Help design a DC motor drive for the 10EE - Page 2
Close
Login to Your Account
Page 2 of 3 FirstFirst 123 LastLast
Results 21 to 40 of 42
  1. #21
    Join Date
    Mar 2016
    Country
    UNITED STATES
    State/Province
    Minnesota
    Posts
    10
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    3

    Default

    Quote Originally Posted by johansen View Post
    The 120-265 VAC PFC front end (the part that turns Ac into 380-400 volts dc) in an Eltek rectifier.. is about 1.25 inches by 3.5 inches.. by about 8 inches.
    Johansen, the Eltek parts look interesting. Although everything I'm finding one ebay is lower voltage, typically 48V. I see that there are DC/DC modules with 380VDC output, but doesn't look like much of anything for used (read inexpensive). But, I'll definitely need to do a bit more research on this as the PFC front end is where majority of the cost is (~$1000 BOM cost). If I can find a better solution for that bit... The motor drive section of the design is essentially as you describe and will be lucky to hit $200 BOM cost.

    The PFC part I'm using on the front-end is from Artesyn, a company in the same market as Eltek. They have a 1600W PFC module that can be easily paralleled ( Artesyn 16W AC-DC Power Supply, PFC Modules, Full BrickAIF4ZPFC Series ). The rectifier section of the design uses 3 of these modules in parallel to deliver 380VDC at up to 4800W. These modules normally go for ~$220 each, making up the majority of the BOM cost. Luckily, I stumbled onto a dozen on them on ebay at $15 each. Seems too good to be true, but they appear to be brand new... Worst case, it would still be worth it to me to buy some new ones.

    Anyway, the overall rectifier stage will be about 10" x 10" x 4". Feed it 230VAC single phase in and get 380VDC out at 4800W with 95% efficiency and 0.97 power factor. Feed it 120VAC single phase in and it will still give you 380VDC out, although limited to only 3000W. DC Drives may not be bleeding-edge technology, but power supplies for server farms most certainly are.

    DC Supply power board
    monarchdc_power.jpg

    Motor Control power board
    monarchmotor_power.jpg

  2. #22
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Virginia
    Posts
    10,264
    Post Thanks / Like
    Likes (Given)
    5843
    Likes (Received)
    4629

    Default

    Quote Originally Posted by Kezorm View Post
    DC Drives may not be bleeding-edge technology, but power supplies for server farms most certainly are.
    And there is your second major ass u me umption. There are at least two of us here that know that bleeding edge PSU's are the ones we've been involved with putting into outer space. Server farm 'bleeding' is done on the books of their accountants, and PSU are sourced to a PRICE 'coz their server has gone obsolete and been upgraded before they fail. They do not have to last long enough to leave the solar system - only the building.

    In any case... rather than get the sort of 'surprises' Jonathan at Beel/BICL got... why not get your hands on a 3 HP large-frame Reliance motor out of an 'upgraded to VFD' (not..) 10EE.

    And find out what you have overlooked. Second-best choice a 'modern' 5 HP Reliance RPM III, type T, straight shunt wound.

    Do keep in mind 'stick and fry' DC at the voltages involved does not take prisoners.

    It can "rectumfry" a body, terminating their dreams rather abruptly.

    Bill

  3. #23
    Join Date
    Aug 2014
    Location
    bainbridge island
    Posts
    1,169
    Post Thanks / Like
    Likes (Given)
    233
    Likes (Received)
    275

    Default

    So... yes you open the eltek and pull power right off the 450 volt capacitors. Ignore the 48v side of it.

    And if you want to dump power into or out of the motorfield coils at 800v... you can do that with an appropriate topology. I don't understand why you need to do that.

  4. #24
    Join Date
    Sep 2002
    Location
    Salinas, CA USA
    Posts
    3,898
    Post Thanks / Like
    Likes (Given)
    209
    Likes (Received)
    177

    Default

    Quote Originally Posted by Kezorm View Post
    I saw the Beel option in the following post - 10ee solid state dc motor retro-fit The Beel drive is an inexpensive chopped AC drive, far from a clean DC output and not really capable of full armature voltage with 230VAC input. Yes, the drive is reasonably functional option, but not what I'm looking for.

    The drive I am working on can be more closely compared to the GE DC300 drives that someone sourced years back (minus the regenerative braking). Or, for $2000, here's a closer new comparison to what I'm working on - DC Drives | Non-Regen.(D), 3 Phase | D2Q1-1-1 - Driveswarehouse . At a glance, it looks like it could be a good fit, but note that it needs 3-phase input. Don't know much about it though. Might be able to get away with the 10HP version running on single phase. Or, the software might refuse to operate if missing a phase.

    Anyway, the $1500 BOM cost is when built one-off. Could probably come down a few hundred dollars if built a few at a go. Yes, it's more expensive than what it could be if I had time to engineer the hell out of it. But, seeing as there's limited time and no profit to be had, the project is better suited for a design-by-overkill approach.
    I bought one of those GE DC300 drives when they originally were found and sold. Still have it, new in the box. If you want it for $1500 let me know, its available!

  5. #25
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Virginia
    Posts
    10,264
    Post Thanks / Like
    Likes (Given)
    5843
    Likes (Received)
    4629

    Default

    Quote Originally Posted by Kezorm View Post
    The drive I am working on can be more closely compared to the GE DC300 drives that someone sourced years back (minus the regenerative braking). Or, for $2000, here's a closer new comparison to what I'm working on - DC Drives | Non-Regen.(D), 3 Phase | D2Q1-1-1 - Driveswarehouse . At a glance, it looks like it could be a good fit, but note that it needs 3-phase input.
    IF ...you have to try to cobble a DC-300 OR the Polyspede to work off single-phase?

    AND ... (think) you do not need 4Q for its regenerative braking?

    THEN ... Just go and use the kid brother of an SSD-514C-16, a 2Q Parker-SSD 512C.

    There is at least one making chips on a 10EE (Peter, in the UK).

    Price, brand-new, with warranty is around $600 and it is inherently single-phase.

    OTOH, another $100-150 gets you the 'smarter' 4Q SSD-514C-16, new, with warranty, so....

    In any case, you'll still have to deal with a regulation method for the 'field weakened' range. Which happens to be "MOST" of a 10EE's menu, even on the 2500 RPM. max machines. Do the math. it is but 'grocery store' 'rithmetic.

    N'er mind the pulley ratios for the moment. Look at the motor, bare even of its gearbox.

    Approx 60 RPM minimum-useful with the best of bog-standard controllers, 120 RPM more realistic under load before choosing to utilize the reduction gears.

    690 RPM base.

    2400 RPM nameplate do-not-exceed.

    Call that a 120 RPM to 2400 RPM range, 1710 RPM delta.

    Full field power, constant reserve torque, (~ 23 foot-pounds) 120-690 = 570 RPM delta. A bit less than 30% of the useful range..

    Weakened field 690-2400, the alleged constant HP (NO reserve..) range = 1710 delta.
    A bit over 70 % of the useful range.

    Getting load-regulation right with a low/no-inertia 'active response-only' solid-state DC Drive wants a 'field regulator'. A controllable field power source with FEEDBACK.

    BOTH Armature and Field then need that tachogenerator feedback - whether directly mechanically derived analog or done with a resolver/slot-wheel/proximity sensor via a frequency to voltage converter.

    Now ... one dasn't actually NEED the full nameplate-power of the motor. Stable RPM under load, OTOH? That we - and our tooling - DO like.

    A nominal 3 HP, 4.5+ actual motor was used so the 10EE could make chips even at partial power of 2 HP. Or even less, so long as delivered at RPM of choice. And could HOLD that RPM reasonably well.

    Close one eye to the great mass and rigidity of the overall 10EE for one second.
    Look only at the #2 MT tailstock and the compound-rest and the size of toolpost it mounts.

    Near as dammit the same size as a 3/4 HP to 1 1/2 HP SB Ten, are they not?

    The 10EE was meant for uber-stability, uber-precision, infrequent and short downtime for maintenance, and looooong life at holding tight tolerances.

    A 10EE wasn't meant for hogging.

    Monarch were happy to sell a DIFFERENT lathe for THAT sort of work. Several of 'em.

    So...for the work a 10EE was actually meant to do?

    Several folks who use Beel/BICL, SSD-51X, even KB-Penta 180 VDC drives make plenty of chips at lower loads, lower-end of the RPM band, or some combination thereof.

    Many more just run restored versions of whatever Monarch shipped it with, be they MG, WiaD, or Modular.

    How many 10EE d'you suppose even NEED a 'modernized' solid-state DC Drive in any given year?

    The one I own, EE 17120. Just because I could do. No significant chipmaking involved.

    The one YOU own?

    What's the serial number of your one?



    Bill

  6. Likes WILLEO6709 liked this post
  7. #26
    Join Date
    Mar 2014
    Country
    UNITED STATES
    State/Province
    Michigan
    Posts
    471
    Post Thanks / Like
    Likes (Given)
    88
    Likes (Received)
    254

    Default

    Another datapoint. I recently purchased, but have not yet acquired, a 1951 10EE that is a former Navy machine. The Navy had at least 6 EE's retrofitted with DC drives based on the Fincor 2601, I think back around 1994. My machine is 480V. One would hope that the DoD contractor retrofit was well engineered, and performs well. But who knows. I look forward to documenting the circuit. I had hoped to get info from the contractor but have been unable to track them down.

    If this drive has failed, or does not perform well, then I will need an alternative DC solution. The M & G are gone.

  8. #27
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Virginia
    Posts
    10,264
    Post Thanks / Like
    Likes (Given)
    5843
    Likes (Received)
    4629

    Default

    Quote Originally Posted by Glug View Post
    If this drive has failed, or does not perform well, then I will need an alternative DC solution. The M & G are gone.
    Where in the 'North East' are you? I have a 'piggyback exciter' MG unit and control Ohmites I've been trying to get rid of.

    Working when removed (in PM threads of the PO), but should have a cleanup, new bearings and brushes.

    DC Panel is already gone to another Pilgrim, but if unobtainable, a functional equivalent can be scratch-built from stock components and the advice of a couple of the grown-ups on PM.

    It also looks as if I'll have a spare 'set' of 514C-16 (Armature) + 50X (field) SSD DC Drives.

    TWO sets of 'spares' is probably overkill for the years I have left..



    Bill

  9. #28
    Join Date
    Nov 2001
    Location
    WAPELLO, IA USA
    Posts
    6,407
    Post Thanks / Like
    Likes (Given)
    29
    Likes (Received)
    1840

    Default

    from my perspective ( be warned my opinion and $7.00 may buy coffee at Starbucks ):
    I have a Beel drive 10EE
    I had a WIAD 10EE
    I have multiple MG 10EE's
    I have an ac conversion 10EE
    Macona did a servo conversion 10EE ( saw the video )
    In actual use I see no real difference between them. I can't pick any one more or less serviceable for whatever reason.
    For what AC drives cost in modern time ( especially surplus, ebay is your friend ) most will likely be following that route. Yes It ditches DC completely, yes it is newer technology, yes it "works", yes its easier to get electrical help with if required ( general electricians, phone support, etc.). I had a good motor on the beel conversion ( was working wiad but began to have more issues than I had time to correct), but if it would have had a dead motor there is no question it would be ac right now. For what I paid for the Beel it was the fast way to operation, I am happy with it too. It does not go 4000 rpm like the wiad, but my ac conversion will peg the 4k tach,,,, I don't recall the last time it was over 1500( if I really need it I have both a 4500 rpm fanuc turning center and an 8000 rm bar fed cnc swiss turning center) but I have never missed higher rpm's on the 2500 rpms machines either in real life.
    In an ac conversion 7.5hp or higher I don't really know if you NEED the gearbox. I put it in mine but I rarely put it in back gear. I use it as an ergonomic collet/ small chucker and I have larger lathes with more torque.... it taps 3/8-16 at low speed in high, thats all I need it to do.
    There is a LOT of experience on the board....I don't see where your drive would be any more functional than any of those I have run ( and seen via video, aka macona's servo machine). Maybe I am wrong, but to me if you can't feel a difference when machining why go through all the development.

  10. #29
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Virginia
    Posts
    10,264
    Post Thanks / Like
    Likes (Given)
    5843
    Likes (Received)
    4629

    Default

    Quote Originally Posted by WILLEO6709 View Post
    from my perspective ( be warned my opinion and $7.00 may buy coffee at Starbucks ):
    Quite the reverse. Your 'real world' observations, rather than unfounded opinion needed quite a few thousand dollars worth of 'lab rats' in the form of 10EE motors, hyndreds of pounds of transformers and 'swinging' chokes, several DC Drives, several new meters - even a new O'Scope.. to replicate "on the bench".

    Whilst the theoreticians and actual experimenters have been arguing, lo these many years?

    YOU have simply been making chips with what yah HAD, not what you WISHED you had.

    Cuppa coffee?

    I'd ha' been serious money ahead to have sent a case of Champagne, had I but listened better, earlier.



    Bottom line is that if all one is asking of the 10EE is 1/2 HP to 2 1/2 HP, per pass, and mostly at 1500 RPM or below, it in fact DOES NOT require DC Drives as sophisticated as all that - nor incoming voltage boosters.

    Most of what you describe could work 'well enough' with a commodity 180 VDC-out 4Q KB-Penta KBRG-255, which CANNOT (safely) be boosted as a Beel D510 or Parker-SSD 51X can be. But also costs but $ 404.40 - new, with warranty - off Galco as-at this instant.

    As a 4Q, it won't need either the braking resistors, nor ON-OFF /FWD <-> REV contactors the 1Q-only Beel D510 drive has on-platen.

    And therein lies yet another available 'good enough' solution that eats into any remaining "market" for another 'new' 10EE-bespoke DC Drive.

    Bill

  11. #30
    Join Date
    Sep 2013
    Country
    UNITED STATES
    State/Province
    Oregon
    Posts
    80
    Post Thanks / Like
    Likes (Given)
    123
    Likes (Received)
    35

    Default

    Quote Originally Posted by Kezorm View Post
    I'm in the process of developing a new drive for my '43 round-dial 10EE. Modern electronics with PFC (Power Factor Correction) input stage and full-bridge IGBT output stage. This means a reversible DC motor drive capable of full armature voltage, field weakening, closed loop speed control, etc. Designed for 230VAC single-phase input, although it will actually be able to run as low as 85VAC at reduced output power. Output is designed specifically with the 10EE 3 and 5HP motors in mind, although it will also be capable of driving 3-phase motors. Clean DC output (high frequency switching of DC bus), not the chopped up AC of the cheapy SCR drives. The initial design won't be regenerative, but it will incorporate control of a braking resistor. Not going to be cheap - I estimate bill of material costs will end up around $1500. But, it keeps the original DC motor and backgear vs. spending money on a VFD conversion or the like.

    So, here's where I need help... What inputs / outputs does the drive need? Think ideal world, not necessarily limited to only direct retrofit of a 10EE. Here's what I have so far:
    • Forward switch
    • Reverse switch
    • Left carriage travel limit switch
    • Right carriage travel limit switch
    • Left ELSR Switch (ELSR-like function without the need for the actual ELSR mechanical bits)
    • Right ELSR Switch
    • Primary Speed Control Potentiometer
    • Secondary Speed Control Potentiometer for rapid reverse
    • Output to drive contactor for braking resistor
    • Encoder input for closed loop speed control


    I'm sure I'm forgetting something or not aware of some functionality on the newer machines...

    Note that this entire effort is open-source hardware and software. I would welcome anyone who wants to contribute to the effort at any level (time and knowledge, not $$$). So far, I have all of the power design complete and am working on the final "logic" board that will have all the smarts and interface I/O - thus the question on I/O needs.

    To the same point, anyone is welcome to build their own drive from this design. Obviously the design is not yet complete and documentation is seriously lacking, but what is there (and what will be there) is free for anyone to use or modify for their own purpose. In fact, I would love if I could get a few more people to buy into building one of these drives for themselves. I've ordered 5 each of all the PCBs, so it would be nice if I could recoup some of that cost. If demand is there, I would even consider coordinating a group order / build. Component costs often come down significantly when buying more than 1.

    I'll work up some more info if anyone wants to get involved in the design at a lower level. Here are a few links to the circuit design so far. All untested at this point, but I do have boards and parts on order for the first unit.

    Input stage power board (essentially a 380VDC, 4800W DC power supply)
    CircuitMaker: MonarchDC - Power

    Input stage logic board
    CircuitMaker: MonarchDC - Logic

    Output stage power board
    CircuitMaker: MonarchMotor - Power

    Thanks!
    Greg

    No tech input Greg. Think about all the functions you expect out of your lathe and the I/O will come to you.

    Nice project....go for it and learn all you can; you'll be all the richer for it!!!

    Bill

  12. #31
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Virginia
    Posts
    10,264
    Post Thanks / Like
    Likes (Given)
    5843
    Likes (Received)
    4629

    Default

    Quote Originally Posted by morganfield52 View Post
    No tech input Greg. Think about all the functions you expect out of your lathe and the I/O will come to you.

    Nice project....go for it and learn all you can; you'll be all the richer for it!!!

    Bill
    Too true.

    Something to be said for "re-inventing the wheel, alone and in the dark" as Dr. Bill Godbout tol' me when I started adapting his 'Disk One' S-100 floppy controller to Lawrie Ackermann's S-100 6809 board...

    .. so long as ... one LEARNS sumthin' from it as will be found useful, eventually.

    Now.. as to 'useful'. No longer have an aged parent to care for, but still arm wrassling with the Estate, so.. my bench time is limited..

    'On MY drawing board'... is a PCB that takes pulses from a Miller welder wire-feeder parts-bin slot wheel read via Keyence right-angle-head fibre optics to a black-box that has in it a frequency (pulse train) to voltage IC, and.. a Johnson counter that is enabled when gearbox in/out is actioned .. and divides the pulses according to the ratio.

    Build that black-box first. Cheap. Easy. Can be tested with any-old variable-speed motor, and any of many sensors.

    Why FO to a remote BB?

    D'yah have any IDEA what the electrical noise a 10EE's motor on SCR commutation does 'up close and personal' to a vanilla 'local' slot-sensor and its single-ended low-level logic signals?

    I surely do!

    Anyway- there ain't no SPACE at the arse-end of the motor to mount any of the many 'real' analog tachogenerators I have under-roof, even if I were to want to open up the no-exposed-shaft end-bell fab a shaft teat, then also an encoder mount at that end.

    Tachogenerator is a 'bolt on' item for any of my RPM III type T motors.

    Not so with their Grandmother, the 10EE's Type T.

    Converter IC should be under three bucks. Rest of the parts under $20.

    You can Go Ogle "Johnson Counter", and why they matter in an otherwise 'powers of two' world.

    Newer tech welcome, of course. So long as it is cheaper, smaller, more robust, cannot introduce a bug, is dead-reliable, and dumb as a box of rocks. IE no programming language higher-level than solder or a wire-wrap tool.

    Pick ALL of the above, of course.



    Bill

  13. #32
    Join Date
    Mar 2016
    Country
    UNITED STATES
    State/Province
    Minnesota
    Posts
    10
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    3

    Default

    All,

    Thanks for all the great info on this thread and throughout the forum in general! Why spin my own drive? Why not?! As mentioned previously, the primary project that kicked this off is actually a 3HP PMAC (Permanent Magnet AC) motor in need of a drive. I could have cobbled something together from off-the-shelf components that would have been relatively close, but never exactly right. Figured it was time for something open source out there. This is what I do and have been doing for going on 20 years. My experience is with fractional-HP brushed DC and BLCD and PMAC up to 10HP. But, I will readily admit that I am lacking experience specific to >1HP Brushed DC motors of the type in these lathes. So, tailoring the PMAC drive to the 10EE will be a fun learning experience with regard to these motors and, best case, will get me a nice running 10EE. Worst case, it fails miserably and I fall back to an off-the-shelf drive. Worst-worst case, someone else thanks my grieving widow for a nice condition round-dial in need of a drive.

    Bill, my 10EE is s/n 18075 originally sold to the Burroughs Adding Machine Company. Seems to have led a relatively easy life. Unfortunately the M/G wasn't operating when I got it 10+ years ago (yes, the lathe has been sitting unused in the corner of my shop for 10 years ). In hindsight, I should have attempted to bring the M/G back to life, but 10 years too late for that now. I know I should just go with whatever is easiest to start making chips, but restoring the lathe and the electronics is a much of a hobby for me as is the eventual projects that will have use for the lathe.

    Thanks!
    Greg

  14. #33
    Join Date
    Apr 2010
    Country
    UNITED STATES
    State/Province
    Wisconsin
    Posts
    6,106
    Post Thanks / Like
    Likes (Given)
    4195
    Likes (Received)
    3758

    Default

    Quote Originally Posted by Kezorm View Post

    Bill, my 10EE is s/n 18075 originally sold to the Burroughs Adding Machine Company. Seems to have led a relatively easy life. Unfortunately the M/G wasn't operating when I got it 10+ years ago (yes, the lathe has been sitting unused in the corner of my shop for 10 years ). In hindsight, I should have attempted to bring the M/G back to life, but 10 years too late for that now. I know I should just go with whatever is easiest to start making chips, but restoring the lathe and the electronics is a much of a hobby for me as is the eventual projects that will have use for the lathe.

    Thanks!
    Greg
    Anecdotal information - my own 10EE was originally sold to Burroughs, as well. Serial 17978.
    The irony of the world and EE17978

  15. #34
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Virginia
    Posts
    10,264
    Post Thanks / Like
    Likes (Given)
    5843
    Likes (Received)
    4629

    Default

    Quote Originally Posted by Kezorm View Post
    Unfortunately the M/G wasn't operating when I got it 10+ years ago (yes, the lathe has been sitting unused in the corner of my shop for 10 years ). In hindsight, I should have attempted to bring the M/G back to life, but 10 years too late for that now. I know I should just go with whatever is easiest to start making chips, but restoring the lathe and the electronics is a much of a hobby for me as is the eventual projects that will have use for the lathe.

    Thanks!
    Greg
    Good on yah, Greg! Welcome aboard!

    As you can see by now, 'tough crowd' - having each and severally paid some serious dues - has a way of arse-kicking a newcomer UP the learning curve.

    Not always a pleasant experience, that.... OTOH? It is FAST, and entails minimal extraneous 'overhead'!

    As more of my time disengages from other demands, I'll be happy to onpass such observations as I have collected and am - hopefully - soon to resume.

    As to your MG unit? First - just fix it. That will let you understand a motor that is sore lacking in documentation. Take and record measurements as well.

    It also makes a great 'baseline' for the target you will be trying to hit - that of uber-smooth 'rotating power' DC source.

    PREPARE to DO the work, and PREPARE to instrument and record your work, ELSE do everything more than once... 'too much' more.

    I did NOT plan ahead as well as I should have done, have not, and will not, post fotos 'til I have cleaned-up my horrible example of a near-as-dammit 'suicide kit'.

    Damned work area looked like an Augat wire-wrap backplane, writ VERY large for a time..

    But it works, and works well, so far - even if still on 'two knob' manual balancing of Armature & Field power rather than 'ganged'. (A plus, actually...)

    Thanks to well-over a hundred pounds or iron and copper in full-isolation transformers and swinging chokes is also uncanny QUIET, electrically as well as acoustically.

    Another goal for your solid-state equivalent in there. Lower shipping costs than all that iron and copper!

    So I'll carry-on for a while yet... tacho feedback implementation the current challenge.

    Good hunting, and stay in touch!

    Bill

  16. #35
    Join Date
    Jul 2006
    Location
    Indiana
    Posts
    550
    Post Thanks / Like
    Likes (Given)
    262
    Likes (Received)
    93

    Default

    Greg - I'd be happy to review your schematics. Do I need to sign up at CircuitMaker to gain access?

  17. #36
    Join Date
    Mar 2016
    Country
    UNITED STATES
    State/Province
    Minnesota
    Posts
    10
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    3

    Default

    Quote Originally Posted by Jim McIntyre View Post
    Greg - I'd be happy to review your schematics. Do I need to sign up at CircuitMaker to gain access?
    Jim,

    Yes, looks like you need to register if you want to download from the site. But, let me know if don't want to register and I can send you a zip of all the design files.

    CircuitMaker is pretty slick for what it is - free Altium Designer at its core. De-featured from Designer and has catch that everything needs to be open source. But, for purpose of basic PCB design, is plenty. Decided to go with CircuitMaker vs. Designer because I really wanted to go open source. Eagle was another option, but boards are too big for any of the free versions. Don't have any experience with KiCAD and don't have time to learn yet another tool.

    Anyway, with CircuitMaker, the community parts library alone is fantastic. Need to pay attention to ensure the quality of the components (lots of people still on the steep slope of the learning curve), but saves loads of time overall. Also has automatic version control built in, although forces you to be connected to cloud to do anything. The software and files all run local (not a web application), but relies on the cloud for components and file repository.

    Once registered, you can download a zip of all the design files. No need to actually install the software if you don't want.

    I would love as many sets of eyes on the design as possible. Keep in mind that the design approach is design by overkill with minimal engineering. Use off the shelf components wherever possible, even if they are larger or more expensive than what could be done. For example, I use relatively expensive Recom power supply brick to get from line down to 15V. It's overkill, but I know it will just work without the need to engineer a cost-optimized switcher with custom magnetics and all that. A fully engineered, cost-optimized design could definitely be less expensive, but at tradeoff of time.

    Thanks,
    Greg

  18. #37
    Join Date
    Mar 2016
    Country
    UNITED STATES
    State/Province
    Minnesota
    Posts
    10
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    3

    Default

    Also, not sure if I made clear yet. The design is very modular. Again, there is a cost tradeoff, but I wanted the design to be as reusable as possible. For the purpose of the 10EE drive, there will be 2 separate modules, each with 2 boards.

    The first module is essentially a 380VDC, 4800W DC power supply. Made up of a power board and a logic board.

    Power board of the DC power supply is no intelligence, just the power components with a 16-pin header to connect to logic board. In this way, the controlling logic for the power board could be easily switched out for whatever purpose. Also, the power board uses 3x 1600W Power Factor Correction modules in parallel. Design and PCB layout is such that it could easily be built with 1x, 2x, or 3x depending on power requirements (1600W, 3200W, or 4800W). In interest of minimal engineering, any required filtering on the incoming line will be off the shelf filter. Again, not least expensive, but I don't have time to design and spec custom magnetics for a cost-optimized filter.

    The logic board of the DC power supply has a small microcontroller to watch incoming line, monitor current, temperature, power good signals from the PFC modules, output voltage, etc. Also has isolated USB on it for purpose of development and configuration and isolated CAN bus because... I don't know, just figured might want to do some higher level communication / coordination between the power supply and other modules (motor controller). Will probably never get that far, but... If someone else wants something different, then they could easily design a logic board for their needs without needing to worry about the engineering of the power board - an area where fewer people are likely to have the necessary experience.

    Similar for the motor controller module. Divided into 2 boards, one for power components and one for controlling logic. The design will accept up to 400VDC input and has triple 50A half-bridge output. For purpose of 10EE, I added a "Ground" terminal to the output so that can control armature with 2 legs and field with 3rd leg and ground. Power board contains all the critical bits for IGBT gate drive so interface to logic board can be just isolated low-level logic signals.

    Working on the logic board for the motor controller now. Will share as soon as in reasonable state for review.

    Thanks,
    Greg

  19. #38
    Join Date
    Nov 2004
    Location
    Webster Groves, MO
    Posts
    7,326
    Post Thanks / Like
    Likes (Given)
    1847
    Likes (Received)
    3396

    Default

    I needed to remove the front cover from my Sheldon R 15 to do some leveling, so I took a picture of its power supply. The three ganged Variacs at the top are motor driven, controlled by the up and down buttons on the panel. They are delta connected to 240 V 3 phase on the tap that gives a boost. The transformers formerly fed gallium arsenide furnaces in the Monsanto plant in St. Peter's. They installed all new equipment and I inherited them along with the contactors at the far left. The diodes are Diesel electric locomotive rectifier rejects. The technicians did a 500 V reverse leakage check and any that showed any at all were removed. Wasteful maybe, but a lot cheaper than towing a dead locomotive home. Running at the lower voltage, they have never given me problems. The connections are interesting. The transformers output something over 90 VAC. Connected in wye, the 90 V is multiplied by square root of 3. Full wave bridge rectifying that gives 6 peaks per cycle so the point where one peak ends and the next picks up the load is 30 degrees off the peak. Cos 30 is .866 and the low point is short, so the effective voltage is about 230 VDC without need for filtering. This feeds a 15 hp DC motor belted to give a 2500 RPM top spindle speed. I have never had occasion to load it to 15 hp at full speed, which would be awesome stock removal, but that gives 7.5 hp at 1250 RPM, 5 hp at 833 RPM, etc., which are very useable. None of this field weakening nonsense, I get full torque across the range. The only IR comp is what is built into the motor, so the speed does vary with load, but it is only annoying at very low speeds. Since the voltage drop with increasing current is the same at any speed but the voltage applied to the motor is less at lower speeds, it is annoying at very low ones, but I can work down to about 50 RPM in direct drive, less than 10 in backgear.

    The meters on the panel read load on the left and RPM on the right. The center one reads armature voltage before closing a contactor and is calibrated to give an idea of the RPM when running. The neon lights show when the Variacs are at zero or full output limits.

    There are no MOSFETS to blow up if I look at them crosseyed, no computers to program. The space occupied by the supply wasn't doing anything anyway and the weight helps hold the floor down.

    sheldon-power-supply.jpgsheldon-control-panel.jpg

  20. #39
    Join Date
    Mar 2016
    Country
    UNITED STATES
    State/Province
    Minnesota
    Posts
    10
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    3

    Default

    Bit slow around here, so figured to share some (very slow) progress on the new controller. 380VDC, 4800W power supply is almost complete. So far, it's been up to about 800W output (3x 250W heat lamps in series). Next step is full 4800W load (3x 1500W water heater elements in series). So far so good. Clean 380VDC+ output, even when powered by 120VAC input (but is limited to 3000W at 120VAC input). Power Factor Correction appears to be working nicely as can be seen in scope capture of input voltage and current at 800W output (yellow voltage, blue current).

    img_0978.jpgmonarchdc.jpgimg_1051.jpg

  21. #40
    Join Date
    Sep 2007
    Location
    Spanish Springs, NV
    Posts
    2,365
    Post Thanks / Like
    Likes (Given)
    155
    Likes (Received)
    292

    Default

    I'm no electrical/electronic dummy but after reading through this thread it's time to go hug my MG. AC to DC when there was no other economical path to it appeals to me.


Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •