Quotient Mistress - making your own

Flicking through the pages of a hobbyist electronics mag, there are stepper motor drivers for 18 and LCD drivers including display, also for 18 pounds. Both kits are based on
PIC microcomputers.**
For those of us who like to "Roll our own", (isn't that all of us?) it would seem that the cost of manufacturing our own Quotient Mistress, using off-the-shelf professionally produced circuit boards, is about 36.
QED?
** Magenta Electronics Ltd, page 240, "Everyday Practical Electronics", April 2003 edition.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Why are we waiting..???
Why are we waiting..???
Why are we waiting..????
Oohh Mr Bean.....
-- Regards,
John Stevenson Nottingham, England.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

You can indeed buy those kits from Magenta (and I bought and assembled them myself in the early days of my experiments with stepper motors - they have long since been discarded, having served their useful but very limited purpose), and separately, each kit will allow you to drive a stepper or an LCD display (but not both together). They are great for learning how PICs work & experimenting with LCD and stepper motor driving, but that's about it, and for anyone that has the interest I would recommend them as a cheap way of finding out more about these fascinating devices.
As I am sure you are perfectly well aware, but conveniently choose to gloss over, there are some significant differences between what you would be able to do with these 2 kits and what DivisionMaster can do:
- Once you have interfaced the LCD to a PIC 16F84 (needs minimum of 5 I/O lines) and the stepper drive (the way they do it, using full phase drive, they use 4 I/O lines), you have unfortunately run out of pins on the 16F84 to drive a 16-key keyboard (needs another 8 I/O lines which you haven't got unless you move up to the bigger PICs, like the 40-pin devices in the 16F8XX series and the 18FXXX series). Of course, you could go for a twin-processor approach, which would give you a total of 26 I/O lines between the two PICs, in which case, good luck to you as you try to partition the functionality you need across the 2 PICs and devise a suitable communications protocol to connect them together (which would cost you at least 2 more I/O lines of course).
- DivisionMaster has a bipolar (not unipolar) output stage capable of driving motors at up to 2A/phase, as opposed to the 1A unipolar drive on the Magenta kit. So you would have to factor in the 2A/phase H-bridge circuitry & decoding logic, and while you're at it, the means of controlling the motor current from the PIC, which DivisionMaster can do; with the simple unipolar drive circuit used on the Magenta kit, the only way to adjust the motor current is via fixed resistors in series with the motor coils.
- The PIC 16F84 they use in the kits is a 4 MHz part (there is a 10 megs part available these days, but they don't use it in the kits to keep the cost down). Consequently, the step rates it is capable of generating will be between 1/4 and 1/8th of the step rates that can be had with the bigger PICs that can be clocked at 20 or 40 MHz. DivisionMaster can generate a pulse rate of up to 10,000 pulses/second (actually, it can do significantly better than that, but that is the maximum that can be programmed in from the keyboard).
- The 16F84 has a very small amount of program memory (only 1K), so you would run out of memory space long before you had created anything remotely like the range of functions and features that DivisionMaster supports. At the last count, my firmware takes 2K to store the character strings for driving the LCD display, and the maths libraries for LCD display, keyboard access, and 8, 16, and 32-bit fixed point arithmetic, and that is before you have written a line of application code.
- The PIC 16F84 has no non-volatile RAM, so (unlike DivisionMaster), there is no way for it to remember any operating parameters from one day to the next. So you will have to factor in a bit of external flash RAM in your design to cope with that, unless you want to enter the operating parameters every time you use the machine, on that keyboard you can't drive because you haven't enough I/O lines. And of course, there's no space on the Magenta boards for adding NV RAM chips or their supporting components.
- The kits from Magenta don't include a power supply, let alone one capable of generating the necessary 1A/phase to drive the motor - you have to source that yourself. Or you could of course use Magenta's bench PSU kit - variable output up to 25V at 2.5A, which is just the job - as the Magenta stepper kit is a simple unipolar drive, the PSU will need to be able to source 2A continuous at whatever voltage you choose to use, as the step sequence involves 2 motor coils being energized at the same time. I have bought & built one of their bench PSUs, and it works extremely well. Kit price for the bench PSU is a very reasonable 65 by the way.
- You have also conveniently neglected to mention the cost of the keyboard hardware, a suitable case, mounting hardware, plugs & sockets, interconnecting cables, keyboard membrane...etc., none of which are provided in the Magenta kits.
- To make any sense out of these kits over and above the trivial sample programs they contain, you would need to provide yourself with the software and hardware tools to program them to do anything other than the rather trivial pre-programmed functions that they perform "out of the box". Magenta sell a variety of programmer kits - you're looking at a minimum of 26 (again, don't forget the cost of the power supply, which is an extra 4 with these kits) for a parallel port connected programmer. However, make sure you don't make the mistake of trying to use any of these tools that are designed for connection to a parallel port of a PC if you are using NT or any of its derivatives (WIN 2K, XP etc), because you will find that they don't work, as NT doesn't allow applications direct access to the I/O ports. Magenta's "Icebreaker" system (which uses a PC serial port & works on NT) is great for breadboarding PIC projects by the way - 39 with power supply, but it only works with their own specially programmed 16F877's and the upper 4K of memory can't be accessed, so it wouldn't help you with the '84's. FED (http://www.fored.co.uk ) will sell you a nice serial port connected PIC programmer for about 45 plus the cost of a PSU, or one fully built for USB connection (no extra PSU needed) for 95. They also do very nice software development tools, but you're talking another 50+ on top of their programmer prices.
- And when you're done cobbling these 2 kits together, don't forget the EMC testing you need to do for CE compliance. And no, the fact that you build it yourself & it is not for resale doesn't exempt you from the legal obligation to comply with the CE regs, as I am sure you understand from your ham radio experience. So if your home-built stepper driving device causes interference that screws up your neighbour's sensitive electrical equipment, you are liable. The most reasonable testing house I have found to date charges around 800/day to use their labs and services; figure on a minimum of 2 days, if you are lucky enough that it passes first time without the need for re-work.
So, all in all, I'm not remotely worried that you are likely to give me any competition with your kit-based "solution" anytime soon. Or anytime before Hell freezes over, for that matter. QE definitely not D, Old Man.
However, yet again, I must thank you for the additional publicity you are giving my company's products, and on behalf of Magenta's owner (who is a friend of mine, by the way), thank you very much for advertising their products too.
Regards, Tony
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
There's no "convenient gloss-over". Anybody who sets out to design kit must put in some effort of his own, that's only to be expected.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

That's just what we said. Are you sure you are in the right skitzo mode and not mirror image.

-- Regards,
John Stevenson Nottingham, England.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
So you choose another in the range, or conceive of functional partition between two of them? All fairly trivial stuff. If you think otherwise, perhaps that is indicative of inexperience on your part which may contribute to your excessive costs?

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
...so now you're changing the rules (as usual). You started this particular thread by stating:
"For those of us who like to "Roll our own", (isn't that all of us?) it would seem that the cost of manufacturing our own Quotient Mistress, using off-the-shelf professionally produced circuit boards, is about 36."
Now, having had it pointed out to you that this was total cr@p, you start talking about using a different chip. Where are the hobby board kits to support that? apart from in your imagination?
Regards, Tony
wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
One really has to ask whether your contribution quoted below is a serious contribution to a discussion amongst hobbyists seeking to produce something by their own efforts?
It seems to be a somewhat childish rant to me.
"you're changing the rules (as usual)"? This is a thread exploring how we the hobbyists might do things for ouselves much cheaper than the "cynical" (your word) traders would charge us for the same thing. You yourself have contributed many articles over the years persuading the hobbyist to do just that.
"changing the rules (as usual)"? Not at all. In a design project carried on by the NG, we will get to our final design by discussion; a proposal is made, a discussion ensues, and a conclusion is reached. In this case, I saw the products advertised by Magenta and proposed them. You brought into play your experience of using them, and so the discussion progressed. There is really no need for you to employ such childish forms of self-expression such as you do below.
If you argue that the Magenta products are, based upon your experience, unuseable, what do you expect to happen? That someone in discussion with you who hasn't actually used those particular products will gainsay you? Of course not! As the inevitable result of your freely contributing to the design project (very brave for a trader who suggests that he is in "competition" with the project, I might add; good on yer!) I will react to what you say and change my proposals accordingly. That is normal progress in a civilsed technical forum.
For you to have introduced your experience and then to castigate me for taking on board what you have said is just silly. "Now, having had it pointed out to you that this was total cr@p, you start talking about using a different chip"? Have you been playing with Mr.Stevenson in the school playground again?
you

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
The use of stepper motors to drive a rotary table in the hobbyist's workshop is a low-power requirement. 1A per phase is more than sufficient. You don't move the table unless the brake is off!
Using unipolar motors at 1A and the excessive (and irrelevant) cost of 2A bipolar drive is no longer justified.
The speeds necessary will more than justify L/R adjustments.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Bollocks. 1 Amp will not move a 6" rotary table with plain bearings against frictional forces. Tables with thrust races in them are well beyond simple hobby import prices.

-- Regards,
John Stevenson Nottingham, England.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
You clearly haven't actually tried driving a rotary table of the kind most hobbyists are likely to use. On the other hand, I have.
Regards, Tony
wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
If you must resort to silly childish jibes, then you can do no better than to take on the services of Mr.Stevenson as a tutor.
Do you mean to say that the very light duty example that you were parading at Donnington was a con trick? Were you not attempting to sell your goods based upon an illustration of how simple clock wheel cutting could become?
QED, once again.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

He does do, the monthly cheques are gratefully received.

NO QED again at all. Tonys Division master was designed to output 2amp so it wasn't a con. If it was only putting 1A out then it could have had problems. You program the output needed, a point you omitted off the third beer mat in your design. For the record I have one of these on a 6" brand new vertex table and it draws 1.8 amps at 24 volts. I also have one running on a large Hoffman table that weighs around 250 pounds. This one takes 75 volts at 7 amps using an external driver box.

-- Regards,
John Stevenson Nottingham, England.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Glad you brought up the clock wheel cutting engine that I demonstrated on the DivisionMaster stand at Donnington. The motor used for the dividing head is 140 oz-in, rated at 2A/phase as it happens. The other "light duty" application was the Sherline rotary table - probably the easiest table to drive that I have seen so far (and actually easier to drive than the dividing head on the wheel cutting engine, as it happens), as it has ball races on the wormshaft and on the table spindle, versus the plain bearings approach in the Vertex tables and on the worm shaft of the dividing head. Sherline themselves recommend (and sell) 140 oz-in motors for that table; in reality, the 100 oz-in motor that I used to demonstrate it works well enough, but it still requires 1.5A/phase. All the other motors I was using on the stand needed to be driven at between 1.5A and 2A. With the general trend away from L/R drives towards using bipolar chopper drives, it is increasingly hard to obtain decent motors that are rated as low as 1A/phase.
Like I say, you clearly haven't tried to drive these sorts of machines with stepper motors. On the other hand, I have.
Regards, Tony
wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
And therefore unnecessarily overpowered, as it happens. The application that you were selling for did not need high powered motors.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
A red herring. 1000 pulses per second is more than sufficient for a hobbyist's workshop. The "Measure twice, cut once" mantra will mean that extra speed (and the excessive costs associated with it) will go to waste.
With a 200 steps/rev motor and a 60:1 worm, that's 12000 steps per rev for the workpiece and a matter of only 12 seconds to go right around. Neither here not there.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
If you're buying in maths libraries then perhaps that shows up an excessive costing and also your inexperience? 32 bit integer arithmetic is all that is needed, no trig or exponentials. Set your minimum "integer" value to be 0.001 (or even a tenth of that) and Bob's your uncle! 8 and 16 bit arithmetic is well handled by 32 bit routines.
I looked at your glossies - the "functions and features" that you supply are trivial extensions of each other and ably handled by a well-factored set of routines..
Sure, there are memory limitations in some microcomputers, choose whichever in the range is appropriate once you've produced your design. All very trivial, again.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Talk is cheap. If you think you can do it, get on with it.
Regards, Tony
wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Sales blurb. Irrelevant. Rarely is the hobbyist doing mass production using rotary tables. Even if so, the number of parameters to be set up is so small as to be a non-problem.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Far from irrelevant. I have several customers now that need to use their DivisionMaster controller to drive multiple tables/dividing heads. They want to be able to store several sets of config values precisely so that they don't have to scratch around setting the controller parameters for different speeds/drive ratios/motor current settings whenever they move to a different device. And as it happens, most of the ones that want that feature are hobbyists. So that capability will go into the next version of the firmware.
Regards, Tony
wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.