Tech Inspection

What is the acceptable voltage for the tech inspection with multimeter positive to battery and multimeter negative to frame.

Not voltage; resistance. You’re looking for at minimum 10 kΩ.

Voltage should be Zero or darn close.

Check resistance: from frame to battery + or battery - MUST be much higher than zero.

We just failed inspection, with a chassis to PDB resistance of about 970 KOhm to ~1 Mega Ohm. I was not there so I was unable to help troubleshoot or get a better understanding of the situation. Last I checked though, 1 Mega Ohm is indeed greater than 10 K ohms. :confused:

I was told (second hand) that we were told it indicated a short somewhere. It’s late and has been a long week at work, but I am currently failing to see the logic that it indicates a short.

Any thoughts? Anything I could be missing? I can certainly help try and isolate branches (remove breakers 1 at a time, visually inspect wires and sensors, etc) but 1M is > 10K so I just fail to see how it is a sign of a problem. I am sure someone will enlighten me here though. :slight_smile:

I see you are from NJ, so I assume you are at a MAR event (Rutgers or Chestnut Hill). Due to the nature of the district system, there are some new folks volunteering, including inspectors. They have been doing a great job but little things slip by here and there.

Choosing a value for electrical isolation can be a bit tricky. Since the robot power system is not my design I’m going to defer to FIRST here… Per the robot inspection checklist that is on the FIRST website and every inspector carries with them:

Isolated Frame – Must be electrically isolated from battery, cRIO and camera must be insulated. (>10k Ohm between
either PD battery post and chassis) <R38>

Bottom line: Before you go ripping things apart, please check with the Lead Robot Inspector at your event.

Make sure your Axis camera and cRIO are electrically isolated from the metal chassis.

Check any Banebot 775 motors for case shorts.

Check any sensors for grounded frames/mounting bolts (Sharp IR range sensors’ mounting screws are grounded).

Again: Go to the Lead Robot Inspector and kindly ask about this issue.
The LRI is experienced and will spot any errors in interpreting the multimeter.

Thanks Guys - would you agree that 1Mohm is indeed isolated? After all it meets the requirement and is greater than 10K. Or if you saw it would you be concerned that it could indicate a lack of isolation somehow?

I just want to make sure I am thinking about this correctly - clearly if we have a problem we need to fix it. (In other words, does it seem more like getting on the same page with the inspection team and Lead inspector, or that we do actually have an issue?). Since 1M is > 10K, I logically thought it indicated we were good to go.

Any idea what is a typical/average reading?

Thanks again

On our robot, it’s about 7V. There’s about 1M ohm between the positive PD lug and the frame, and about 1M ohm between the negative PD lug and the frame. Nothing wrong with that.

The rules say “isolated” and the cheklist says “10k ohm” so I see absolutely no reason that 1M would be illegal.

For the purposes of FIRST, I’d say 1Mohm is indeed isolated, especially since they’ve spec’d it. I’d say your first priority would be get on the same page as the inspector and lead inspector and see what, if any their concerns are.

I couldn’t guess a typical/average reading but I know on 1712’s robot this year and last, we couldn’t get any resistance reading (out of range) on the meter. It was a good Fluke and out of limits is >40Mohm on it.

If I had to guess, there is a sneak path somewhere on the robot that you are picking up your ~1Mohm resistance. Possibly a shorted motor winding. It probably isn’t a problem but I would either want to find the root cause or keep monitoring it even if it passes inspection, but that is just me.

You aren’t using a Banebots RS-775 motor are you? They were notorious last year for having shorting issues, some of which started around where you are and got progressively worse.

We do have some 550’s on there whose motor housings are not isolated (as none of the motors are for obvious mechanical reasons). We can check those tomorrow, recheck the camera isolation, etc. But first and foremost as you have all stated, we need to get on the same page with the inspectors to make sure we understand their concern.

550s have never been reported to have a problem with a case short. The 775s on the other hand…

The LRI is the first person you want to talk to tomorrow. Not an inspector waiting at your pit (if there is one). If he’s holding you to a tighter spec than the inspection checklist (which is based on the rules), then ask him to call HQ for a ruling.

970K indicates a potential problem that needs tracked down.

Can be EXTREMELY difficult to trace… better done BEFORE competition starts.

I’ve spent half day helping a team disconnecting everything until culprit was isolated… a good team too… 207 from LA!!

When doing RESISTANCE test: swap probe leads to check BOTH polarities.

Shorts often exhibit diode behavior due to intervening multi semiconductor paths.

The Lowest R reading rules and must be >10K to pass… BUT–>

Anything around 100K is a problem in spite of 10k spec. if you don’t want sporatic random robot resets during competition!!

Good robots will read infinity!!
there should be nothing conductive at all to chassis. >2M (Harbor Freight) to >40M ohm (industrial grade)

Do you use a 1M potentiometer by chance??
a bit high to use for position or most sensor feedback applics, just a thought

RS-775 tab to case shorts are usually around 1 ohm AT motor

… however when routed thru H-bridge of Jag’s or Victors can show up as ~100k… I believe Team 207’s last yr was 200-300k & asymmetrical when ohmmeter leads were swapped… they had 2 case to tab, 775 shorts!
And all new ones left in Spare Parts were shorted! (6)

GP: Attention teams possessing good 775’s & NOT using them
PLEASE bring unused RS-775’s to competition: Pitt announce they’re available

To troubleshoot…
Remove both leads from one motor / device at a time until R changes
… use lower of swapped values as it is more solid
(higher reverse resistances often waver around a bit making ya crazy)

Good luck…

7V? Ignoring any resistance checks, that should be throwing major red flags. If your frame is completely isolated, you should not be able to get any voltage between the battery and frame.

That being said, as many others have said here, talk to the LRI. Though we on CD can debate back and forth for ages on what is/isn’t an acceptable level of resistance, our decisions carry no weight. You should never cite a post on CD when talking to officials about rules.

As I said, our frame has a little over 1 megaohm to both leads – it is not completely isolated, nor does it have to be. I was trying to point out that voltage means nothing, it’s about resistance.

Opinion from a 10 year 3 events/yr RI & 15 year FIRST Mentor:

NOT so! the voltage on the frame is stored in the frames capacitance.

Capacitance (Joules) is like a charged battery with a pretty low internal resistance ready to dump its energy to anything that will take it independent of resistance that charged it. (R affects time to charge it to 7v or whatever)

The charge time is ~3xRC. your R is 2x1meg in parallel= 500K C= TBD but will be significant charge on such large metal structure so as to blow out or damage (cause indeterminate intermittent) to CMOS & TTL inputs and maybe even outputs (even protected TBD)!

there is surely potential to upset digital logic:

Future problem scenario: robot reset via intermittent pinched wire contact to 7V chassis! Of course it happens near beginning of playoff match or crucial qual. Murphy is mean!

How many resets, raspy control scenarios can your team endure and still win?

>10K per se means little; LRI likely to give provisional pass but wise to suggest team keep looking for the offending continuity and clear it, for team & alliance performance benefit.

More importantly teams should view this as a symptom thus give it priority to avoid future unexplainables: painful resets, jerky response, unhappy partners

I am sure I could justify Risolation being >500K or even 1M due to chassis charge stored and its ability to upset logic under many circumstances of vibration, impact, etc.

If R isn’t infinity, something is wrong. There are no insulators used on the robot wherein it is normal for chassis to either battery terminal to be <infinity

Some coatings & anodization are normally insulators may exhibit diode behavior polarity sensitive R one direction high, other very low (Copper oxide rectifiers = diode used 50 yrs ago)

(so depending on your meter…infinity = >1m, >10M >20M >40M is normal)

see my earlier post this thread for 2011 team 207 Rbatt-to-chassis= 200k-300K offender RS775 <1ohm measured at motor

Good luck… Go for infinity!!!

Thanks again for the replies. We got there this morning armed with 3 different multimeters. All three read ‘0L’ (no load), and we measured 0V from battery to chassis. I am not sure what was happening with the meters and tests last night, but it was smooth sailing today and the final inspection was quick. ::safety::

That would lead me to want to check it again after each of your first couple of matches just to make sure you don’t have some kind of intermittant problem. If it does come up again, isolating which circuit it’s on while you can measure it may help in some diagnostic later in the day should other symptoms arise.

Most robots will read OL (over load or immeasurably high resistance). This should be your ideal target, and I’d insist on it before bag’n’tag. If you are measuring hundreds of kOhms at a regional, you likely have more important things to do. In any case, keep an eye on it and re-measure every time you do something that might risk a short.

OL: You are done! Measure again before your next match.

0-10 Ohms: There is a direct short somewhere in your system. You are likely to light on fire the next time you get rammed.

90-110 kOhms: One of the devices you have pulled a breaker for has a short close to its positive lead.

1MOhm + : You might have a short down stream of a piece of silicon. The actual value read will heavily depend on the exact model multimeter you are using. This is often also polarity sensitive.

Here are a few that I’ve measured with my multimeter, YMMV:
Grey Jag Power Output:
2M Ohms to PD + or -. Takes 15 seconds to stabilize, invariant with meter polarity.

4 slot cRIO chassis:
10kOhm to PD - with positive meter on PD.
9kOhm to PD + with positive meter on PD.
Oscillation with negative meter on PD.

Could others add to this list?

Can we agree that we disagree on that? FIRST specifies >10k, 970k is almost 100 times that.