Connect with us

Transducer AC mv/V to 0-5 VDC Conversion........... I think?

Discussion in 'General Electronics Discussion' started by Peter Hurley, Mar 8, 2016.

Scroll to continue with content
  1. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016
    Hello All,

    I am in the process of building a dyno for my motorcycle shop. I have acquired an AC powered rotary transducer (RT) to measure bike torque. The RT is a Himmelstein 28004T and I have its matching exciter which powers it with 6v RMS at 3kHz. It's output is 1.5 mv/V.

    What I want to do is..... use a DAQ system with integrated computer GUI from Performance Trends (PT). PT's DAQ allows for a slew of other pertinent test-cell componentry, has integrated computer software, etc..... Hence the reason I wish to use their system. BUT... the only transducer output their DAQ can read is a 0-5vdc circuit.

    It appears to me that I can excite/power the RT and pick up its output signal, convert it to the 0-5vdc and plumb it into the DAQ.

    First.... is this possible?
    Secondly.... If so... with what and how?
    Thirdly.... any thought?

    Thank you
    Peter
     
  2. duke37

    duke37

    5,343
    753
    Jan 9, 2011
    Is the output of the transducer AC or DC?
    1.5 mV/V?
    What is the Voltage? Is this the input AC voltage?
    Is 1.5mV what you get at full scale torque?
    Sounds like an LVDT type of transducer.

    It should be possible to amplify an AC signal and then use a precision rectifier to give a 0 to 5V dc output.
    There may be trouble with noise pick-upnext to a petrol engine. This can be reduced by using a synchronous detector locked to the 3kHz source.
     
  3. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016
    Hi Duke,

    Thanks for chiming in so quickly. I suspect since the input power to the RT is AC (6vrms input), it will output AC. I am 99.9% confident is does NOT have a built in modulator/rectifier/etc.... so it should be outputting AC also.

    The specs on the RT positively state 1.5mv/V. So the full scale output should run from 0-9mv. But then again.... I am a neophyte when it comes to electronics and look at like this: If 6 volts goes in and they specify 1.5mv per v.... I believe its full scale output will be 9mv. If my thinking is wrong on this, please correct me.

    It is sort of an LVDT.... except it's an RVDT. And, the dyno is being built under the floor and away from all electrical noise issues.
     
  4. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016

    One other thing. I was just going over the Signal Conditioner (exciter) specs and find it amplifies the signal to +-5v. Finding where to pick up this signal off this board may be beyond my capabilities. But at least it's there somewhere and would make converting the AC signal to a DC signal a lot easier.
     
  5. duke37

    duke37

    5,343
    753
    Jan 9, 2011
    I had a look at the torque meter specs and see that it is a strain guage type.

    The signal conditioner will doubtless produce the excitation and also demodulates to DC. It needs quite some effort to read through the details of what you have.
     
  6. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016
    That's what I was thinking. To complicate the whole even more........... the signal conditioner is part of a whole console (Himmelstein 66032) that houses it, a speed card (frequency counter) and a digital LED readout. It all works but I want to run the signal into the computer's GUI. There are also outlet cards that couple to a computer via ASCII but one may(?) be an analog output. I am reading up on it more but very complicated. I would really like to get rid of the console. It's huge.... like 18" x 18" x 5" and made for rack mounting.
     
  7. Alec_t

    Alec_t

    2,788
    742
    Jul 7, 2015
    Without the manufacturer's conditioner, how would you calibrate the torque sensor? The '1mV/V' spec seems meaningless to me in the absence of data showing the relationship between that figure and torque.
     
  8. hevans1944

    hevans1944 Hop - AC8NS

    4,459
    2,074
    Jun 21, 2012
    There is a Himmelstein Signal Conditioner Model 721 that will provide ±5 and ±10 VDC outputs. It doesn't appear to require a rack-mounted console either. You should look into it.

    It would help if you provide us the model/part numbers on the "Signal Conditioner (exciter)" you are currently using. Did you acquire this rather fine instrument as part of a larger package, perhaps at auction or surplus sale or on eBay? Is it this one?

    Your torque sensor is a steel shaft with four strain gauges applied around the periphery of the shaft in such a manner that axial and bending loads are canceled and only torsion in the shaft is measured. The strain gauges are no doubt arranged in a Wheatstone bridge configuration with 3 kHz excitation applied to two opposite corners and the resulting strain gauge unbalance signal taken from the other two opposite corners. Because the shaft is rotating, the excitation and output signals must be transformer-coupled to the shaft using (no doubt proprietary) rotary transformers without ferrite cores. Whatever the implementation details, you simply apply AC excitation and receive millivolt AC signals with 1 mV/V sensitivity, which is typical of strain gauge bridge instrumentation.

    A signal conditioning chain will amplify the millivolt output signal by a gain of about 1000 or so, turning the 1 mV/V sensitivity into 1 V/V. Then this high level AC signal will be demodulated, typically by synchronous rectification, and then filtered to produce ±5 V DC or ±10 V DC analog outputs. More sophisticated processing may also be employed to yield digital results for display on a digital LCD panel display, or as serial data for interfacing to a computer. There might even be an IEEE 488 parallel interface. You pays your money and you makes your choices, but it sounds like you don't need a bunch of bells and whistles for your application.

    On strain gauge signal conditioning and calibration: if you can get what you need at an affordable price, it will be less expensive and less bother in the long run to purchase a commercial product.
     

    Attached Files:

  9. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016

    Yeah Alec, you have a good point there. Himmelstein's conditioner has a resistor specific shunt calibration circuit built into it. The resistor can be changed to match the RT itself. We can also (will need to) dead weight calibrate the two once everything is installed and operative.
     
  10. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016
    Good Stuff Hevans. And yes, a full Wheatstone bridge as you mentioned. And yes.... eBay on the RT and signal conditioner/console. Also received some good operation's manuals from Himmelstein themselves. I was surprised in consideration of their age. Himmelstens 721 would be great.... BUT they are many thousands of dollars and beyond my build budget, hence the eBay purchases of various old, but very nice, components. What we are building is a hydraulic steady dyno specific to motorcycles and being installed in our new bike shop. No way to install a commercial unit here and their cost is 10's of thousands of dollar, especially a steady waterbrake/hydrualic type of dyno.

    Anyway. I am including photos of the front and back of the console. It does have input/output cards as shown. The console back has IEEE488 stamped on it but I think that is generic. As Himmelstein has "checked" the check boxes for two ASCII installations. I have mocked this up with its speed sensor (frequency counter) and the Rotary Transducer. The console readout works fine and shows the speed works and it appears the transducer is operating too. It's hard to apply any torque to the transducer but we were able to get it to read up and down with what little we could apply. The RT is a 450NM (4000 lbf-in) unit.

    [​IMG]

    [​IMG]



     
  11. hevans1944

    hevans1944 Hop - AC8NS

    4,459
    2,074
    Jun 21, 2012
    Yikes! You need to add a ten foot lever arm and hang some weights to calibrate that puppy. I found a spec sheet that says your signal conditioner does provide ±5 V DC analog outputs. Your mission, should you choose to accept it, Mr. Hurley, is to find where on the edge connectors the analog signals appear. I couldn't find a user manual for the SHC 66032 signal conditioner, but I sure wouldn't give up now. I tried to uploaded the datasheet for this (and other models) that indicate the analog output is available, but the server for this site tossed an error... You can find it on the Internet here.

    Please let us know if you find where the analog signal appears.
     
  12. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016
    Duke,

    You should see the hydraulic pump we're running!!! It's a set of two piggybacked Rexroth Piston Pumps weighing in at over 200 lbs and will absorb around 225 hp. Obviously, we will never need to run it at that level. But 150 may be in order at some point. Mostly in the 30 to 90 range and more so for engine break-ins and fine tuning, etc... It has been and will continue to be a fun project. I've been designing it for months now but some aspects are a bit frustrating.... like this one.

    I have decided it will be best to set it up with the Himmelstein conditioner/amplifier/readout for the moment. I will tinker with the output signals from the console and post my findings. The Operator's Manual, along with others, shows how to build a proper Star Bridge for conditioner troubleshooting. I may do this in order to set up the whole unit initially. Tomorrow, I will upload the 66032 operator's manual and the RT Installation Manual to my site and post a link here............ if you care to look it over.

    Meanwhile, I am still interested in minimizing the conditioner interface and finding a proper (short path) means of adding it into the computer GUI in order to tie it together with the other sensors. I also will be seeking some help with establishing a DAQ for thermocouples, speed sensors, O2 sensors, etc.... in the near future. That will be another thread.

    Thanks
    Pete



     
  13. hevans1944

    hevans1944 Hop - AC8NS

    4,459
    2,074
    Jun 21, 2012
    When you first posted this I was thinking, "Hmmm... large relay rack, maybe with a table, whirring and spinning magnetic tape drives, blinking lights all over the rack panels, white-coated geek-priests in attendance... very Star Trekkie and not exactly my vision of a custom performance MC shop style... not even close to Orange County Choppers, but maybe aspiring." So then I read up on the Himmelstein product line and realized these folks provide some serious instrumentation for anyone with deep pockets... like maybe 10,000 BHP marine engine manufacturers. What's a small MC shop gonna do?!

    Well, you could build something, if you are electronically so inclined and visit forums like Electronics Point, and have more time than actual cash available to work with. We can help you design and build a custom 3 kHz sinusoidal exciter for the strain gauges on your torque transducer. You will also need an instrumentation-grade differential amplifier circuit to get the signal amplitude (1 mV/V full-scale) up to a reasonable level prior to rectification. I doubt you need anything more than an "absolute value" full-wave rectification circuit, followed by some low-pass filtering. The only concerns I would have about "home brew" versus "commercial" product are accuracy, drift, and calibration, all of which depend on the competence of the constructor.

    One of my first jobs (late 1960s to early 1970s) involved signal conditioning for strain-gauge load cells, fairly large ones that had full-scale ranges of 500,000 Lbf for instrumenting fatigue-test machines at Wright-Patterson AFB in Dayton, Ohio. A company in California had won a large contract to build a fatigue test machine from scratch. At that time there were several highly qualified companies that specialized in the art (Instron and MTS Systems Incorporated come to mind). The Air Force was using ancient machines, captured from the Germans after WWII, and wanted to procure a large new fatigue test machine through competitive bidding. The German stuff was very good technology for its time, but hard to maintain because all the docs were written in German and replacement parts were virtually unobtainable.

    The "new and improved" fatigue tester failed miserably after it was installed because the California engineers had decided to build their own high-performance servo valve, hydraulic actuator, and custom electronics. When I was assigned to help salvage the project, about the only things that actually worked were the hydraulic actuator, the in-line Lebow commercial load cell, and specimen grips. Even the upper load frame, which was mounted between two large jack screws with Acme threads, was often jammed in place because the driving nuts that were supposed to raise or lower it got out of sync. The contractor eschewed a commercial hydraulic power supply and plumbed together a real kluge that operated on regular motor oil instead of commercial hydraulic fluid... probably not a problem if all you want to do is hoist a car up and down on a lift, but this was a big problem for the mechanical engineer in charge of the project. He replaced the ersatz "servo" valve with a high-performance Moog servo valve and IIRC overhauled the hydraulic power supply and used mil-spec hydraulic fluid in place of motor oil. About a year's worth of work to finish that.

    By comparison, all I had to do was get the servo electronics working so dynamic fatigue tests at frequencies up to 50 Hz and ±500,000 Lbf could be performed while the results were recorded on strip-chart recorders. I should perhaps mention that the load cells were DC excited, as were most of the extensiometers used to measure strain on the test specimens. We did have some long-stroke LVDTs to control piston extension from the hydraulic actuator, mostly used during set-up while positioning specimen grips, but these also used commercial signal conditioners IIRC, so no problemo with the LVDTs.

    So, if you want to build your own signal conditioner with DC output... I am game to help. OTOH, if you find the DC analog output signals, you can just shove the SHC 66032 out of sight someplace after setting up the gain and offset to comply with your computer data acquisition system.

    BTW, where is your shop located? Can you post a link to your web page if one exists? Or PM that information if the moderators deem that too commercial. We normally offer advice for free here because Electronics Point is a hobbyist site not intended to promote commercial enterprises.
     
  14. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016
    I am hopeful you are correct about stuffing the console away and I suspect you are correct. I will post the Console Op's Manual for your review. I am located north central WV. My site is for my wood products manufacturing business....... highly unlikely anyone here would be interested in wildlife feeders. The bike business is going to be the end of work life biz for myself. I have about 13 years until formal retirement and I wish to do something new and far more interesting than building wildlife feeders, although the feeder biz will always operate.

    I agree. Even the old Himmelstein items are still killer products and will suit what I am doing very well...... if I can work all the bugs out. Hence the reason I think it best to use the whole set up together on the front end. I am positive the console will allow outgoing data to be collected into a GUI. If it actually does offer a 0-5vdc output signal, I can simply plug it into the DAQ/computer program I am looking to purchase.

    Maybe Duke, you can find this info within the ops manual I'll attach. I have yet to read through the whole thing. I will have to test the console either way to find the signals.

    Thanks
    Pete
    PS: if you care to see the best wildlife feeders anywhere.... hurleybyrd dot com.
    PSS: The 6-488B file is the console ops manual. The mv_install..... is the Transducer install manual.



     

    Attached Files:

  15. hevans1944

    hevans1944 Hop - AC8NS

    4,459
    2,074
    Jun 21, 2012
    The 6-488B Manual was quite informative... and it was like a way-back machine that took me to the 1980s and Z-80 microprocessors... it is really quite amazing what creative engineers and programmers could wring out of that "primitive" machine for a dedicated application... floating-point arithmetic not a problem! LED numeric and alpha displays... not a problem! Human-machine-interface (HMI)... ugh. Users gotta remember a lot of cryptic commands and data entry protocols. Gotta use, or work around, functions you may or may not need. One size fits everybody, like it or not, but what's not to like? Well, the only output seems to be cryptic serial ASCII characters or obscure parallel printer (Centronics standard, perhaps?) data. The IEEE-488 parallel data interface was available as an option but is not installed. Who speaks IEEE-488 nowadays anyway? For that matter, who still uses parallel printer interfaces or serial RS-232 communications? Well, I still do, for some things, but like most folks I have also moved on to USB 3.0 for every-day machine-to-machine communications.

    Fast forward thirty-something years. All the 1980s tech now fits in your Apple Smartwatch and can (probably) be downloaded as an app on your Android or iPhone. But the SHC 66032 is still hanging in there, unwanted features and all... except for the 3 kHz excitation, phase demodulation and DC amplification circuits. This analog processing normally feeds into a high-speed analog-to-digital converter and the rest of the signal chain is all-digital floating-point arithmetic with plenty of bells and whistles added. If you don't need or want the bells and whistles, there may be a high-level analog signal available on edge connector J106 (System Input/Output) pin 12 (Analog Signal Ground) and pin 19 (Analog Module Signal). At least according to the 6-488B Manual, page 96.

    Please get out your trusty multimeter and measure the signal on those two pins. You need to put some torque on the transducer of course. Please let us know what you find out.
     
    duke37 likes this.
  16. Peter Hurley

    Peter Hurley

    28
    1
    Mar 8, 2016
    Duke,

    Just sent you an email from my biz email address.

    Pete
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-