Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Noise coming from PMAC card when powering on amp
#1
Hello all,

I am working on a project upgrading a system from an old PMAC1 card to a newer version. I have very little information (read: none) on the wiring of the amps at the other side of the JMACH1 cable. I am operating under the assumption that the JMACH1 pins are consistent between versions - as such I assume that if I configure the the new card the same as the old one, there shouldn't be a problem. (please, correct me if I am wrong).

I have connected the new card, however when I power up the amps, a loud buzzing sound comes from the back of the PMAC card. Similar to a fan spinning with an obstruction.
I am not sure what could be the cause of this. Is it likely to be mis-configured jumpers? It is a very concerning noise.

OLD PMAC CARD:
VERSION: 1.16D
TYPE: PMAC1, ISA/VME, FLASH, PID, CLK X3

NEW PMAC CARD:
VERSION: 1.17C
TYPE: PMAC1, ISA/VME, FLASH, PID, CLK X2

Regards,
Chris Herring
Reply
#2
Hi Chris,

The noise could be from the watchdog relay. If the 5V supply is too low, the relay will activate. Then, if the 5V subsequently increases or revives, the relay can de-activate, and so on. What is the watchdog LED doing?

It could be mis-configured jumpers. I would see that they are the same as on the old board, and also make sure that the socketed IC drivers are the same, either 2803A or 2981A, and use the same backup configuration file.

The JMACH1 pin connections are the same. I suggest connecting the pins one at a time to isolate one that causes the problem.
Reply
#3
Hi Gregs, Thanks for the reply.

The watchdog LED is inactive, unless it is switching on for a very short period of time, the ticking is of fairly high frequency so perhaps this is possible? I look a long exposure picture and couldn't see any red in the LED.
(The watchdog led is right next to the power led on the top board, correct?)

The IC driver in the new board is 2803A, E1 and E2 are both configured to connect pins 1 and 2, which I believe is the correct configuration. Unfortunately the old board is currently inaccessible, is there a way to tell the type of driver via software?

I have all the I-variables configured the same (aside from homing and Jog speed). Are there other variables that need to be configured that may cause this problem?

I will try connecting each pin one at a time. Are there any pins that could cause damage if connected independently (like a voltage rail with no ground connected?)

Which jumpers may be a likely culprit?
I am looking through the documentation so see which jumpers may affect the 5V rail;
I have E40-E43 all on. (although the card is not installed in PCI 0 slot, but that doesn't matter, does it?)
E28 connects ping 2 and 3.
E85 and E87 are OFF.
E89 is ON and E90 connects pins 1 and 2.


Regards,
Chris
Reply
#4
Hello Chris,

See responses inserted below:

The watchdog LED is inactive, unless it is switching on for a very short period of time, the ticking is of fairly high frequency so perhaps this is possible? I look a long exposure picture and couldn't see any red in the LED.
(The watchdog led is right next to the power led on the top board, correct?)

The watchdog LED is normally green and turns red when a watchdog occurs.

The IC driver in the new board is 2803A, E1 and E2 are both configured to connect pins 1 and 2, which I believe is the correct configuration. Unfortunately the old board is currently inaccessible, is there a way to tell the type of driver via software?

That is for the JOpt (J5) I/O connection. I would try disconnecting that too. No way of which I know of knowing which IC was in place based on the software configuration.

I have all the I-variables configured the same (aside from homing and Jog speed). Are there other variables that need to be configured that may cause this problem?

The I-vars should be OK. Does this happen when the amps are powered even if PMAC is not enabling any axes?

I will try connecting each pin one at a time. Are there any pins that could cause damage if connected independently (like a voltage rail with no ground connected?)

No, but I would try connecting 5V and 0V (pins 1-4) first, then +analog power (pins 58 and 59), then -analog (pin 60), and then individual signals.

Which jumpers may be a likely culprit?
I am looking through the documentation so see which jumpers may affect the 5V rail;
I have E40-E43 all on. (although the card is not installed in PCI 0 slot, but that doesn't matter, does it?)
E28 connects ping 2 and 3.
E85 and E87 are OFF.
E89 is ON and E90 connects pins 1 and 2.

A fairly likely culprit is conflicting supply voltage connections along with jumper settings. The 5V supply comes from the bus if PMAC is plugged in a computer in which case 5V must not be supplied on the JMACH connector. It is supplied on JMACH (pins 1-4) in a standalone setup. Which do you have?

With your jumper settings, +/- 15V should be supplied on JMACH1 (J8) pins 58, 59, and 60. Note that the unmentioned E88 must also be off.
Reply
#5
Hi gregs,

The watchdog LED is normally green and turns red when a watchdog occurs.

I see, it is not green nor red.

The I-vars should be OK. Does this happen when the amps are powered even if PMAC is not enabling any axes?

Yes, as soon as the main power to the amps is enabled the noise starts. The emergency stop switch is held all the while - so no amps are actually active.

A fairly likely culprit is conflicting supply voltage connections along with jumper settings. The 5V supply comes from the bus if PMAC is plugged in a computer in which case 5V must not be supplied on the JMACH connector. It is supplied on JMACH (pins 1-4) in a standalone setup. Which do you have?

The PMAC card is connected in a PCI slot, as was the old one. Perhaps the old setup is not actually drawing from the bus and 5v is subsequently being applied via JMACH, I will investigate this. If I see 5v across 1-4 then that would indicate that this is the case?

With your jumper settings, +/- 15V should be supplied on JMACH1 (J8) pins 58, 59, and 60. Note that the unmentioned E88 must also be off.

I was unable to find E88, is it likely to be underneath the CPU board? I assumed as much, in which case it is most likely to be default (off). The diagram of the board is different to any that I could find in the manuals (ASSY No. 603588-103)

Thanks,
Chris Herring
Reply
#6
Sounds like a problem with the 5V supply, which you might see with a scope. Make sure that 5V is not being supplied to pin 1 or 2 on the JMACH. The old computer would have supplied 5V on the bus, so 5V should not have been supplied to JMACH pins 1/2 there either. Try disconnecting any other connections on pins 1/2 as well. If that is not it, move on to other JMACH connections.
Reply
#7
Hi Gregs,

I have isolated the problem to the analogue 15V supply. When both the +15v (pin 59) and -15v (pin 60) are connected (not individually) the sound starts. AGND is connected too of course.
Pins 1-3 are not connected. Pin 4 is GND.

I read a voltage of approximately ±11V over these pins (58-59 and 58-60) however. So it seems that this was sufficient for the old card, but not the new card.

I am currently in the process of obtaining a power supply to provide sufficient voltage to test if this solves the problem, does this 1v discrepancy sound like a likely cause? Is there any other troubleshooting I may be able to do whilst I await a new power supply?

Thanks
Chris
Reply
#8
Hi Chris,

Does the power supply measure 12-15V before being connected to the PMAC?
I don't think that the 11V level would cause the noise, at least if it is a steady 11V. Can you check it with a scope?

Try re-initializing the PMAC either with E51 jumper or $$$***. Maybe there is programming that is causing the noise.

Check the ribbon cable for intermittent or poor connectivity in the IDC connectors.
Reply
#9
Hi Gregs,

Yes the power supply reads only 11V always.
I tried re-initalizing with $$$***, same problem.
The IDC seems to be fine as when I reconnect the old PMAC, the problem is not there.

Something interesting happened today however. When I came into the lab to test these these things, the problem only exists for 5-10 seconds on power-up and several seconds on power down of the amps (also, it appears to last less time with only 58, 59, 60 attached than it does with the full JMACH connected). It is looking like this is an issue with the amps.

Approximately what voltage would 59 & 60 need to fall to in order to trigger then watchdog?

Also, I rely on the HMFL flags for homing, and LIM flags for safety. If the voltage drops during operation, will these flags be affected? When the watchdog is tripped, does it kill any motion that may be running and open loops such that it will not continue when the watchdog de-activates moments later?


Thanks
Chris
Reply
#10
Hi Chris,

I am inclined to think that the PMAC needs repair if the other PMAC does not have this problem in the same setup, but I would get a power supply that provides a good 12-15 volts first.

Watchdog is triggered by the 5V supply not being between 4.5 to 5.5 volts (JMACH pins 1,2,3,&4), not by analog voltage levels on JMACH pins 59 and 60.

The LIM flags are fail safe, meaning they will trip if something such as insufficient supply voltage disrupts the current flow thru them. I would expect homing to completely fail if there is a problem.

It all depends on how PMAC and control software is programmed as to what will happen on recovery from a watchdog. PMAC would behave the same as if it were reset.
Reply
#11
Most PMAC controllers with analog outputs have an "interlock" circuit to confirm that the analog supplies are solid. The key concern is that if you lost one of the analog supplies altogether, the output would pull to the other rail, possibly resulting in a runaway. So if any of the supplies drops below a threshold, a relay drops out all power to the output op-amps. The threshold for the nominal +/-12 to 15V supplies is about +/-10.8V (-10% from 12V).

What can happen with an overloaded supply is that the load will cause the supply voltage to droop below the threshold, causing the relay to open. This reduces the load on the supply, so its voltage increases and the relay closes, increasing the load, etc., etc. You can hear the relay "chatter" as this happens. If it is fast enough, you might even describe this as a "buzz".

So I would follow Greg's advice to see what happens with a better power supply.
Reply
#12
Hi Gregs and Curt,

I have tried with a better supply and it works. When I turn down the voltage I hear the relay open at around 10.88V as expected. When this happens I do not see any red watchdog LED. So it must be the interlock circuit.

When not connected, the supply from the amps provides around +10.9v and -11.1v. So it seems that when connecting to the PMAC, the load due to being connected to the sensors as well as analogue circuitry is sufficient to drop it low enough to trigger that interlock relay. The only reason I can think of why this happens on the new card and not the old card is perhaps the old card has a little wider tolerance.

Does that sound likely? Or is there a possibility that a PMAC hardware fault may be causing the voltage drop?

Thanks
Chris
Reply
#13
Hi Chris,

It does sound likely that the old card has a little wider tolerance and not likely that there is any hardware fault on the new card, considering that the amp voltage is already only about 0.02 volts above the minimum.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)