Kind of a generic electrolytic cap question

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View
   One of my CNC machines has a circuit board that has failed 3 times
now. The control is a FANUC 3T. It is an older control so failing
boards don't surprise me. But it is just the one board. This board has
many custom hybrid ICs, some other of the shelf ones, 6 EPROMs, and
one electrolytic capacitor. I have checked this cap on two boards with
my meter and the cap has the stated value and is not shorted. But I
don't have a way to check ESR.
   The failure mode for all the boards is the same. First the turret
on this lathe will occasionally index in the wrong direction. After
the machine is powered up for a while it will then start to index
properly. I can control the turret rotation from a program but when
the board starts to fail it will only index in one direction.
   Furthermore, then turret should always rotate in the direction that
is the shortest to the next tool called up, but when the board starts
going bad the turret still only indexes in one direction.
Unfortunately this is the wrong direction.
   After this failure other things start going wrong, like the speed
control. Changing the boards has always fixed the problem. Note that
when I change the boards I have to swap all the EPROMs so that the
programming for the machine operation will be the same.
   Since all the boards seem to fail the same way I was wondering if
somehow a failing cap could cause this failure. Even though the cap
value reads fine with my VOM, that does not measure ESR.
   In any case I'm gonna order some caps and see if anything changes
but I would like to know if it is probably gonna be a waste of time.
Thanks,
Eric

Re: Kind of a generic electrolytic cap question
On 2019/11/14 12:20 p.m., snipped-for-privacy@whidbey.com wrote:
Quoted text here. Click to load it

I'd say it is unlikely the capacitor is the problem - normally these are  
used to smooth the Vcc to the chips. You say there are custom chips on  
the board, so depending on if you can get schematics or not then you may  
be looking at the rest of the circuits.

Regular electrolytic caps are rarely used for critical timing, rather  
they would use a tantalum cap if the value was important.

If you want to check the ESR then Bob Parker's Blue ESR meter kit is not  
a bad way to go...available here (flippers.com - shameless plug!) and  
there...

However I don't think the cap is responsible for your problems.

Have you tried simply reseating the chips and connectors to the PCB and  
any readouts, sensors, etc. in case it may be that simply reseating  
plugs fixes the problem?

If the problem is only apparent after the machine has been on a while  
have you checked the temperature of the logic? Chips don't like being  
more than 50C by and large, and the cooler they run the better. A heat  
gun and some cold spray may help find the sensitive IC as you say the  
problem goes away when the system is warm.

John :-#)#

Re: Kind of a generic electrolytic cap question
wrote:

Quoted text here. Click to load it
Greetings John,
   Actually the machine starts to work after it has warmed up for a
while. I have tried the whole contact cleaning route several times and
it has had no effect on this machine.  
   Today I tried out two boards that I got used from eBay. One worked
and the other, whil it indexed then turret OK would not allow
programmed speed changes.  
   The speed change fault is mysterious to me and it has happened on
another board. The mystery is because of what the fault does.
Niormally speed changes while running a program can only happen from
programmed speed changes or from the constant surface speed command.
But when the board goes bad then the spindle speed can only be
controlled by the dial used for speed changes when in jog mode. And
that dial normally does not work when running a program.
  I checked all the parameters and they have not changed. And the
EPROMs that hold the ladder programs for the machine are not affected.
In fact, I have to change them from board to board.
   Speaking of EPROMs, I am ordering a GQ-4x4 programmer so I can copy
my EPROMs. I am really afraid I will destroy one one of these days. I
have lots of 2716 and 27C16 EPROMs that I can erase and re-program.
The machine has in it some 2716 and 2516 devices and I was told that
the 2516 devices should not be used. I am hoping the programmer can
read the 2516 devices.
Thanks,
Eric

Re: Kind of a generic electrolytic cap question
On 2019/11/14 4:01 p.m., snipped-for-privacy@whidbey.com wrote:
Quoted text here. Click to load it

Hi Eric,

If those are TI TMS2516 then they are equivalent to regular Intel 2716s.  
The TMS2716 by TI had a triple supply and A10 was displaced.

TMS2716 vs regular 2716s

https://www.dropbox.com/s/7xaz0b829bhha6b/TMS2716vs2716.jpg?dl=0

TMS2516 (Thanks JRock for hosting):

http://www.jrok.com/datasheet/TMS2516.pdf

I don't know about the dependability of the TMS2516, but I would back up  
ALL EPROMs just for safeties sake!

I hope your new EPROM reader can handle the 2716s, if not if it can read  
2732s then you will get the data twice and you can split off the high  
and low sections.

John :-#)#

--  
(Please post followups or tech inquiries to the USENET newsgroup)
                      John's Jukes Ltd.
We've slightly trimmed the long signature. Click to see the full one.
Re: Kind of a generic electrolytic cap question
On 11/14/2019 6:01 PM, snipped-for-privacy@whidbey.com wrote:
Quoted text here. Click to load it



  If this is a consistent response, you may be able to locate the problem
by heating/cooling the board. The easiest way is to have the board  
warmed and working. Buy yourself a can or two of 'Freeze Mist' spray  
half the board to cool it down, while checking operation. If it still  
works, let it warm up completely so you don't confuse the issue. Now  
cool of the other half of the PCB and check operation. If it  
malfunctions, you know which end to continue troubleshooting. Warm it up  
until proper operation, then start cooling individual parts on the bad  
1/2 of the pcb, one at a time until the problem occurs. You can often
isolate the problem part this way.

  Or you can do just the opposite, cool the pcb in the fridge, then use  
isolated heat until the machine works. It is harder to isolate the heat
  than to cool an individual part. So, I recommend the first procedure.
   If you decide to use heat on the pcb, use a hair dryer, not a heat gun.

                                   Mikek

Freeze Mist, pick the best price with shipping.
Quoted text here. Click to load it
  Skip the CRC automotive product, I suspect it has other chemicals in it.




Re: Kind of a generic electrolytic cap question
snipped-for-privacy@whidbey.com says...
Quoted text here. Click to load it

The capacitor probably only costs a few dollars at most.  Just replace  
it and see if the board comes back to life. That is see if the machine  
acts normally.



Re: Kind of a generic electrolytic cap question
On Thu, 14 Nov 2019 17:35:48 -0500, Ralph Mowery

Quoted text here. Click to load it
Yeah, I am gonna do that. I just won't get my hopes up.
Eric

Re: Kind of a generic electrolytic cap question
On Thursday, November 14, 2019 at 3:20:47 PM UTC-5, snipped-for-privacy@whidbey.com wrote:
Quoted text here. Click to load it

I work on a lot of industrial equipment for two local machine shops, and electros do go.

They bring me the suspected boards but I have them well trained to first heat the boards with a heat gun (without crisping them) and report if the normal operation is restored when heated.

If so, the first thing to get checked is the electros.  I pull every one and test them for value, ESR, and dielectric absorption.  Testing ESR in circuit on many modern circuit boards is a waste of time.

99% of the time a board that works normally when heated has a lazy electrolytic.

If the electro replacement doesn't fix it, try cooling individual parts to see if you can get it to act up.  

Re: Kind of a generic electrolytic cap question
snipped-for-privacy@gmail.com says...
Quoted text here. Click to load it

Once pulled, you might as well not test them, just replace them.

You may want to test the new ones before you put them in.



Re: Kind of a generic electrolytic cap question
On Thursday, November 14, 2019 at 11:14:21 PM UTC-5, Ralph Mowery wrote:
Quoted text here. Click to load it
e and test them for value, ESR, and dielectric absorption.  Testing ESR in  
circuit on many modern circuit boards is a waste of time.
Quoted text here. Click to load it


If you know the board *will* be fixed and put back in service, then yes, by
 all means replace them.  But if I pull an electro during testing and it's  
good, I put it back and continue working on the board.  No sense changing g
ood caps on a board that may have to be replaced, like OP is dealing with.

Lots of these industrial boards use custom ICs, ICs and semis that have bee
n intentionally defaced, and processor ICs for which there is just no data.
  Sometimes you just have to replace a board.

Re: Kind of a generic electrolytic cap question
On Friday, November 15, 2019 at 5:30:14 AM UTC-8, John-Del wrote:
Quoted text here. Click to load it
one and test them for value, ESR, and dielectric absorption.  Testing ESR i
n circuit on many modern circuit boards is a waste of time.
Quoted text here. Click to load it
by all means replace them.  But if I pull an electro during testing and it'
s good, I put it back and continue working on the board.  No sense changing
 good caps on a board that may have to be replaced, like OP is dealing with
.
Quoted text here. Click to load it
een intentionally defaced, and processor ICs for which there is just no dat
a.  Sometimes you just have to replace a board.

'lytic caps go bad eventually. After removing the caps for testing you've a
lready run the risk of damaging the board. Why put the old ones back? In br
oadcast gear small value caps are the most common failures. I've replaced t
housands of 100uF 25 V caps. Unit price 32 cents, 1000 for $83. Buy in bulk
 and replace in bulk. It costs less than testing and re-installing.



Re: Kind of a generic electrolytic cap question
snipped-for-privacy@yahoo.com says...
Quoted text here. Click to load it
less than testing and re-installing.
Quoted text here. Click to load it

Right, at the small cost of the capacitors, it is foolish to put backin  
an old one even if the board may have other problems and not used.  

Now if it was a $ 10 or even maybe a $ 5 part it may be worth putting  
the old part back on.  

At the rate capacitors fail, the one that tested good may go bad much  
sooner than a new one.  Then you have to do it all over again.

It takes often takes more labor time and effort to test than to just put  
in a new one.  


Like where I work, a man from the factory came in to repair a 480 volt 3  
phase 200 HP motor speed control.  He found 2 bad dioides that were  
rated at something like 200 amps.  I asked him to replace the 3 rd one.  
He told me those things cost arond $ 200 each.  I told him that with 2  
bad it may have weakened the 3 rd one.  With the machinery costing the  
company over $ 1000 per hour for down time, and the cost of getting him  
back in the plant, I am willing to pay $ 200 for insurance and piece of  
mind.  That old diode may pop in a day or 10 years, but I know that I  
have done my best to make sure the machine keeps on running for minimal  
extra cost.

Re: Kind of a generic electrolytic cap question
On Saturday, November 16, 2019 at 11:09:21 AM UTC-5, Ralph Mowery wrote:
Quoted text here. Click to load it
ve already run the risk of damaging the board. Why put the old ones back? I
n broadcast gear small value caps are the most common failures. I've replac
ed thousands of 100uF 25 V caps. Unit price 32 cents, 1000 for $83. Buy in  
bulk and replace in bulk. It costs  
Quoted text here. Click to load it
  
Quoted text here. Click to load it


I've read that sentence like 5 times and all I can say is that you wouldn't
 be working for me..

Honestly, sometimes I wonder if people just post because they want to say *
something*.

Re: Kind of a generic electrolytic cap question
snipped-for-privacy@gmail.com says...
Quoted text here. Click to load it

I doubt that I would work for a tightwad like you.  If those capacitors  
are less than one dollar, it will take more time to test and replace  
them than the labor would pay.

Lets say I just make $ 20 per hour.  It may take me 5 minuits to setup  
the capacitor checker and test it.  That would be about 33 cents.  If  
the capacitor was a dollar, you have saved 77 cents.  

However if the board is good, I have to pull that capacitor and replace  
it anyway.  How long would that take and what would it cost ?

You have just lost money on that capacitor you thought you were saving.



Re: Kind of a generic electrolytic cap question
On Sunday, 17 November 2019 00:43:09 UTC, Ralph Mowery  wrote:
Quoted text here. Click to load it


Takes me about 10 seconds. I like to know whether I've found a fault or need to keep looking.


NT

Re: Kind of a generic electrolytic cap question
On Thursday, November 21, 2019 at 3:17:43 PM UTC-5, snipped-for-privacy@gmail.com wrote:

Quoted text here. Click to load it

Takes me about 10 seconds with the ESR meter, containing a 12 volt battery. Takes a bit longer on the big checker that operates at up to 550 VDC.  

Which do yo think will give the more reliable results?

Peter Wieck
Melrose Park, PA  


Re: Kind of a generic electrolytic cap question
On Thursday, 21 November 2019 20:47:02 UTC, snipped-for-privacy@aol.com  wrote:
Quoted text here. Click to load it

550v will tell you more about leakage, not more about ESR.


NT

Re: Kind of a generic electrolytic cap question
On 11/16/19 10:09 AM, Ralph Mowery wrote:
Quoted text here. Click to load it

Not even with $5-10 parts.
My time is worth more than that.

Quoted text here. Click to load it

Exactly, and then you KNOW they aren't the problem.


--  
"I am a river to my people."
Jeff-1.0
We've slightly trimmed the long signature. Click to see the full one.
Re: Kind of a generic electrolytic cap question
:
  
Quoted text here. Click to load it
, by all means replace them.  But if I pull an electro during testing and i
t's good, I put it back and continue working on the board.  No sense changi
ng good caps on a board that may have to be replaced, like OP is dealing wi
th.
Quoted text here. Click to load it
 been intentionally defaced, and processor ICs for which there is just no d
ata.  Sometimes you just have to replace a board.
Quoted text here. Click to load it
 already run the risk of damaging the board.  



Anyone who would ruin a board by removing a cap isn't a professional and sh
ouldn't be playing with it.  There is ZERO chance of damaging even a multi-
layer board.



Quoted text here. Click to load it
 common failures. I've replaced thousands of 100uF 25 V caps. Unit price 32
 cents, 1000 for $83. Buy in bulk and replace in bulk. It costs less than t
esting and re-installing.



You actually quoted my reason (thanks for not snipping it) but either didn'
t read it or didn't understand it.  So I've copied and pasted my previously
 posted reason below:

"If you know the board *will* be fixed and put back in service, then yes, b
y all means replace them.  But if I pull an electro during testing and it's
 good, I put it back and continue working on the board.  No sense changing  
good caps on a board that may have to be replaced, like OP is dealing with.
"

"Lots of these industrial boards use custom ICs, ICs and semis that have be
en intentionally defaced, and processor ICs for which there is just no data
.  Sometimes you just have to replace a board."  


So if I have to replace a board that caps won't fix, why go through the tro
uble of recapping it?  On low to mid percentage shots like these one-off in
dustrial controls, the plan is five stages:  

1)I troubleshoot and repair the board first.  If it's not repairable, I go  
to step 5.  If I get the board repaired, I go to step 2.

2)recap it

3)Clean and deflux the board, and inspect under a bright light and loop. Re
flow any questionable solder and reflow all the solder on edge connectors,  
transformers, and high heat generating components.

4)Deflux again and spray with conformal coating if the board originally was
 built with it.

5)Get paid for my time.

Re: Kind of a generic electrolytic cap question
On Sat, 16 Nov 2019 15:46:48 -0800, John-Del wrote:

Quoted text here. Click to load it

There is when *I'm* around! Here's how I got an inaccessible cap out of  
this network analyser a couple of days ago: https://postimg.cc/bZYGSvnr

But as you will probably gather, I'm no professional. :-D




--  
This message may be freely reproduced without limit or charge only via  
the Usenet protocol. Reproduction in whole or part through other  
We've slightly trimmed the long signature. Click to see the full one.

Site Timeline