Thread: The joy of having a TOS.... ( FNGJ32 mill )

Hi Steve, John, Guys,

Adding to what John and I said earlier. Those long wires to the diodes will radiate lots of interference that can be coupled to any other wires in that cabinet. We spoke about modern silicon diodes, I would simply add one directly across the coils and leave the existing one there. They won't interfere with each other unless one goes short circuit.

As far as "Twisted Pair" cables is concerned, you can use any Ethernet cable you have, they have four twisted pairs in them for exactly the same purpose, to reduce coupling between them. The other trick is to strip the woven braid from some co-axial cable and feed the Ethernet cable through it, and use that as a screen. But remember to only connect it to the earth at one end only, usually the head end. If you need to fasten it to the machine body with a clip, use a wrap of insulation under the clip.

I've ordered a handful of the 1N5408 diodes. Thought I might have had some in my supplies, but no luck (have about 20 other rectifier diodes of different ratings but nothing close).

I'm hearing you about the long distance to the current diodes, but the issue I've got with trying to put the diode directly across the feed clutch coils is that those coils are inside the feed gearbox.
I'd have to connect the diode to the connector on the brush housing, and find somewhere to earth the other end to (likely have to drill and tap the housing for that).
To be honest I'm also not convinced how the diodes will handle being bathed in oil all the time.
I could cut the distance to the diodes in half by putting them in the first layer of the cabinet where the timers and contactors are. I'd easily fit half a dozen of those 1n5408's on a small piece of board down in the bottom LH corner next to the earth.
Do you think there's any value in doing that or is there no point messing with them unless I'm going to go directly across the coils?

Made good progress on the VFD tonight. Have the FWD/REV working off an external switch, and a remote pot working for the frequency control.
What I found with the pot though, is that the 75Hz upper frequency limit I'd previously set and was working when changing the freq from the panel isn't active any more, and the max frequency has reverted back to 50Hz.
So the pot can vary the frequency from 0-50Hz.
I'm guessing there's a different set of limit parameters when the external pot is in use to when the control panel is in use.
I'll have to delve back into the manual and dig them up.

Steve

Hi Steve,

Thankyou for your comments re: diodes, I don't think that the oil splash would bother them at all, but the oil could migrate up the cables. If you can get them closer comfortably fine otherwise simply replace the existing one and see if you get any problems with contact burning or radiated interference later on.

I've ordered a handful of the 1N5408 diodes. Thought I might have had some in my supplies, but no luck (have about 20 other rectifier diodes of different ratings but nothing close).

The 1N5408, will be much faster than the existing diodes and probably a higher voltage rating as well. Do you have a part number for the existing ones ?

Re: VFD. Yes I agree that there is probably a different set of parameters for internal/external pots.

Sorted the parameter for the external pot limit frequency.

Existing diodes are p/n KY705F

Steve

Hi Steve,

Originally Posted by OxxAndBert

Sorted the parameter for the external pot limit frequency.

Existing diodes are p/n KY705F

Steve

Hmm, 700 volt 0.5 amp ! Czechoslovakian germanium. I would just replace it and see how it goes. I'll bet its OC.

Had a successful day getting the VFD working from the original control panel on the mill.

Removed the wires running from the spindle motor fuses to the contactors, and from the contactors to the overcurrent relay.
Connected the input supply for the VFD to the motor fuses and used the (now vacant) contacts on the FWD and REV motor contactors to switch the VFD FWD/REV.
All good, and thanks again to Shedhappens for suggesting to do it that way rather than running new wires to the control panel.

I had to pull the back off the contactor cabinet to be able to remove the wires as they were clipped in:



I still need to run some new cable out to the control panel for the speed control pot.
Had a scratch around in my wire box and found some nice instrument/control wire that I'd salvaged a while back. Brand new (still in its plastic bag) German made 5 core + shield, 5m long, with a very nice little plug on one end.
Should be perfect.
Pity to have to cut the plug off, but still better than the whole thing ending up in landfill.



Also had a look at where I can mount the transducer for the spindle tacho. Unfortunately the only thing protruding from the top of the spindle housing is the quill spline, so the only real option I can see is under the spindle housing, with the signal magnet just above the drive dogs.
If the transducer is mounted at the 1.30 o'clock position when looking from above it should be reasonably out of the way.
I can mount the display on top of the head or ram - will see which suits best.



Steve

New relay for the dodgy timer arrived today. Fitted it to the timer and powered it up on the bench to test.
Tick, tick, tick.....
Not what I wanted to hear. Whatever should be latching it - isn't
Any suggestions on how to troubleshoot it further? Guessing it needs me to finish off drawing up the circuit diagram I started on last week....??

Steve

Originally Posted by OxxAndBert

New relay for the dodgy timer arrived today. Fitted it to the timer and powered it up on the bench to test.
Tick, tick, tick.....
Not what I wanted to hear. Whatever should be latching it - isn't
Any suggestions on how to troubleshoot it further? Guessing it needs me to finish off drawing up the circuit diagram I started on last week....??

Steve

Shooting in the dark here, if the ticks are .8 seconds apart it might just rely on an electrolytic capacitor hooked up between the
+ and - to keep the relay coil energized between the pulses.

Originally Posted by shedhappens

Shooting in the dark here, if the ticks are .8 seconds apart it might just rely on an electrolytic capacitor hooked up between the
+ and - to keep the relay coil energized between the pulses.

Good thought.
No obvious electrolytics at all, and only a few caps total. I probably have enough junk around here to remove and replace them all one at a time if it comes down to it.
I'll do some work on the circuit diagram - hopefully its to someone electronics-savvy once its drawn up.

Steve

Hi Guys,

Originally Posted by shedhappens

Shooting in the dark here, if the ticks are .8 seconds apart it might just rely on an electrolytic capacitor hooked up between the
+ and - to keep the relay coil energized between the pulses.

That would be one way of slugging the timing !

My original thought was that one of the contacts on the relay might have been used to latch it at the start of the time period. But since all the contacts were wired in parallel and a new relay with a single heavy contact doesn't latch, but swapping the timer produces the correct behaviour, the fault has to be in the circuit of the timer.

Without being able to look at the circuit or the board it is hard to tell. Certainly any capacitors in there would warrant examination.

OK guys, this is what I've managed to come up with for a circuit diagram.
Hopefully its accurate, but if you understand this sort of magic and tell me that it can't possibly work because of X then I'll double check that part again.

Note that the timer doesn't actually run until pins 3 and 4 are connected.
I've omitted the 110:22v transformer that supplies the bridge rectifier.

TX100 timer.png

Steve

Hi Steve,

Carefully check D1, then the capacitors C2, C3, C4. I'm not sure your circuit is correct, because D1 feeds back the capacitor voltage on C2, C3 and C4 to keep Q2 turned on via the diode.

Opening pins 3 and 4 breaks the supply. The delay time is set by the value of the pot and C2, C3 and C4.

I've redrawn your circuit, inside the red circle is where it gets weird. Both transistor bases are shown tied together.

TX100.png

The transistor driving the relay, I'm not sure whether the relay should pull in immediately and drop out after the delay or pull in after the delay and hold in. I think it should be the latter, which would also explain the switch on 3 and 4 used to reset the timer.

Thanks John - much appreciated and heaps clearer than mine!!
I'd like to blame the late hour when I did it, but truth is electronic circuits brick wall my brain....

Definitely the later - the relay should pull and latch when the timer expires.

Q1 is KC307 PNP https://alltransistors.com/transisto...ansistor=37186
Q2 is KC507 NPN https://alltransistors.com/transisto...ansistor=37195
Q3 is KT509 NPN https://alltransistors.com/transisto...ansistor=38893

Striped end of D1 definitely goes to pin 3 on Q1 (emitter?)

Yes - I've screwed up my connections onto Q2
Q2 base is connected to Q1 collector
Q2 collector is connected to trim pot and C4

Here's a couple of photos that might help clarify. The one from the top of the board has backlight do you can see the tracks through the board.



Steve

Good Morning Steve,

That makes more sense ! A transistor version of a thyristor.

The three capacitors combine to provide the delay. If D1 is OK and I think it is, then the two transistors are suspect. I think one has gone short. I would replace both with modern silicon ones. BCxxx series will work OK. You may also find that the timing constant changes as well. I'll do more analysis later. Honey Do time !

how-thyristor-works.gif
A little animated picture should help !

Removed D1. No ID markings on it other than the blue cathode band.
Tests OK on the meter but I might as well replace it. I've got a bunch of 1N4004 diodes here that I'm guessing would be OK.

C2 is 2u2 40v and its got one leg marked with + (didn't expect that)
C3 is marked 1M 40v - is that 1uF ??

EDIT: Looks like they are both tantalum type, so both polarised - I can *just* make out the + on C3.

C4 looks like 2/35 - here's a photo of it.



Steve