Drone kit -Building a ZMR 250

Preamble

I got hold of a clone of the ZMR250 and this is my build log. See also Drone kit -ZMR 250. See also Drone kit -Building a Tarot Mini 250 (TL250A).

Please note that this is an on-going blog, and things will change from day to day, as extra information is added, and errors are retrospectively corrected. Please bear that in mind, when reading this – This is not a finished blog, as yet, by any means.

See also Drone kit – ZMR250 PDBs, regarding the three ZMR250 PDBs that came with the kit and the blog on the 3D printed spacers.

Links

A good build guide to follow is the Dronetrest QAV ZMR250 assembly build guide.

Builds

I made a number of iterations during the building of the frame:

  • Built without PDB, only two bottom plates;
  • Built with PDB and 12 mm spacer, and two bottom plates. (PDB was the lower plate, spacer was inverted (so that the wire exit was furthest from the arms and the arms exit/gap was at the top) – I, basically, bolted the spacer and the PDB onto the bottom of the previous build without the PDB);
  • Built with PDB and 12 mm spacer, and one bottom plate (PDB is the middle/upper plate, with the LEDs on the top of the middle deck);
  • Built with PDB and 12 mm spacer, and one bottom plate (PDB is the lower plate, this is evidently the correct method, such that the LEDs are mounted on the bottom – see Anyone use the Diatone PDB? I’m stumped.), and;
  • Built with only PBD and one bottom plate, no spacer, with PDB as the lower plate.
  • Built with the bottom plate, a 20 mm spacer,  and the PDB as the middle plate, with the LEDs on the top of the middle deck.

I, finally, settled on the PDB being the lower plate – as it should be – with the pads for the LEDs facing down, used a 20 mm spacer, and then with the middle and smaller top plate joined with the 35 mm spacing rods.

After assembly I noticed that I had some triangles left over. They are legs, see ZMR250 remove triangles on arms?

ZMR250 leg
ZMR250 leg

Here they are in situ

To fit the legs to the arms, they must be carefully clipped on at the narrowest point of the arm and then push up to the thicker part of the arm, so that they are wedged in position.

Fitting the Motors

I got hold of four Emax MT2204-2200KV  motors and Flycolor 12A/450Hz ESCs (2S-3S LiPo 12.6V Opto) for £25 + £3.99, however, I was disappointed to see that they did not fit the arms that I had. Two of the motors fitted, while the other two did not. This seemed to be due to the holes being inconsistently drilled (in an ellipse rather than in a circle: 19 mm x 22 mm). Then I realised that the motor’s screws are not set in a circle but also an ellipse, and that they do actually fit.

However, one of the arms was most definitely drilled such that the large center hole was not in the centered between the four screw holes, and thus out of alignment with the motors axis. See photos.

So the axis was impinged by the plastic, when fully screwed in and because of this the motor would not turn.

Motor axle is hindered by the pastic
Motor axle is hindered by the pastic

However, from ZMR250 Motors and ESC, a user mentions that not all arms are compatible with 2204 motors, as they are drilled for 1806 motors.

Your ZMR arms might not fit 2204 motors though, as many of them only come with the screw hole pattern for 1806 motors.

So you may need to order up the upgraded arms for 2204 motors, or even the 4 mm thick 2204 arms which are extra durable.

Note that if you go for the 4 mm arms, you will need to go to a hardware store and pick up 1 mm longer arm screws to secure the arms to the frame (so 8-16 of them depending on if you run 2 screws per arm, or all 4 screws per arm) and you will need slightly longer motor screws as well to make up for the 1 mm thicker arms.

If you go with the standard thickness arms, then you can reuse the same arm screws. new motor screws should come with the motors.

I ordered some 3 mm M3 spacers to give the misaligned motor some clearance from the arm.

3 mm M3 spacers
3 mm M3 spacers

and mounted them between the motor and the arm

Motor mounted with M3 3 mm spacers
Motor mounted with M3 3 mm spacers

Mounting sequence of motors

The sequence of CCW and CW motors is as follows, note the numbering of the motors:

Quad rotor direction and numbering
Quad rotor direction and numbering

Determining the direction of motor rotation

Now if you find that you haven’t actually labelled your arms/motors, as to which are CW and which are CCW, but the motors have the propellers mounted then it is easy to tell, without having to switch them on. In fact there are a couple of indicators:

  • The way the propeller nut unlocks is the direction the motor should spin in. I.e. the motors in the boxes labeled ‘R’ should spin clockwise and their propeller nuts unlock by unscrewing them clockwise.
  • The highest edge  of the propeller blade is on the side to which the propeller will rotate

    Highest edge indicates direction of rotation
    Highest edge indicates direction of rotation
  • The straightest edge of the propeller is on the side to which the propellor will rotate

    Straightest edge indicates direction of rotation
    Straightest edge indicates direction of rotation

See also Wolf Paulus’ Journal – Propeller, for some interesting information. The image below is from that blog:

Straighter edge on rotating side
Straighter edge on rotating side

Rough Assembly

Here are some photos of the inital frame, ESC and motor assembly:

Using the KK2 mini flight controller

I had obtained a seriously small KK2 mini flight controller for £10. Connecting it up seemed easy enough, as the pins are labelled on the rear of the board.

Rear of the KK2 mini flight controller
Rear of the KK2 mini flight controller

The servo outputs 1, 2, 3 and 4, with the three pins (V, GND and signal), go to the ESCs, using the numbering system as labelled in the diagram above.

Note that the inputs from the receiver (AUX, RUD, ELE, AIL) have only one pin. The throttle connector (THR) can be connected using all three pins. See Mini KK2 Board Wiring ?

… only the signal wire for AIL, ELE and RUD will be attached to the board. (Usually white or lighter color). All three pins of the THR from your receiver will be attached.

See also, KK2 Mini help!!

For receiver wiring, follow the silk screening on the back of the board:

From top;
aux (self level on/off channel)
Rudder
Elevator
Aileron
Throttle
5v+
Ground

So, use a standard three wire (S, -, +) servo lead on the throttle channel, and single wire leads (signal only) for the other four channels.

Power the KK via the M1 output. If you’re using opto ESCs, insert the red wire from a 5V BEC into the servo lead coming from the M1 ESC into the lead plugged into the M1 pins on the board, and ground the negative lead from the BEC.

Also of interest, KK Mini manual???, which links to the video, HK KK-Mini Multi-Rotor Flight Controller – HobbyKing Daily:

The simplest set of connections for the KK Mini are shown in the diagram below:

Minimal connections for KK2 Mini
Minimal connections for KK2 Mini

In the same KK mini thread, there is a secondary issue:

Originally had opto’s connected to my kk mini on my DJI F330. Power was fed thru the RX via ubec. Everything working fine, flying super fast. Brought it in and decided to connect the voltage input from the lipo so I wouldn’t have another half mile recovery from low battery issues. long story short, connected positive from lipo solder point to LONE SINGLE BATTERY input on kk mini below m8, board screen flashes and text disappears. Needless to say I now have a dead board. LCD is illuminated but that’s it. Youtube is useless for a fix.

I was thinking possibly this was due to the fact the m1 pin was not receiving power from any ESC but thru the RX input.

Solution:

Run your +5v from the UBEC to the +5v on M1 and try the Batt. V again..

Soldering up the PDB

I put bullet connectors on the basic PDB and the ESCs power leads, so that I could easily dismantle the board if need be.

[photo]

The added weight was something that I could life with, for the moment. If I was to start racing with it then I would consider soldering the ESC power leads directly to the PDB.

Which size batteries?

Looking at a battery, I was thinking about a 3S, but a chap on eBay was talking about 4S batteries making his 250 racer fly along. So I looked into it and it does have to do with the KV rating of the motors. See here, Looking for guidance for first time quad builder. Thanks. Some comments from Holtneil:

Battery need to be as you said but 4 cell not 3

Now the motors are 2- 3 cell LiPo, 250 racing quads are 4 cell, I can’t find any 4 cell high Kv motors, and, on a 3 cell, I don’t know if your get a good flight time 1300-1500 mAh 4 cell is the norm. On a 3 cell I think you’re going to need 4000 mAh. Now the problem is it’s heavy and bigger then the 1500 mAh 4 cell , I can’t find any motors with high Kv for a 4 cell set up on this site.

and

OK, motors KV = rpm per volt so 650 Kv on a 11 V = 7150 RPM the same motor 650 Kv on a 4 cell 14 V=9100.
Now racing 250 quads scream with the rpm they pull so you need lots of battery also as your running the FPV off the same battery too.
So a 3 cell would need to be 4000-5000 mAh this will not fit in a 250 frame , so then we reduce size ( in your hand size) to a 4 cell short and dumpy 1500 mAh this will be man enough to power quad and FPV, not add to much weight , 250 don’t fly long but 3 cell will not give you enough power for a normal race

A note about the ESCs

Also the 12 amp ESC are a little low… 20 amp minimum is required, really, if your going to run 6 inch props

I ended up purchasing a cheap Chinese “Zip Power” 3S 1500 mAh 40C battery, with XT60 connector for £8.99, because of the higher C rating, and lower cost.

Zip Power LiPo 3S 1500 mAh 40C
Zip Power LiPo 3S 1500 mAh 40C

Although I was sorely tempted by a Turnigy 3S 1500 mAh 20C, with RC3 connector for £9.97, from the UK

Turnigy LiPo 1500 mAh 20C
Turnigy LiPo 3S 1500 mAh 20C

Things ground to a halt for a month, while I got engrossed in researching 3D printing. When  I returned my attention to the quadcopters once more, I tried to remember why I had lost interest. It was due to confusion regarding the powering of the flight controller (a KK mini), as well as powering/wiring the ESCs.

When I had tried to wire up the KK mini, the issue was that the KKmini should draw its power from the servo controller cables from the ESCs, as it has no actual separate power pins. However, it can only do that if the ESCs have a BEC. I received an email from the former owner, confirming the same:

I believe the KK2’s designer intended for the flight controller to be powered by motor 1’s ESC/BEC to prevent all of the other BEC’s outputs ‘fighting’ each other, possibly causing them to fail prematurely thus taking out the flight controller and Rx in the process.

So the real question is what ESC’s are you using? If they have a built in BEC then you can use that to power the KK2. If not and you are using Opto ESC’s with only signal and ground, how about plugging your external BEC into a spare PWM channel on you receiver?

Also, as a heads up, if you plan to use CPPM, the input pin is the AIR pin, not the THR pin for some reason.

Also don’t power the board directly from the LiPo, I think these comments you have read may refer to the battery monitoring feature only. There is a potential divider connected to the red pin that converts the battery voltage to a processor friendly level (Apologies if some of my comments are a little simplistic). Connecting a LiPo to the +5V rail will most likely kill the FC immediately.

Unfortunately, my FlyColor 12A ESCs do not have BECs:

Specifications:
Continuous Current 12A
Burst Current 18A(10S)
Battery Cells 2S-3S Lipo
BEC NO
Size 37*18*5mm
Weight 9g

Neither does the basic PDB that I had wired up above. So, I needed a separate BEC. Luckily the PDBs that had come bundled with my 250 racing frames included BECs, so I took a Diatone v 2.1, and after watching this amusing teardown video about it

I, then, soldered on 2 mm gold bullets (female) and four single (white) header pins for just the ESC servo signal lead. I also added four inline (white) header pins for the subsequent ESC servo control connections to the KK Mini and two red header pins for the power (5V and GND). Additedly, I should have used red for the 5 V and black for the GND, but structurally, two header pins joined by the inline plastic seemed stronger.

The KK mini was not the only flight controller that I had to choose from – there was also the Naze32 that required soldering onto the RCTech flight controller board, along with the Micro MinimOSD. I was awaiting some right angles single row header pins for this purpose. Insufficient right angled pins had been supplied with the MicroMInimOSD and two more were required for the power (GND and +5 Vin/Vcc) solder pads.

[photo of soldered KK mini]

I obtained some from eBay, but annoyingly they were no a good fit. The right angle pins that I received were bent in the middle (with equidistant legs either side of the bend), but the ones that I require (and that came, in insufficient numbers, with the Micro MinimOSD) are longer on the downward header pin side, and thus, crucially shorter by just a few milimeters, on the solder pad side

[photo of right angled headers – required and obtained]

I also decided against soldering the Naze32 directly to the RCTech PDB and instead opted for some 3 x 2 female header sockets for the PDB, for the 6 x 3 input ESC servo control matrix (V, GND and signal) and the Telemetry, Buzzer and voltage monitor pins.

[photo of Naze pins and sockets]

As an aside, the only thing that I was missing now was the XT60 plugs. It was incredible how frustrating it was trying to win an auction for 5 pairs of XT60 plugs for under £1. I knew that it could be done, as I had seen them in the Completed Items listings. I had for weeks been lining up endless auctions ready to be sniped at £0.97 but all of the auctions ended up at £0.96 or over, days before the end.

Finally, however, an auction went through at £0.97 for five pairs of XT60s

I purchased a clone of a Boscam 48 channel 5.8 GHz video receiver for £16.30

I realised that the 2.5 mm gold bullets can not be soldered directly on to the PDB as there would be not enough ceiling space (between the top of the bullet and the bottom of the second chassis board, even with the 20 mm spacer, to actually plug the ESCs in, so I removed the direct soldered bullets and replaced them with bullets on wires, that were then soldered on to the ESC solder pads.

[photo of wired bullets]

I also soldered on two wires, using multi strand (not single core) mains cable, for the XT60 plug. This pair was soldered onto the rear solder pad for the power, rather than the front.

I also added a bare (with no plastic stopper) right angle header pin to the positive solder pad, for the battery monitoring lead, which goes to the KK Mini. See KK-Mini Voltage Pin.

[photo of power and header]

The ESCs were put inside the 20 mm spacer, and looked somewhat cramped together, and hopefully there were not be any overheating issues. If so, then the ESCs may require moving to the arms, with only wiring within the spacer area. The second plate was fitted.

[photo of fitted second plate]

The KK Mini was connected to the M1-4 ESC control leads, and the +5V and GND lines. The buzzer was attached to the KK Mini. The Battery Monitor lead was attached, and the CPPM, +5V and GND pins on the single row, on the left, were connected to output #1 of the D4R-II receiver. See KK Mini.

The receiver has pins 2 and 3 jumpered together to enable PPM, like so:

D4R-II enabled for PPM output
D4R-II enabled for PPM output

A CMOS camera was added, to the front mounting plate.

I trimmed the power leads for the battery so that they were flush with the port in the 20 mm spacer for the XT60 connector.

I soldered on the XT60 female connector.

I then charged the 3S  LiPo, and plugged it it. I noticed that the regulator was getting rather warm on the PDB, and the KK Mini did not come on. The microcontroller on the KK Mini also got very hot, too hot to touch in fact.

I then realised that I had connected the power supply (+5V and GND lines) incorrectly and that the polarity had been reversed. Again, see KK Mini, and refer to the sections on reverse polarity of the power supply. The KK Mini was, most likely, fried.

When I connected the power supply correctly, the LCD backlight illuminated, but there was no information on the LCD. The MCU was in fact dead.

I was fortunate that the RC receiver, the D4R-II, that had been connected at the time of the reverse polarity was not broken, even though it had been supplied with a -1.54/-1.83 V supply, via the receiver connections on the left hand side of the KK Mini.

I managed to verify that it was still operational, by pairing it up with the XJT module in my Turnigy 9XR Pro transmitter, following this video:

Moving on with CC3D

As this section has got rather large, it was moved to a separate page. See the blog CC3D

Setup the Naze32

As this section has got rather large, it was moved to a separate page. See the blog Naze32.

D4R-II and Telemetry

From FrSky Telemetry

D4R-II and Telemetry
D4R-II and RS232 Telemetry Interface

Overly Discharged LiPo

Whilst running through the transmitter configuration and spinning up the motors a few times, I had an issue was my LiPo discharging overly. One cell was reduced to around 1.20 V while the other two were at 2.4V. The LiPo refused to charge, or rather the charger refused to charge the LiPo with the SkyRC B6 Mini giving the “Cell Error Invalid Voltage”. I tried using the NiMh/NiCd charge methods as outlined in the Instructables link. But they gave similar error messages. Then I followed the video, Solution■Charging Low voltage LiPo using iMAX B6 HD, but still no joy.

Disgruntled, I left it for around 14 hours and then after watching, How to recharge a fully flat lipo (lithium polymer) battery,  I tried again, but I got an error stating that the Balance connector was not connected. I plugged in the balance connector, expecting to see the Cell Error message, but, miraculously, the LiPo cells had re-adjusted themselves to around 3.1 V each, give or take, and the charger no longer gave the Cell Error message. This apparently is normal behaviour, see Is my LiPo pack bad?, which implies that the cells self-balance over a period of time:

Batteries that have sat loose balance….

I, then, slowly did a balance charge at 0.1A. After 175 minutes the voltages plateaued at 3.80, 3.80 and 3.79. Shortly after, I got the “Over time limit” message and the charging stopped at 33% capacity.

I started another balanced charge, upping the charge current to 0.5 A. After a little over an hour, the charging was complete.

Frame Orientation

I had the camera at the wrong end, the correct orientation of the camera, AND the anti vibration plate, is shown in the photo below, which is taken from CHAPTER 1 – ASSEMBLY OF THE ZMR250 V2 FRAME

Correct orientation of frame parts
Correct orientation of frame parts

I also had an issue with the battery, where the XT60 lead is not long enough for the battery to sit on the top and reach the rear XT60 port (in the 3D printed spacer). The solution seems to be to slide it in between the middle and top plates, although it makes it somewhat rear heavy and unbalanced. It will still need to be secured.

Somewhat annoyingly, the snuggest fit was at the front, under the anti vibration plate, as bottom part of the rubber bungs, held the battery in place, by themselves – but obviously, firstly the battery port in the 3D printed spacer was at the other end (although I could have turned the spacer around such that the XT60 port is at the front, although this would upset the venting of air through the spacer), and secondly, that would leave no space for the camera plate. A solution could be to re-orient the top plate, such that while the camera remains at the front, the anti vibration plate is at the rear.

Propeller guards

As my flying was rather inept, I thought that some propeller guards might come in handy, so I found a 3D printable set: ZMR250 prop guard normal and slim for 5×3 prop by Lentech.

There is also the ZMR250 Prop guard Shroud by ppoinha,

ZMR250 Prop guard Shroud
ZMR250 Prop guard Shroud

Spare props

I ended up crashing into a wall and breaking a propeller:


I obviously needed some spares. The cheapest set, is this eBay item for 10 pairs, £2.97 for black and £3.30 for green, 10 Pairs 5040 CW CCW Propellers Props For RC Quadcopter Multi Copter Green F7. Note that shaft adapters are not included with these propellers.

I was unsure as to how strong these replacements would be, as the previous props had really taken some bashing and were more of less intact. So I contacted the seller of the ESC/motor/Prop combo that I purchased it from on eBay and he pointed me towards these 4 PAIRS Unbreakable CW & CCW Propellers Mini Props Blades 250 280 for Quadcopter. However, they were out of stock of the 5040 and I ended up erroneously purchasing 6040 props (in white, orange and green) which were too large for the frame. So I contacted the seller and they offered me a discount of £1 on my next purchase.

Fitting the prop

The prop was a CCW one, so to remove, you turn the retaining nut also CCW.

The prop, from the second batch was an exact match.

Basically, you should hold the nut and the motor turning, should tighten the nut automatically. You untighten in the direction that the motor turns. In summary, if:

  • The motor is CCW you tighten CW, and untighten CCW;
  • The motor is CW you tighten CCW, and untighten CW

Also:

  • Silver nuts – CCW
  • Black nuts – CW

Flying again, after six months

After a six month hiatus, the quad wouldn’t work, so I had to open it up and discovered that one of the power leads from the XT60 connector to the PDB had become detached. I don’t know why that happened – poor soldering?

I also had a problem remembering how to re-arm it. I following the binding video, and re-bound the receiver and XJT transmitter. After re-reading OpenPilot – first flight,  I remembered that it was settable in OpenPilot, but I couldn’t remember which combination I had chosen. I played with the arms for a bit and discovered that I had chosen:

  • Left lever – down
  • Right lever – SW.

Also the flight controller needs to be properly seating flat and in the correct direction, not just hanging off some wires/cables/leads.

I tightened everything down, so as to avoid losing bolts in the grass and nyloc nuts in to the chassis, with the risk of a short circuit. Then I started flying again, on concrete. I moved it on to some long grass, and the grass was impeding the props, so I moved back to the concrete and shortly after one of the props stopped turning. Was this due to an ESC having overheated? I left it for a while, to see. There was also an intermittent beeping coming from the FC. Turns out that the motor control wires on the FC had worked themselves loose.

Jerky flight

Also, from time to time, I saw very jerky, wobbly flight, especially in the wind (I think).

This started happening after a crash. Sometimes the front left motor (only that one) will stop momentarily and, depending on the height of the mod at the time, the mod will either crash, or dip down, until the moor comes back on and regains height. Seems to happen on hot days (I think).

It could be related to the antenna damage.

Cut antenna

The antenna from the D4R-II receiver gets damaged by either:

  • being hit by the prop, or
  • being “guillotined”, for want of a better word, by the sharp edge of the bottom plate/PDB and the ground.

It has not been cut but the plastic shrouding, and the co-axial shielding is cut through and the wire seems a bit scored and folded at the cut

Antenna Damage 1
Antenna Damage 1

and close up

Antenna Damage 2
Antenna Damage 2

To fix, I thought of using insulation tape or heat shrink, for the moment, but I was concerned as to how to replace the antenna should it get trashed, as there is some epoxy over the soldered connection of the end of the antenna to the D4R-II PCB. The epoxy would need to be gently removed and fix a new antenna: 1pc 2.4G Antenna Replacement For FUTABA/FRSKY/JR Receiver Antenna 150mm Black TD, $0.99. The antenna is attached to the receiver via a uFl connector.

Meanwhile, a makeshift repair to the remain length is shown here: How to repair a broken receiver antenna. The length of the exposed antenna is related to the quarter wavelength of the 2.4GHz signal.  Using the formula,

\lambda ={\frac {v}{f}}\,\,,

λ is wavelength (m), ν = the speed of light (3×108 ms-1), and ƒ is frequency (2.4 x 106 Hz), which gives

wavelength = 300/frequency

and

quarter wavelength = 75/frequency

Using mid-band frequency of 2.44 GHz, the wavelength is 122.95 mm and the quarter wavelength is 30.737 mm.

Cut off the plastic out insulation, to reveal the braided insulator. Remove braid and expose the wire, and verify the length

Improving on the technique, More precisely repair a broken receiver antenna, 31mm insulation needs to be paired off, to remove the correct length of wire for the wavelength of the signal. It maybe be better to pair back 33 mm, or more than the 31 mm, and then snip the end to length remove the excess exposed wire, then bringing the length of the exposed wire down to 31 mm.

Note this message from FrSky:

Hi, i contacted Frsky because original X4r Antennas are 33mm long, Which is not in the calculated spectrum for Frsky 2.4GHz. This is their answer: Dear Friend, Thanks contact FrSky, yes, generally considering the antenna length is 1/4 of the wavelength, but we consider other circuit losses and gains, and make multiple comparison tests, then the length 33 mm is more appropriate. Best regardsRay

Turnigy Design flaws

The Turnigy 9XR Pro LiPo has some “issues” with the balance charge power connector used to power the transmitter, when using LiPo/LiFe power packs:

  • The balance charge power connector is too close to the top of the battery compartment – the balance plug is hard to take out, you almost have to pry it out, with your fingernail, first teasing it out one side, and then the other, and back and forth, until it is loose enough to pull straight out. This is made more difficult by a plastic guide tab, on the immediate left hand side of the balance connector, which means you can only get your fingers around the right hand side of the balance plug when pulling it out – you are almost forced to pull on the leads themselves. Recommend using hot glue on the leads and connector join. However, this cause cause more problems with the battery door;
  • The balance charge power connector is way too close, in height, to the battery compartment door, you have to push down on the balance cable wires, in order to be able to clip the battery cover in place, when sliding the battery compartment door closed, and once the plug is in the socket, the leads are crushed by the battery compartment door

Photo of balance plugged in and the black tab

Black tab impedes connector removal
Black tab impedes connector removal

Photo of bent balance wires

Balance cable interfers with edge of battery compartment door
Balance cable interfers with edge of battery compartment door

Interesting video guide of the Turnigy 9XR Pro

From Mark Harrison, comes these great video series, starting with:

 

Advertisements

7 thoughts on “Drone kit -Building a ZMR 250”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s