X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f
X-Recipient: geda-user AT delorie DOT com
Subject: Re: [geda-user] (Off-topic) Question wrt driving HV piezo transducer
To: geda-user AT delorie DOT com
References: <c3aa1897-3806-3f72-feef-85551a63c2db AT linetec DOT nl>
 <alpine DOT DEB DOT 2 DOT 00 DOT 1802091235090 DOT 12270 AT igor2priv>
 <6221de93-bb98-d397-79b4-eecf3975c2e6 AT linetec DOT nl>
 <alpine DOT DEB DOT 2 DOT 00 DOT 1802091324350 DOT 12270 AT igor2priv>
 <f2259553-7015-8fc8-33b2-8ca8be6c5a45 AT linetec DOT nl>
 <alpine DOT DEB DOT 2 DOT 00 DOT 1802091427450 DOT 12270 AT igor2priv>
 <20180209161532 DOT 6AF57804CEFA AT turkos DOT aspodata DOT se>
 <alpine DOT DEB DOT 2 DOT 00 DOT 1802091735590 DOT 12270 AT igor2priv>
 <20180209204934 DOT CAA35804CEFA AT turkos DOT aspodata DOT se>
 <alpine DOT DEB DOT 2 DOT 00 DOT 1802100406500 DOT 12270 AT igor2priv>
 <b125c1d5-48a1-c1dc-3ee1-db9875d50bf6 AT linetec DOT nl>
 <c49d734f1b1f6373733e6f9935afff35 AT iae DOT nl>
From: "Richard Rasker (rasker AT linetec DOT nl) [via geda-user AT delorie DOT com]" <geda-user AT delorie DOT com>
Message-ID: <da435a2d-0bd0-f6d7-7d9a-05394665621c@linetec.nl>
Date: Thu, 15 Feb 2018 23:02:28 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.6.0
MIME-Version: 1.0
In-Reply-To: <c49d734f1b1f6373733e6f9935afff35@iae.nl>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id w1FM2WJS024191
Reply-To: geda-user AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: geda-user AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

Hi Marcel,

> The piezo transducer will electrically look like a big capacitor and with
> a 1:100 transformer you'll introduce a huge capacitive load on the 
> H-bridge
> MOSFETs. At 40 kHz you're sure to destroy them with hard switching, maybe
> already within the first few HF cycles.

Yup, that's one of the problems alright. Transducer capacitance is 
somewhere in the 5 nF range, so that would transform to a primary load 
of 500 nF.

However, this afternoon, the product dealer was kind enough to send me 
the schematic of a 500 Vpp development driver kit, featuring a 
center-tapped symmetrical push-pull transformer(*) circuit with two 
FDD16AN08 MOSFETs, each with a 1 uF capacitor across drain-source. These 
big DS-caps seem rather weird already.
And on the secondary side, a 5.6 nF/1.5kV capacitor is effectively 
connected in parallel to the transducer, doubling the capacitive load to 
10 nF. Also, there's  a 680 ohms damping(?) resistor on the secondary 
winding as well -- I wonder how one calculates the power requirement for 
that one; the load is 500W during a burst, but that lasts only a quarter 
of a millisecond. The resistor doesn't appear overly large, physically.
The whole thing is powered with 12 volts, through a small SMD LM317 
device with a 4700uF/16V low ESR buffer cap. Nothing unwieldy there.

*: The transformer is 4+4:140 turns, so effectively that's a 1:35 turns 
ratio.

(I can't post the original schematic here for reasons of copyright, but 
I could create a gschem version for posting here -- that would be 
lightly less conspicuous... apart from what's described above, there's 
the driver circuitry, based on a humble 556 dual timer, driving the 
MOSFET gates in a way I haven't fully figured out yet -- it almost seems 
as if only one MOSFET gets the actual burst signal, whereas the other 
one merely gets the LF pulse rate, but not the burst frequency).

> Your idea of adding a small inductor in the primary of the transformer
> (or design the XFNR to have defined leakage) can counteract that.
> The switching
> frequency should be higher than the resonance frequency of leakage +
> transformer and the piezo capacitance. Added benefit is that you drive 
> the
> transducer with sinusoidal current. It will not like the harmonics of a
> square wave, and it will probably foul up the distance measurement.

Yes, I also read something like this from the sparse information in the 
Internet. I definitely think that an extra series inductor is a good 
idea, probably something along the lines of a few uH.

> I would start with 450V dc (double rectified line) because the 
> transformer
> will be a lot easier.

The whole thing is supposed to be powered by a 24VDC truck battery, so 
no line voltage available. Amps galore though :-)

> You can look at datasheets for CFL / TL lamp drivers,
> they are all using HV H-bridges and probably the notes will contain some
> magnetics ideas, too.

Ah, yes, that's something I hadn't thought of! Excellent idea!

> Oh, and buy ear-muffs :-)

I will -- although I think that for the time being, I'll use a $1 
4700pF/6kV ceramic capacitor instead of the $350 real deal, until I got 
the whole thing ore or less figured out, calculated and under control.
But I'll certainly keep your advice in mind when the time comes for live 
tone burst tests in my rather smallish workshop :-)

Thanks for your helpful suggestions and other thoughts, this goes quite 
a long way pointing me in the right direction!


Thanks again, best regards,

Richard