Spindle and vacuum advice on a power budget

[tacogrande]

Hi all,

I want to cut melamine panels on a 4x8 router I've had in our garage for many years. There is currently no spindle or vacuum and there is 220V, 30A single phase service. The pieces I wish to cut are relative large, about 2ft^2 in area and of course non-porous. I consulted my electrician and he says I can upgrade to 220V 50A for about $2000, but beyond that I would need to dig up my patio and get a new service entrance.

I need recommendations for..

1. A spindle. I would like the ability to change tools quickly with indexing, and maybe set up ATC in the future, so maybe something like this? Obviously this would blow my power budget, but I don't have to run it at full power, right? I'm most interested in the ATC.

https://www.aliexpress.us/item/3256807418185855.html?

2. Recs for a vacuum hold down setup?

3. I assume I should upgrade my power to 50A? Is that going to be enough?

Similar Threads:

• PACER 2515 - No power going to the Spindle and the bed vacuum
• PACER 2515 - No power going to the Spindle and the bed vacuum
• Seeking Advice on Best Budget Friendly CNC Machine
• Newbie- Sieg X2 conversion on a budget - looking for advice on stepper drivers, please...
• $10k budget need some advice

[awerby]

30 amps seems sufficient to run your machine and perhaps a small spindle. You need to estimate its power consumption at maximum draw, plus a little overhead for safety's sake. But if you want an ATC, those usually require an air compressor, which would require more power, depending on its capacity. And a vacuum pump capable of holding that panel down will also have considerable power needs of its own. But I doubt you'd want to run all these devices on a single 50 amp circuit. More likely, the compressor will want a circuit of its own, and so will the vacuum pump. These should be sized to the machines you decide on - enough to give them more power than the nameplates say they need, but not so much more that the circuit breakers won't click off in event of a serious overload. Ask your electrician if there's room in your existing panel and enough power in the service for a couple more heavy circuits.

[joeavaerage]

Hi,
I would have thought 50A at 220V would be ample but 6kW??? from single phase, no that's not realistic. The effective practical maximum for single phase is 2.2kW to 3kW.
VFD driven induction motors draw huge amounts of current intermittently. If you do not allow for those short duration high draw situations you'll get nuisance circuit breaker trips.

You are right to think about ATC. I have recently fitted a new (3.5kW (S1), 6kW (S6), 3.4Nm (S1), 10000rpm rated, 40,000rpm max, 400V, HSK32 tool interface) spindle and even just manually changing
tools (still, until I build an ATC) has changed my CNCing markedly. Should have done it years ago. I would urge ou consider ATC from the outset. I wish I had.

Its not unreasonable to assume a spindle of 2.2kW to 3kW would take half of that, say 25A (peak) despite the average being much lower than that, say 10A.
As Andrew has already posted a compressor will take a fair chunk too, but it will be intermittent.

The big unknown is the vacuum pump. If you have a very good and vacuum tight set-up then the pump will need run only every few minutes and for twenty or thirty seconds at a time.
More usually what you had thought of as non-porous turns out to leak like a sieve!

I used to do a lot of vacuum bagging of both model aircraft parts and boat parts. Personal experience is that MDF is NOT vacuum tight, in fact its as porous as hell.
That will in turn mean that you need a high volume vacuum pump and that it runs nearly continuously, and they can draw a lot of current, probably more on average than the spindle.
With my vacuum bagging setups I tended to go for high vacuum, say 740mmHg, but because the materials are genuinely non-porous the pump would run only intermittently. I have a 75l ballast tank
that evens things out nicely. At first I used an old refrigeration compressor that was being discarded from the local butchers shop. As my setup advanced I found a single stage vane pump to be the
right choice, and 1hp or less was easily big enough

In your situation look for a more modest vacuum, say 500mmHg but much MUCH higher volume. A liquid ring pump is probably the right choice and they can absorb several hp at least.
This is not liquid sealed but otherwise could be the ticket:

https://www.ebay.com/itm/23511422273...Bk9SR7Sx7qi-ZA

This is a small liquid sealed pump:

https://www.ebay.com/itm/37501897394...3ABFBM8py0qb5k

This one is much larger, and probably more size appropriate. Note it is three phase so you'd require a VFD:

https://www.ebay.com/itm/15518382409...Bk9SR9bM5qm-ZA

Craig

[tacogrande]

Ok thanks,

I should have made it more clear that the most I am going to get is 220V @50 amps total, for the whole garage! I was originally wondering if there is any chance of getting away with 30 amps, but it sounds like I should just upgrade the power.

What is the cheapest spindle with either automatic tool changing or quick tool changing spindle? At $1200 shipped, the 6kW spindle I linked to was the cheapest I could find. It seems like I could just run the 6kw spindle at a lower power, since the power is really determined by the load, right? Also from what I understand, I'm going to want to run it at a low speed anyway since I don't think my machine will travel fast enough to have a decent chip load at high rpm.

Regarding the vacuum, I have 2 surplus ring pumps that are 1.5 HP each or 5 amps. I am going to cut the same 2 or 3 routines over and over, so I should be able to make something with low leakage specific to what I am cutting.

[joeavaerage]

Hi,

I'm going to want to run it at a low speed anyway since I don't think my machine will travel fast enough to have a decent chip load at high rpm.

That is a mistake. The sort of spindles that you are looking at (including the one you linked to) are high speed/low torque types. They need to be run at rated rpm or close to or
they will likely overheat. Depending on who you talk to; some say never run below half of rated rpm, while others say one quarter of rated rpm.

The spindle I pictured above is rated at 10000rpm despite having a max of 40000rpm. I have set-up my VFD to not run at anything below 2000rpm, so 1/5th of rated speed.
I have not done a lot of cutting down that low to date, the spindle has been in service for only a few months, so at this stage I cannot tell you whether I am pushing my luck. What cutting I have
done at that speed has been fine.

As for the power demand of a given spindle, by in large you are correct, the cutting load determines the current. The problem with VFD's is they have a very poor power factor at partial
loads, that is to say they absorb a lot of current for only a modest motive power output. Typically they have a power factor of 0.55 to 0.65

2200W at 220V you'd expect 2200/220 =10A right? But no,..... with a power factor of 0.6 the current drawn by the VFD will be something like 10 / 0.6 =16.67A

I would suggest that a 2.2kW or a 2.5kW or even a 3.0kW spindle run at or near to rated rpm will give you the best results.

Additionally most VFD's will run with single phase input...up to about 2.2kW. Above that they really need three phase input. A lot of cheap Chinese VFD's claim that their VFDs
run on single phase input at 3kW, 3.4kW, 4kW and more. Must be that physics are different in China....or they could be BullS*****ting!!

If you want to run at slow speeds, or at least speeds well below rated then asynchronous spindles are not the best choice. There are motors that are more like a servo,
a type of brushless DC motor. They have a wide speed range and tolerate running at low speeds while still producing rated torque without problem. While they may appeal
performance wise, they are more costly.

I rather think a four pole 12000rpm asynchronous spindle of 2.2kW to 3 kW will prove to be the most cost effective. There are tens of thousands of CNCers using spindles of this
type and power output. The four pole 12000 rpm type have nearly double the torque of a two pole 24000 rpm type, and thus the 12000rpm type is preferred for low, or at least,
lower speeds.

Craig

[tacogrande]

Ok, thank you, this is helpful. Jianken makes a 2kW 12000 rpm motor with iso20, I will ask what that costs.

https://m.jian-ken.com/permanent-mag...c-spindle.html


What you say about VFDs is surprising. I wouldn't expect 2 phase power to be an issue and I wouldn't expect the power factor to change that much. But what I "would expect" doesn't mean much!

[tacogrande]

Here's another generic 12,000 rpm spindle with BTC30

https://www.aliexpress.us/item/3256806328192820.html?

[joeavaerage]

Hi, I
Jianken has a good name and that particular spindle is 'permanent magnet synchronous' type which is a better candidate for part speeds than regular asynchronous spindles. I'm not sure that it does not benefit from
a specialised VFD....it is worth enquiring.

With the severe downturn in the Chinese economy these types of spindles are as cheap as I have ever seen....so much so that I'm surprised they have not attracted large tariffs....and
they yet may do so.

What you say about VFDs is surprising. I wouldn't expect 2 phase power to be an issue and I wouldn't expect the power factor to change that much. But what I "would expect" doesn't mean much!

All VFDs have a rectifier at the input followed by smoothing capacitors to derive a DC supply for the VFD proper. Note how the current (the red trace) is drawn in very large spikes despite the average current (blue trace)
being much lower. In electrical terms we'd say that the current waveform is very distorted. Ideally it would have the same shape as the input voltage (green trace).

Most people are familiar with power factor describing a current waveform that 'lags' the voltage waveform, typically because of an inductive load....but not so here. The current is in phase with the voltage just very distorted.

For nonlinear circuits, the power factor calculation needs to account for harmonic generation by nonlinear components. The definition of power factor relates the current in the primary harmonic to the RMS current contained in all harmonics:

Power Factor = 1/[1 + THD²]^1/2

This tells us that harmonic distortion degrades power factor. Harmonic distortion is very much load dependent.

Here in New Zealand there has been an explosion in rural areas of the use of high power (100kW, 200kW and more) VFDs for down-hole water pumps for irrigation. The harmonic distortion of such devices has degraded
the quality of the rural power supply that the power companies now insist that all such VFDs must have active power factor correction....at considerable cost.

Seldom do we as hobbyists ever have VFDs big enough to create objectionable harmonic distortion, but any sensitive electronics on the same circuit as a VFD could suffer from harmonic distortion. If you have a limited
power supply, especially if you have to share it with your wife's PC on which she works....then mitigating harmonics may become a priority. In the meantime just be aware that VFDs are a difficult load and can draw
remarkably large currents.

Craig

[routalot]

I suspect it might ultimately be less complicated to buy a generator with more than enough capacity to run the entire machine plus vacuum pump.If you build a decent zoned vacuum table-no small task itself-and limit yourself to a 4Kw pump,you may well have enough for holding panels.Craig is quite correct in suggesting a high flow,low vacuum pump and it will need to run constantly.

If you anticipate cutting several panels each session I'm not sure that a 2.2Kw spindle will allow you to approach normal commercial cutting rates unless you keep the DOC moderate and cut in several stages.Gong too slowly will wear the cutting edge away quite quickly and if the material is melamine faced chipboard,you need to anticipate the occasional foreign body which adds to the rate of consumption of tooling.

A modest compressor to run an ATC will add an intermittent load but by the time you have added a generator,vacuum pump and compressor,the workshop will have become a quite noisy place and you could make it a bit more pleasant by building something like a large kennel on the other side of the wall to contain them.Then you have the noise of the spindle and an extractor so I hope it is evident why I make the suggestion.

[joeavaerage]

Hi,
routalot is probably right, a genuine production machine would probably be equipped with a spindle north of 7.5kW, and that is not very practical with the supply you have.

Lets say you have 50A available. At unity power factor that is 11kW. Surely that would be enough? The key here is power factor. Any real VFD it will be about 0.6, and so the motive power available from a 50A
supply is not 11kW but 6.6kW. Of course that leaves nothing left over for the machine, the compressor or the vacuum pump. Aside from anything else there is no VFD that will run a 6kW spindle from single phase.
There is a sound technical reason why that is so, and its not insurmountable to overcome that limitation, but as a hobbyist I'm not sure that is a good course of action.

Lets say that you allow 10A for the machine and vacuum pump and compressor, which should be doable if you have a fairly vacuum tight setup, and that would leave 20A available for a spindle from a 30A supply. That is doable and
practical. It suggests that a 2kW spindle is about the max.

The same calculation but with a 50A supply would mean about 40A for a spindle. With some harmonic reduction, which in effect means power factor correction, you could possibly run a 6kW or 7kW spindle, but would
require that you exploit some of the techniques that I suggested you not entertain in order to run on single phase.

The same 50A supply, with a 10A allowance for the machine, pump and compressor, would leave 40A. Without any harmonic reduction techniques, or at least some simple power factor correction
measures, nor any of the provisions that extend the ability of a VFD to run on single phase, would suggest a 3kW to 3.5kW spindle.

You have to make a decision about the cutting speed and duty cycle. If you want genuine production speeds then you want a big spindle....and take whatever complexity and or cost is required to exploit that power.
The other choice is a lower power spindle which would be slower than production speeds, but could with a little care be driven from your existing 30A supply. The middle road is to up your supply to 50A, take some
cost effective but simple power factor correction precautions and run a spindle of 3kW to 3.5kW.

Craig

[joeavaerage]

Hi,
see the two attached pics.

One is a single phase supply, while the other is three phase. They are both suppling the same output voltage (308VDC approx) and the same current (3A approx).

Note how the single phase supply...the capacitor gets 'charged' at twice the line frequency (100Hz) but results in a ripple voltage of over 30V. The three phase supply gets charged at six times
line frequency (300Hz) and see how much smaller the ripple is.

The bottom line is that a three phase VFD will work fine with only a single phase input BUT the ripple voltage increases markedly, eventually causing 'Low Voltage' alarms on the VFD.

If you were to increase (triple) the capacitance then the three phase supply would work equally well on single phase. The electrolytic capacitors used are quite big and expensive for good ones.
Typically Chinese VFDs use cheap capacitors, and not many of them whereas good VFDs use more capacitors and much higher quality, but cost five times as much.

Depending on the VFD it is possible to add more capacitors externally. Extreme care is required. These things charge up to 350VDC, if you take a hit from one it may well be your last.
BE WARNED THEY ARE POTENTIALLY FATAL!!!!! They need a pre-charge circuit for when power is first applied and a discharge resistor for when the power is turned off.
Unless you are familiar with moderate DC voltages, and by moderate I mean up to 700VDC or so then I'd not recommend it.......not that its not possible or even hard, but if you get a zap
its bloody dangerous.

By adding some extra capacitors say four, five or six or so, and good ones are likely to cost $30 each or more, you can easily convert a three phase 5kW VFD to 5kW VFD with a single phase input.

I would not trust any cheap Chinese VFD...I don't care how cheap they are. For twenty years I've used Delta (Taiwanese brand made in China), both on my own machines and supplied them to customers
and never had one fail. The Delta MS300 series has the required terminals on the outside for you to add capacitors in the manner I have described, whereas cheap VFDs don't bring those terminations
to the outside.

Craig

[tacogrande]

Joe,

Thanks for all the info. I agree delta is a good brand, but the allure of cheap Chinese VFDs is too strong. I bought some 3kW Huanyangs from the official HY store since they were in sale for $115. We have an HY or HY knockoff VFD running a lathe and it's fine.

I'm not sure what the practical benefit of reducing the ripple on the VFD is? Less harmonic pollution? I don't think the breaker will trigger with short spikes. Maybe with multiple VFDs there would be beat frequencies...

I didn't order the spindle yet. The Jianken is $1300 + shipping. I'm thinking I will buy a "g-penny" 2.2kw 12000 rpm spindle and see how that works, then maybe later upgrade to ATC. I'm leaning towards 12000 since I'm not sure I will be able to push the machine fast enough to take advantage of the higher speed spindles.

The 2 pumps I have are actually 3 HP, not 1.5. they are air tech 9671645 with 3 HP baldor motors and I'm guessing they will do 150 cfm each and 80 inches of vacuum.

Routalot - the compressor would be miserable, I hope to avoided that.

[routalot]

I agree that having a lot of machinery running at the same time makes the workshop a pretty miserable place to be,which is why I recommended building a kennel type building outside for it.Having the vacuum pumps out there would also make the place a bit more pleasant.I think we can be pretty sure that you won't be seeing 80 inches of vacuum and you ought to be delighted if the actual level is 15 inches. If the table isn't zoned you can cover the unused area with plastic sheet to reduce the effect of the leakage.

[tacogrande]

Routalot - thanks, cfm and vacuum estimate was based on a current product on the airtech site that has a matching pump casting. But good to know 15 inches is something to shoot for.

The other reason I want to avoid the generator is I am in a pretty dense residential area.

The panels I am cutting are for commercial purposes but I'm not too worried about reaching commercial speeds. I'll probably average a couple of panels a day or less, maybe 1000 inches of cutting per panel. More important is the edge quality since the panels will need to be run through a edgebander after routing.

[joeavaerage]

Hi,

I'm not sure what the practical benefit of reducing the ripple on the VFD is? Less harmonic pollution?

The Power Factor equation makes it plain.....the benefit to reducing harmonics is to improve power factor which in turn allows for a bigger spindle at less current.
At the start of this thread you said that you wanted a big spindle but was concerned that your power supply would not support a big spindle. Reducing
harmonics is the way to get the most from the limited supply you have.

By the way a perfect vacuum is 30 inches of mercury (10.2m of water) so you'll never get 80 inches of vacuum.
Craig

[joeavaerage]

Hi,
for seven years I worked for a company servicing welding equipment.

A fair fraction of that work involved repairing inverter welders, and on the basis of price a lot of those were Chinese. Like a VFD they have a direct off-line
rectifier, capacitive smoothing followed by the inverter. They suffer the exact same harmonic distortion that VFDs suffer from.

Typically a cheap inverter welder would claim 180A (more or less) but would in fact on do so on a very limited duty cycle. We would often load the output on a DC load bank,
100A output at 20V. This is similar to running a 4mm electrode. According to the nameplate this should be at 100% duty cycle, although very few ever actually did so. This is
a 'quality' test....does it actually do what the nameplate says it can???

I had a three phase current metered test board. In the case of these single phase welders I would obviously use only one phase. The vast majority of welders (of this type) would
draw 20A via an AC ammeter from the 240VAC supply . 240VAC is nominal here in New Zealand.

That would suggest that 20 x 240 = 4.8kW going into the welder....while only 100 x 20 =2kW coming out???? If the 'missing' 2.8kW were being dissipated as heat the whole thing would turn
into flames and smoke within seconds. What is of course happening is that it has a poor power factor. The welder would get warm, but only modestly so, lets say 200W of heat.
The power factor is:

(2kW + 0.2kW) / 4.8kW =0.45

So while it appears that the welder is consuming 4.8kW from the supply it was delivering only 45% of that to the load. I think you'll agree that's not very good, in fact abysmal...but that is what they are.

There are better welders, typically US, Japanese and European made that have active power factor correction. As an example a little (about the size of a shoe box) welder by Lincoln. Its rated output was 150A
at reduced duty cycle but 100A at 100% duty cycle. When we load tested it it did indeed produce 100A at 20V and did so for well over an hour without ever overheating, but more importantly while drawing
only 10A (through the exact same ammeter I mentioned above) from the 240VAC supply. Its power factor was:

( 2kW + 0.2kW)/(10 x 240)= 0.91

So 91% of the apparent power going into the welder is actually delivered to the load. This is a vast improvement.

The Lincoln welder we sold for $1500NZD while we sold the Chinese type welder for about $500NZD.

So this improvement in power factor was not cheap.

You might ask well why would anyone buy a Lincoln welder being that much more expensive?

As a result of a severe earthquake here in Christchurch in 2011 there was, in fact still is, a large amount of reconstruction going on. Small, portable but capable welders are very much
in demand, and often run on a generator. In order to get the best possible result from a fixed size generator then you would use the most efficient welders, ie with the best power factor.
Anyone who has done any on-site welding will know just how important power factor is.

Craig

[tacogrande]

The airtech pumps are rated in inches of water. I was guessing -80 to -88 inches of water based on a current airtech product with the same pump housing and horsepower. That's 6 to 6.5 inches of mercury. So maybe not much better than a shop vac? The CFM of a single pump seems OK so I'm wondering if I could chain the two pumps together to make a two stage pump and get more vacuum?


With the VFDs I will see how it goes. We happen to have a 255 amp Lincoln tig welder from the 90s that has gone 30 years without a repair. The DIY water pump just failed yesterday, but that's it. I certainly would expect off brand anything to last that long, from China or elsewhere.

[joeavaerage]

Hi,
Lincoln used to, and in some cases still do produce such equipment.

They tended to be large and heavy, big copper wound transformers, puck type SCR's and technologies of that type. They have a superb reputation
for lasting, but the TIGs particularly drew enormous input currents.

The downside is that if you were to buy such a device today it would be prohibitively expensive, and thus very few models are still in production.
The modern market is all about inverter welders, and while not cheap they are very much cheaper than the 'belt and braces' style of yesteryear.
Try as they might, even Lincoln have yet to make an inverter welder that is as reliable as the old style.

Craig

[routalot]

The airtech pumps are rated in inches of water. I was guessing -80 to -88 inches of water based on a current airtech product with the same pump housing and horsepower. That's 6 to 6.5 inches of mercury. So maybe not much better than a shop vac? The CFM of a single pump seems OK so I'm wondering if I could chain the two pumps together to make a two stage pump and get more vacuum?

I have run with 10 inches of (mercury) vacuum and it works with large panels very well.With small parts it doesn't give enough clamping force for aggressive feed rates and deep cuts so you will need to cover the unused parts of the zone with something impermeable and probably do your cutting in stages.The only way to determine the optimum combination is trial and error unfortunately.

[tacogrande]

I have run with 10 inches of (mercury) vacuum and it works with large panels very well.With small parts it doesn't give enough clamping force for aggressive feed rates and deep cuts so you will need to cover the unused parts of the zone with something impermeable and probably do your cutting in stages.The only way to determine the optimum combination is trial and error unfortunately.

Thanks, that's good to know since I'll be cutting larger piece and probably won't get much more than 10 inches. Were you drawing vacuum through an MDF spoilboard with the 10-inches-of-vacuum machine? Or did did you have a setup with channels and gaskets?

I did a little digging and it seems like these regenerative type pumps (I may have called them something else earlier) can be plumbed in series to get something approaching 2x vacuum. Becker pumps' website recommends 175 CFM for 4x8, so I'd have close to 150 CFM and vacuum of up to 13, but let's just say 10+. Seems workable.