What is this BS Castle is feeding this customer?

[Pdelcast]

WELL, nice of everyone to keep everybody else honest. :)


HERE IS THE STATUS of our investigation.

1. It seems to be an infant mortality problem with the switcher control chip, a National Semiconductor LM3485
2. There is already a TVS (transient voltage suppressor) on the circuit that feeds the LM3485 -- so it may NOT be a voltage spike related failure (28V TVS on LM3485 power line, LM3485 part is rated to 36V operational,) but we are not ruling out the possibility at this time.

We are working with National Semiconductor to determine the cause of the failures. It seems that making a couple minor changes to the circuit make everything work fine -- but we haven't determined the cause of the failures yet.

Patrick

[bdebde]

Originally Posted by Pdelcast

The MMM uses regen braking -- has to, there is too much energy to try to dissipate as heat in the ESC, so the braking energy goes back to the battery.

Besides, the diode would just make the capacitors work HARDER, not less -- the issue comes when the power side impedance is so high that the capacitors have to work really hard to keep the ripple voltage low. When the capacitors have to work hard, they get hot -- if they get too hot they fail.

SO -- no resistors, no diodes, nothing at all in series between the battery and the ESC. The lower the battery to ESC resistance, the better for the ESC, and the cooler the ESC will run.

Thanx!

So Patrick,

Can we use a resistor on the battery to prevent the big spark with 6s setups? Plug in neg, plug in resistor to charge caps, then plug in main positive.

[JerryF504]

Originally Posted by Pdelcast

WELL, nice of everyone to keep everybody else honest. :)


HERE IS THE STATUS of our investigation.

1. It seems to be an infant mortality problem with the switcher control chip, a National Semiconductor LM3485
2. There is already a TVS (transient voltage suppressor) on the circuit that feeds the LM3485 -- so it may NOT be a voltage spike related failure (28V TVS on LM3485 power line, LM3485 part is rated to 36V operational,) but we are not ruling out the possibility at this time.

We are working with National Semiconductor to determine the cause of the failures. It seems that making a couple minor changes to the circuit make everything work fine -- but we haven't determined the cause of the failures yet.

Patrick

That's waayyy over my head lol, but I know Castle can and will make it right. Does that mean that it is a problem with all of the ESCs sold so far?

[lutach]

Originally Posted by Pdelcast

WELL, nice of everyone to keep everybody else honest. :)


HERE IS THE STATUS of our investigation.

1. It seems to be an infant mortality problem with the switcher control chip, a National Semiconductor LM3485
2. There is already a TVS (transient voltage suppressor) on the circuit that feeds the LM3485 -- so it may NOT be a voltage spike related failure (28V TVS on LM3485 power line, LM3485 part is rated to 36V operational,) but we are not ruling out the possibility at this time.

We are working with National Semiconductor to determine the cause of the failures. It seems that making a couple minor changes to the circuit make everything work fine -- but we haven't determined the cause of the failures yet.

Patrick

It might just be some faulty units. I see that a lot in the electronic industry. Your failure rate is for a new product is not bad at all. Big electronic companies would love to see that kind of failure rate for their new designs.

[skellyo]

Originally Posted by Pdelcast

WELL, nice of everyone to keep everybody else honest. :)


HERE IS THE STATUS of our investigation.

1. It seems to be an infant mortality problem with the switcher control chip, a National Semiconductor LM3485
2. There is already a TVS (transient voltage suppressor) on the circuit that feeds the LM3485 -- so it may NOT be a voltage spike related failure (28V TVS on LM3485 power line, LM3485 part is rated to 36V operational,) but we are not ruling out the possibility at this time.

We are working with National Semiconductor to determine the cause of the failures. It seems that making a couple minor changes to the circuit make everything work fine -- but we haven't determined the cause of the failures yet.

Patrick

Thanks for the information Patrick. I just have a hard time swallowing information from someone other than those who actually work at Castle. It's one thing for someone to speculate upon the failures, but it's entirely another for one to dissemenate information about failures from Castle as a 2nd hand source. If indeed someone at Castle is speaking with RURC regarding these failures, you guys should keep an eye on how your PR looks with someone like him speaking like he works for you. However, if he's just some crackpot trying to solve the world's problems, then you guys really should make it known that he is not affiliated with your company because he's really not giving anyone warm fuzzies about the whole situation.

[BrianG]

Could it be the LM3485 is really sensitive to ANY spike over 36v and the TVS isn't responding quickly enough to prevent the spikes?

Another possibility (I'm reaching here); could the high voltage and high currents on the motor leads be inducing a voltage somewhere in the BEC circuitry where it is "unhealthy"? I wouldn't think that the chip would be that touchy.

[TexasSP]

I think it's rather obvious the guy has no connections with CC. I have run across his posts/rants/hallucinations many times reading on forums and usually choose to ignore him. I get real suspicious when people choose to ignore questions they can't answer. I find it rather impressive Patrick and Joe both respond to us on this forum rather rapidly all things considered and would only take there words in these matters and ignore the others.

[phatmonk]

Originally Posted by TexasSP

I think it's rather obvious the guy has no connections with CC. I have run across his posts/rants/hallucinations many times reading on forums and usually choose to ignore him. I get real suspicious when people choose to ignore questions they can't answer. I find it rather impressive Patrick and Joe both respond to us on this forum rather rapidly all things considered and would only take there words in these matters and ignore the others.

Diddo he is repeating it today almost word for word.

[jfruge]

Cound it also just be an infant mortality problem of the LM3485 and nothing more? So some of the units will just die in the first 10 minutes of use...The rest will work just fine for ever.....?

Has anyone heard of one running longer than 10 minutes and then stopping?

I am just thinking out loud here.....

[Pdelcast]

Originally Posted by jfruge

So some of the units will just die in the first 10 minutes of use...The rest will work just fine for ever.....?

Yep -- that's what it looks like. If it works OK for more than about 20 minutes or so, it doesn't seem to have a problem after that.

Patrick

Are you going to do a recall on the current Monster Mamba Max's and hold off the sale of anymore until you locate & fix the problems?

I mean it wouldn't be fair to someone like myself who has to wait a couple of weeks for international shipping to receive your product only to have it fail within the first 20 minutes, wait a month to get a replacement only to have the same thing happen again.

[Pdelcast]

Originally Posted by Longhair

Are you going to do a recall on the current Monster Mamba Max's and hold off the sale of anymore until you locate & fix the problems?

I mean it wouldn't be fair to someone like myself who has to wait a couple of weeks for international shipping to receive your product only to have it fail within the first 20 minutes, wait a month to get a replacement only to have the same thing happen again.

We have shipping hold right now, pending final determination of the issue.

Please note that this is (so far) a MINOR issue -- we have had 6 failures total out of 1000 pieces shipped.

[tc3_racer_001]

yes. im from australia, and i just hope that mine works... international shipping isnt very nice... costs a bit too!

[bdebde]

Originally Posted by bdebde

So Patrick,

Can we use a resistor on the battery to prevent the big spark with 6s setups? Plug in neg, plug in resistor to charge caps, then plug in main positive.

Patrick, any answer for this?

[Pdelcast]

Originally Posted by bdebde

Patrick, any answer for this?

No problem,

Use about 50-500 ohms or so (value isn't that important), 1 watt minimum, surge rated resistor.


You will lose that "MANLY" pop when you plug in the battery though... it's very cool on the SHV with 20S and 4000+ microfarads of capacitance... sounds like a gun going off... (We added an inrush current limiter circuit on the SHV, 'cause connectors would die in about 3-4 plugins...)