Today's Message Index:
----------------------
1. 06:36 AM - Re: cheapo 24 -> 12V dc-dc converters (Robert L. Nuckolls, III)
2. 02:14 PM - Re: Re: DAS system (Steve Williams)
Message 1
| INDEX | Back to Main INDEX |
| NEXT | Skip to NEXT Message |
| LIST | Reply to LIST Regarding this Message |
| SENDER | Reply to SENDER Regarding this Message |
|
| Subject: | Re: cheapo 24 -> 12V dc-dc converters |
At 11:07 AM 5/28/2020, you wrote:
><colyncase@earthlink.net>
>
>I have a 28V aircraft with a few critical items that run on 12V. I
>previously used a Vicor VI-M11-ES but it became damaged and that
>model is no longer available.
>
>What Vicor has available are mini-packs which require building up
>your own circuit board and heat sinks. cost is $400 and several
>days of work for me.
>
>So I found this thing on the net. data sheet attached. for
>$22. Includes heatsink, leads etc. The worst thing I could find
>about it is .140V ripple. but that beats the Vicor spec.
>
>Any reason I shouldn't just use this thing?\
Not at all . . . the ripple value is of no
consequence given that the ripple values
on the operating DC bus of an airplane are
on the order of 2500 mV pk-pk.
If you have any difficulties with this
device it will be interference with receivers
in the a/c due to radiated and conducted noise
in the radio frequency ranges . . . small
chance but not zero and easy to spot if they
do occur.
Bob . . .
Message 2
| INDEX | Back to Main INDEX |
| PREVIOUS | Skip to PREVIOUS Message |
| NEXT | Skip to NEXT Message |
| LIST | Reply to LIST Regarding this Message |
| SENDER | Reply to SENDER Regarding this Message |
|
On 5/27/2020 12:21 PM, Robert L. Nuckolls, III wrote:
>> Bob and Eric, thanks for pointing out a much more elegant solution
>> than a shunt farm!
>>
>> Bob, I re-read the section of Chapter 7 on Hall effect sensors. I'm
>> sure I read it around 30 years ago, but I had forgotten.
>
> They've become a LOT better and cheaper
> since!
I can see that!
>> One way this is more elegant is that it can be inserted into the
>> ground path without disrupting the ground, which makes the physical
>> wiring safer. (The shunt and fuse block in my van's wiring closet are
>> a bit vulnerable right now.)
>
> You still have to break into the ground
> path which involves terminations/splices
> of some sort.
Yes, but no injection losses. A shunt on the low side would make the
ground float around, which is never good.
>> The microcontrollers I'm planning to use have lots of A/D pins, so
>> that's not a problem. They're Arduino compatible, but at 3.3V, so I
>> wonder whether the LTS 6-NP must have a 5V supply voltage? If so,
>> that'll factor in to the multiplier produced by the parallel wires, to
>> make full scale no more than 3.3V.
>
> Which boards are you planning?
I have one of each of these coming from Adafruit next week for
experimentation:
Feather M0 Express: https://urlzr.mp/dqb (adafruit.com)
Feather nRF52840 Express: https://urlzr.mp/eqb (adafruit.com)
Each has six ADC pins, which should be plenty. The ADCs can measure
0-3.6mV.
The nRF52840 is faster than the M0, has a lot more flash and RAM for
firmware, and Bluetooth low energy, in case I want to use a smartphone
app to display the data.
There are lots of other Feathers with various capabilities:
https://learn.adafruit.com/adafruit-feather
I'll also need this "Featherwing" (like a shield) to carry the SD card
and provide an accurate realtime clock:
Adalogger FeatherWing: https://urlzr.mp/fqb (adafruit.com)
> I've got
> an UNO DAS shield that was crafted to
> mate with an Arduino R3. It's got one
> dedicated LTS6NP sensor. The LTS6 has
> three internal passes through the core
> brought out on six pins. I've arranged
> to exploit those internal passes with
> terminals that place 1, 2 or all 3 in
> series for 6, 3, and 2 amps full scale.
>
> Of course, the drilled aperture is available
> for customized passes on the monitored
> conductor.
And thanks for sending the schematics separately. I reviewed the
schematics closely, and I have some questions I may send you.
But I understand the LTS6NP better after reading the data sheet more
closely. I can use your parallel-wires method to scale to more than 6A.
>> I assume you specify a 1 foot length for each of the parallel 22AWG
>> and 10AWG wires because 10AWG has a cross section about 9X 22AWG? Why
>> 1 foot? What about just a shorter length of each, from one side of
>> the sensor to the other? (I'm picturing a printed circuit board with
>> the 22 AWG soldered near Fast-Ons or threaded studs to attach the
>> 10AWG and the wires to ground and load/battery. Or, heck, just a bolt
>> through to the chassis, like the ground block.)
>
> Nothing magical about the 1 foot length.
>
> The goal is to take 10% of total current through
> the aperture while taking 90% around the aperture.
> 10AWG is 1 mohm/ft; 20AWG is 10 mohm/ft. What we're
> looking for is a total paralleled resistance
> of 1 mohm (10% of the 20AWG thru conductor).
> If you want to parallel 10 mohm with something to
> give you 1/10 the total you'll need about 1.1 mohms.
> This would work out to about 13.2 inches of 10AWG.
The LTS6NP requires 5V and outputs 0-5V, whereas the Feathers can
measure only 0-3.6V. The LTS 6-NP outputs 3.6V at 2.64A. So I would
adjust the parallel wires to carry only about 2.5A through the aperture,
if that makes sense.
But here's the next gotcha: I want to measure separate currents FROM the
charging sources and FROM the battery, so I don't need any of the
negative current range of the LTS devices, which output 2.5V at 0A.
That wastes most the resolution! The Feathers use 12-bit ADCs.
Measuring only over 2.5-3.6V leaves only about 1,250 bits of resolution.
At 60A (from the alternator) that's about 50mA per bit. Is that
enough resolution?
Ideas for using more of the range?
> My DAS systems just record raw, uncompensated
> data. So 1 foot of 10AWG paralleled with 1 foot
> of 20AWG gives me a total of 0.0091 ohms or
> about 10% low for being a calibrated shunt.
> You CAN calibrate the shunt-wires but it takes
> careful cutting, splicing and verification.
Yes. It'll be interesting to see how repeatable the assembly can be.
> My data analysis software allows me to apply
> offset and scale factors to raw data to derive
> calibrated data. Hence, my shunt wires need only
> be in the ball park.
Yes, there's no reason not to just measure each assembly on the bench
and put the empirical numbers into firmware.
In fact, since each of my desired measurements have different expected
ranges, I might use different lengths of wires for each, or even just
use the internal passes for lower currents.
Heck, I might go up to the LTS 15-NP for some of the lower current
sources (like my 8A marine charger), and just use the internal passes
for simplicity.
> You haven't shared how many apples are already
> in your 'cart' . . . don't what to toss off
> too much effort but be aware of availability
> of the DAS shield illustrated. The assembled
> Arduino boards includes an SD memory card
> socket. The software allows user selected
> sample rates and writes data to comma delimited
> ASCII files. If you'd like to explore this
> hardware in more detail, I'll get you the
> schematics. AEC-Lister Paul Fisher has put
> considerable effort into the supporting
> software. I'm pretty sure he'd be interested
> in spooling this project up again.
Thanks! I really want to explore the smaller Feather devices, but your
DAS shield is an excellent starting point, and much of Paul's software
may be usable since the Feathers support the Arduino IDE. (They also
have "Circuit Python," but I don't yet know the differences.)
Other Matronics Email List Services
These Email List Services are sponsored solely by Matronics and through the generous Contributions of its members.
-- Please support this service by making your Contribution today! --
|