Log in to the online community

Want to post a reply? You'll need to log in
DC injection braking and a burned out transformer
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts
Question
I have been fettling a lathe which was fitted with DC injection braking. I was a bit unsure about the condition of the components in an enclosure on the back, but having looked up a bit about motor control circuits and having learned about ladder diagrams, I have satisfied myself that I know what has happened.

At some stage, a transformer (240/130V) has burned out. That leaves two questions: (1) why? and (2) what is the spec of the old one?

The horizontal surfaces of the enclosure and its contents have been covered by a thin brown layer, which I assume is vapourised enamel from the transformer windings, but there is no evidence of any further damage.

Father taught me never to replace a fuse without finding the fault which blew it. Along the same lines, I want to be satisfied that none of the other components has caused a short circuit. However, it appears that it is the primary which has overheated. So my first question is what would happen if the secondary of a transformer is shorted. Would it be fried? Woud the primary be fried? Or would both be fried?

My second difficulty is determining the spec of the transformer. The circuit diagram has "N" and "240" on the primary side, and "Z" and "130" on the secondary side. I assume that this means that the secondary provides 130 V, which after rectification, is fed to the motor. The problem is that there are no markings on the transformer. The manufacturer of the unit ceased trading 10+ years ago, so no help there. The motor plate specifies 6.4/3.7 A. The transformer wires (both sides) are about 4.2 sqmm CSA so good for 3 - 5 A (?). Based on a weight of 4.0 kg and the transformer's dimensions, and comparing them with current models; I am guessing that a 320 VA transformer is required. My second question is whether my above reasoning is sound.

Before anybody else says it, yes, I think that I could get a modern module which will do the job, but isn't it more satisfying to fix rather than replace?

eabde54935c495ab8feb4fb009758b5e-huge-20
 
30 Replies
Paul Skyrme 75887465
Joined 12/02/2003 - 134 Posts
To be honest Chris, the drive would be much kinder on the old motor than the DC injection, and may even give you a greater duty cycle.
As far as comms go, you'll need the hardware from Eaton, the comms cable is over £100+vat.
Though if you can figure it out it's a USB to RS485 unit, so if you have an RS485 port or card, or can make one to comply with the modbusrtu protocol then you're fine, make it up into an RJ45 and connect to your rs485 port.
If not you're a bit stuck, mind they keypad whilst long winded, is much cheaper, I've done loads of drives with keypad setup and it's a pain, but as long as you are meticulous then that's fine.
It's a one off so that's not so bad, if you had used a Siemens I could have lent you a keypad perhaps if you were close, I also have some Schneider comms stuff and a few others, but alas no Eaton.
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts
Epilogue

Spoke too soon!

I managed to find a replacement motor on good old fleabay, but it didn't solve the problem so there was no alternative but to replace the VSD. With a bit of ingenuity and some changes in the control circuitry, I fitted an Eaton DE11 into the old enclosure and preserved the original controls. What a difference! The motor is now very much quieter. The learning point here is had I known how quiet it should be, I would have realised that something was amiss. (I had a similar experience with an old Rolls-Royce, but that is another story.)

The only thing that I need to do now is to set the parameters for min and max speeds. Min = 0 is fine, but at very slow speeds it wouldn't take the skin off a rice pudding, or at least I can stop the spindle by grabbing hold of it. Problem is that I need to buy some proprietary hardware, either to set the parameters directly, or by their software which is available for nothing. Unless of course somebody in here knows how to speak to these devices (via the RJ45 socket and what is described as "The internal RS485 interface transmits Modbus RTU and for DE11... devices as well CANopen).

P.S. Injection brake is working well, although this device could do it too!
mapj1 80733779
Joined 22/07/2004 - 1910 Posts
sounds good.
'dropping out' could be an intermittent  high series resistance, and fits with the observed  improvement after the Brasso treatment. An open is unlikely to be mid winding - there you may expect a turn to turn short or perhaps a turn to core short, over time the buzz of loose windings rubs the enamel insulation right through. The modern stuff is coated with a thermally activated glue, so the windings  can be set solid once in place. (for the funny shapes of motor windings, they are usually wound on a cylindrical bobbin off the motor, then the winding  bundles are massaged into place, so the setting phase has to be the last step)
shorted turns are best found by comparing the inductance of the identical phase windings  the one with noticeably less inductance has shorted turns.

Well done on the perseverance. I now wonder if that local specialist has scrapped many serviceable motors.

I have had intermittent problems on a solenoid actuating a contactor where half heartedly prepared end of an enamelled wire went into an crimp lug and became a rather useless end to a winding - in that case a scrabble with the glass fibre brush to expose some copper and a touch of the  50W soldering iron put the problem to bed for ever.
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts
Update

I think that I have fixed it. 🤞

I took the motor to a local specialist who kindly bench tested it. They told me that the windings were "dropping out". They also told me that given the number of man hours required, it isn't cost-effective to rewind anything below 7.5 kW.

Before I binned the motor, I thought that I would check the resistance of the windings and IR test them. I couldn't get a stable resistance measurement. So I stripped down the little pile of washers and nuts on the terminals and cleaned them up with some fine sandpaper. Then I got nice stable readings (4.91 Ω per phase using both my MFT and a multimeter). IR was > 999 MΩ all round. I still couldn't find and signs of damage to the windings. So I thought that I would give it one more try. Much improved!

There is also a contactor involved and it's contacts are not making perfect contact, but I shall see how it goes 'cos it is not very accessible.

As for the DC injection brake, I have settled on a 10 Ω resistor on the DC side. This gives a breaking current of 8.8 A. With the chuck mounted, at a slow speed (300 rpm) it brings the spindle to a rapid stop; and at full speed (2000 rpm) it brings it to a stop in about 2 seconds (which, assuming linear deceleration, is 17 revolutions).

Thank you for everybody's contributions.
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts

Sailbadthesinner:
Unless you have three phase available you won't be able to plug the motor in DOL.  Even then you'll need to swap it to 400v (star) from 230v (delta).  

I just knew it wasn't going to be as simple as wiring up a plug and switching it on! I do have three phase, but I'd have to remove the motor and bolt it down 'cos that would be easier than shifting the whole lathe. Sometimes I wonder how we got it in place originally! 😄

I ordered a replacement capacitor (actually, minimum order 5) from RS at 17:50 last night, and they arrived at about 09:00 this morning. How do they do that? 🤔

Any road, today I have been rewiring a Victorian cottage, so I have no energy left for the electronics. 😞
Sailbadthesinner 81805338
Joined 13/11/2004 - 21 Posts
If it is connecting to the earth then it is likely to be a Y suppression capacitor.  VFDs are very noisy for EMC and will generally have a pair of Y caps and an X2 cap as a delta on the supply usually along with a common mode choke.  Not surprised it has failed in the last 30 years.  Given one good voltage surge on the mains and considering plenty of locations get hundreds of >1000v surges per year it's no surprise it's given up.  There is a slim chance failed suppression could be allowing noise to upset the VFD control circuit at certain frequencies.  

My starting point for old pcbs it to reflow (or hand touch) the soldering to eliminate cracked and dry joints, replace the electrolytics or test every capacitor. 75% of the time it then works fine.

Unless you have three phase available you won't be able to plug the motor in DOL.  Even then you'll need to swap it to 400v (star) from 230v (delta).  

 
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts

mapj1:
I'd be wary of that cap - what was it wired between ? Usually means it has flashed internally, producing hot gas, and cracking the case.

One end appears to go to nothing other than the earth connexion, which I don't quite understand.
 
I am beginning to think that it would be easier to update the electronics, but I much prefer repair to replacement whenever possible.
mapj1 80733779
Joined 22/07/2004 - 1910 Posts
I'd be wary of that cap - what was it wired between ? Usually means it has flashed internally, producing hot gas, and cracking the case.

it was one of these metalised paper devices, falling from favour nowadays    PME261EC6100KR30 is the original maker's part no.

 modern near equivalent  would be metalised polypropylene, this one is rated to 305VAC,  higher voltage types are available but not so common. Temp ratings not so hot.  

there are better modern metalised paper - price and large size reflects the lack of demand.
 
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts
Mike and Kelly, thank you for your help.

I have tested the three big capacitors, which are indeed in parallel, and they are within spec. They are 470 μF and 385 V. However, the capacitor below is clearly not happy so whilst the board was out, I thought that I should replace it. It gives a reading of 0.5 μF against a spec of 0.1 μF. Whether it is the cause of the problem remains to be established. I cannot see any other obvious faults.

I had thought of plugging in the motor direct. It would be a bit of a faff. Is it safe to plug it in and switch on?

be5ea5c38056a46421b6fd8e43e511c1-huge-20
 
mapj1 80733779
Joined 22/07/2004 - 1910 Posts
The caps may be losing capacitance, but unless  they are bulging alarmingly then it is no biggy to substitute a similar part or if that is hard then leave them in and add some more uF in parallel. However, It looks like they are in parallel, what is the value/ voltage ?

I have also had solder joints fail where fat legged components are held to the board, or fracture the track close to the solder blob, especially parts that run hot like the switching transistors. A bit of prodding with an insulated stick may reveal more - the old TV service technique of hitting things with the plastic handle of the screw driver to narrow down the are of a poor joint requires some nerve but can be revealing.
Beware of damaging those orange encapsulated hybrids that seem to be doing the HV isolating the drive to the transistors- under the goo  is a small substrate and very small components, and some models of the things can be as rare as hen's teeth.


IF you want to be sure it is the VFD, run the motor off 50Hz direct for a bit, or put 3 lamps on the VFD and see if they flicker when it misfires.
kellyselectric 11001191399
Joined 20/08/2016 - 69 Posts
That was the year I started there the date cone means 1989 week 43 we counted the weeks from January the first so 43 is around September ish at a guess
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts

kellyselectric:
It could well be a one or more of those caps failing they were made in 1989 if I read the datecode correctly  they look like the ones we made here in Weymouth at BHC components it I would be interested to know if they are from that manufacturer brings back not so happy memorys

Kelly, you are spot on. Yes BHC. If 8943 means 3 April 1989, it fits with the date of the lathe which is circa 1991.

Unfortunately RS says that this particular component has been discontinued.
kellyselectric 11001191399
Joined 20/08/2016 - 69 Posts
It could well be a one or more of those caps failing they were made in 1989 if I read the datecode correctly  they look like the ones we made here in Weymouth at BHC components it I would be interested to know if they are from that manufacturer brings back not so happy memorys
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts

mapj1:
Judder is not a normally a winding symptom on induction motors.

If it's not the motor, it must be the VSD. Come to think of it, the juddering is rather like an intermittent misfire in a car. Could one of those big capacitors be failing?
eb5645d9ed45225a0303181fe3358e7e-huge-20
 
mapj1 80733779
Joined 22/07/2004 - 1910 Posts
Interesting background, but as you say, not that helpful. I wonder how much of a 'development model' that brake really was.
Certainly cut the DC as soon as you no longer need it, any more just heats things and eats into the restarts per hour budget.
Judder is not a normally a winding symptom on induction motors.
The VSD may have certain critical speeds that give resonances I suppose.
 
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts

mapj1:
Glad you have restored it to operation, though it sounds like a bit of a beast, perhaps, given they could not draw a bridge rectifier correctly, the maker's datasheet was muddied with figures for a larger model. Thinking in terms of not cooking things, what stops the DC flowing  when it is all stationary and  no longer needed ?

Mike, thank you for your typically helpful contribution.

This was the company's only lathe which was fitted with DIB. Unfortunately, they went out of business 10 years ago, so no help available there.

A series of contactors ensure that the brake cannot be applied when the motor is drawing power, but that it is applied when the motor is switched off. Then there is a device which allows the braking current to flow for up to 20 seconds (I have it set to 2 sec) after which there is the reassuring clunk of a contactor opening.

It would be useful to know what happened in the past. I am a bit concerned by the motor. Sometimes it runs reasonably smoothly, sometimes it vibrates, and sometimes it has intermittent brief judders. I have replaced the bearings so I wonder whether the windings have deteriorated. Or could it perhaps be the variable speed drive?
mapj1 80733779
Joined 22/07/2004 - 1910 Posts
Glad you have restored it to operation, though it sounds like a bit of a beast, perhaps, given they could not draw a bridge rectifier correctly, the maker's datasheet was muddied with figures for a larger model. Thinking in terms of not cooking things, what stops the DC flowing  when it is all stationary and  no longer needed ?
The stopping force is proportional to the B field, in turn proportional to the DC current so yes  some peak current limit in the form of a resistance on the DC side would do. I'd be less sure of using any resistance in series with the transformer primary, as while the brake is not applied the transformer load is almost nil, so the voltage reduction is not very much at the instant the brakes come on, and yet may be too much once the braking current has started to flow. Equally having a sharp initial brake, and then a more gentle slow-down may be quite acceptable.

Or, as a quick try out, if you have 110v in the shop, how well does it work with that on the primary instead of 240? That would show you the effect of halving the voltage, and the voltage division would not be so load dependant. I might try a lamp dimmer, but not all designs of dimmer are happy on an inductive load. The simplest is resistance on the load side, and wire wound resistors of the metal clad kind are very forgiving of gross overloads (being little more than a heating element cast in ceramic in a metal tube ) Resistors that are a light evaporation of metal or other conductor on a ceramic rod (metal film / carbon film) tend to be self-fusing in overload, so I'd avoid those.

Estimating the resistor values may be a bit odd, as you may think that one looks at the voltage and the windings resistance and there are the amps, but the  winding is a bit of a complex thing, and really the current will build slowly though the inductance.  You may think you need 10A, and at 110V, a total of 11 ohms will do, but that is not the full story,  so some experiment may be needed. You  may want to design a dropper with a no of higher value resistors in parallel, both to allow  some adjustment, and to provide  some be robustness against any single failure.
 
MHRestorations 11001209997
Joined 03/03/2019 - 252 Posts
Maybe a reactive impedance (choke or the primary of another transformer in the primary ? Or even as it's not loaded very often... a heating element or high wattage lamp used as a ballast in the primary?
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts
I have stripped down the lathe, replaced bearings and seals, etc., and turned my attention back to the electrics this weekend.

I have replaced the transformer and the DIB works, but it is terribly fierce and brings the spindle to a dead stop almost instantaneously. (This brief clip shows a similar lathe stopping quickly - just before the end.)

The new transformer is rather bigger than the old one, so I may have over-egged the pudding, but would the lower resistance in the secondary (0.49 Ω as opposed to 1.16 Ω) make much difference? I selected a 230/110 V transformer and it is giving an output of 116 V today.

Then I wondered whether the wiring diagram might be wrong - was the output voltage in error? So I unwound the old transformer, which was a messy and tedious job with all the scorched paper and so on. The secondary had 123 turns of 21 SWG. The outer part of the primary had 29 + 55 + 268 turns of 28 SWG plus a further 352 turns of 25 SWG. It was this innermost part with the thicker wire that was connected. With 240 V input, that gives 84 V output, which is not consistent with the wiring diagram's figure of 130 V, but not grossly different from the new set up.

Looking online, I have found a few references which suggest that the DC current should be 2 - 2.5 x the full load current. The motor's plate gives 6.4/3.7 A at 240/415 V respectively. 6.4 x 2.5 = 16 A and in fact I measured the max current as 17 A over about 2 seconds when the brake is applied, so again, not grossly excessive.

Now what should I do? I think that the brake needs to be tamed somewhat. I don't want to go to the expense of buying another transformer if I can help it, but I was wondering whether to put something like a 4 Ω resistor in the braking circuit.

Sensible advice gratefully received.
mapj1 80733779
Joined 22/07/2004 - 1910 Posts
of side interest , there are other ways of slowing an AC machine, including to put a capacitance and a load  accross it so it becomes an induction generator - this is most effective at higher speeds, as the generated voltage is speed dependant, and DC injection is better at stopping the shaft dead when it is already slowing.
various patents have expired .
 DC braking 1958  dc we understand
combining the bet of two methods DC braking which unlike normal DC injection,  still works in a powercut.
 
 1948 auto generator DC breaking 
In accordance with the present invention, direct current for dynamic braking of the rotor 11 following interruption of the line circuits at the contacts 14 is derived from the kinetic energy of the rotor itself thereby avoiding the use of external power for dynamic braking.  
I suspect with modern polymer film capacitors for generation and electrolytics for the DC smoothing capacitor this has become a lot more practical now than it would have been when first proposed - a set of  60uF paper capacitors for half horse power at 220v phase to phase would have added a significant volume and cost.
Tony S 98502724
Joined 21/11/2006 - 23 Posts

Paul Skyrme:
Just remember that there is no requirement per se for a DC injection brake on a metal working machine.
In fact, I know of one example where by the clever persons fitted a brake and unscrewed the lathe chuck the first time they tried it!...
Also, note that if this is not a DIY project that there is legislation around this which may need to be complied with that you need to be careful of.
A DC injection brake is part of a safety function, maintenance is one thing, refurbishment is another, and requires careful consideration to ensure compliance.

You beat me to it. I’ve seen a lathe chuck come off, two brick walls didn’t stop it. It went in a nice graceful arc across the workshop before vanishing though the first wall.
 
Chris Pearson 11001208764
Joined 05/12/2018 - 1367 Posts

Paul Skyrme:
Just remember that there is no requirement per se for a DC injection brake on a metal working machine.
In fact, I know of one example where by the clever persons fitted a brake and unscrewed the lathe chuck the first time they tried it!...
Also, note that if this is not a DIY project that there is legislation around this which may need to be complied with that you need to be careful of.
A DC injection brake is part of a safety function, maintenance is one thing, refurbishment is another, and requires careful consideration to ensure compliance.

I take the point, but no risk of unscrewing the chuck. I have considered switching the supply to the brake. When in use, it would be energized, but before routine stopping it could be isolated so that the lathe could slow down gently. It would appear to be better for the drive and bearings, etc.

Of course the easiest approach would be to remove the whole thing, but having turned my attention to the purpose of the box on the back, I find myself driven to repair it. Isn't that what we do? 🙂
Paul Skyrme 75887465
Joined 12/02/2003 - 134 Posts
Just remember that there is no requirement per se for a DC injection brake on a metal working machine.
In fact, I know of one example where by the clever persons fitted a brake and unscrewed the lathe chuck the first time they tried it!...
Also, note that if this is not a DIY project that there is legislation around this which may need to be complied with that you need to be careful of.
A DC injection brake is part of a safety function, maintenance is one thing, refurbishment is another, and requires careful consideration to ensure compliance.
mapj1 80733779
Joined 22/07/2004 - 1910 Posts
Well spotted- the AC inputs should each see one kathode and one anode, and the positive output should be 2 kathodes, and the negative output should be 2 anodes.

The  data sheet for the rectifier indicates it is a 25 amp part, rated at  a mere 200v for reverse breakdown.  A rectifier rating of 25A forward feels like overkill, given what you said about the wire sizes, though that rating assumes you have it on a 25 degree C cold plate. However the 200V reverse, (noting a 110V RMS is more like 145v pk) seems to leave very little margin for the sort of 'flash and pop' surges often seen when contacts close on motor shuddering to a halt.  (the data sheet notes that 140V RMS is the limit for the 200V part.) However, if it passes the 'am I a diode' tests with the meter, then you are OK to keep it in there. (there are plenty of similar parts made of varying ratings  examples of bridges from the CPC catalogue. )


The current rating of wire when wound into a transformer is a funny function of the wire packing density, and its ability to cool via conduction through the rest of the winding, and the fact that the skin depth is far more serious than normal, because the magnetic field vs current is higher than the free space case. However the key thing is that the BS7671 wire rating values are miles off.
Luckily others have done the work for us , and there is a good reference for wire size versus cu rent in winding for low power 50Hz mains transformers here,  though in summary about 3-4 A per mm2 seems about right and if your wire sizes are  CSA in  mm2, then suggests only an amp or 2 for the windings you have. (I did say the BS7671 values were miles out for transformers didn't I..) This is more like 250VA, so not fitting that well with the 4 kilo weight. It may never have been fully optimised, if it predated CAD, that is quite likely.
Indeed if you are very brave, his "how to roll your own" article from the same website is very informative too see here.  Nowadays I only would only ever countenance this if I had a need to replace some special part for for some very odd voltage, as life is too short, and there are plenty of transformers in the catalogues, but I remember winding replacement transformer of a couple of hundred watts  with about ten secondaries to restore an oscilloscope many years ago, and it taking forever (probably about 2 weeks of evenings) just to wind the thing. I have used a lathe on freewheel as a manual winding machine, but even so it quickly gets painful, and hand winding never seems to manage to get quite as many turns in as the original manufactured item.


 
Sailbadthesinner 81805338
Joined 13/11/2004 - 21 Posts
I hope it was never wired as shown in the drawing.  The bridge is incorrect!  As shown it gives a short across the secondary for one half cycle.
It can't have been wired this way originally as it would fail immediately but if it was rewired to the drawing that would explain its demise.

Log in

Want to post a reply? You'll need to log in