# Vibe Table: Increase Peak Force Output?

I'm driving a small, benchtop vibe table (shaker) with two Crown DSI-1000 PA's, and also cleaning my signal with a DBX 231s EQ. Currently running the system at 6.8 GRMS, but I'm looking to increase my overall driving level.

With the test article mass/ test levels I'm shooting for, I'm conflicting with the specified peak capability of my shaker.

To be more precise, I would like to drive a ~5.5 lb object to ~10 grms (random vibe). According to my back-of-the-envelope calcs (Eqn 1), I don't think there is much I can do besides purchasing a larger shaker. This shaker tops out at 200 lbf. However I figured it would be worthwhile to probe the community for ideas in case there is some sort of electrical engineering voodoo I can employ. Will also post in engineering SE.

TL:DR, Is there any way for me to increase the peak driving force/ capabiltiy of my shaker table without the risk of damaging it?

My PA's and EQ are also getting close to their maximums, at just 6.8 grms.

Desired profile: NASA GEVs STD-7000a:

Eqn 1: GRMS = [PEAK_SHAKER_FORCE / (ARTICLE_MASS + ARMATURE_MASS)] / 2 . --> GRMS = (200 lbs/ 13.5 lbs) / 2 ...................................................................... --> GRMS_MAX = 7.4

• Increase by 1 g using a horizontal slider axis and use accelerometer feedback to flatten response, what is the limit voltage or current? – Sunnyskyguy EE75 Aug 13 '18 at 19:04
• @TonyEErocketscientist : By "horizontal slider axis" do you mean a linear translation stage? I'm hesistant to move forward with a solution that adds mass, since more mass will only lower my max output ceiling. Also, really only interested in driving the signal in 1 axis at a time (this is a vertical shaker, so a horizontal stage would add some complexity to my feedback/ control scheme). Good suggestion though! – Austin Prater Aug 13 '18 at 19:18
• V or I limit? On driver – Sunnyskyguy EE75 Aug 13 '18 at 19:24
• – Sunnyskyguy EE75 Aug 13 '18 at 19:41

'Can I drive it beyond specifications' is a question that comes up frequently.

The answer is 'it depends on what the 'it' is, and what its failure mechanisms are'.

In the case of motors (a shaker is a generalised motor) operating electromagnetically and wound with copper wire, the answer is 'sometimes'.

A motor has several failure modes. The first one to bite kills your motor. Thermal, your windings should not exceed their rated temperature, and must not exceed the insulation decomposition temperature. Mechanical, the force must not exceed the strength of the motor components. Magnetic, the field from your armature must not demagnetise the permanent field magnets.

The output force depends on motor current (for PM motors) so you need an increase in current. For a short enough period, where the windings heat adiabatically (without losing heat to the surroundings), the increase in temperature goes as $I^2T$. Double the current, for 1/4 of the time, same heating. If you can identify this constant for your motor, then you can avoid over-temperature by over-driving for only a short time.

Fortunately with shakers and BLDC motors where you have direct access to the windings, and to some extent with brushed motors, you can measure the winding resistance and so estimate temperature. The resistance of copper increases by about 10% for every 25C temperature rise. So you measure room temperature resistance, blast the windings with a short pulse of current, and quickly measure the resistance again.

Unfortunately while that does give you temperature rise, you are still left to guess what the maximum temperature is you can tolerate. What sort of winding insulation did the motor manufacturer use? That tends not to be in the specifications. Would he tell you if you asked him?

If you look at the specifications for your shaker, the 200lbf maximum already says 'intermittent', they've already factored in that the current needed for this is too much to pass continuously. You could use this vague specification as a way to go back to the manufacturer and ask him to clarify 'intermittently' by giving you an $I^2T$ value. It's worth a shot.

However, when you come to estimating the mechanical strength of the shaker, or its demagnetisation current, then you're probably on your own. Probe the supplier about these, and he'll know what you're trying to do. His answer will be 'use it within specifications, or the warranty is void'.

When you want to use something beyond its specifications, you take on all responsibility. You could reverse engineer a shaker to see how its built, and estimate strength and demag current. Or you could test one to destruction, and hope that the second one is similar. It might be, or it might not, as long as it meets the original spec, you can't really demand any more of the manufacturer than that.

Or you could just buy the next bigger shaker.

• I appreciate the very thorough answer. Unfortunately, the company that owns this product has a customer support team that consists of few-to-none human beings, so I don't expect it'll be very easy to get details on the internal shaker design. But like you said, it's certainly worth a shot to get my hands on that I2T value. Another suggestion that was brought up in engineering SE is that I divide my bandwidth into multiple segments and reduce the overall GRMS that way -- however, I'll need to do some more research before I can label that as a plausible solution. – Austin Prater Aug 14 '18 at 16:57
• I'll give your method a shot, regardless, and if my progress seems to hit a wall, then perhaps I'll bite the bullet and find a larger shaker. My vibration testing knowledge is self-taught, for the most part, so it's definitely helpful to get advice from pros every once in awhile. Thanks again! – Austin Prater Aug 14 '18 at 17:00
• PS: In addition to mechanical strength, de-mag, and thermal runaway, I believe another common failure mechanism for shakers is displacement top-out. Trying to drive the shaker beyond it's force limitations for a given frequency will result in an 'armature strike', as it tries to translate beyond its range of motion. – Austin Prater Aug 14 '18 at 17:03