states with a current of 1mA the output can go between 4.64 Volts (source) Min and 0.25 Volts (sink) Max. Not sure if this will be a problem though, depending on the current required from it or the current it is expected to sink. The figures shown are the worst case.

My concern is the A1302 current draw according to the spec sheet could be a maximum of 11mA I measured mine at 7mA at rest. So may be with four of them it might be wise to run them off another five volt regulator.

[jhsa]

or replace the radio's regulator as part of the mod..

[MikeB]

Currently, the 5 volt supply is used both for the sticks supply and the 5V reference for the analog to digital converter in the processor, so they should really be the same source.
The idea of upgrading the existing regulator is therefore probably best.

Mike.

[jhsa]

Is there a better regulator in the same package?
I have some AMS1117-5.0 but they are bigger. thez are low drop..

Joao

[MikeB]

That may be a different pin-out. One problem with using the same package is heat dissipation. A higher current draw means more heat in the regulator.

Mike.

[jhsa]

yes, that's true.. so, anyway if we need a bigger package the pinout doesn't matter as we can't replace one with the other.. would have to put it on a stripboard for example and connect with wires. don't even have to be on the current reg pads but somewhere easier to solder I guess..

João