Registration Notice

Due to increased spam, forum registration must be manually approved by a moderator! Please see this post for instructions.

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #15 on: January 28, 2020, 12:57:49 PM »
Hi Luke,

Well, good news, and terrible news......

The smoker now works as advertised, thanks so much.

My Sabertooth 2x5 is now a smoker...

Here is the sequence of how I proceeded with the testing of the recoil.

I adjusted both velocity and duration up until I got some recoil.  Had to go to maximum on both settings.  Note that this tank has a sabertooth motor controller which may be an issue with getting a good pulse for the recoil.  So the recoil was still not robust enough, only a short burst, and it looked more like someone was pushing back on the tank rather than a sharp gun report was happening.  I then removed both motor leads from the motor controller, and hooked up the scope to the output of one motor and got the pulse shown in the video:

Then I reconnected the motor leads, and disconnected the scope and USB cable, noting that the USB end that plugged into the TCB was warm, and the ground lead alligator clip to the scope was also warm....hmmmmmm.   Upon starting the tank, the saber tooth controller would not move the motors, and so I cycled the battery switch, with no effect so I reflashed the TCB....all flashed okay.  So tried again, and the tank moved fwd on first try but subsequent tries the red light on the saber tooth went on, and smoke issued from underneath the controller, between the aluminum lower and the circuit board apparently.  I guess something has happened to the drive FETs on the board? 

I must have created a ground loop between the USB, the controller, and the scope?  First time ever this has occurred, as I thought the scope would be isolated from input grounds, and the USB connection also.....So it looks like I'm out a controller for my trouble....ouch!! 

Jerry

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #16 on: January 28, 2020, 03:34:46 PM »
Hi again Luke,

So apparently the probe ground lead is common with the input plug ground pin, ugh....

I removed the back of the sabertooth 2x5a and I see a fried 4838 GA7G11 by Alpha & Omega. 

I've ordered a few AO4838 by A&O, to replace it....hopefully it will take.  The right hand motor still works, so very possible.

Two steps back...drat.  I hope the video above is useful?

Jerry

*

Offline LukeZ

  • 1014
    • View Profile
  • France
Re: Track Recoil fine tuning
« Reply #17 on: January 28, 2020, 04:10:48 PM »
Sorry to hear that, especially as I know the Saberteeths are not cheap. There is a decent chance you will be able to repair it by swapping out the FETs as the Sabertooth processor seems to still be working.

It's not entirely evident to me what happened, but it's true the output of the Sabertooth is running at a different positive voltage (v-batt) than the logic side on the TCB (5 volts). Sharing a ground between all that is not a problem and indeed is a good thing, but somehow a differential must have been created. I'm surprised that what fried was the Sabertooth rather than your scope or something else. The Sabertooth has various onboard self-protection features so it is hard to kill them, and in fact I think this is the first time I've heard of anyone doing so.

I have fried several cheap scopes in my own TCB testing, but they are powered by USB and I have found it is not good to have them plugged into my computer at the same time my TCB is plugged in, for whatever reason. However your scope is presumably not connected to your computer.

As for the signal you saw on your scope - it is quite likely the oscillation is a reflection of the PWM pulse the Sabertooth is using. So I'm not sure it tells us much and I think a scope is not the best debugging tool in this particular instance if you think the recoil is not working correctly.

I can say that in my own testing conducted just yesterday I am using a Sabertooth 2x5 and track recoil appears to work, but my miniature travel tank is an excessively over-geared Tamiya toy so no matter what signal I send it the response is going to be very anemic.

What might be a good test, if you have not done so already, is to connect your Clark to this tank and see what the recoil response looks like with these gearboxes. If I remember correctly they are not the same gearboxes in your other Clark tank which is the one whose recoil action you are hoping to emulate. If your Clark does not give you the crisp response in this tank as it does in the other tank then you can deduce the issue is really the gear ratio for which there is not much that can be done. If however it is still quite a bit more punchy than what you were able to get out of the TCB, we can dig further.

NO SUPPORT THROUGH PM - Read why
Need a forum account? Read here
Open Panzer FAQs

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #18 on: January 28, 2020, 05:18:20 PM »
Hi Luke, Thanks for the quick response.  I have one of those Tamiya gearboxes in my 38T Panda.  Smoothest tank I have, but I had to find the right motor for it since the 3v ones included didn't last too long at 7.2v..... ;D

So in connecting up the Tektronix scope, I essentially grounded one output lead from the Sabertooth, and since the ground from the USB is common with the scope it must have been firing to ground each time I did the recoil.   I also noted the mouse on the computer, also plugged into the USB port, stopped working for a short period ....near miss? 

I will try the relocation of the Clark, maybe it will be anemic, but I don't think so.....Shouldn't take too long to dummy it into the other T55A since all I need is motor connections and battery.  Good idea to prove viability.

Jerry

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #19 on: January 28, 2020, 07:59:25 PM »
Hi Luke,

I just tried the new version on my King Tiger, with 390 motors/sabertooth 2x5a/Taigen steel gears 4:1 ratio, all metal beast.  And noticed:

1) can't get any recoil until 200ms duration. 

2) at max settings it has some recoil, but somewhat erratic.  Sometimes  nothing, sometimes a good hit, but with the new Taigen BB gearboxes, they have now keyed the final drive gears to the axles, instead of a press fit so there is a noticable rotation slop in the axle to final drive.  So with this arrangement, when a good recoil happens, you can hear a snap when the key hits against the gear or shaft.  Not too sure I like this issue.

3) I was observing the Sabertooth Motor Driver too, given my recent experience, and noticed that on recoil the red light would come on at recoil. This concerned me also. 

I will have to look into what the red light means, other than a short, but I expect the locked rotor current on recoil is setting the light off.  Have you any experience with that?

Jerry

*

Offline LukeZ

  • 1014
    • View Profile
  • France
Re: Track Recoil fine tuning
« Reply #20 on: January 29, 2020, 09:14:19 AM »
Well I have just performed a test on my Sabertooth 2x5 and I can get a response even at 5mS (barely visible, of course, but I can tell that it is working). My Sabertooth doesn't exhibit any of the red LED behavior you describe.

Using serial print statements I can confirm that the TCB is sending the signals at the correct intervals, or in other words, at whatever the user has set them to.

The 2x5 manual says the red LED on the Sabertooth indicates either overheating, overcurrent, or depleted battery. Overheating would not be an issue in this instance, but possibly the inrush current of a sudden jump from stop to full throttle could do it. My little model is so small it wouldn't have a problem but possibly on your larger gearboxes it might. That or else you could check your battery, but I suppose that would be too easy.

You could try testing the track recoil with the tank on blocks to reduce current, or just remove the tracks. If it starts to work then we could probably assume an overcurrent condition.
NO SUPPORT THROUGH PM - Read why
Need a forum account? Read here
Open Panzer FAQs

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #21 on: January 30, 2020, 12:23:50 AM »
Hi Luke,

Okay, did some more testing. 

On the KT w/2x5 sabertooth I found I could get the red error LED to light by jamming full throttle up quickly, consistently.  So perhaps the current is causing this?

So, thinking that possibly the current limiter of the 2x5 is getting in the way of the high current of the recoil I changed the sabertooth for a pair of the HK 45A ESCs.  Ahh, now we are talking...... adjusting the recoil settings down, to 100 speed, 30 duration, and I'm getting a sharp recoil, but short duration, perfect.

So perhaps the smaller sabertooth isn't up to dealing with the recoil current on the 390 motors?  However, the T55A had the smaller 380 motors, but still didn't have a sharp, quick recoil, so perhaps the serial motor controller cannot do sharp, quick recoils?  I'll have to get additional HK 45A ESCs for the T55A for testing.... or perhaps I can put in a 2x12 to test with my Pershing... I'll see what I can do tomorrow......

Jerry

 

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #22 on: January 30, 2020, 01:26:09 AM »
Hi again,

So I tested the 2x12 sabertooth with 380 motors on the Pershing with the updated firmware and glad to report that I can achieve a quick, sharp recoil at 100/35 setting.  So in conclusion it looks like the 2x5 sabertooth simply cannot conduct the quick, higher, locked rotor current of the recoil successfully.

I'll order up some ESC's for the T55A and see if that cannot solve the issue.....

Jerry


*

Offline LukeZ

  • 1014
    • View Profile
  • France
Re: Track Recoil fine tuning
« Reply #23 on: January 30, 2020, 10:29:20 AM »
Good work Jerry, another problem solved through persistence and experimentation. I am not surprised that even with the smaller 380 motors the overcurrent issue remained, it is probably not so much a function of the size of the motors but of the amount of inertia they are being forced to overcome so quickly. Needless to say, track recoil is pretty hard on the drivetrain, gearboxes, and the electronics.

The 2x5 work quite well in normal operation but I suppose if we had a sufficient meter (I have a Fluke in the States that could do it), we would see that inrush current when going from full stop to full speed is many multiple times higher than normal operating current. When driving around normally if we were to encounter some obstacle that caused a momentary current spike we probably wouldn't even notice it, but since the track recoil action is so brief anyway, any hesitation or interruption during that small span of time is going to make for a sluggish response. I feel confident, as you have found with the 2x12, that the Sabertooth controllers can easily accommodate very quick changes via serial commands, given they can handle the currents necessary.

I hope you can get your 2x5 repaired and find a suitable model for it!
NO SUPPORT THROUGH PM - Read why
Need a forum account? Read here
Open Panzer FAQs

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #24 on: January 30, 2020, 12:40:05 PM »
Hi Luke,

So I resumed testing this morning by installing, temporarily, the 2x12 into the KT.  The ESCs worked well, at least from a recoil standpoint, but they suffered drivability issues with the heavy, wide tracked, all ball bearing KT.  It would either do wide turns or 90 degree turns on the high traction carpet unlike the sabertooth which would regulate the inner track speed quite nicely. 

So at the same recoil settings of the ESCs, 100/30, I now have some recoil.  Not as sharp or crisp as the ESCs, but at least something.  The driveability is restored  though, and the red error light does come on with recoil, as it did with the 2x5 controller. 

Hmmm, mayhap I'll try the Scout in the T55A to see how the recoil works.  It is not too heavy, and has plastic tracks.  Back in my previous life,....we used to allow  5 to 7 times running current for motor start.  But its hard to capture that initial spike as rotor position, and other mechanical factors affect it, which may be why the recoils are not consistent. 

It would be nice to know what the current limiters in the Sabertooth's are set at.....

jerry

*

Offline LukeZ

  • 1014
    • View Profile
  • France
Re: Track Recoil fine tuning
« Reply #25 on: January 30, 2020, 02:15:55 PM »
I can't be sure about the shutoff limits of the Sabertooth controllers, but the specs do list the peaks at double the rated current (so 10A for the 2x5 and 24A for the 2x12).

The turning problems for heavy wide-tracked tanks is a known issue with the ball bearing Taigen gearboxes which simply don't have enough inherent friction to force the turn. The Scout attempts to overcome this by braking/dragging the inner track, and possibly the Sabertooths do something similar but I am not sure.
NO SUPPORT THROUGH PM - Read why
Need a forum account? Read here
Open Panzer FAQs

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #26 on: January 30, 2020, 02:34:44 PM »
Hi Luke,

I noted the peak current numbers the Dimension Engineering listed in their descriptions, but wasn't sure about spikes.   Maybe the peaks are the spikes, since the error light does come on for the duration of the recoil.  At least that is the way it seems.  Too bad we can't 'tune' that aspect. 

As for the turning, yes, the KV1 chassis seems to do better on the turning with ESCs than the King Tiger.  The KV chassis has narrower tracks, and is lighter with bushing gearboxes, only BB on the drive axles.  The shorter the chassis, the lighter it is, the wider the wheelbase, and the narrower the tracks all apparently contribute to the turning ability.  The KT is maxed on just about all that, and is the worst turner of all my tanks.

So I tested the Scout in my KV1 just now with the new settings.  It doesn't have the current limiting since the tank really responded to the recoil.  I eventually settled with 100/7 for settings since it only has a 75mm gun, and is a massive tank in reality so thought the gun would not show much recoil.  Nice. 

So I guess I now should do ESC's or build that new Scout I've been putting off for so long for my T55A, to get that to a better recoil reaction...which started all this.   :-\

jerry

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #27 on: January 30, 2020, 04:59:20 PM »
Further progress: Tiran4 (IDF T55A) with HK ESCs:

Thanks for all your work Luke!!  Looks like a winner!!

jerry

*

Offline LukeZ

  • 1014
    • View Profile
  • France
Re: Track Recoil fine tuning
« Reply #28 on: January 31, 2020, 01:16:40 PM »
That's a beautiful model!
NO SUPPORT THROUGH PM - Read why
Need a forum account? Read here
Open Panzer FAQs

*

Offline Lotuswins

  • 155
    • View Profile
  • Roseburg, oregon
Re: Track Recoil fine tuning
« Reply #29 on: February 01, 2020, 01:25:54 PM »
Hi Luke,

Thanks for the compliment!!

I was able to repair the 2x5 sabertooth by replacing the left channel AO4838 IC's, two each at 60 cents per. 

I was also updating all my other TCB's with the new firmware and tuning the recoils.  When working with the 2x5 sabertooths, which I have 5 in service now, I found that if I set the duration at a nominal amount, say 20, and then run the speed up until the red error light on the sabertooth just starts to blink on recoil, then back off a little, then run the duration up until I achieve the appropriate recoil for the model, it works better.  Double check the speed setting again after all this, and it doesn't do too badly.  Charge the battery before testing though as this can have an effect on the current the controller can draw. 

Thanks again!! 

jerry

 

bomber-explosion