Spot the connection

[Steve PaveyParticipant @stevepavey65865]

Judging from what you've written in the above posts I think you're right, I will need some help with the wiring.

I've just checked the coils and I get 3.2 ohms for each of the coils.

[Mark CParticipant @markc]

That sounds to be ball park to me. More important is that they are all the same.

Mark

[Neil WyattModerator @neilwyatt]

What make of inverter?

If you go for an IMO one, see here for some ideas on wiring up.

Neil

[Mark CParticipant @markc]

Neil,

I use all sorts including the cheap import ones. Probably my favorite is the small GS1 offered by Lamonde Automation, they are a good price, very compact and reliable.

Mark

[KWILParticipant @kwil]

Lamonde Automotion are also a good source of push button switches if you are making up your own control boxes.

[Neil WyattModerator @neilwyatt]

Posted by Mark C on 06/10/2015 14:48:04:

Neil,

I use all sorts including the cheap import ones. Probably my favorite is the small GS1 offered by Lamonde Automation, they are a good price, very compact and reliable.

My question was aimed at Steve!

Neil

[Mark CParticipant @markc]

Yes, I knew that – I was just adding a spurious comment in case Bill or anyone else was interested! Perhaps I could have worded it better?

Mark

[Steve PaveyParticipant @stevepavey65865]

I now have 6 wires appearing out of the connection box, all the new wiring connected, sleeved and tied in. The connection box itself has three terminal posts, and I have labelled the three original wires 1, 2 and 3. The other ends of each of the coils are labelled 1B, 2B and 3B respectively, and these are the ends which were originally connected together to form the star point.

Presumably I can now connect the wires up to the three terminal posts as follows: 1 and 3B together, 3 and 2B together and 2 and 1B together, my thinking being that if one or two of the coils are reversed there won't be 120º between the phases.

As to the inverter Neil, I have yet to buy one. I keep looking on eBay and see models by Huanyuang at around £100, compared to around the £250 mark for a known name like Hitachi or Mitsubishi. I have no idea which to go for.

[Mark CParticipant @markc]

Steve,

Sounds good to me!

At 3 HP you are going to be spending a bit more than the normal 1HP drives cost. I have a 4kW Huanyang from ebay. The cost for anything else was well over 300 quid for one this size new. I would be happy to buy another with the reservation that you will probably "need" rather than "want" a drive reactor as I mentioned. I get mine from Lamonde as they are the only place I have found that sell the correct spec reactor for the application (voltage and power) at a sensible price. the only down side to these inverters is the size which – they are bigger than comparable "named" drives. If you are sticking it on the wall up in the air that should make little difference.

They are also easy to set up if you get some help – Along with others who have been down this route, I would be happy to assist with the parameters that you need to change for a lathe/mill installation.

Mark

[MuzzerParticipant @muzzer]

Nice looking Yaskawa V1000 2.2-3.0kW drive for about £200 inc vat and delivery?

[Steve PaveyParticipant @stevepavey65865]

Never heard of a load reactor before – is it to iron out the peaks or to provide some sort of suppression?

Also I've seen some people mention that I might need a vfd with a higher rating than the motor size. Is this true?

Thanks for the Lamonde link, which may come in handy in a week or two.

[Mark CParticipant @markc]

Muzzers prices are better than I could manage – wish I had found that outfit a while back (I have spent quite a bit of money on drives for my products).

You only need a drive rated for your application, you can always use one with a higher power output and use it de-rated should you choose.

Drive reactor (choke) produces an output that is closer to the mains than the HF output from the inverter. You need to look at the output with an oscilloscope to see and appreciate the difference – it is a drastic improvement over the unfiltered output. If you get interference up the line onto the ring main, you will also want a filter on the input side but they are much cheaper.

Mark

[MuzzerParticipant @muzzer]

What matters to a motor is the current in the windings which ultimately is what generates the torque. If you are going to look at the motor with a scope it's more instructive to use a current transformer to observe the phase current than to look at the PWM voltage. The motor inductance filters out the high frequencies so that the current is highly sinusoidal – often more sinusoidal than when the motor is connected directly on line. The mains waveform is often fairly distorted.

The main benefit of a filter choke on the output of the inverter is to reduce the interference due to the voltage transitions resulting from the PWM waveform. The voltage transitions on a single phase input VFD are typically of the order of 340V in a few hundreds of nanoseconds. This can generate significant recirculating currents in the motor wiring. After the filter choke (where fitted), the voltage edges are almost eliminated – at a cost obviously. Keeping the wiring short and using shielded cables can help too.

Murray

[Mark CParticipant @markc]

Murray, I don't know about the AC theory involved particularly but I can speak from practical experience that once I fitted a choke all my problems with analogue signals went away and looking at the screen, it went from a mass of traces all over the screen to a nice tight steady voltage trace with a few volts of ripple. I did notice that the signal (once smoothed through the choke) showed a distorted form from a sinusoidal trace – it seemed to be pulled down on the falling side of the trace, equal on all three phases? Is this what you meant when you say the motor can distort the line voltages?

Mark

[MuzzerParticipant @muzzer]

No, what I meant was that the waveform generated by VFDs is generally very sinusoidal, as it's digitally generated. In contrast, when you look at the mains voltage, it's often distorted, particularly at the crest, due to the non-linear loads connected to it. A lot of equipment only takes power from the crest of the waveform, leading to a flat top there. Doesn't affect the motor power really. You'd have to post a picture of what you meant to understand better what you mean.

The output filter significantly reduces all the switching frequency edges, so it's not surprising you have got much cleaner analogue voltages. If you imagine being one of the analogue inputs and trying to figure out what the "real" signal is amongst all the noise, you can understand why filtering the noise at source is the best fix. Engineers often think they can simply filter out noise on the inputs with low pass filters and / or software filtering (averaging etc) but it generally doesn't work like that.

My background is power electronics, so designing with this mixture of noisy power and quiet analogue signals is what it is all about for me. Bottom line is that each installation will be different and must be commissioned carefully to ensure it works as intended.

Sounds as if you've got a good, solid setup on your systems. Hopefully this will avoid strange unintended behaviour later on!

Murray

[john fletcher 1Participant @johnfletcher1]

After reading the above regarding interference, I wonder if one of those input filters as fitted to washing machine would be of help. After all they are readily available and appear to be well made. I have fixed several front loaders for the family and have never seen a faulty one. Just a though.John

[Mark CParticipant @markc]

John, the input side filters are a good idea according to the drive manufacturers but they did nothing for the output noise I had problems with. The main advantage as I understand it is preventing problems with other items connected to the same ring main, perhaps Murray can expand?

Murray, "filter out noise on the inputs" – guilty as charged! Problem I always seem to have is feedback loops that require psychic ability. Its also hard a lot of the time to be able to change the circumstances so you have to go the filter route rather than fixing the problem at source but then I guess you know that already.

To explain about psychic feedback, I am talking about dynamic control (mostly speed control in response to some process variable) which involves trying to maintain a steady state in the process. Often, responding to a change is too slow and you have to try and anticipate how much change is needed before it happens – an operator can judge and anticipate but electronics (PLC or industrial PC's) are not so clever – yet!

In my case, the process would best be controlled by responding to the magnitude of a specific frequency out of a "spectrum analysis" output but that wont work so we have another work around that seems to work for the job.

(For Murray and anyone else daft enough to be interested; I actually tried to use an FFT implementation on an Arduino using "n" buckets and using the first as the PV for speed control but it generates a signal that is always lagging "n" number of machine cycles. I ended up programming the PLC to perform a rolling average of the output and look for a change greater than a preset magnitude to indicate the degree of speed change to make and just accept a severely damped response)

Mark

Mark

[Steve PaveyParticipant @stevepavey65865]

Still waiting for the Huanwang VFD to arrive, and in the meantime I'm sorting out the original wiring. A couple of questions spring to mind:

1. The DOL starter on the lathe is an MEM ads7, probably a replacement as it is mounted in its own enclosure and there are signs that the original starter was a panel mounted one. In any case it obviously has a 415v coil and a 3.7-5.6A overload so is not going to be a lot of use to me. Do I need a new DOL starter wired into the supply to the VFD, or is there sufficient overload and no-volt protection provided by the VFD itself?

2. The original (rotary) switches on the lathe (a main on/off, a FWD/OFF/REV and an ON/OFF for the coolant pump are all good quality and positioned in sensible places. They are what I would call latching, as opposed to momentary, and I'm hoping that they are suitable to connect to the VFD as remote controls.Can someone confirm? I think the answer is yes from what Mark C has written in his post at the top of page 2 but I'd sooner ask another dumb question than risk damaging the VFD.

ps I have downloaded the manual and am slowly ploughing my way through it, but as everyone has said in previous posts on this topic, the writing takes a bit of getting used to!

[Neil WyattModerator @neilwyatt]

Any switch should be OK, but whether its momentary or latching means different programming required.

A down side of big switches carrying tiny control currents is they are prone to poor contact, compared to small signal switches, which can interfere with operation.

Neil

[Mark CParticipant @markc]

Steve, As far as I can tell there are no such thing a stupid questions, just stupid people who live in ignorance rather than ask the question!

Bearing in mind that it is my private workshop, I do not use any switching in front of the drive (mains side) other than the main isolator switch. I do have an e stop wired in (no volt release) on the input control wiring/switching. If you are bothered you can use an interlocking switch for this but the inverter does this function if you set the input logic up properly. In reality, if you want a true e stop function you would need a relay in front of the motor as the inverter will continue to run while it still has DC bus voltage to do so and this takes a few seconds to drop off as the charge dissipates from the capacitors. As a point of caution, an e stop disconnection will cause the motor to coast to stop but it might also require you to buy a new inverter if the output throws the towel in as a result of the (suddenly) open circuit output.

I will take a note of the settings on my inverter (same as the one you have ordered – all but the size at least) and post them for you. I will bear in mind that your switching will be latching rather than momentary.

Mark

[Mark CParticipant @markc]

Neil,

You are correct of course, the answer is to give them the occasional squirt with contact cleaner/lube. I had one of my other inverters swapped over thinking there was a problem only to find that the rotary switch was playing up! I think there might be a saying about hasty action?

Mark

[MuzzerParticipant @muzzer]

The e-stop input on VFDs is usually a simple hardware override that inhibits the switching devices. That's in case the software or the control inputs have a wobbly. Generally you don't plan to cater for 2 faults happening together, an approach that is taken in industrial, automotive and other environments – look up IEC 61508 or ISO 26262 if you need to get to sleep.

As you say, some (many) VFDs will get upset if you disconnect the motor from the output when it is getting busy with the motor. It won't come to rest any quicker and you are probably better off using the regen / braking resistor / DC current injection functions to stop it quickly.

Industrial sensorless VFDs aren't really designed for closed loop speed control. No idea what frequency response you will find if you characterise yours but as you know conventionally you would use a servo drive which will generally contain a tunable PID controller and is designed for that application.

Murray

[John RuddParticipant @johnrudd16576]

My HY inverter on my Chester Eagle mill, comes to an almost immediate stop when the primary supply is removed…,Not sure if thats the dc bus decaying fast or something else….

Edited By John Rudd on 08/10/2015 20:39:37

[Mark CParticipant @markc]

Murray,

Deciphering the Chinglish took me long enough (and that was given I am reasonably conversant with the more mainstream makes, so I had an idea of what I was looking for). trying to find out if there was a suitable hardware e stop function was not top of my priority list by then! As for meeting IEC directives, they couldn't manage a working implementation of ASI bus, I don't know how that bodes for safety items but I think I can guess. Having said all that, the drives work OK and at bthe price and given the job – that was all I wanted.

As an aside, many older machines found in our caves have screw fitting spindles (thinking lathes here) and sudden stops can be problematic with these.

On a more technical note, the drives on my recent machines (production jobs) require tight speed control (and good response) as the feedback loop is very damped (can't do much about it). I am getting very good results using sensor-less vector drives with them and they do have good(read flexible) PID routines but I am not using the drive PID's relying instead on my own routines tuned to the process and implemented in the PLc. My experience of drives has been the increasing replacement of servo drives by vector drive AC motor combinations probably due to their rugged/robust and economic offerings.

John, I have never had occasion to shut one of my drives down in that way so I don't know how long they will run on when driving a load but they take ages to die when you pull the plug at the end of the day!

Mark

[Neil WyattModerator @neilwyatt]

Posted by Mark C on 08/10/2015 21:10:26:

John, I have never had occasion to shut one of my drives down in that way so I don't know how long they will run on when driving a load but they take ages to die when you pull the plug at the end of the day!

Interestingly, my inverter monitors the health of the bus capacitors and tells you when they need replacing (hopefully in many year's time at hobby duty cycles).

Neil

[Author]

Posts