Thread: DC-7 3HP Mods

I was thinking about modelling this in Solidworks and was curious about why an iterative or genetic back-test wasn't done yet.... give it the basic problem and let the computer solve it through a million small changing simulations. A form of artificial evolution.

Curious that something like a DC impeller is so.... primitive.

Turns out this company has already done it. https://www.youtube.com/watch?v=3j4An-heg18 It would be interesting to make one of these.

Obviously you wouldn't want to ingest any big chunks!!!

I was thinking about modelling this in Solidworks and was curious about why an iterative or genetic back-test wasn't done yet.... give it the basic problem and let the computer solve it through a million small changing simulations. A form of artificial evolution.

Curious that something like a DC impeller is so.... primitive.

Its been done. There's been a huge amount of research done on high pressure pumps that move air including impellers. There must be 1000's of impeller designs out there especially considering every aircraft turbine and turbo pump is effectively a high pressure air fan.

The problems with anything that's pressurised, are safety, efficiency, cost, and especially in anything involving air, noise.

Efficiency is very important because folks with only 10A circuits in their sheds already struggle to access enough power without going to pressurised systems which result in a loss of efficiency.
As the pressure rises the power requirements increase by approximately the square of the pressure.
This is demonstrated when using a VFD to drive impellers faster. Just increasing an impeller speed from 50 to 60 Hz requires some 40% more electrical power.

There are ways of increasing pressure without using much more power provided flow is sacrificed which is not really what we want. Vacuum cleaners generate about 3 times the pressure of DCs but only move about 1/10th of the air for the same power expenditure and just look at how much noise they make. If this was to be scaled up by say a factor of 10 imagine the noise.

Once the pressure goes over a certain amount safety becomes a real issue.

In the early 1980's I remember our first turbo molecular pump. It was about the size of a 4L can of paint, 50k rpm, 750 CFM, amazing pressure etc etc. It was encased in a 20 mm thick shell of SS because if it let go it would destroy the lab and anyone in it. It ran so hot it was water cooled like a CNC spindle. It cost over $20k and the controller was $5k. It screamed like a banshee especially when the bearings were approaching the end of their lifetime when we had to wear muffs in the lab. I remember a Uni OHS officer coming in with a sound meter and threatening to shut us down unless we did something about it which is when we decided to buy one with magnetic bearings.

A pressurised air impeller is not really something for a home workshop. Apart from pump safety issues, if a high pressure impeller was attached to 4" stormwater ducting and throttled machinery, every piece of ducting would need to be upgraded to pressure pipe and the noise though the ducting and choked machines would be horrendous. Hit one of those pressurised pipes with something hard and the resulting explosion would not be pretty.

The issue for dust extraction is not really impellers, e.g. the BP designed impeller moves more than enough air, the issues are still mainly in the ducting and machine port size court. Attempting to get around this by upping the pressure alone is not going to be very productive.

Once again, i was thinking out loud. I enjoy the time on the PC noodling with toys I used in the past. Sketchup, 3DS, Solidworks, etc... Severe time thieves

HeadScratchers thoughts on making a blower box had me thinking about what to do as an upgrade path for my own is all. Like all semi-pro woodworkers I'm now hitting the limits of my amateur gear and upgrades in everything is becoming important. It's a treadmill from hell cos-wise and tool budgets are stretched to buggery.

If I could piggyback on the designs here it would make upgrading the housing on the DC here more interesting than simply spending money on a system you say is from the 80's (Im not being rude or belligerent, but over the last year I've suggested a few commercial systems and they are all quite inadequate/obsolete from a design point of view).

Its a vexing problem. I despise dust (DESPISE it!!!) and I'm sincere in wanting the best I can afford.

Apologies to HS for the hijack.

That's perfectly fine woodPixel discussion and ideas are great for coming up with new and improved ways of doing things. So far I have an idea, but I think it will take some prototyping and testing to see if the idea translates into something better than standard.

I'm all for tossing around ideas and playing with any available tool - that's how the modified 2HP thread developed and resulted in a useful mod - provided any claims of improvement can be adequately tested.
What worries me a bit is folks designing and making stuff and then claiming improvements based on a wet finger.

I'm not a fluid dynamicist but I know enough basic stuff and perhaps more importantly if I do guess, have some reasonable gear to test things out.
When I was working there was a small team of fluid dynamicists in my department that designed mainly nautical stuff and watching how they worked with high specialised CFD software and how it it still took them months to years to come up with changes to even small things tells me this is nor a game for amateurs. There is some highly regarded open source CFD software out there (look up OpenFoam) but CFD and CFD software is a highly specialised discipline requiring years of training to make sense out of. It's not something you can take off a shelf and make sense of like CAD.

Even with access to reasonable testing gear it can still be tricky and vert time consuming to clearly measure small improvements - ask Lappa how much work he has had to do to measure his air flows.
In my visits to some sheds I see folks making Q&D changes they think are an improvement but often what they have done has made no difference or are as likely to make things worse.

I wonder if anyone with the means to measure air flow correctly has actually tested the loss of this Y junction.

One thing to consider with these units, they “may” have different impellers, mine has a 14” impeller (I measured it) some have 13”.

So, I have one of these designs with the Y outlet, rated at 3HP and turning a 14” impeller with 4 cloth bags – no plastic bottom bags – so even less back pressure from the bags.

The measurements I made were with a Volt meter, Current meter and Manometer – YES I know this is a crude test and the results may not indicate the true effect when measured with correct equipment – however the results may still be relevant for one particular reason.

I tested this DC with a 150mm 1.5m length of pipe connected to the intake.

The static pressure (test pipe blocked) was a shade under 10” @ 243vac @4.2A

The open port pressure was 5” @ 142vac @ 8.9A

Removing the bags made almost no difference to the open port readings except for slight reduction in current of a hundred millamperes or so. – Obviously closed port static readings won’t change.

Removing the Y piece resulted in absolutely no difference to the readings.

When I said that the results may still be relevant for one particular reason – the reason is this:

If I connect the system and open the port at the Lathe – this port is very close to the DC – the readings are 5.5” @ 8.8A @ 243vac.

Now If I start to make a small bowl on the lathe, within 30 minutes the static pressure has dropped by a 1/4 ” and the current has dropped by 300mA, as more work is done on the bowl, especially sanding and as more work is carried in the workshop - especially cutting MDF on the BS, the current and pressure slowly decrease until after a few weeks I’m running at 3.2” of pressure and 7.2A because of the increasing restriction in the FOUR dust bags?

My point is, if I can easily see even a small amount of dust causing a decrease in flow rate, if I can see a change between new clean bags on and off, then surely I should see a change when disconnecting the Y piece “if the Y piece was indeed causing a significant loss in my DC”, is it really worth going to all this trouble if the gains are minimal?

Shouldn’t we at least try to verify the loss this Y piece may cause? Yes I know they bought out the “improved design”, but how much of that improvement was really to do with flow rate and not the possibility that it’s simply easier and less costly for the manufacture to make the DC without the Y piece - Cynical? ok - guilty!

Now maybe if I had a 3 phase unit spinning at over 3400 RPM, then some loss in the Y piece may show up at the higher flow rates – but I don’t.

When you add machine port restrictions, installation losses and the ducting losses into the equation, I feel there are way more gains to be had in fixing these issues before ever bothering about that ugly Y piece - Just putting it out there.

Originally Posted by MandJ

Shouldn’t we at least try to verify the loss this Y piece may cause? Yes I know they bought out the “improved design”, but how much of that improvement was really to do with flow rate and not the possibility that it’s simply easier and less costly for the manufacture to make the DC without the Y piece - Cynical? ok - guilty! .

Not cynical - you make a good point - and I agree with you about cost factor.
Based on what you see it looks like there is little or no difference and I would indeed just leave it alone for a 6" system.
I'm surprised you don't see any difference in the current maybe a better resolution ammeter would show this?

My understanding is that there are at least two Y designs/sizes - one that is larger than the other - do you know which one you have?
I think the bigger ones were used on the 14' impeller and smaller one on the 12" impeller (DC7).

BTW Im surprised the 3HP motor only draws 8.9A with a 14" impeller. This suggest throttling somewhere in the system

The only thing throttling the system would be the design of the impeller or the impeller housing. I even tired a larger bell mouth style for the impeller housing intake, made no real difference.

This would be the larger Y as it is really quite large and well finished inside with good mating alignments, BTW There seems to be a lot of clearance around the 14" impeller, but I lost the measurements I took.

It would have been nice if I could have carried out the flow testing on a standard system before I made the first mods but the flow meter was a later acquisition.

Having the flow meter has been great as I can make mods and see changes but I agree with MandJ that these can also be seen with other measurements eg. Changes in current flow was one I also observed, although they were not as clear cut (much smaller changes) as the changes in air flow for small changes. I still use a manometer to check the condition of the filter (fitted between impellor outlet and filter)
One recent mod was after I found my table saw cabinet had dead areas. I was playing around with changing the air inlet size and shape and noticed a sloping piece of timber was redirecting the air flow and clearing the areas. So I mounted it on a hinge so the inlet is closed at rest and opens with the air flow when the DC is turned on. I checked the change in airflow and modified the size of the hinged section bit by bit so it made no difference to air flow than what is was without the flap. Works great and still clears the cabinet.
The extraction system for my SCMS is a weird design so the Flow meter will be an essential item during the construction and mods. I am sure will be needed.

Before and after measurements are really important as it's the only way to check what is going on.

To verify changes, it's also important to build things so they can be put back to how you had them initially and repeatedly retest the final and initial states.
Its a bt of a PITA but proper testing protocol requires measurements be undertaken alternating between sets of conditions/construction just in case there is an unknown bias creeping into the readings.

For example when comparing the meters against each other I measured the air speeds up and down the air speed range a couple of times and averaged the readings - it takes a while but its well worth it

For a set of ~20 air speed measurements at around 10 m/s in my 240 mm test duct it looks like the reproducibility/repeatabilty of my Testo is about +/-2% compared to the TSI meter which is about +/- 1%.
I was hoping the Testo would be as good as the TSI but it looks like what it lacks in precision it will make up for in convenience.
It will be interesting to see how it works at higher air speeds and smaller ducts.