Benchtest.Com - Workshop - ShumaTech DRO-350

ShumaTech DRO-350
February 12, 2006

Digital Read-Out.
A DRO for a mill is a nice addition for a couple of reasons. One, it does away with having to count the number of revolutions of the cranks. Two, it lessens the chance that you will make an error in measuring. Not to say that you still can't make mistakes in measuring or forget to account for the size of the cutting tool, but it helps cut down on some stupid errors. Last, but not least, it just looks cool!

When I purchased my mill, my long-term intentions were to adapt it to full Computer Numeric Control (CNC) at some point in time. I also wanted to keep the hand-wheels for manual milling. When I started reading about CNC conversions, I came across a project to build a moderately low cost, but pretty accurate DRO. The project was the baby of Scott Shumate. He set up a website (http://www.shumatech.com/) with sales and construction information and started a Yahoo group (http://groups.yahoo.com/group/ShumaTech/)for support and related issues. Scott's involvement in the sales and support of the ShumaTech DRO-350 has lessened in the past year or so, but the DRO-350 seems to have a life of its own and is now being marketed in Europe and Australia,with some talk of a US supplier as well.

The DRO-350 was introduced on the CAD_CAM_EDM_DRO Yahoo group around the end of July in 2003 message63408. There may have been earlier announcements, but this was the first I could find. The ShumaTech Yahoo group started in August of 2003.

I was fortunate enough to find out about the DRO-350 while Scott still had a few of the original circuit boards in stock and bought enough for2 units. This included 2 DRO-350 boards and some QCC-100 adapter boards. The adapter boards allow the use of standard quadrature type scales, rather than the inexpensive Chinese scales or calipers that the project was originally designed for. After I had read about half of all the posts on the Yahoo group (there were about 7000at the time), I purchased the parts. This was a good move because in another 4 weeks, there were no more DRO-350 boards available. I understand that a new batch of boards will be available from the new suppliers soon.

The project is quite a professional piece of work. The board is very well thought out and nicely constructed. The documentation is first rate and the evolution of the software was done with input from the Yahoo group members, so the functions are geared to the people who use it. The one negative point that is brought up most often about the DRO-350 is the resolution. Because the project was based on the inexpensive Chinese scales/calipers, measurement is to two places to the right of the decimal for metric and three places in inches. There is an additional decimal point to the right of the last digit which indicates 5/1000 of a millimeter or 5 -ten-thousandths of an inch. The DRO-350 mirrors the functionality of the scales/calipers in this respect. If you, as I have, use a scale with a resolution finer than that of the DRO-350, you cannot read the last digit, except for: no decimal = 0 and decimal lit = 5. However, what those who complain about this don't take into consideration is the fact that very few of the "hobby-class" mills and lathes are able to work to this resolution. The exception might be the Z1/Z2 axis of a lathe, where the measurement across the part is actually double that of the reading. With my mill, I consider producing a part with .001 inch accuracy as nothing short of amazing. If I require that kind of accuracy, I cut a bit over-sized and "sneak up" on the exact dimension. There aren't many projects I do where it is critical to be within less than a single thousandth of an inch.

Getting started
After I ordered the circuit boards from Scott, I put together my orders for the components. At that time the lists on ShumaTech were a little outdated, so I needed to go through each item and make sure that it was still available and if it wasn't, I needed to find a component to substitute. All of the parts except for an eight channel source driver #UDN2983A were available from Mouser. The UDN2983A was available from Digikey as a TD62783AP-ND. Unfortunately, Digikey hits you with a penalty for small orders, so it is more cost effective to order a few things from them. I ordered enough from Digikey to meet their minimum requirements and the rest from Mouser. The fact that I was ordering enough for 2 boards + a few spares of each of the ICs, I was able to get price breaks on some of the parts. Even if you are only ordering for one board, you can save a bit of cash by taking advantage of quantity discounts. There are nine 7-segment display LEDs needed for each board, but if you up your order to 10, you save a few cents and get an extra display for "free." Spending a little time going through the on-line catalog saved a few dollars on each board.

Jenix Scale

Scales
After doing quite a bit of reading on the Yahoo/ShumaTech site, I came to the conclusion that I would use the Chinese calipers for my scales on all axes except the long one - the X axis. For that I would use the Jenix 18" JSN5L-450 scale from Tim at Linear Measurement. Tim has been offering the Jenix scales a good price to the members of the ShumaTech group for quite a while and though it is about one and one half times as expensive as an 18 inch Chinese caliper, it is more rugged and more accurate. As the X axis gets more use than the other two, I thought that it would be money well spent in the long run. Unfortunately, Tim was out of stock in the size I needed and he wouldn't be placing an order for a while, so I picked up a 12 inch Chinese caliper for temporary use on X. Why 12" you ask? Well, the 18" Chinese scale is about $85 at the lowest Ebay price, but they get you with outrageous shipping at that price. About $100for either the caliper or scale is more usual with a "normal" amount for shipping. On the other hand, a 12" caliper was $35 from Harbor Freight (on sale). I picked up two of the 12 inch for $35 each and one 6 inch at $10 for the Y axis. I would modify the 6"and one of the 12" to work as scales and leave one of the 12" as a caliper and add a cable so I could read it from a switched port the Z axis of the DRO. The 12" that I would initially use for the X axis would be moved to the Z axis once my Jenix scale arrived.

Building the project
All my components arrived within a day of each other and a couple hours later, I had made sure that I had everything I ordered. I broke out a new tip for my soldering station and got to work. Soldering for a couple hours a night, I had both boards full of parts in about a week. I hadn't done this much soldering in a very long time. Both boards gave me a boot screen upon completion, but one wouldn't change modes when the clear button was pressed. After a couple of days of troubleshooting, I was getting frustrated. I realized that I hadn't cleaned either of the boards after soldering them, so I washed the circuit boards down with isopropyl alcohol and let them dry over night. Problem solved! It turns out that in the process of soldering, some of the solder must have spattered to short a couple of traces and a thorough cleaning with the alcohol and a tooth brush was all it took to rid the board of the short.

Modifying the calipers to scales was pretty straight forward. I used a Dremel tool and a fine cut-off wheel to cut the jaws off of the caliper. An eighth inch end mill made the mounting holes and cleaned up the rough edges from the Dremel work. One nice thing about using the calipers is that you gain a little measuring length from the area that used to be occupied by the jaws. This made the 6" able to measure about 6. 65" and the 12 inch able to almost hit 13". The 6.65 is almost exactly the amount of movement on my Y axis, however 13" is too short for the 14.5" of actual X axis travel. Grizzly claimed the X travel as 15.625 on the spec sheet,but the end caps on the table prevent it from moving that far. The table is 26.125 inches in length (not counting the 1.5 inch end caps) and the saddle is 11.625 inches wide, which leaves us with 14.5"of movement. I am planning to machine down the end caps to allow the table to move further when I install the Jenix scale. This will allow me to gain about another inch on each end, making the total available travel on X to be 16.5 inches.

The Chinese calipers are positive ground and seem to be somewhat susceptible to picking up noise that may make the readout jitter or jump. For this reason, many of the ShumaTech group members have electrically isolated the scales from the metal on the mill. I considered this a good warning and mounted the scales using HDPE (High Density Polyethylene). I cut up some plastic cutting boards for the occasion. This stuff is great to work with, but it has a major draw-back in that it is difficult to glue.

After a few evenings of work, I had a working setup on the X and Y axis, with my 12" caliper attached to the Z axis. What a nice thing having a DRO is. It has already saved me from making some errors as I milled the enclosure for the DRO. Setting the "home" location as the lower left hand corner of the work piece, it is easy to move to all of the measurements on the piece and make sure that I didn't add incorrectly or transpose any figures. These are the kinds of errors that "bite me" on occasion.

Finishing up for now
The last item to finish before I could take a break from the DRO and wait for the Jenix scale to arrive was adding the tachometer. The design of the tach is pretty simple and works by a Fairchild infrared photo transistor/emitter pair that "reads" a dark line on a shiny surface. I mounted the device under the spindle pulley after buffing the pulley surface to a high gloss. When switched on, the photo transistor sends a pulse to the DRO every revolution and appears to be pretty accurate. On the test motor I checked, it showed1007 RPM when my automotive tach/dwell meter showed 1000. Close enough for what I am using it for. It did show me that the estimated RPM of the Grizzly pulley chart were mostly too low, but the difference never amounted to be more or less than 10%. It is unlikely that being 10% off by using the mill's speed chart will cause you to ruin a cutter or work piece.

With the DRO finally built to a point where I can use it until my Jenix scale arrives, I will leave you with some pictures of building the DRO-350.

The circuit board is partially assembled.

The remainder of the ICs get installed.

The tactile switches and displays are installed.

Soldering a cable to a Harbor Freight digital caliper.

A close-up of the same. A 100µF electrolytic capacitor is substituted for the standard button type battery.

With 3-1/2 of the 4 caliper jaws removed, the caliper is transformed into a digital scale and mounted.

Holes are drilled free-hand in the saddle. It appears that before it was beige, the saddle was painted red! Well, at least the other side is beige.

The Y axis scale is partially mounted.

Control arm and cover are installed.

Working on the X axis. This is a temporary scale that will be used until my new Jenix scale arrives.

The caliper jaw is embedded into the plastic. Having this much surface holding the jaw made for good support.

Bolt holes are added.

Aligning the holes for the cover. An extra drill bit holds the cover in place while I measure for the next hole.

The limit switch is remounted and the bumper is put in place. With the temporary scale, my X axis was limited to just under 13 inches until the new scale arrives.

Testing the read-out while still working on the bumper for the right side of the X axis.

Using the DRO for its first task. Cutting its own housing!

Wiring the housing. The white circles are switches to switch between two scales on the Z axis and to choose either RPM or electronic edge finder.

This is the Fairchild photo transistor for measuring spindle speed, housed inside a liquid and swarf proof HDPE box.

And mounted. Looking up at the under-side of the spindle pulley (which is a lot more shiny than it looks here).

An arm holds the DRO in place and adjusts in length as well as allowing the unit to swivel.

Scott was out of over-lays (and I wasn't crazy about the blue lettering anyway), so I made my own.

Done for now. Step back and admire one very cool DRO.
My thanks to Scott for a useful and fun project!

A little update:
I just received my Jenix scale and already have it mounted. The accuracy of the Jenix is better than the calipers and the movement is very smooth. I am in the process of modifying the X axis to gain a couple more inches of travel and designing a new adjustable setup for the limit switches for the power feed. Once I get it all working as it should, I will finish up this article.