| Author |
Comment |
 safiel
Online!
Registered: 3/26/05
Posts: 4
Edit
| Delete
|
hello,
During one of our initial robotics tests last Saturday our speed controller “let out the magic smoke.” That day we emailed you, then again on Sunday with pictures attached. Tuesday, March 29th we filled out your form hoping you'd respond to that faster then the emails we sent during the weekend and attempted to call you during your business hours twice with no answer. We have yet to receive a reply from you.
We are currently working under a really tight deadline and we need to know what happened to this speed controller as soon as possible so we can act accordingly. A detailed explanation of what happened has been emailed to you and can also can be found online: http://www.eigenspace.net/robot/build44.html
We’re not electrical engineers so we’re somewhat poorly equipped to diagnose this problem, but the nature of the damage is extremely concerning. One of the MOSFETs is completely lifted away from the heat sink and the thermal compound has been burned away.
We need information right away if we are going to have any chance of meeting our deadline.
thanks
Safiel – Mike and Christina
| cosma
Moderator
Registered: 4/10/04
Posts: 88
|
When "magic smoke" appears, the damage can already be quite extensive and it is no surprise that you see burned or exploded MOSFETs. Unfortunately, it is imposible to diagnose what happened looking at the damage. Typically, smoke will come after one MOSFET started conducting when it shouldn't have and causing the bridge to short. Undisired conduction can happen when there is a voltage spike that reaches >55V (which could happen during regeneration if battery is not attached) or if the MOSFET overheat. It may also happen that a flaw in the MOSFET driver causes the MOSFET to fire at the bad time. Controller damage is quite infrequent. We'll call you to discuss what could have caused this in your particular case, and arrange for a repair
| safiel
Online!
Registered: 3/26/05
Posts: 4
Edit
| Delete
|
Thanks for your quick reply. I've sent a private message with contact information.
| safiel
Online!
Registered: 3/26/05
Posts: 4
Edit
| Delete
|
I have an update on this situation. RoboteQ did finally get in touch, though by “arrange for a repair” they actually meant “we’ll give you 60 bucks for the scrap metal and you can buy a new one.” For a while, it didn’t look like I was going to get any information at all out of them, but finally Cosma did contact me by phone.
During this conversation, we talked for over 40 minutes about my setup. You see, I’m using Briggs and Stratton Etek electric motors, which are quite demanding on the system. What’s more, I was forced to cut the master power during the test in question. I’ve been told that neither of these things should have caused damage. Despite this, they apparently do not believe that my unit was defective, and never offered any assistance in its replacement save the parts salvage mentioned above.
We also talked about possible prevention measures for next time. See, at that point, I was under a huge time crunch. I ordered a replacement right away, not happy about being out another $500 but without really any other choice if I wanted to graduate on time (this is for my senior design project).
He suggested that I add a kill switch (using the controller’s signal wire), update the software on the controller to the newest version, use longer wires leading to the motors (coiled, to add inductance and resistance), and crank up the acceleration limits. I did all of these things.
I put everything back together (with new wiring of the thickness and lengths Cosma suggested, new software, a new kill switch, and acceleration limitation all the way up), and took it in for my final presentation. It survived the speech, but driving it back out to the car, my new controller smoked.
We never sent it a command to raise the power over 25 out of 127. We did not cut the power or batteries. We did not suddenly change direction. I've had the controller for less than 48 hours, and it has already failed.
I have spent, to date, over $1000 on this product, and have nothing but high blood pressure and two foul-smelling paperweights to show for it. RoboteQ: you have on your hands an extremely unhappy customer. It would behoove you to make a serious and immediate effort to back up the claims that you have made about the capabilities of your equipment.
| cosma
Moderator
Registered: 4/10/04
Posts: 88
|
I understand your disapointment, which I share.
There are two parts to your post.
The first is about the repair policy. Since you bought the controller from one of our reseller, we had no record of the sale and could not immediately verify the warranty status. Normal procedure in these case is to contact the reseller and arrange for the repair/return. We have since established contact with the reseller, briefed him on the procedure. The furtunate (unfortunate in your case) fact is that we almost never run in failure situation from any of our sales through NPC - so the procedure has little practice.
The second is about the repeated failure. As you may recall, I was not all that at ease when I learned you purchased the AX2550 to drive an Etek motor. This motor has an extremely low impedance and can draw up to 1000A. The controller has current limiting, however this feature cannot be depended on to overcome such a mismatch. The 2550HE, with twice the number of MOSFETs, would have been a preferred choice. Beyond this, very high current switching can cause all sorts of transients with often destructive consequences.
Please rest assured that we stand behind our product and reputation and that we will be contacting you off line to work to a satisfactorily resolution.
| tonysantoni
Registered: 4/16/04
Posts: 11
|
Mike As far as the warranty issue I sent you on March/31/2005 the following e-mail: QUOTE Mike
As far as the warranty, since you purchased the controller at NPC. W ask you to contact them to make arrangements through them.
We have no records of your purchase since you did not buy from us.
Tks
Tony
UNQUOTE
I reconfirm that you should contact the reseller since they have the details of your order (date of the order). If the controller was within the 90 days warranty period as of March/1 then we will arrange for a free replacement unit through NPC as per warranty terms. I am also sending this message directly to your e-mail address. Tks Tony --
__________________
Robotony
| tonysantoni
Registered: 4/16/04
Posts: 11
|
I forgot to say that Apr/1/2005 I notified NPC that you would be calling and gave instructions that if according to their records your controller was under warranty they had to replace it for free under RMA 50401. Everything had been arranged for the moment you would call NPC. Tks Tony
__________________
Robotony
| safiel
Online!
Registered: 3/26/05
Posts: 4
Edit
| Delete
|
I have been in contact with NPC Robotics with regards to this issue. I emailed them shortly after the first failure, and they never responded to that particular email. I called them on April 5th about buying a replacement (at the time, I had every indication that unless I just paid for a new one, I’d never get it in time), and they said nothing about the arrangements you mentioned (setting up the RMA 50401 with them). Unfortunately, I knew nothing of this at the time, otherwise I would have been able to leverage it, but this is the first I’ve heard of it.
I emailed them again yesterday about the second failure, I will email them again tonight now that I know that you’ve notified them of the situation, and I will call them on Monday to talk about my options. I will let you know how they respond.
As for the problem with using Eteks, I recall that you told me that you had not tested the controller with these motors, but that you had tested Perm motors, which have the same characteristics (low impedance and high current draw). It was based on these test results that you suggested some of my preventative measures (longer, coiled wires and updated software).
In your FAQ and other documentation, you do not indicate an upper bound on the size of the motors that this controller is capable of handling. You say that, thanks to current limiting, “the controller will be able to work safely with practically all motor types and will adjust itself automatically for the various load and environmental conditions.” Also from your FAQ, “The whole idea of active current limiting is that the controller will never give more current than it is safe to. If the motors are hungry for more current because of peak demand, the controller's maximum current is all they are going to get.”
From this, I concluded that using the AX2550 to drive Etek motors would simply result in motors that could deliver no more than 5 horsepower (roughly, 36 volts and 120 amps). This was sufficient for my application. I decided not to use the high efficiency version because this would deliver unnecessary amounts of power to my drive train and would have a negative impact on my battery life. According to everything that I could find on the controller, I had no reason to believe that they would not be up to the task.
Besides, as you’ve said, an Etek can draw in excess of 1000A. The high efficiency version doesn’t even approach this, so I can only assume that it would have smoked as well. From what I understand of the electronics involved, the problem is likely a shorted bridge, which will destroy a HE controller just as quickly as a regular one. I was depending on the current limiting capabilities to protect the controller as advertised, but they clearly did not.
Now you say that the current limiting cannot be depended on to safely drive these motors. It’s good to finally know what’s causing the failures, though it would have been even better to know it before I bought the second controller. I will work with NPC in order to reach a satisfactory solution to my own damages, but I strongly suggest that you update the documentation that you make available to your other customers to reflect what we’ve learned. I think that anyone looking into your products should be made aware of these “transients with often destructive consequences,” before they decide to buy a controller from you.
| cosma
Moderator
Registered: 4/10/04
Posts: 88
|
We will start putting a clear notice that our controller cannot be used with the Etek motor, pending further investigation. As I said, I do not know of any of of our controllers connected to an Etek motor, but I suspect there should be a few. We will try to find out where they are and which successful configuration has been used.
| |
3/30/05 at 03:26 AM
