UcD for DC-amplification: magnet valve drive

[tiki]

Hello,

a colleague of mine researches on the bassoon in a University project and needs a mechanical stimulation of the air vibration. We "slaughtered" a loudspeaker drive, the Aurasound NS12
The valve assembly
The Aluminium tube in the middle acts as the guidance and contains little holes (just below the foam) to let the pressurised air out, that is the valve in combination with the sliding bushing.

Until now we used a linear DC coupled amplifier, which cannot deliver enough power.
The UcD400(OEM?) would provide much more power and output current capability. Is there a possibility
- to drive the UcD DC-coupled?
- to drive the UcD as an DC-current source?
- Would you prefer the SMPS180 to supply the UcD400 for that application? The current capability is not adequate for long term high power applications, I know. Is it possible to drive two or more SMPS180 in parallel?

Thank you for your help.

Regards,
Timo

[Cro maniac]

I'm new to class D so I might be wrong, but it seems to me that if you need to produce continuous power at or near DC you might run into serious bus pumping issues if you don't use full bridge.

[tiki]

Of course you're right, a friend remembered me on that issue yesterday. Shame on me, because I'm not that new...
One could try two modules in bridged mode, if the small remaining differences between the two channels do not integrate, causing the same behaviour.

[Eva]

Does the valve require one quadrant, two quadrant or four quadrant drive? You can easily build a half bridge or full bridge driven from a microcontroller with proper feedback. This does not seem like a very linearity demanding application.

[tiki]

Thank you for response Eva,

the assembly acts as a speaker motor, because it is really one, see the links in the first post. So it needs a 4-quadrant drive, right? The "efficiency" of the motor should be significantly higher than the original driver because of the lower moving mass.
The fo of the application is wished to be in the 1kHz-range with an amplitude of +/- 1mm. Dependent on the moving mass this would need driving powers well up to the kW-range for at least one minute to take the measurements. We also want to control the shape of the flow time function.

Because the HD pattern characterises the bassoon sound and is used to classify the effects of the changes done, the drive should not induce lots of harmonics into the volume flow, generated by the valve movement. That's the reason for thinking of audio amp usage.

Many years ago I built a PWM-drive for a similar linear motor, but much slower. This 1-quadrant "half bridge" showed the pumping effect too. If I remember right, the pumping was limited to u=sqrt(L*i²/C) - if the same supply reservoir is discharged and charged again, not the case in classD-half brindges.