Fun, Learning, Friendship and Mutual Respect START  HERE


Unregistered
Go Back   HeliFreak > R/C Helicopter Support > Blade Helicopters (eFlite) > 130X


130X Blade 130X Helicopters Information and Help


Reply
 
LinkBack Thread Tools Display Modes
Old 01-06-2014, 06:41 PM   #1 (permalink)
Registered Users
 
Posts: 1,667
 

Join Date: Mar 2013
Default 130x Linear Servos - Demystifying the Smoking Elevator Servo Failure

This thread is prepared as a complement to 130x Linear Servos - The Unauthorized Tech Manual

A frequent 130x servo failure involves a component on the servo board burning up, often glowing red hot in the process. An informal search for 130x servo failure discussion threads leads to an observation that this failure is typically with the elevator servo, and that the servo actuator is found driven to the lower, gear end of the servo during the failure. The part that burns up is normally the one just to the right of where the 130x A-gear passes on the electronic component side of the servo board.



The reader is forewarned that this is a lengthy post. An explanation of the component failure requires a broader description of the servo design. For those readers that don’t want or need the servo circuit description or the technical details of the failure analysis, post #2 of this thread provides a brief summary. Post #3 addresses some obvious questions related to the proposed failure scenario.

The component failure is in the motor drive circuit. The 130x servos use an H-bridge design to provide reversible control over the motor. The H-bridge gets its name from the “H” appearance of four switches and the motor when drawn on a schematic. The motor is driven by turning on diagonal pairs of switches in the H-bridge. The rest of the servo circuit discussion requires an understanding of this. Spend time visualizing the current flowing in the motor for the switch positions shown in the truth table.



Some additional combinations of switch states are worth mentioning. If all four switches are off, or if only any one switch is on, the motor is defined to be in a “coast” state since it is not being driven in any way.

If S1A and S2A are on, or S1B and S2B are on, the motor is declared to be in a “brake” state. DC magnet motors turn into a generator if they are mechanically forced to turn. By turning on either of these pairs of switches, the motor is shorted out and any attempt by the motor to generate voltage will be presented with a heavy load, hindering its ability to turn.

Finally, it should be obvious but simultaneously turning on switches S1A and S1B, or S2A and S2B, would be very, very bad since it would create a short circuit from V+ to ground. Revealing this now should lead some readers to suspect where this thread is headed.

Let’s refine the H-bridge to use bipolar transistors or MOSFET devices instead of the mechanical switches. MOSFETs will be more efficient at higher currents and typically require fewer parts, so we’ll use those. For those familiar with the difference, P-channel devices are used switching from V+, and N-channel devices are used switching to ground in the H-bridge.



Gate signals for the MOSFET switches in each leg can be tied together because of how both P-channel and N-channel devices are used. The pullup resistors R1 and R2 will ensure there’s always a valid control signal on the MOSFET inputs, even if the input is floating.

Now let’s expand this to the actual design of the 130x servo. The components on the servo are either proprietary or from an obscure manufacturer, but the general motor driver design can be determined with an ohmmeter and by observing circuit traces.



Switches S1A and S1B, as well as S2A and S2B, are combined into 6-pin components that I’ve marked with dashed lines as S1 and S2. The servo microcontroller operates on 2.7V from a voltage regulator on the servo. NPN transistors Q1 and Q2 provide translation between the 2.7V microcontroller signal levels and the BEC voltage being fed to the servo from the 3-in-1. In the truth table, “UP” and “DOWN” labels for the actuator refer to the actuator direction for cyclic servos when mounted on the stock frame.

Let’s map this design to the components on the servo circuit board so we can see what typically burns up. The board layout doesn’t include any component reference designators, so I’ll use the ones from our schematic.



Comparing this to PHOTO 5535b earlier, component S1 is the part that typically cooks on the elevator servo. In PHOTO5535b, you can also somewhat see that there seems to be two areas of damage to the part. My premise is that both switches S1A and S1B inside S1 are put into a conducting state, and we’re presented with the very bad short circuit situation mentioned in the introduction of the H-bridge.

Several possible failure scenarios have been investigated. Only one seems to pass scrutiny. It has to involve a failure internal to S1 since the common gate signal cannot turn on both S1A and S1B at the same time. On at least the older V1.5 servo layout, the A-gear is passing very close to the circuit node formed by R1, S1 pin 1, S1 pin 3, and the collector of Q1. This node is marked “SEE TEXT” in PHOTO 5561. To be specific, the A-gear passes extremely close to the collector pin on Q1. So?

My premise is that for the MOSFET devices to operate on the low BEC voltage available in the 130x, the S1 and S2 components are designed with ultra-sensitive gate inputs. The potential bad side of this is that the components may also be unusually sensitive to static discharge, compared to higher voltage MOSFET devices.

My conclusion on the smoking elevator scenario is that static is being passed from the A-gear to the collector pin on Q1, where it subsequently passes to the gate inputs of S1 and blows out some substrate layer in the S1 MOSFETs, shorting them out. This then causes S1A and S1B to act as a short circuit across the BEC output of the 3-in-1. With no provisions for current limiting anywhere, component S1 burns up.

For a plastic A-gear, the static could be caused by the A-gear flying extremely close to the plastic body of Q1, or possibly the two occasionally touching (as in a Van de Graaf generator). For a metal A-gear, more possibilities exist. Static could discharge from the tail boom & torque tube or the main shaft, onto the A-gear, and then onto the Q1 collector pin. Static could be coming from, say, touching the tail boom after walking on carpet. It could also be generated locally on the 130x if plastic tail or main blades are spinning up on carpet in a tipover scenario, or blades passing on a nylon jacket, etc.

MOSFETs are inherently sensitive to static discharge, and typically include discharge protection diodes in their design. We don’t know exactly what parts are used in the servo manufacture, so there’s no way to know how much protection they have. It is also possible that any protection diodes provided degrade with multiple exposures to static and they eventually break down, or that a substantial static discharge is beyond the protection built into the components.

The updated V1.6 servo layout improves on this by moving transistor Q1 below the A-gear.
__________________
Kevin B.
"Fun, Learning, Friendship and Mutual Respect" - good words to live by
RC: 300x, 130x, DX6i, iCharger 106b+, Phoenix Sim
Other: Atmel AVR microcontrollers, electronics, 3D printing

Last edited by helibus; 01-10-2014 at 11:47 PM.. Reason: Final edits for clarity
helibus is offline        Reply With Quote Quick reply to this message
Sponsored Links
Advertisement
 
Old 01-06-2014, 06:42 PM   #2 (permalink)
Registered Users
 
Posts: 1,667
Thread Starter Thread Starter
 

Join Date: Mar 2013
Default The short summary

This post provides a short summary of the proposed failure scenario for the 130x smoking elevator servo. Post #1 provided the technical details behind this conclusion.

The component that often burns up on the 130x linear servos is one of two dual-MOSFET components forming the switches in a reversible motor drive circuit. The MOSFET component is forced into a situation where both of the internal MOSFETs are conducting, essentially shorting out the BEC voltage from the 3-in-1. This leads to high current that burns up the component.

The design of the servo is not capable of turning on both MOSFETs in the component, so the short circuit condition must be caused by a failure internal to the component.

My conclusion is that the most likely failure scenario is static being discharged from the A-gear onto a circuit trace that connects to the gate input pins on the particular MOSFET component that typically fails. This results in a catastrophic failure of the component, leading to the short circuit condition. Without any current limit protection, the component gets red hot during the short circuit and burns up.

The updated V1.6 servo layout improves on this by moving the component picking up the discharge away from the A-gear.
__________________
Kevin B.
"Fun, Learning, Friendship and Mutual Respect" - good words to live by
RC: 300x, 130x, DX6i, iCharger 106b+, Phoenix Sim
Other: Atmel AVR microcontrollers, electronics, 3D printing

Last edited by helibus; 01-10-2014 at 11:49 PM.. Reason: Final edits for clarity
helibus is offline        Reply With Quote Quick reply to this message
Old 01-06-2014, 06:42 PM   #3 (permalink)
Registered Users
 
Posts: 1,667
Thread Starter Thread Starter
 

Join Date: Mar 2013
Default Q&A related to the 130x smoking servo failure analysis

This post answers some obvious questions regarding the proposed failure scenario leading to servos with burned-up parts.

Isnít the failure caused by the A-gear causing physical damage to a component on the board? Not likely. The A-gear would rub on a drive transistor (Q1 in the circuit description), not the actual MOSFET part that burns up. While the drive transistor could cease to function, physical damage to the transistor alone canít cause the MOSFET device to short out since the signal coming from the transistor can't turn on both switches involved in the failure at the same time.

Couldnít the component burn up if the motor is stuck, say with grit in the gear teeth? Not likely. The failure is also often associated with the servo actuator being driven to an extreme, indicating the motor is rotating fine.

Well, couldnít the failure be caused by the motor being jammed when the actuator hits that extreme position? Unlikely. In this scenario, equal power would be dissipated in both of the two MOSFET components (S1 and S2 in the circuit description). As a minimum, both components would be getting equally hot. In reality, the motor is actually dissipating most of the power, so it would be getting hotter than the MOSFETs, at least until the motor windings burn up and short out.

Could a glitch in the microcontroller cause this? What happens if the microcontroller is dead? Because of the way the drive circuit is designed, signals from the microcontroller cannot turn on both of the MOSFET switches involved in the failure at the same time.

Could lube, metal bits or dust from the A-gear, or dirt collecting on the circuit component side of the servo be causing the failure? Not really. If the right mix of lube and contaminants gets conductive, it could mess up operation of the servo (most likely motor not working or motor stuck in on state). The servo design, however, really has no way for it to turn on both of the troublesome MOSFETs that lead to the short circuit condition.

Could this just be an infant mortality or other random issue with the part? No. Two of the identical parts are used on each servo. The failure is almost always with the one part location. Random failures would be expected to occur in both part locations equally. The failure also occurs on the elevator servo more than the other servos.

Why is the servo actuator usually in the full down position when this failure happens? Referring to the earlier circuit description, the default state of the circuit has switches S1B and S2B turned on, pulling both legs of the motor to ground. When component S1 shorts out, the voltage applied to the red motor wire will be somewhere between ground and BEC V+, or positive with respect to the blue motor wire being held to ground by switch S2B. Positive voltage applied across the red-to-blue motor wires causes the motor to rotate CCW, driving the actuator down.

Why does the failure typically happen right at initialization? The damage from static discharge could have actually happened as part of a prior flight or crash, repair work, packing and unpacking for transport, or even just as the battery is being connected or the canopy installed. Failures that occur as part of a tipover at startup could involve static from blades passing over static generating material such as carpet.

Could the component failure happen at other times, say in flight? The failure could occur any time there is a high voltage difference between the A-gear and the servo circuitry. A few have reported the failure occurring in flight. This would imply that at least part of the problem is static being generated in flight.

If the failure could happen at other times, could it include a situation where the servo actuator is found driven to the top end of the servo? Unlikely to maybe. Referring to the circuit description, S1 could fail when the servo is being instructed to move the actuator upwards. Switch S2A would be on in this case, supplying BEC V+ to the motor blue lead. When S1 fails, the motor red wire will likely have Ĺ BEC V+ applied, causing the motor to rotate CW. This would drive the servo actuator up. If there is enough voltage for the servo microcontroller to continue operating, however, it should detect when the actuator passes the desired position and then attempt to drive the actuator downwards. The problem is that in the failure scenario where the 3-in-1 BEC is shorted out, we donít really know what could be going on with anything connected to the BEC output.

Can this failure damage anything other than the servo? Definitely. Youíre looking at short circuit current flowing from the battery, through the battery wires and connector to the 3-in-1, circuit traces on the 3-in-1, components in the 3-in-1 BEC circuit including the troublesome D882 output transistor, the servo connector on the 3-in-1, and the cable to the servo. Any of this could be damaged. This failure is likely a frequent reason for a stressed or damaged BEC output transistor on the 3-in-1.

Is a damaged servo repairable? Maybe, if the damage to the circuit board isnít too extensive. Other than the failed FET component, the rest of the servo appears to survive OK. Replacement MOSFET chips for the 130x servo are listed under the DIY section of the Megasmicros web site. DIY repair would require experience soldering surface mount components, good eyesight or magnifier, and a very fine soldering tip. While FET replacement isnít a repair listed by Mega, you could ask him about it but the repair may not be cost effective. Suitable MOSFET replacements from places like Digikey havenít been identified as of this time.; dual MOSFETs in a 6-pin package appears to be unusual.

How does the V1.6 layout improve on this? In the V1.6 layout, the transistor component likely picking up the static discharge from the A-gear has been moved. It would now take a higher voltage potential to jump the gap and cause the MOSFET damage.

Is the V1.6 layout immune to this failure then? If the failure is associated with the A-gear getting close to the elevator servo, are the servos on the frame sides safe? Not necessarily. All it takes is a strong enough static potential from whatever source applied to a sensitive enough spot anywhere on the 130x and damage can occur. Just as with any electronic device.

What is safer in this regard Ė a metal or plastic A-gear? Hard to say. A plastic A-gear might be more likely to build up static as it spins close to the drive transistor for the MOSFET component that fails. A metal A-gear would be more likely to pass static coming from the tail or main shaft.

Would it be better to use a carbon fiber frame than the plastic frame? Hard to say. A carbon frame would do better at maintaining an equal charge potential across the 130x. The CF frame would, however, better conduct an external static discharge from one part of the frame to another.

------------------------

What can be done to minimize the failure from happening?

I personally prefer to cover the back of at least the elevator servo with clear packaging tape to protect it from lube and other contaminants. Whether this would actually hinder the static discharge isn't clear, but it still seems like a good idea. Yeah, I know the tape I use isn't static safe. I'm more concerned about the lube gunk than static from the tape. Other materials like nail polish or a thin layer of silicone sealant would likely also work.

We don't have the option to fly in a static-safe environment, and few of us have proper static-safe workbenches, but some common sense can be easily applied.

Be wary of taking off from static generating material such as carpet or plastic landing pads. An inadvertent tip over with plastic or plastic-covered blades running could generate substantial static.

If you notice static buildup when you're walking around (sparks are felt touching doorknobs, for example), be careful handling the 130x as you would be careful with other electronics. Discharge yourself before picking up the heli, or at least first touch the surface where it is sitting. When carrying it, try to discharge yourself before setting the heli down, or at least touch a hand to the where you're going to put the heli before you set it down. This works for handling components on the workbench, too.

Consider the fact that the pluck-n-pull foam in your transport case is likely not static safe... although you can buy some that is.

While we can't readily eliminate sources of static build-up, the goal here is to at least equalize charges and not have static discharge through the 130x. Again, this would be good practice with any electronics.
__________________
Kevin B.
"Fun, Learning, Friendship and Mutual Respect" - good words to live by
RC: 300x, 130x, DX6i, iCharger 106b+, Phoenix Sim
Other: Atmel AVR microcontrollers, electronics, 3D printing

Last edited by helibus; 01-11-2014 at 12:00 AM.. Reason: Final edits for clarity
helibus is offline        Reply With Quote Quick reply to this message
Old 01-06-2014, 08:00 PM   #4 (permalink)
Registered Users
 
Posts: 1,166
 
My HF Map location
Join Date: Apr 2013
Default

Again, an excellent piece of work. . Question. If one coats the back of the servo board with someting to insulate it, like clear nail varnish for example, would this help prevent the static discharge damage? What other preventative measures can one implement?
__________________
Allan
TREX 250 PRO,Talon15, GPRO, Zeal ----TREX 500 EFL PRO, GPRO, RAILS ----2x TREX 450PRO V2, 1xGPRO&1xBRAIN, Talon35, RAIL , Zeal ---- SAVOX powered TREX500 EFL in Roban scale UH1-N ---- TREX 600 EFL PRO, GPRO, RAILS ---- DX9 ----
TR450ZS is offline        Reply With Quote Quick reply to this message
Old 01-06-2014, 08:19 PM   #5 (permalink)
Registered Users
 
Posts: 1,667
Thread Starter Thread Starter
 

Join Date: Mar 2013
Default

Quote:
Originally Posted by TR450ZS View Post
Question. If one coats the back of the servo board with someting to insulate it, like clear nail varnish for example, would this help prevent the static discharge damage? What other preventative measures can one implement?
Fair question! I would say worst case is doing nothing. Lube from the A-gear will coat the circuit component side of the elevator servo. The lube could have any kind of conductivity to it, plus it will collect dirt and possibly wear residue from the A-gear that could add it's own conductivity. So, covering the back of the elevator servo with tape, nail polish, or similar is probably a good idea. Smearing that small area of the board in front of the A-gear with silicone comes to mind...

I wish I had a static meter to determine whether there are static-generating sources on the 130x itself. Short of knowing that, I would suggest always handling the 130x with the same caution you should with any electronic device - be careful after walking on static generating surfaces, discharge yourself to a ground somehow after sitting down at the workbench but before touching electronics, when walking with the 130x in hand always touch the ground or worksurface yourself before setting down the heli, etc.

EDIT: Earlier posts have been updated
__________________
Kevin B.
"Fun, Learning, Friendship and Mutual Respect" - good words to live by
RC: 300x, 130x, DX6i, iCharger 106b+, Phoenix Sim
Other: Atmel AVR microcontrollers, electronics, 3D printing

Last edited by helibus; 01-07-2014 at 03:55 PM..
helibus is offline        Reply With Quote Quick reply to this message
Old 01-07-2014, 12:08 AM   #6 (permalink)
Registered Users
 
Posts: 5,156
 

Join Date: Aug 2011
Default

What is this V1.6 you speak of? Is there an updated servo version?
jmtyndall is offline        Reply With Quote Quick reply to this message
Old 01-07-2014, 02:13 AM   #7 (permalink)
Registered Users
 
Posts: 1,667
Thread Starter Thread Starter
 

Join Date: Mar 2013
Default

Quote:
Originally Posted by jmtyndall View Post
What is this V1.6 you speak of? Is there an updated servo version?
Yes, an updated version of the SPMSH2040L cyclic servo has been out since late 2012 or early 2013. The replacement servos I purchased after obtaining a 130x in April 2013 were all the newer version. The version found on a new 130x will depend on when the 130x was actually built.

Pictures of the two versions are in 130x Linear Servos - The Unauthorized Tech Manual. You'll especially be interested in Part 1 - Functional Overview, Versions
__________________
Kevin B.
"Fun, Learning, Friendship and Mutual Respect" - good words to live by
RC: 300x, 130x, DX6i, iCharger 106b+, Phoenix Sim
Other: Atmel AVR microcontrollers, electronics, 3D printing
helibus is offline        Reply With Quote Quick reply to this message
Old 01-07-2014, 12:16 PM   #8 (permalink)
Registered Users
 
Posts: 5,156
 

Join Date: Aug 2011
Default

Ah thanks, I've had it bookmarked, just need time to go read it. I got mine late last year, so I'm probably okay
jmtyndall is offline        Reply With Quote Quick reply to this message
Old 10-15-2019, 12:15 PM   #9 (permalink)
Registered Users
 
Posts: 591
 

Join Date: Jul 2018
Default

Do you have the part numbers for the FET servo arrays?
__________________
130X parts wanted! The Precious! You has the Precious!
Mrowka is offline        Reply With Quote Quick reply to this message
Old 10-15-2019, 12:28 PM   #10 (permalink)
Registered Users
 
Posts: 1,667
Thread Starter Thread Starter
 

Join Date: Mar 2013
Default

Quote:
Originally Posted by Mrowka View Post
Do you have the part numbers for the FET servo arrays?

No. I remember looking around a bit back in the day but nothing jumped out as a suitable replacement. Can't say the search was exhaustive, though.


EDIT: Megasmicros has them listed on their website, but out of stock. Maybe they'd give you the part number if they don't plan to restock them.



https://www.megasmicros.com/product_p/fet120130x.htm
__________________
Kevin B.
"Fun, Learning, Friendship and Mutual Respect" - good words to live by
RC: 300x, 130x, DX6i, iCharger 106b+, Phoenix Sim
Other: Atmel AVR microcontrollers, electronics, 3D printing
helibus is offline        Reply With Quote Quick reply to this message
Old 10-15-2019, 01:04 PM   #11 (permalink)
Registered Users
 
Posts: 591
 

Join Date: Jul 2018
Default

Thanks, I already texted Rudy but have not heard back.
__________________
130X parts wanted! The Precious! You has the Precious!
Mrowka is offline        Reply With Quote Quick reply to this message
Old 11-02-2019, 01:11 PM   #12 (permalink)
Registered Users
 
Posts: 591
 

Join Date: Jul 2018
Default

It just occurred to me - if static electricity were the cause of the smoking elevator servo, would a ground wire be the cure?
__________________
130X parts wanted! The Precious! You has the Precious!
Mrowka is offline        Reply With Quote Quick reply to this message
Old 11-02-2019, 09:49 PM   #13 (permalink)
Registered Users
 
Posts: 10,618
 

Join Date: May 2013
Location: Pittsburgh,Pa.
Default

In all the posts I have read since the first release of the 130X, I have never seen anyone mention static issues.
I pretty much follow every post in this section.


HW.
__________________
Many helis from the 70s to 2000s.
Align,Blade,GMP,Hirobo,Kyosho,Schluter.Flybars cool,FBL too. Currently flying the little ones.
Heliwhisperer is offline        Reply With Quote Quick reply to this message
Old 11-03-2019, 12:38 PM   #14 (permalink)
Registered Users
 
Posts: 591
 

Join Date: Jul 2018
Default

Unless I am mistaken, this thread....
__________________
130X parts wanted! The Precious! You has the Precious!
Mrowka is offline        Reply With Quote Quick reply to this message
Reply




Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the HeliFreak forums, you must first register.
Please enter your desired user name, your REAL and WORKING email address and other required details in the form below.
User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself. Use a real email address or you will not be granted access to the site. Thank you.
Email Address:
Location
Where do you live? ie: Country, State, City or General Geographic Location please.
Name and Lastname
Enter name and last name here. (This information is not shown to the general public. Optional)
Helicopter #1
Enter Helicopter #1 type and equipment.
Helicopter #2
Enter Helicopter #2 type and equipment.
Helicopter #3
Enter Helicopter #3 type and equipment.
Helicopter #4
Enter Helicopter #4 type and equipment.

Log-in


Thread Tools
Display Modes

Posting Rules
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On




Copyright © Website Acquisitions Inc. All rights reserved.
vBulletin Security provided by vBSecurity v2.2.2 (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.

SEO by vBSEO 3.6.1