PDA

View Full Version : Help me to choose transformer



Gytis
28-06-2019, 05:02 PM
Hello guys,

Can someone check out if I correctly sized transformer for the drivers and motors (4axis)?
and do I need inrush current limiter for that size transformer ?

Transformer I chose is - 1000VA 230v to 2x36v 13,9amp
airlinktransformers.com/product/chassis-mounting-toroidal-transformer-standard-range-cm1000236

Drivers-LCDA808H
Voltage input range: 50V ~ 80VAC
Maximum peak current: 8A
servo.xlichuan.com/prod_view.aspx?TypeId=70&Id=167&FId=t3:70:3

Motors-LC86H298
nema 34
6,5 nm
6 amp
4,3 mh
servo.xlichuan.com/prod_view.aspx?TypeId=75&Id=191&FId=t3:75:3

and in the end, if someone thinking about LICHUAN
motors & drivers then check themout on
www.ebay.com/str/Shenzhen-XinLiChuan-Electric?_trksid=p2047675.l2563
I saved 171€, compared to AliExpress

Thank You for your time
G.

Doddy
28-06-2019, 05:30 PM
To simply answer your question and not get dogged with the stepper/driver selection. For a 1kVA toroidal transformer I would seriously recommend some form of inrush protection. There's various solutions for this from the simple to the less-simple. I prefer the NTC thermistors which are designed for this application, others on here will offer other solutions.

I would question why you're going for a 36+36 secondary, for the same price and range of transformer you can get a 70+70 secondary option. This would allow you to have two independent split supplies (rather than wiring the two 36+36 in series). It's a subtlety sometimes lost that it will help manage large instantaneous switching currents if all four axis step at once, for a given wiring solution for the control box. Your copper losses increase as a square of the current so reducing the overall current through the PSU is generally a "good thing". It costs you an additional rectifier and you can then split the smoothing capacitors across two supplies rather than bulking up on one supply, then try to intelligently mix the supplies for the four axis depending on your anticipated use. For me I split X and Y across two supplies (as the two most active axis) then marry X with Z and Y with A. How effective this is, I've not (nor will) done the analysis, but it makes some sense to my fuddled brain.

Gytis
28-06-2019, 05:52 PM
Hi and thank you for your input Doddy

but those drivers are AC not DC..,
or I don't understand here something?

Doddy
28-06-2019, 05:57 PM
Good point well made, but the argument still stands.

Another point - the drivers typically have their own onboard rectifiers (they are DC-driven devices even with a AC supply), if you look at the physical size of the drivers vs a PSU supply capacitor then you can expect that there is a compromise with the onboard rectification (that needs to be balanced against the losses introduced by the additional rectifier). But, yes, you can still opt to use the AC supply straight from the toroidal. In that case I'd chose to wire each driver straight back to the transformer, and not to daisy-chain them.

Gytis
28-06-2019, 06:25 PM
thank you for clarification,

I was more worrying about Current, if I understand it right then
4 steppers requiring 16 amps, 1200 VA transformer can provide 16,7 amp but I guess it overkill a litle bit or no?

routercnc
29-06-2019, 11:22 AM
I am running 5 NEMA 23 motors from a 750VA toroid and there is a fair bit of inrush when it switches on (lights dim momentarily). I needed a 10A type D MCB in the end to avoid it tripping out.
1000 or 1200 VA is going to be interesting.

Do you need that much as the steppers do not take anything like the rated current added together

Gytis
29-06-2019, 12:55 PM
Thank you for you're input,
I will take your advice and begin with a 750VA toroid, I can always upgrade if it will be necessary.

btw. you have an impressive machine and great vidoes!

routercnc
30-06-2019, 03:54 PM
Thank you. Had fun making the machine and videos and the best thing was there was great feedback and it has helped give ideas to other builders.

On the toroid sizing question- I’m only repeating what others have said on here in the past and the Gecko website but the guide is 1/3 x number of motors x maximum current. So 1/3 x 8A x 4 =9.6 A. I think I would start somewhere there.

Sometimes people use a larger toroid for future proofing, depends if that needs to be considered. I suspect pulling small currents from very large toroids is not very efficient.

Neale
06-07-2019, 10:41 AM
Good point well made, but the argument still stands.

Another point - the drivers typically have their own onboard rectifiers (they are DC-driven devices even with a AC supply), if you look at the physical size of the drivers vs a PSU supply capacitor then you can expect that there is a compromise with the onboard rectification (that needs to be balanced against the losses introduced by the additional rectifier). But, yes, you can still opt to use the AC supply straight from the toroidal. In that case I'd chose to wire each driver straight back to the transformer, and not to daisy-chain them.

It's certainly true that onboard capacitors are probably smaller than those that someone building their own PSU would use. However, we are talking stepper drivers here, not audio amplifiers, and a bit of PSU mains frequency ripple really makes no practical difference. I've measured around 10V ripple on my own PSU with the machine running a test toolpath and it really doesn't seem to worry. Maybe you'll lose a tiny bit of the torque theoretically available, but that's about all. The stepper driver itself is chopping the motor output current at a much higher frequency, and it will adjust dynamically to whatever the input voltage happens to be. After all, even with the ripple, the supply voltage still stays within the nominal working range.

As Doddy says, though, try not to daisy-chain them and wire each driver back to the transformer independently. That makes sure that any voltage drop in the wires to one driver does not affect the others.

Doddy
06-07-2019, 10:46 AM
It's certainly true that onboard capacitors are probably smaller than those that someone building their own PSU would use. However, we are talking stepper drivers here, not audio amplifiers, and a bit of PSU mains frequency ripple really makes no practical difference. I've measured around 10V ripple on my own PSU with the machine running a test toolpath and it really doesn't seem to worry. Maybe you'll lose a tiny bit of the torque theoretically available, but that's about all. The stepper driver itself is chopping the motor output current at a much higher frequency, and it will adjust dynamically to whatever the input voltage happens to be. After all, even with the ripple, the supply voltage still stays within the nominal working range.

Completely agree with the sentiment - and perhaps I can be a bit too precious about vagaries around electronic design. If you're talking about reduced torque due to ripple - again, I'd agree, and this would be an instantaneous effect (i.e. not related to holding torque - which *could* result in a bit of wobble on a stepper under load, particularly on a micro step). A bigger capacitor can help help to absorb energy from back-EMF from the motor.