Oh Blimey I bought a CNC Lathe !!!!

[Pete.]

One thing I always wondered is - when you are using this test equipment to probe circuitry, how do you know that a) the power you're using isn't going to damage something and b) your test doesn't find another path and give a false result.

Yes, I'm an electrically/electronically-challenged noob. I had to look up resistor circuits yesterday

[awemawson]

Well you choose your probing technique to suit the object you are working on. So when fault finding on a 24v dc relay logic set up a 24v bulb is suitable, but when probing delicate electronics, a high impedance oscilloscope probe is more appropriate and remember to take anti-static precautions. Touch an earth before touching any pcb you are working on to discharge any static on you. When doing it commercially I'd wear a wrist band connected to ground via a 1 Meg Ohm resistor and work on an anti-static mat.

As for the other path issue - yes it happens and you need to be aware of it. If you are lucky and have an accurate circuit diagram you can predict the other paths. The circuit that I have for this CNC Lathe is 'largely correct' but I've found a few discrepancies as when it was originally sold there were all sorts of feature options, and different safety options for different countries. Also it had a bar feeder fitted originally, but I don't have it, nor do I have it's circuit so undoubtedly there will be issues there as it was interlocked with the lathes safety features.

[awemawson]

Today started with a bang. Literally 

Went to the workshop to resume chasing relay faults, turned on the power and there was a crack /  bang like a fuse blowing. Turns out that the dual servo drive for the X & Z axis had blown up  Looking at it it was showing no signs of life. 3 phase going in, no supplies on the board. Here it is in situ:

[awemawson]

Obviously going to have to get it out and on the bench - loads of connections:

[awemawson]

So got it on the bench and need to pull it to pieces for a very close examination:

[awemawson]

I was expecting to find blown up power components in the input side but everything looked (and smelled - don't over look the nose in diagnosis !) ok - It's the usual input stage of the three phases coming in to a bridge rectifier and charging up a whacking great reservoir capacitor, with inrush current limiting by having power resistors in series with the three phases, shorted out a second or so after power on. However all looked well to my trusty Avo!

Then I started looking at the driver board and found this:

[awemawson]

There is a bit of pcb track blown open circuit, and a flash mark next to a 180v working electrolytic capacitor. My guess is that the capacitor has gone short circuit blowing the track but yet to prove that theory as it's time to feed the livestock.

[vtsteam]

Today started with a bang. Literally 

                

[Pete.]

Didn't you say before that capacitors that have been dead a long time have to be started slowly to 're-establish' them or something?

[awemawson]

Yes Pete, but this one had been powered up for a long time so I don't think it was a re - forming issue. Anyway the capacitor would seem to be innocent of all charges, not only does that track not connect to it, it tests out ok. This section of the circuit is as far as I can tell (by staring at it !) is a switched mode power supply generating all the internally needed voltage rails for the logic and to drive the 'H Bridge' driver circuits. However they are notoriously problematic to fault find - particularly without a circuit diagram and this one has Meldas proprietary lsi modules on it 

If I wanted a divorce I could always buy this one:

http://www.ebay.co.uk/itm/360637652653?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

(link to eBay item of similar servo drive for £3,538.18 for those reading after the ebay item expires)

[Fred Bloggs]

Awemawson

I have been following this thread with interest and as an electrical/electronics engineer with 30 years tinkering professionally, I take my hat of to you sir  as you've got more balls than I have to buy 4 plus tons of gear which keeps kicking back at you  (If my SWMBO copped wind of a toy like that papers would be served within 24hours).

On a more serious note might I suggest that you rip all the 'ebay'able electronics out - flog them and then use the cash generated to buy some huge stepper motors and associated control gear, at least that way you understand how it would work. although there would be a lot of work fitting the motors

Good luck

Fred

[awemawson]

Thanks Fred.

Started in earnest trying to find WHY that track burnt out. It runs from one pin of the transformer so I decided that I needed to pull the transformer so I could hopefully work out its windings and thus the circuitry around the blown track

A/ Transformer in way - remove it!
B/ Clear Board where it was
C/ Transformer underside
D/ That blown track
E/ Equipment used to remove the transformer
F/Transformer Diagram

[awemawson]

Removing the transformer was a bit of a pig, as there was masses of solder on the component side that I couldn't get at. Ended up using the "SMD removal" low melting point solder to dissolve it and sucked out ok then! Turns out a single ended winding on the transformer (which presumably is driven as an oscillator at a few kHz)  is being used as a very simple voltage supply as per diagram below. I cannot for the life of me see how this circuit can produce sufficient current to blow the track, as all other components in the cct are ok.

[awemawson]

To prove the circuit I lashed in a lab supply in place of the transformer, which allowed me to measure that the (unmarked) zener diode is a 5v one (probably 4.7 actually)

SO HOW DID THE TRACK BLOW?????

A/ Lash up to power it
B/ Lab supply

[Fred Bloggs]

Sir

Just noticed your comments about what caused the track to blow & how!!

My pennys worth "Is the transformer primary connected to a high voltage or is the transformer only to isolate grounds" either way have you tested the transformer at high voltage (Flash or 'Megger' insulation resistance test) .

Also can you test the transformer primary drive electronics with a load resistance in place of the primary and use a scope with Differential probes to observe whats going on??

Best regards

Fred

PS you've probably done all this but sometimes a different pair of eyes can sometimes spot the obvious (more to the point I'm not worrying about divorce papers )

[awemawson]

Fred, as far as I can tell, remembering I have no circuits of this AC servo drive, two of the three phases are brought into this board, rectified, and chopped, the chopper / oscillator using this transformer as a load (and possibly feedback on one winding). As you can see from its size its a high frequency transformer with I suspect a ferrite core rather than iron. I agree Meggering it is a good idea that I hadn't thought of as presumably the primary is at a few hundred volts.

[Fred Bloggs]

Sir

As I thought the Power supply is to generate on board LV DC supplies and as such the DC level on the primary on the transformer will be floating around 325 volts (SQR 2 (1.414) x 230V AC)  Be careful!!!

You could wire in an external supply to power the on board electronics and power it all up on the bench first then fit the board back into the system to see if the rest of the system works okay!!- Either way a bit of a risk, I have done it a few times myself.

I have fitted supplies to equipment before, got it to work and decided to leave well alone as it did the job, this included doing a very quick rush job were I left the 0-30V DC Thandar lab supply inside the equipment as they wanted to start production within that day, finally got the supply back when the company shut the line down and moved production abroad 3 odd years later

Best Regards

Fred

[lordedmond]

yep been there done that

most risky one was a lash up supply for 80vdc ( dont ask why the funny volt try the manufacture instead ) all it was doing was to supply the sync panel for the 441hz supply for a number of IBM mainframes for a large UK bank it ran for a year , the risk was the downtime £40 million every five mins 


Fred I do doubt that the logic dc volt would be floating that would be very dodgy indeed any disconnected plugs around the system would float up  BTW i though it would be root 3 (1.732) ?

Anyway back to the plot good luck with sorting what is becoming a rats nest of problems with no drawing and prop parts Sir you have my condolences


Stuart

[awemawson]

The machine internally works on 100v AC - there is a three phase transformer bringing my 415v three phase down to 100v three phase. This particular board on the AC servo drive pick off two of the three phases to use as it's source - so probably about 170 volts phase to phase. When I've plotted out more of the circuit of the switched mode converter I'm going to power it from a Variac and bring the volts up slowly as I watch & measure.

Fred: I megger'd the transformer winding to winding with my (ex GPO - shows how old it is!) 500v megger and confirmed that there is no leakage - it was a good idea though - thanks.

This little bit of the circuit is generating quite a few independent supplies some of which I suspect are used on the drivers for the 'H Bridge' output power FETs and as such need to float as the H bridges (two of them as there are two AC servos in the unit) are powered by the rectified 100v three phase. Not really practical to supply external replacements as I don't know what they should be anyway!

I must spend some time this evening plotting out more of the circuit - I got diverted this afternoon sorting out my poor dog that had to be spayed this morning and dismantling the penning in the lambing shed and converting it to a shearing shed ready for when the sheep have their next haircut. Don't you find life gets in the way of doing what you want to do!

(anyone want some sheeps wool?)

[Fred Bloggs]

Stuart

SQR of 3 is used when working with 3 phase power systems.

For example with a star connected 400 volt phase to phase system the phase to neutral voltage (single phase) is 400/(SQR3) = 230volts

 

SQR2 is the value used to convert a measured AC sinewave voltage (RMS or root mean square) value to its peak especially when you are charging capacitors up. The peak voltage of the sinewave will be present across the smoothing capacitors, the voltage measured will fall as load is put onto the circuit and will have an increase AC ripple component as the capacitor is discharged more heavily. That is why you have to use some form of electronic regulator (either series or shunt regulators or switch mode regulators) when you power up electronics via mains transformers with rectifiers and filtering/smoothing capacitors.

Sorry ladies and gentlemen of the forum if my explaination is not very good - I'm one of lifes doerrs not teachers

Best regards

Fred

[Fred Bloggs]

Opps

As I was typing awemawson commented!!

Fred

[lordedmond]

Sorry Fred

guess I worked with 3 phase to much

I will get under my blanket and let Andrew continue with his epic adventure


Stuart

[awemawson]

And it seems I was wrong - it's 200v 3 phase running around so the input will be about 340v I suppose. Just has a quick look again and the input from two of the phases goes through the usual common mode rejection filter, a 3.2 amp fuse (which is ok) to the bridge rectifier and across the large electrolytic capacitor. There then sadly is a custom hybrid thick film potted circuit so the actual functions gets hazy, but there is a 2SK847 power fet in series with one winding that I presume is the excitation for the transformer. I've dug out my ancient open frame variac, powered it up and checked that it works but now called back in on dog sitting duty while the wife goes out selling eggs! 

[John Stevenson]

but now called back in on dog sitting duty while the wife goes out selling eggs!

Jesus H Christ on a push bike.
Tie the damn mutt to the chip conveyor and get on with the damn thing.......................

[awemawson]

John, your soft side is showing again .

Actually she's still totally zonked from the operation (the dog not the wife) so best not left on her own until she comes round poor thing.