Beginner attempt at DC - thanks to forum members

[Lappa]

Nice post and great results. They mimic mine though I'm measuring flow which has increased. I think your point re the particle counter is something I'd like to do at the outlet on the Thien.
Guess that's my next project

I've got my eye on a 3hp 415v at the moment which has two separators and two huge bag filters that won't fit in the DC enclosure so a cyclone or Thien inside and one outside would be ideal.

[MandJ]

Thanks, yes I half expected it to be a total fail, but turned out great. I'm glad I went with a low cost 3HP twin bag right from the start (actually it's got 4 filter bags) no plastic bottoms bags, but they are over 4 years old and seem to reduce flow after not much work, really the only reason I was looking at a cyclone one day is not having to clean the bags so often. If the output is reasonably clean (and no solids) and for the amount of work we actually do, I feel venting out into the open without a filter should be fine, at least I'll have a definite reading of particle size and concentration as soon as I can get back to some cutting and test it, I'll then know exactly what I'm putting out there, I'll post results when I test it. Hopefully I won't have to scrap the idea.

[MandJ]

Lets me say right at the start, this is badly designed cyclone and not that good for dust separation, especially fine particles. There are a number of design problems with it, the short cone is not efficient for fine dust sepertion and the size makes it inefficient for flowing anything more than a small 1HP DC.

I made a couple of cuts on the BS with just the dust bags, and then with the cyclone venting. I measured the particle counts "inside the DC housing", as expected the particle count with the cyclone was quite a bit more that with just the dust bags. I'm on my own at the moment so I couldn't do much more that a quick test.

The interesting thing is that later measurements at 1M from the exhaust port of the DC enclosure showed particle counts only slightly elevated compared to the bags alone, so obviously the small amount of fine dust dissipates rapidly in the atmosphere. FYI the exhaust flow was blowing straight over the sensor.

So considering the small bin, the size of the cyclone and the quick test setup it's not bad, I'll leave it as is and do some wood working over the next few weeks and see how it goes, especially as the dust bags start to clog again and more flow is directed to the cyclone.

If it still looks good then a few thoughts on moving forward:

1: Make use of the two exhaust to dust bag housings and convert each into the top part of two larger cyclones, the housings are a lot bigger that the cyclone I'm testing and will reduce back pressure even more.

2: Remove the motor from the base plate and mount it towards the top of the cabinet.

3: A higher mounted motor allows for a much longer cone, better separation, and a larger bin.

4: I'd take a pipe from each cyclone straight through the roof of the enclosure and put exhaust flap rain caps on them - or something like that.

The test setup:

Yes I know there are number of things wrong with respect to dust flow, efficiency, cyclone feed input, size, the moon phase and whatever, I'm aware of them all, this is a quick "in principle" test setup.




Made this small bin from scrap to get the cyclone at the correct height - yes it's absolutely air tight.




The dust in the cyclone bin is from two cuts through this small piece of particle board on the band saw, a lot of the dust in towards the back of the box..

[woodPixel]

You could detach the right hand side (Block the end with a bolted-on endcap) and force all the air to the cyclone.

The greater volume of air will increase velocity and force more particles to the side of the cyclone.

i was watching quite a few videos on cyclones and fluid dynamics modelling and the straighter and more non-turbulent you can get it going in the better, plus the greatest velocity... So particles essentially hit the wall and looses inertia.

perhaps zip-tie the old bag to the top of the Cyclones out-hole?

[MandJ]

If you read a few posts back you see why I tried this small cyclone.

1: You loose virtually all air flow if you block one side.

2: That Y exhaust splits into two smaller outlets, not only would you force ALL air flow through a small cyclone, you now also force it through a single 130 mm duct.

As stated in my last post, I fully understand every loss aspect of a cyclone design and the feed point and exhaust loss.

However I'm talking about something smaller than a single big cyclone designed for 3 to 5 HP on the vacuum side of a DC. The idea is something low cost and simple to build because of its size and the fact that it does not have to handle the vacuum from a big DE, which is why I'm using the split output on the exhaust side through two smaller cyclones to overcome resistance loss of a single small cyclone.

If you are going to build a large expensive cyclone then you would not put it on the exhaust side, even a big expensive cyclone looses around 2.5" of vacuum in a 4 HP system. A single small non optimised cyclone looses close to 5" of vacuum, in other words, virtually all flow is lost in a 3 HP system. You need 4 to 5 HP (and big impeller) to overcome the loss of a single high performance cyclone (2.5") and get the flow rate back to the equivalent of a true unrestricted (spotlessly clean low loss filters) in a quality 3 HP DC.

IMHO separation is more about the design of the cyclone for the intended task, not just about flow in those ideal on line videos, otherwise you would be saying that a cyclone can never be used on a small DC (loss calculations not withstanding) but again this is off track for what I'm trying to test here.

[MandJ]

I mentioned in post #93 that if this works out I could raise the motor, just realised I don't have to do that, just need to insert a 1m extension between the impeller exhaust port and the Y piece.

FYI, I have tested the Y piece for loss on a number of occasions, there is none.

1: This is a 3 Hp DC designed for 60 Hz.

2: The impeller turns around 20% slower in Australia (50 Hz)

3: Impeller speed @ 60 Hz = 3450 rpm

4: Impeller speed @ 50 Hz = 2850 rpm

5: A slower turning impeller means: A: Less flow rate. B: Less input power.

Now can we at least assume that this thing was designed to flow correctly @ 60 Hz without getting into discussions about manufacturing optimistic calculations and miss leading testing results.

So with less than the design flow rate I'm not seeing any impact from the Y piece at this stage, that would likely change if I installed a 3 phase motor and VFD and ran it at 60 Hz, but that's a discussion for another day.

EDIT: As far as I know all imported single phase DC's are designed and made overseas for 60 Hz.

[woodPixel]

I had a look for the other thread but couldnt find it. Search on this site is terrible.

Which machine is this?

I ask as I don't think my DC journey has ended yet....

[MandJ]

Hi, if you are referring to my DC then here are two pictures of the model and 14" impeller, if not, my apology for miss understanding.







EDIT: This unit is sold and re-badged by many importers, it's an old design and can have impellers sized from 12" to 14". I got this for around $480.

[MandJ]

This test setup started at Post #90:

Some results at last.

Normally after turning and sanding 2 small bowls, the DC current and vacuum have dropped significantly, after the third bowl it's pretty much useless as far as I'm concerned. - 5 year old bags seem to clog even faster now with sanding and lathe work.

With the test setup and a compromised small cyclone on one side, we have turned, sanded and finished 6 bowls. The current has only dropped 0.6A and vacuum shows a 0.8" drop.

There was 3 litres of what looks like coloured flour in the Cyclone bin and 3 litres in the filter bag on this DC with dual exhausts: It normally runs dual sets of filter bags. So a total of 6 litres of very fine powdery wood dust.

Why no large shavings? For this test we only used the lathe with sharp tools, the large curls and chips end up on the floor as they are thrown away from the DC pickup and shielded by the blank and close mounted tool rest, however all fine dust is collected by the 150mm bell mouth pickup and this was verified by dual particle counters, one is a body worn dust sensor at head height.

The floor shavings are collected by an incredibly effective small cyclone separator mounted on a big deep collection bin, this feeds an externally mounted "Shop Vac" vacuum cleaner – the bin on the vacuum cleaner is still spotlessly clean after 6 months of vacuuming the floors around the lathe, general workshop floor clean-ups and bench, grinder, drill press cleaning. Cyclone bin is only half full and suction has not changed.

Back to the 3HP DC: Yes there is a bit more fine dust inside the DC cabinet. The crappy 1/2 size test cyclone has many design faults that compromise fine dust separation. However as the cyclone is on the exhaust side of the DC it’s easier to build a cyclone, it can be of a lighter weight construction. Because there are two cyclones taking up almost the same space as the dual upper/lower bag area, they will fit in my existing cabinet. As they are each running half the flow rate of the DC, the intake construction can be simpler and as long as the correct cyclone size in maintained then the loss when compared to single cyclone should not be a problem for a 3HP DC. This test indicates that it should be equivalent to "slightly" seasoned bags and won't change with use.

Since I already have an external DC cabinet and a low cost 3HP DC, I only need the addition two light weight cyclones venting through the roof of the cabinet, no bags to clean, and in the future I can fit a 3 phase motor with VFD and run it at its 60Hz design speed if and when needed.

The particle count 1m from the DC exhaust was only slightly higher than normal background, for the small amount of timber that goes through our hobby workshop and the fact that even now, it’s virtually unmeasurable at 3m from the DC exhaust with this dirty inefficient 1/2 size cyclone, it should be even better with correctly sized and constructed units.

[BobL]

Thanks for the update - very interesting results there.

Why no large shavings? For this test we only used the lathe with sharp tools, the large curls and chips end up on the floor as they are thrown away from the DC pickup and shielded by the blank and close mounted tool rest, however all fine dust is collected by the 150mm bell mouth pickup and this was verified by dual particle counters, one is a body worn dust sensor at head height..

It depends on the size of the work involved but I still get a some large shavings captured by the BMH, especially for small spindle work.

Have you tried to see how far shavings are flung with and without the BMH running. I find the chip scatter is reduced from about 3 back to ~2m which helps reduces clean up time..

The floor shavings are collected by an incredibly effective small cyclone separator mounted on a big deep collection bin, this feeds an externally mounted "Shop Vac" vacuum cleaner – the bin on the vacuum cleaner is still spotlessly clean after 6 months of vacuuming the floors around the lathe, general workshop floor clean-ups and bench, grinder, drill press cleaning. Cyclone bin is only half full and suction has not changed.

This again confirms all the fine dust is being collected by the BMH. However, instead of running a completely separate system I use a 3m length of flexy connected to a fixed 150mm PVC duct and use the flexy as a vac hose to suck up the shavings direct into the DC.

[MandJ]

The lathe shavings are contained to the floor area below and slightly in front of the lathe, it's pretty big curls and some large chips, most come from the 180 mm partly rounded stock and the roughing gouge and the initial bowl shaping.

I use a separate floor clean vacuum system to stop shoving that stuff into the DC filter bags, I won't have to worry about that when I get the new system made. It will be interesting to see how it goes but it's certainly looking like a really doable setup after these tests and one I could fabricate myself from light sheet metal.