*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
XRAD'S Elefant
« on: June 26, 2018, 04:48:24 PM »
Hello All,
Been working on this 1/16 predominantly scratch build for a while.  Built many arduino IDE based trinket and other microprocessor projects, including writing code for some basic sound and light control, both button press and RC for laser blasters, glowing crystals, live steam, and many scale models (if interested: https://www.youtube.com/user/goldenarch3/videos )  Also, created many 3D printed parts for these. This is my first build with the OP TCB. So far, all the basic functions work, cannon, barrel pan and elevation, cannon flash (HL), and MG1. I am using a RadioLink 12 channel with the 12 channel sbus Rx.  I plan to use the Benedini sound card as it, IMHO, offers the most realistic engine throttle sound algorithm.

The OPConfig program worked like a champ. I did have to update the board firmware twice but after that, no issues.  Waiting on two ESc's for the motors. 

Thank you OP for creating these boards. I know it is MUCH work and time.

Some pics... more to follow.
IMG_5149.JPG
XRAD'S Elefant IMG_5149.JPG
Views: 186
IMG_5199.JPG
XRAD'S Elefant IMG_5199.JPG
Views: 178
IMG_2423.JPG
XRAD'S Elefant IMG_2423.JPG
Views: 183

*

Offline LukeZ

  • 564
    • View Profile
  • Kansas, USA
Re: XRAD'S Elefant
« Reply #1 on: June 27, 2018, 05:01:01 PM »
Hi Rad, you certainly do have a lot of interesting projects on your YouTube, you must stay busy. The Elephant is looking good and I'm glad you like the TCB so far. Thanks for mentioning your radio setup, I don't think the R12DS receiver was available when I first wrote the Wiki so now I've been able to update it. Looks like RadioLink now has telemetry options even, seems like a good radio.

Keep us posted.
NO SUPPORT THROUGH PM - read why
Open Panzer FAQs

*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
Re: XRAD'S Elefant
« Reply #2 on: June 29, 2018, 01:23:43 PM »
Thx Luke!  I have many more builds under XRAD on rctanwarfare.com and modelboatmayhem.com.    Yes, the radiolink Tx/Rx seems like a quality unit for the price. Easy navigation and set-up.

Back to the TCB.  Only 2 issues:

1)
Had some issues 'connecting' to the TCB with the recommended FTDI VCP driver and set-up. The baud rate, com port and driver are all correct. In my case, I used com4, but tested on com3 as well. I have used two different windows computers, one w/12GB ram and AMD 3.3GHz processor win10 with all updates. Same issue both computers. 

I am able to snoop and update firmware no issue. however, opconfig failing to 'connect.'  So I tried snooping several times and then connecting, and connecting is still hit or miss but better after a snoop. I tried updating the firmware(newest version, easily loads 100% each time), and then connecting, and this usually works. But I don't want to have to update firmware each time.  In snoop mode, i can see all switch functions (gun fire, engine on, MG, smoker) working fine. I can see also forward throttle but not reverse left,right, or turret turn -gun elevation.  USB com4 works fine and detects the TCB. OPConfig just not making easy connection. once connected, I can download and upload to TCB no issue and connection is pretty solid (does not drop).

2)
Second issue is that I have uploaded my config settings (after setting stick centers and limits on channel 1,2,3,4) but the driving stick functions are a bit wonky. Turret and gun elevation work fine.  but with driving, stick straight back is not reverse, but rather back and slightly left.   forward is up and slightly right. Weak right turn, strong left turn. I changed up motors/esc's and tried 're-centering/limits' on drive channels, no change. maybe the radiolink Rx puts out a bit different sbus sequence for these channels than expected.   Using hobbywing 40amp esc's(with 5v red wire removed).

Anyone w/thoughts or ideas?

*

Offline LukeZ

  • 564
    • View Profile
  • Kansas, USA
Re: XRAD'S Elefant
« Reply #3 on: June 29, 2018, 08:12:39 PM »
Hi XRAD,

1)
Had some issues 'connecting' to the TCB with the recommended FTDI VCP driver and set-up. I am able to snoop and update firmware no issue. however, opconfig failing to 'connect.'  So I tried snooping several times and then connecting, and connecting is still hit or miss but better after a snoop.

As you are probably aware several users have reported a similar problem, though I have not been able to replicate it on my end, and it doesn't affect all users. There is a note about this in the Troubleshooting guide. The solution as you have found is to first open a Snoop connection, and only after click the Connect button. This should always work, but it is important after you open Snoop to wait a few seconds for the TCB to finish spitting out whatever it is going to spit out (either "waiting for radio" or if the radio is found it will dump a bunch of settings). After that is done then click Connect.

You also mentioned in another thread that your TCB came with a faulty dipswitch #5. This switch is directly related to the TCB's ability to connect to your computer. So I don't know in what way you determined it was broken or what you might have done to fix it, but if it is at all intermittent or loose or whatever, that is also going to cause you lots of problems.

I tried updating the firmware(newest version, easily loads 100% each time), and then connecting, and this usually works. But I don't want to have to update firmware each time.
I agree, this is a lot of extra work and indeed there are times when you specifically do not want to be updating firmware repeatedly over and over (some updates will overwrite prior settings) so I don't recommend this approach. Just use the Snoop first, then Connect trick.

In snoop mode, i can see all switch functions (gun fire, engine on, MG, smoker) working fine. I can see also forward throttle but not reverse left,right, or turret turn -gun elevation.
You definitely should see both forward and reverse messages, so if reverse isn't showing somethings is wrong with your radio setup. You can also use the red and green LEDs on the TCB to confirm forward and reverse movement, see the LED key here. Left and right turns, as well as turret movements are not presently listed by Snoop so that is normal.

2)
Second issue is that I have uploaded my config settings (after setting stick centers and limits on channel 1,2,3,4) but the driving stick functions are a bit wonky. Turret and gun elevation work fine.  but with driving, stick straight back is not reverse, but rather back and slightly left.   forward is up and slightly right. Weak right turn, strong left turn. I changed up motors/esc's and tried 're-centering/limits' on drive channels, no change. maybe the radiolink Rx puts out a bit different sbus sequence for these channels than expected.   Using hobbywing 40amp esc's(with 5v red wire removed).
I don't know what the issue is here, but it is not SBus. SBus is a standard and by definition it is implemented the same by all, but even if there are variations that is what saving the stick centers and end-points is for, it would adjust for those. If that process has been performed correctly then the TCB truly does know the center and end-points of your channels. The fact that you weren't seeing reverse in your earlier comment makes me wonder if that process completed successfully.

What you want to make certain of is that you have no mixing, dual rates or other modifications defined for those channels in your Transmitter.

The other thing to check is whether you have excessive mechanical resistance in one of your tracks, possibly in the left track since it wants to turn left. If you have a servo Y-cable, eliminate the TCB for a moment and hook up both your ESCs to the forward/reverse channel directly on your RC receiver. You won't be able to turn, but you should be able to test driving forwards and backwards. Since both ESCs are receiving the exact same signal, if your tank keeps turning left that would indicate you have something hanging up your left track. You can also try swapping the left and right ESCs to see if that makes a difference (it should not).

If you don't find any problem there go back and re-visit your Transmitter settings and re-do Radio Setup to make sure everything really has been saved correctly. While streaming the radio, watch the values for the throttle and turn stick channels. Do they fluctuate a lot even when the sticks are still? If so you may want to increase the Deadband to a value higher than the highest static discrepancy.

Also try disabling Neutral Turns temporarily (on the Driving Tab) and see if that makes any difference.
NO SUPPORT THROUGH PM - read why
Open Panzer FAQs

*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
Re: XRAD'S Elefant
« Reply #4 on: June 30, 2018, 08:10:12 AM »
Thx for reply Luke,

1) Fixed the connectivity issue. I purchased a second TCB board. It connects without issue.

regarding asymmetric drive issue:

Tried TCB different turn modes and unchecking of zero turn. Currently it's in mode 1, one side driving only on turns. Seems to work the best and is what I originally used.

 Tx: tried mode heli, quad, and acro (plane) and various modes withing each vehicle mode.  No channel mixing or linked channels. All mixing nulled or off. All unused channels nulled. tried changing left and right stick for driving. Tried second TCB board with same ELEFANT file from first board.  All radiolink Tx channels centered and with full travel according to the radio output and my servo testing.   Still the same issue.

Tried a second radiolink Rx (which works fine on all servo channels in standard PWM mode). same issue with drive asymmetry. I do not have another sbus or ppm Rx/Tx to try.

Side note: I did change up ESC's and drive motors, right side still weaker in Sbus.  When ESC's running off Rx in PWM mode, drive strength is symmetric.

Unfortunately, at this point I accidentally deleted my radio file and will have to reprogram the Tx.  I think your idea of not recognizing end points (and therefore the signal confusion) is valid (even though the limits were all the same on all four stick movements in the OPConfig set-up, and all saved) and will be my next step once radio reprogrammed. I will look for reverse in snoop. Even though sbus is supposed to be a uniform code, it may be possible that the conversion to sbus by the Rx is somehow off due to the suboptimal signal recognition. My other option is to try my RMILEC (which defeats the purpose of sbus), but I can't find it! to many projects...

These issues are not drawbacks. I like the challenge. Easier than trying to shave off a few bytes of code to get a program to fit on a trinket......

 


« Last Edit: July 02, 2018, 07:50:46 AM by XRAD »

*

Offline LukeZ

  • 564
    • View Profile
  • Kansas, USA
Re: XRAD'S Elefant
« Reply #5 on: July 03, 2018, 06:30:10 PM »
Side note: I did change up ESC's and drive motors, right side still weaker in Sbus.  When ESC's running off Rx in PWM mode, drive strength is symmetric.

I think this is a crucial piece of information. For one, it would seem to rule out mechanical drag, so we can ignore that. Theoretically it should also rule out any inadvertent mixing or weird settings in the transmitter, since those would affect PWM and SBus equally (they are just a different means of communicating the same signal from the transmitter).

I can say that inside of the TCB firmware, if you give forward throttle and no turn command, the precise same PWM signal is sent to both the left and right motors. The fact that your tank seems to want to turn implies this is not happening, and we can only conclude some kind of turn signal is being introduced even when you don't intend for it to be.

The thing I would usually look to first of all is radio setup. If radio setup didn't get saved correctly for some reason, and the TCB thinks the center point of your turn channel is not actually where it is, that would certainly cause this behavior.

The other thing you can do while reading the radio signal in OP Config is to watch the turn channel signal and see if it is jittery or drifting. Being SBus it should not be, in fact it should be fairly well locked right near 1500. Also while reading the radio signal, you can try advancing your throttle stick without turning to the left or right, and watch the screen in OP Config to see if the turn signal shows any deviation. If so that would tend to imply a mix in the transmitter. 

Basically some more elimination needs to happen until we can get to the bottom of it, but something unusual is definitely going on. Given your experience in RC and electronics I am certain you will pinpoint it eventually.
NO SUPPORT THROUGH PM - read why
Open Panzer FAQs

*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
Re: XRAD'S Elefant
« Reply #6 on: July 07, 2018, 08:59:00 PM »
Hi Luke, Thx for your help! 

Away for a bit while designing all the 3D parts for the Elefant.  If anyone wants them they are at:

https://www.thingiverse.com/thing:2993799

Yes, more elimination occurred.  Don't know how I missed it but one of the ESC's was set to F-B-R instead of F-R.  Once that was corrected, I spliced in my RMILEC SBUS decoder ring and tested it in parallel with the TCB.  So I get all TCB functions (lights, MG, Cannon, gun elevation-turret rotation) working fine. And, running the motors SBUS decoded back to PWM, I have full forward and reverse to the drive motors using the RMILEC decoder. All linked functions are off in the transmitter as far as I can tell. Full stick movement = full drive motor power, channel 1 and 2, right stick only, full independent motor function. No mixing. 

However, this still does not explain the 'clocking' of stick motion that occurred with the drive motor radio stick on initial TCB use (two different TCB's).  Like you say, I was expecting both motors to turn when drive stick is in full straight up 12 position, not when stick at 1:30 position for F, and the 7:30 position for R. And why I did not see reverse on snoop...

So, now I know that the SBUS signal hitting the RMILEC is correct from the Radiolink Rx, and therefore from the Radiolink Tx.  Next will be to drive the motors through the TCB....  I will reset the TCB Rx channel travel and functions to get back to baseline , and then monitor the TCB radio inputs like you say. 

'Moving' in the right direction...... :)

IMG_2446.JPG
XRAD'S Elefant IMG_2446.JPG
Views: 132

*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
Re: XRAD'S Elefant
« Reply #7 on: July 08, 2018, 01:30:56 PM »
Had some time today to fart around with the OPConfig.  So...had some trouble again connecting with the new board.  Checked my physical USB connection, cables,   baud rate...com port, all device settings and error hx....etc..etc..etc..  then in middle of this search while in the OPConfig firmware section, OPConfig found an update and I installed it. Not an update to the TCB firmware, but an update to OPConfig...

I was not looking for it but OK, so I installed it.  I do not recall the version number offhand, but the version I was using before the update was just downloaded off OP page 2 weeks ago or so. I was able to connect to the TCB after the update after a snoop on first try.  Then, when I opened my saved file for the Elefant to the 'new' OPConfig, it only loaded 219 out of 220 settings.  OK strange. I ignored the warning as it looked like everything I wanted was there and unchanged.  I made no other changes to the TCB,  to my original file, or on the radiolink Tx. I uploaded this new 219 setting file to the TCB, and what happened??

I now have FORWARD and REVERSE in snoop!  I knew that was a good sign. Drive motors now working both forward w/drive stick 12 o'clock and reverse w/drive stick at 6 o'clock.  Everything now works like it should!  :D 

OK, so root cause is still unknown. I know I did not fix anything. 

And for completeness:

1) I am using my original OPconfig settings recorded in OPConfig earlier version, via a previously saved Elefant file on the laptop. I did not re-center or re-record stick movements again. No other changes made to this file.  My guess is that when writing to the TCB, an error occurred in the file or during file transfer to the TCB, which is not reproduced when saving the file from OPConfig to the laptop or loading this file to OPConfig from the laptop. 

2) I put the same ESC back into F-B-R mode.  The only difference was that I had to hit the stick twice  to get both tracks to go into reverse at same time(which is expected as one track was in the F-B-R mode) BUT they both moved fine once brake was deactivated by second stick movement, and turning, forward and reverse were all 'clocked' appropriately on the Radiolink transmitter drive stick position.   

3) one other thing I recall from last week is that even though I reversed several servo switch positions via the OPConfing check boxes, they did not reverse once downloaded to tcb.  This included more than one aux switch and the throttle stick position.  I had to reverse them via the Tx.  Today, all switches are oriented correctly.


video with esc's both set F/R:


forward reverse snoop.jpg
XRAD'S Elefant forward reverse snoop.jpg
Views: 106
« Last Edit: July 08, 2018, 02:51:31 PM by XRAD »

*

Offline LukeZ

  • 564
    • View Profile
  • Kansas, USA
Re: XRAD'S Elefant
« Reply #8 on: July 08, 2018, 04:23:51 PM »
Good news, sounds like you've gotten it all straightened out. It seems to me the fact that one ESC was in F-R mode and the other in F-B-R explains entirely the issue you had with the vehicle not driving straight forward and back. For reference, it is better to put them both in F-R mode, otherwise known as crawler mode. Let the TCB handle the braking, which it can do much better than the speed controllers because there are two of them and they don't know what the other is thinking. If you still prefer double-tap reverse you can accomplish that with the TCB as well by setting a "Change Direction Delay" on the Motors tab of OP Config.

The message about not loading all the settings (319 instead of 320 for example) is completely normal. One of the reasons for a firmware update is to add new variables that didn't exist before. Obviously if you import an OPZ file that was saved prior to the update, that OPZ file isn't going to know anything about the new variable that was just added. OP Config will still import all the existing variables from the OPZ file and any that are missing (because they are new) will just be set to default values.

As for whatever other issues you had, since they are gone now there is no point in worrying about them unless they return. You tried a whole bunch of different things so it's kind of hard to identify what was related to what, if it's working then it's working and whatever the issue was has been resolved.
NO SUPPORT THROUGH PM - read why
Open Panzer FAQs

*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
Re: XRAD'S Elefant
« Reply #9 on: July 09, 2018, 04:02:39 PM »
Did a bit of modding today. Received a TARR III smoker for RC tanks, etc. Consists of a few diodes and a 7805 or similar transistor , a DC 5v fan, a cast metal box, and a plug to go into Heng Long smoke control RX18 or similar.

But the smoke output was minimal. Not enough current to the heating resistor. However, 3 amps supplied by the TCB should be enough to get a small coil hot. So I tested a spare Burn Effect V3 coil. Sure enough, got hot enough to make much smoke. The power transistor on the TCB did not get hot, but the power wires did warm up a bit.

So next plan, mod the system.  Using throttle forward ESC output 1 on the TCB to burn effect esc control module. Then to the heating element. The fan switches on and off via the TCB 'Smoke' plug. The fan speed is controlled by the 7805.  3D printed a mount for the heating coil and put her all back together. Works like a champ.  Fairly decent proportional control of the 5V fan. However, the coil only heats up in forward.  No big deal as it makes much smoke.

dgm. is of the control board for fan speed.  NOTE: I had to use the positive out on channel 1 off the TCB to power the burn effect smoker board. It does not run off the esc. That means middle red wire is connected at the servo out on the TCB. However, on the other end, where the signal passes through the burn effect board, the ESC red wire is obviously disconnected. see pic of messy wiring.

Vid:

psu.jpg
XRAD'S Elefant psu.jpg
Views: 36
IMG_2458.JPG
XRAD'S Elefant IMG_2458.JPG
Views: 43
IMG_2460.JPG
XRAD'S Elefant IMG_2460.JPG
Views: 36

*

Offline johnnyvd

  • 137
    • View Profile
  • Netherlands
Re: XRAD'S Elefant
« Reply #10 on: July 09, 2018, 04:09:11 PM »
3D printed a mount for the heating coil and put her all back together. Works like a champ. 

Nice work. Did you use ABS for the 3D printed support by the way? I can imagine PLA not be too happy with the heat?
* E-75 / E-100 PAK44 "monster" - in progress
* Sturmjagdtiger PAK44 - in progress
* pz.kpfw KV-2 754(r) - in progress
* T-34 88mm "Kurland Tiger" - in progress

*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
Re: XRAD'S Elefant
« Reply #11 on: July 09, 2018, 04:18:25 PM »
ABS. The heating element is fairly well isolated of the mount and the red aluminum piece does not get hot. Just have to make sure the heating element does not touch the metal box. Also, to make more smoke and to more uniformly disperse heat from the element, I improved the glass fiber wick wrapping. this distributes more fluid to the heating surface. Also, I drilled two more fan inlet holes.

I plan to find another source for 5v to the burn effect board. Running it off the servo out on TCB is temporary.

*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
Re: XRAD'S Elefant
« Reply #12 on: July 09, 2018, 06:11:38 PM »
Power to smoker problem solved. using a 3906 transistor and 18k resistor, I tapped into the power from the TCB at the smoker (currently 7.4v) and using the smoker on-off ground, I am able to power the burn effect board on and off , at same time as the fan, using a single radio Tx switch. I thought that I had to decrease voltage to the burn effect board to 5-6v to make all components happy , but I did not.  And I now at 7.4v, I have smoke in reverse. I am guessing that the smoker control board needs at least 6v to work well.

The reason I could not use one of the other TCB switchable power ports is that there is that there is no ground 'out' on the burn effect board, so I could not close the 'load' loop. I could tap into the 5v supplied throughout the TCB ports for 5v  if needed.  And I could configure say a 'secondary' light port to activate ground at same time and with same trigger source and same trigger action. But I am still left with only a 'ground' turning on or off, while the smoker board requires 'power' turned on or off. Power remains constant at these spare ports, so I needed to build the 'grounding' trigger with a 3906 transistor, which then supplies power to the burn effect board when ground in activated through really any TCB port I choose. I chose the smoker port because it makes sense.

I did run a test with the second light port on the TCB. Nice regulated 5v. But not enough to run the Burn Effect board in reverse for some reason.  Back to 7.4v off the smoker tap and all is well again.  And yes, I could have used a 5v relay on TCB aux out to power the burn effect board, but what's the fun in that :o......

 :)
IMG_2468.JPG
XRAD'S Elefant IMG_2468.JPG
Views: 40
« Last Edit: July 09, 2018, 07:06:07 PM by XRAD »

*

Offline LukeZ

  • 564
    • View Profile
  • Kansas, USA
Re: XRAD'S Elefant
« Reply #13 on: July 09, 2018, 10:34:38 PM »
Wow, lots of stuff there. Not sure I followed all of it, but a few random comments: the 7805 is a voltage regulator, it takes an input voltage and spits out 5 volts. It usually requires a minimum of 7 volts to work, if you feed it 5 volts that won't be enough for it to turn on (or turn on fully). Presumably the fan is a 5 volt unit and they are using the 7805 to step down the higher voltage that would normally be taken from the motors in the original design. The diodes of course are to convert the forward/reverse movements of the two motors to a constant positive voltage for the smoker, that is the same system that Heng Long uses. But you don't need any of that diode stuff since you aren't driving this from the motors.

One thing you could have done is attach the heating coil to the smoker output, and the fan to the Aux output. The smoker output is full battery voltage, which you'd want for the heating element. The Aux output is 5 volts, which would match the fan. The heating element would be controlled automatically by the smoker function (or if you always wanted it full hot you could just assign the same trigger as you use to turn on/off the engine to the "Smoker - Manual On/Off" functions). For the fan you would then assign the "Engine Speed" trigger to the "Aux Output - Set Level" function, so the fan speed would automatically match the engine speed. No transistors or external circuitry required, and no extra transmitter switch either, the effect would be automatic.

But there are probably a dozen ways to skin the cat, and what you have is fine too.


NO SUPPORT THROUGH PM - read why
Open Panzer FAQs

*

Offline XRAD

  • 31
    • View Profile
  • Ohio, USA
Re: XRAD'S Elefant
« Reply #14 on: July 10, 2018, 06:34:59 AM »
Hi Luke. Yes right, many ways to get the same thing done.  Yes, 5v DC brushless fan.

I did try running both heating coils off the TCB smoker tap. Too little voltage to get the TARR coil smoking (which is apparently what others have exerienced..it measured 15 ohms and others say it needs 10v or so). I tried to run my burn effect coil(measures 3 ohms) off the TCB smoker tap too, but it was too much current, and the coil got very red hot (unless there is a way to limit voltage via OPConfig?). The Heng Long smoker coils, motor noise, and smoke output is terrible, so using one of these was out.

I did try to decrease power from the 'battery' level TCB smoker tap to the burn effect coil, but it did not work as well.  Found out this was a direct battery feed when I accidentally plugged battery power line into the smoker tap. Ooops. Luckily, no damage.

So that is why I had to use the original Burn Effect driver board. yes, more wiring and boards.... But the nice thing about the burn effect board is that there is an easy analog adjustment to set smoker function on/off point and it has a very important 9 sec latch, then times out so that I do not over heat the coil.  I am sure you could add a few of these functions to the TCB smoke config section.

I really wanted to modulate the fan speed via the TCB, but these DC brushless fans need a certain turn on voltage, and then if I am limited to 5v on one of the aux channels, I can't overclock the fan to get a more proportional smoke effect, which I may try by eliminating the TARR voltage regulator and adding a resistor.