Anybody familiar with Drukov grinder? - Page 4
Close
Login to Your Account
Page 4 of 10 FirstFirst ... 23456 ... LastLast
Results 61 to 80 of 181
  1. #61
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    So I've looked for accelerometers that are sensitive in the 60Hz range. The apps for my Android only go up to 20Hz from what I could find. I'm looking for accelerometers with a large signal per G. I think these piezo types, require a special charge amplifier. Before you know it, you end up with a full electronics lab around your grinding wheels.

    It seems like an interesting way to balance a wheel. The weight adapter does not look difficult in making. I already have issues with the Sopco hubs I got off ebay that prevent the cover of the wheel guard being installed. So I need to cut a large hole for the protrusion of these deeper hubs.

    eKretz; have you found any source for grinding wheels that fit the 3mm and 6mm internal arbors? I see one on eBay but are CBN that fit the 3mm arbor (8x10x3mm (GRIT240 = 63/50um)), seems very fine, but the alum oxide that came with the grinder is fine as well. Is vitrified CBN a dressable wheel?
    5 PCS CBN(BORAZON) GRINDING WHEELS 1A1 8x10x3mm (GRIT240 = 63/50um) | eBay

    One task I have is the internal bore of a MT4, and that requires a very long internal arbor. I wonder how this is done commercially. Seems if heat treated, grinding is the only way to complete the manufacture.
    .

  2. #62
    Join Date
    Mar 2005
    Country
    UNITED STATES
    State/Province
    Indiana
    Posts
    5,872
    Post Thanks / Like
    Likes (Given)
    1997
    Likes (Received)
    1944

    Default

    Lucky for me I've already got a pretty full electronics lab, haha.

    I haven't had to look for any wheels yet since my grinder came with quite a stack of all sizes of wheel. If you get regular AlOx wheels with smaller bores than you need you can readily size them to whatever you need. I have used a plain drill of the correct size to drill out a grinding wheel bore with no issues. Doesn't seem like it'd work but it does.

    The vitrified CBN wheels should be able to be dressed, probably in a similar fashion to metal bond I'd guess.

    As far as MT adapters, yes in industry I'm sure they're ground after heat treating, but probably not with a toolpost grinder, heh. You'll just have to make an arbor if none that you have are long enough.

  3. #63
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    I purchased some MMA7361 3axis accelerometers from ebay. And now the wait for shipping from China. These put out 800mv/g. I hope there's enough vibration to see some analog signal well above the noise floor. We shall see.

    I did regrind the lathe's headstock adapter, as well with the tool post angle set, I changed out the spindle for the long internal one, and reground the spindle. Real hard steel, and the unknown gray grinding wheel had difficulties.
    High spot bluing shows it now contacts at the entry for over 3/4" and at the bottom for about 1/2" of contact. The spindle taper seems to have a hollow in between these, and it was more material then I wanted to remove. Now when I push the adapter in, with just a tad bit of velocity, it takes some major hammering with my brass rod to remove the adapter.
    20210206_192228.jpg

    https://youtu.be/p35xlK0gBhI

    I don't know why I have that wood grain pattern in the grind. It is smooth to the touch, and I don't see ripples.

  4. #64
    Join Date
    Mar 2005
    Country
    UNITED STATES
    State/Province
    Indiana
    Posts
    5,872
    Post Thanks / Like
    Likes (Given)
    1997
    Likes (Received)
    1944

    Default

    Cool. Update us on the accelerometers once you try them out. Do you have some sort of analyzer or are you going to read the output with an oscilloscope? Digital photos are pretty notorious for adding weird artifacts during compression, probably explains the woodgrain effect.

  5. #65
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    I have a Rigol Oscope that can do FFTs. I've never have used its math features, I'm big on troubleshooting seeing the time varying waveform. I hope it can give a decent measurement, and I'm interested if the FFT will single out the low frequency (55-60Hz) resonances from the accelerometer. I probably need to amplify the signal to scale it for the oscope input signal range (I think the scope has an 8 bit ADC). I don't expect the vibration to be detected beyond a few percent of 1g, so sub millivolt output is my current thinking (and lots of noise). The scope is the DS1054Z model, 4 channel, so I can look at the X, Y, Z output from the accelerometer. I'm thinking I should just double stick tape the accelerometer circuit board to the spindle near the wheel. I'll also see if just a good handheld DVM will see a good AC signal in the low range that again is the signal of interest, and not noise.
    Then I need to make that balancing adapter to mount with the grinding hub. I don't have a scale to measure sub grams, so just what I needed was a reason to buy another tool.


    I can see the wood grain pattern in the grind, unless my eyes are going along with the rest of what used to work 15 years ago. It looks like a Damascus steel pattern from forging, but someone familiar with grinding probably can say it's a wheel balance or other resonance issue that leaves this pattern.

  6. #66
    Join Date
    Mar 2005
    Country
    UNITED STATES
    State/Province
    Indiana
    Posts
    5,872
    Post Thanks / Like
    Likes (Given)
    1997
    Likes (Received)
    1944

    Default

    Yeah must be resonance of some sort. I was thinking it was only in the photo. I think your scope will pick up the vibration just fine. I have a couple Tektronix bench scopes, but I've been eyeing some of the newer models. I'd really like to get my hands on a fully isolated handheld. I have an Ohaus digital scale I bought 15 years ago for reloading, it does milligrams, I should try this balancing out too, seems I already have just about everything I need.

    Hopefully the double sided tape will work, I'd be sure to use as thin a tape as possible so it doesn't have that "cushion" to move around erratically.

  7. #67
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    I have two Tek 545B Oscopes (space heaters with minions of tubes) and a newer one, with a CRT but custom chip sets (but pre-nonstorage). Amazingly, I powered up both the antique tube scopes last week, and they powered up, and generated a sweep. I would have thought as old as these are the electrolytic's would have dried out. But I got spoiled with the scopes at work before retiring, so the Rigol was purchased as a large LCD 4 channel. I need to read up on how to store waveforms to download to a PC. I'll post screen shots of the accelerometer output. The Z axis is perpendicular to the circuit card the IC is mounted on, and that has a smaller full scale resolution in that axis. But I don't expect that the level will peak out. And as I believe, I'll have issues with getting enough scaling of the analog output, if not amplified. We shall see, but it's not a big deal to insert an opamp.
    I did do some research on accelerometers that use the piezo effect. The signal needs to be amplified by a FET input opamp, and setup for a charge amplifier topology. Which looks like an integrator amplifier type. They just want too much for these accelerometers. Used ones on eBay have odd ball connectors, that I didn't want to mess with. I did some research on the piezo disk type microphones they use for cigar box guitars. That's a backup if the accelerometer IC fails to pick up the signal adequately. I trust it will work.

  8. Likes eKretz liked this post
  9. #68
    Join Date
    Mar 2005
    Country
    UNITED STATES
    State/Province
    Indiana
    Posts
    5,872
    Post Thanks / Like
    Likes (Given)
    1997
    Likes (Received)
    1944

    Default

    Mine are Tek TDS754D hardware modded to TDS784D at 1GHz, so respectable, decent but not very portable. I am looking forward to seeing how good this works for you.

  10. Likes ignator liked this post
  11. #69
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    What are you fixing with that scope? Gets real tough to keep ground inductance path (for the probe) from giving a clean waveform. I only used that sort of scope at work, when seeing how bad the PC board signal integrity was, on digital waveforms, where 'reflections' became an issue.
    Memory of needing a scope for that was edge rate that caused upset of a late 1980's board layout (10MHz 80186 microprocessor serial bus data switch product). Traced the problem to newer digital microprocessor bus components that had a rise/fall time 20x better then their original specs. Still was old +5 TTL logic, but ACT vendor improvements. The only fix was to specify components that had a 25ohm output resistor in the totem-pole drive, or find slower part types. This softened the edge rate. I was seeing reflections that were causing the bus voltages to be 6volts below ground and +7 above the 5VCC power rail. That does odd things to ICs. In this case the chip select to a custom (ARINC429) UART caused it to randomly set the parity error which caused it to drop the current data being received (idea was the negative edge sucked current from the substrate which caused the parity flip-flop to be triggered). This resulted in flight director flags on the customers aircraft from the data loss. Don't know why I remember all this detail from 2004. The board needed to be relayed out, but it was certified product, and all the testing would have taken too long, and this was causing flight cancellations.

  12. #70
    Join Date
    Mar 2005
    Country
    UNITED STATES
    State/Province
    Indiana
    Posts
    5,872
    Post Thanks / Like
    Likes (Given)
    1997
    Likes (Received)
    1944

    Default

    They are complete overkill for any application I'll use them for. I don't generally do much that I couldn't do with a plain old CRT scope, but the storage is nice now and then for troubleshooting. I do mostly little electronics repairs for myself and a little tinkering with H-bridge motor drives and such. I got them both for a song so I didn't have much outlay. Needed some repairs but no biggie. One was having button troubles with the membrane keypad and the other had some relays in the signal path that were starting to have troubles with high resistance, just needed replaced. Also one of the CRTs was very dim and out of focus but I found a lot of info about the adjustment pot locations and functions on the old Tek message forum and was able to fix it right up.

  13. #71
    Join Date
    Mar 2005
    Country
    UNITED STATES
    State/Province
    Indiana
    Posts
    5,872
    Post Thanks / Like
    Likes (Given)
    1997
    Likes (Received)
    1944

    Default

    Did you get around to trying out your vibration analysis yet? I just snagged a nice Hofmann accelerometer for cheap, but still need to pick up a charge amp for it.

  14. #72
    Join Date
    Sep 2011
    Country
    UNITED STATES
    State/Province
    Virginia
    Posts
    18,940
    Post Thanks / Like
    Likes (Given)
    7561
    Likes (Received)
    11648

    Default

    Quote Originally Posted by ignator View Post
    I purchased some MMA7361 3axis accelerometers from ebay. And now the wait for shipping from China.
    "While you wait".... won't be good enough to quantify it, but should be good enough to DETECT a resonance source, yah use it in a "stethoscope" manner, wander the machine-tool a-hunting...

    A "contact" audio microphone, AKA "inertial transducer" AKA "bone vibrator", hearing aid industry. Early 1970's Esterline Medical put me onto a long string of "go play detective" airplane rides when the Old-Line hearing aid firm they had bought was found to be shipping about 18,000 more bone vibrators a year than we - and the competitors we supplied - made as hearing aids for folks that even NEEDED bone-vibrators.

    Where where they GOING?

    Jim Helle's "Hellephone", "wireless" diver's voice network, Exxon's Deep Water Development - monitoring flow and valve opening and closing on seafloor oil rigs, Bev Morgan's depth-record holding "Kirby-Morgan Band mask" for free diving... Jacques Costeaus' "US Divers" communcations SCUBA masks..."etc".

    Three of the brighter minds under-roof bought it up, made a good run out of it off the transducers, without the hearing aids:

    About us - RadioEar

    Experienced human brain did the rest. "Scope optional.

    Audio band, MOST "puter's usually have "spectrum analyzer" display in their multi-media suites, hand-held "puters included.

    Simple. Effective. Rugged. Easy to find. FAST... might be "there, already?", most ordinary homes, not just "labs"... ergo Cheap.

  15. #73
    Join Date
    Sep 2020
    Country
    NAMIBIA
    Posts
    199
    Post Thanks / Like
    Likes (Given)
    73
    Likes (Received)
    34

    Default


  16. #74
    Join Date
    Sep 2020
    Country
    NAMIBIA
    Posts
    199
    Post Thanks / Like
    Likes (Given)
    73
    Likes (Received)
    34

    Default

    Quote Originally Posted by ignator View Post
    So I've looked for accelerometers that are sensitive in the 60Hz range. The apps for my Android only go up to 20Hz from what I could find. I'm looking for accelerometers with a large signal per G. I think these piezo types, require a special charge amplifier. Before you know it, you end up with a full electronics lab around your grinding wheels.

    It seems like an interesting way to balance a wheel. The weight adapter does not look difficult in making. I already have issues with the Sopco hubs I got off ebay that prevent the cover of the wheel guard being installed. So I need to cut a large hole for the protrusion of these deeper hubs.

    eKretz; have you found any source for grinding wheels that fit the 3mm and 6mm internal arbors? I see one on eBay but are CBN that fit the 3mm arbor (8x10x3mm (GRIT240 = 63/50um)), seems very fine, but the alum oxide that came with the grinder is fine as well. Is vitrified CBN a dressable wheel?
    5 PCS CBN(BORAZON) GRINDING WHEELS 1A1 8x10x3mm (GRIT240 = 63/50um) | eBay

    One task I have is the internal bore of a MT4, and that requires a very long internal arbor. I wonder how this is done commercially. Seems if heat treated, grinding is the only way to complete the manufacture.
    .
    We used to keep an eye on some very expensive bearings ( and re-balance things when needed ) using ADXL335 and MMA7361. An oscilloscope does not have enough resolution and processing power for this ( FFT !) task. Any half decent sound card will do just fine though. We use an OLD ( 2003) Julia ESI 24bit running at 96Khz and around 16k kilosamples ( to get enough resolution ). Software is SpectraPlus which is available for free for 30 days( SpectraPLUS Downloads : Acoustics Software : SpectraPlus.com ) . Suppliers of Arduino things in your area should have both accelerometers. I think ADXL335 has been discontinued but ample stocks exist for the moment.

  17. #75
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    Quote Originally Posted by eKretz View Post
    Did you get around to trying out your vibration analysis yet? I just snagged a nice Hofmann accelerometer for cheap, but still need to pick up a charge amp for it.

    I looked at going that path as well, but the charge amps were crazy in price. If you search for charge amplifier, you can find some pretty simple ones built with cheep opamps.. They seem to all have a feedback capacitor used to control gain.
    I have played with the accelerometer for the past week. I have a treadmill motor on the bench with a piece of modeling clay to cause imbalance with the flywheel. It still needs some work. I built it up with the solderless type breadboard. 60Hz noise is a real problem, so I need to get it all put in an aluminum box to shield it. I have a gain of 10 right now, but I need to get a high impedance amp on the output of the accelerometer, as it is 32Kohm output impedance as there's some sort of switched capacitor circuit they use to filter the output. It also has a DC offset, as it's measuring the earths acceleration. So I'm coupling it like audio and only amplifying the AC component. In fact that's the circuit I'm using with a TL064 opamp. I'm using a 9v battery for power, as I want to get away from things that inject 60Hz into the air with their power cords.
    I'll post the circuit I come up with, after it really finds the noise.
    Then I need to build that weight hub.

  18. Likes eKretz liked this post
  19. #76
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    Quote Originally Posted by Orbital77 View Post
    We used to keep an eye on some very expensive bearings ( and re-balance things when needed ) using ADXL335 and MMA7361. An oscilloscope does not have enough resolution and processing power for this ( FFT !) task. Any half decent sound card will do just fine though. We use an OLD ( 2003) Julia ESI 24bit running at 96Khz and around 16k kilosamples ( to get enough resolution ). Software is SpectraPlus which is available for free for 30 days( SpectraPLUS Downloads : Acoustics Software : SpectraPlus.com ) . Suppliers of Arduino things in your area should have both accelerometers. I think ADXL335 has been discontinued but ample stocks exist for the moment.

    I was seeing that same issue, and am working with an amplifier to get enough resolution. The vibration of interest is below 60Hz in this case, as it's from a AC motor driving the spindle with direct drive in my surface grinder. The tool post grinder is another project as that is belt drive. And I'm experiencing wheel imbalance problems there as well.

  20. #77
    Join Date
    Mar 2005
    Country
    UNITED STATES
    State/Province
    Indiana
    Posts
    5,872
    Post Thanks / Like
    Likes (Given)
    1997
    Likes (Received)
    1944

    Default

    Yeah the good charge amps are a little expensive. So are the good accelerometers! I'm not in a rush, figure I'll wait patiently until a good deal comes along. Some of the better charge amps have adjustable frequency limiting built in, probably going to try to snatch one of those up. That one Conrad has is nice, but I don't see any of those up for sale anywhere. For our purposes, we can pretty well just monitor RMS output of the accelerometer using the spreadsheet method, can't we? No need to get too carried away worrying about actual amplitude values, only relative ones. Any residual values like gravity or spin of the Earth should cancel out I'd think.

  21. #78
    Join Date
    Sep 2020
    Country
    NAMIBIA
    Posts
    199
    Post Thanks / Like
    Likes (Given)
    73
    Likes (Received)
    34

    Default

    Quote Originally Posted by ignator View Post
    I was seeing that same issue, and am working with an amplifier to get enough resolution. The vibration of interest is below 60Hz in this case, as it's from a AC motor driving the spindle with direct drive in my surface grinder. The tool post grinder is another project as that is belt drive. And I'm experiencing wheel imbalance problems there as well.
    Off the top of my ( old ) head there are no res. limitations with 335 or 7361 as they are both analog voltage output. I don't remember us using any amplifier but I might be wrong. I'll test and get back to you. 60Hz is right where they like to be and if you overload them a quick solution is to reduce the shaft speed. Bit of a problem is the "oscilloscope" part of SpectraPlus which is not quite there - real irritating to use. I think we eventually used Sigview ( also 30 days free... ) . The right thing to do though is to re-do all the software using Matlab, digital filters and all. I am too old and senile for such thing but a Matlab expert should be able to put something together is days. Though most aggravations fall apart if you look at only one period.

    For your grinding problems I suggest testing for vibrations at different shaft speeds.

    For balancing you need an index signal - some sort of photodiode thingy. If you find ( or make ) anything which works well I'd like to get some feedback.

  22. #79
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    I'm more fighting all the other noise on the scope screen, I want to be sure I'm measuring the vibration signal from the out-of-balanced wheel, and not electric noise from the scope electronics, or the 60Hz magnetic fields from the motor. If I can just get the RMS of the wheel balance accelerations, life will be easy.
    I'm unimpressed by the RIGOL DS1054Z, I've run into one software bug that locks it up playing with the FFT. On power cycle, it's supposed to keep all the same previous powered settings, but changes the filters for the vertical gain to the default of no filter vs the 20MHz (which would be nice if this could be limited more, as there is all the darn AM broadcast noise being mixed in).

  23. #80
    Join Date
    Oct 2012
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    403
    Post Thanks / Like
    Likes (Given)
    156
    Likes (Received)
    92

    Default

    I'm stuck at the phase speed minus slip. I wish I had a 3phase motor on both the tool post and the surface grinder.
    The method of balance is outlined in
    Grinding Wheel Balancing
    And only requires relative RMS measurements of the mis-balance acceleration. That and the spread sheet, which I've not reverse engineered.


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
  •