All things hang gliding. This is the main forum. New users, introduce yourself.

Moderators: sg, mods

User avatar
By rkswb
#405554
Looking to see how much of a need the community has for a cheaper variometer using a raspberry pi kit. I was planning on making one just for myself, but based upon how much interest there is, I might change how I approach this. I might put more effort into making it a standard kit if people are really interested in something like this, and other people might tag on which could help get things sped up or make it more slick. I'm estimating it'll come in at under $150 (w/ screen).

I'm not looking to make it very fancy or feature rich. No GPS. First phase will just be beeping like other varios, and the second phase will be a display with some basic readout on altitude and climb/sink rate.

Also, I've found other people have made projects like this project (https://github.com/f0n/variometer http://www.paraglidingforum.com/viewtopic.php?t=36799) but if anyone knows of any active or bigger efforts to do something like this, that'd be helpful!

Would this be something that would probably help make flying more accessible to beginning / budget pilots?
#405558
Like this, maybe?

https://www.blueflyvario.com/product-ca ... riometers/

$100 Australian, which is about $72 US. Bear in mind that a $50 bill of material doesn't mean you can sell one for $50.
As a general rule-of-thumb the multiplier for parts cost to retail price is seven. Four if you're in a high-volume business
with a direct channel to customers. I've been in both environments, and everywhere in between over my tenure in this
business.

Here's one for $100 retail, which implies about a $25 BOM cost.
https://www.amazon.com/gp/offer-listing ... dition=all

Here's one for $110 from France, solar powered.
https://www.lebipbip.com/buy/index.php? ... &id_lang=1

Or this one for 59 Euro.
https://skybean.eu/shop/skybean-vario

Not to burst your bubble too much, but it's been done, multiple times.
MGF
#405560
mgforbes wrote: Tue Nov 13, 2018 7:24 pm Like this, maybe?

https://www.blueflyvario.com/product-ca ... riometers/

$100 Australian, which is about $72 US. Bear in mind that a $50 bill of material doesn't mean you can sell one for $50.
As a general rule-of-thumb the multiplier for parts cost to retail price is seven. Four if you're in a high-volume business
with a direct channel to customers. I've been in both environments, and everywhere in between over my tenure in this
business.

Here's one for $100 retail, which implies about a $25 BOM cost.
https://www.amazon.com/gp/offer-listing ... dition=all

Here's one for $110 from France, solar powered.
https://www.lebipbip.com/buy/index.php? ... &id_lang=1

Or this one for 59 Euro.
https://skybean.eu/shop/skybean-vario

Not to burst your bubble too much, but it's been done, multiple times.
MGF
Thanks! Those are some things I hadn't even come across. No need to be sorry to "burst my bubble," as my main interest is that if it can be done, someone is doing it (even if that someone is me). I'm also not looking to sell anything. I would post the code and instructions as open source as well as the housing design for 3D printing.
#405561
Funny. Probably most of you are to new to HG / PG to know anything about the Colver Vario. It was the first aircraft instrument manufactured specifically for hang gliders. That was back in the 70's. It initially sold for $69.

The detector was my version of a dual "hot wire anemometer".

It was very quick acting and audio would indicate a rise or fall of a couple of inches. It could detect the flushing of a toilet in a closed room.

In the early 1980's I turned the whole rights over to Wills Wing who had been distributing the vario for me. At that time the retail price was somewhere around $250. Eventually it reached the product lifetime end and I had no desire to upgrade to a newer design to keep up with competition.

I started out with audio only but after many requests for a meter I added that and the price had to go up a significant amount. And of course the weight and size also. Over the years I have decided that I should have kept it as an audio only device, adding the meter I now think was a mistake.

After i left hang gliding and flew a foot launched, single person, hot air balloon and my personal vario served me well on the balloon.

I estimate that over the life of the product about 5,000 were sold. I've attached a photo of ealy production before the meter version (in my bedroom).

Frank Colver
Vario Production Line spring 1974 R.jpg
Vario Production Line spring 1974 R.jpg (229.73 KiB) Viewed 1867 times
#405562
rkswb wrote: Tue Nov 13, 2018 6:09 pmLooking to see how much of a need the community has for a cheaper variometer using a raspberry pi kit. I was planning on making one just for myself, but based upon how much interest there is, I might change how I approach this. . . . Would this be something that would probably help make flying more accessible to beginning / budget pilots?
rkswb,

In the USA, there may not be much demand, because of the alternatives out there, mentioned earlier. I see little personal benefit to you, then, for your efforts to build complete units. Still, I can see some global needs there, because some countries have insane import tariffs. Would it be reasonable for you to produce a complete list of components, which distant pilots can assemble locally? You may decide to sell the list or kit to individuals, maybe with a 3-D printed housing, giving you some reasonable return for your good work. I would recommend a straight "physical" vario design, like Frank Colver's design, rather than a unit which requires any programming to be done for each unit.
.
#405567
Frank's hot-wire bridge was great , but it ate AA batteries like nobody's business.
The inexpensive sensors available now tend to have high speed serial outputs, like I2C, or SPI.
Processors and development boards are much less expensive than they used to be. You can get a 32-bit embedded ARM for under two bucks, quantity one.
An Arduino approach might work. There's clones available for ten bucks on Amazon. A sensor board for a couple bucks more.
Make a custom enclosure for board, batteries and a resonant chamber for a piezo sounder.

But it's going to be like rec.aviation.homebuilding used to say: If you want to build, build. If you want to fly, buy.
#405571
A simple vario does not require a powerful MPC like a Pie. Just put together an atmega 328p or something similar with a bosch BMP388 pressure sensor (I2C / TWI comms), or the like. Program in C, or if you must you can use the arduino IDE, which will really bloat the code but makes it slightly easier for beginning programmers. Problem with using Python (with a pie) is that it is super processor-intensive for the supposed ease of programming, and it requires more processing capabilities to do simple tasks, which require more battery power, larger size, etc.
I'd design and print your own board (PCB) implementing a simple chip (328p or other AVR) for the controller and add the other simple peripherals and power source. Distribute that board along with the components for people to put together. PCB development is like falling off a log nowadays. Nearly every minor city in the US has board printing businesses, and it's cheap to get 100 simple boards printed off, and it's quick.
For even lower power consumption, you could go into an ATSAMD21 (ARM Cortex M0+) if you have some experience with programming those. A bit more involved than the simple AVR platforms, but you can use essentially the same code body with some tweaking of the I/O comms. The bloated arduino IDE makes that a bit easier, but then you can't really dig deeper when in that limited programming environment, and it requires a USB connection to the chip, which will make everything more complicated than what is required for such a simple application. I really appreciate the Atmel Studio 7 environment.... Excellent debugging and programming environment that allows you to see everything that is being attached to your code. Very user friendly and far more advanced than that super-basic arduino IDE, and far easier to use to boot. Invest $50 in a Segger J-link for super-simple and lightning-fast comms to pretty much any chip.

In C or C++, the code would likely be less than 1.5KB in an AVR (~3KB in an ARM) for a simple, acoustic vario, which would inevitably help keep the power requirements down as well if you sized your controller to the code size.

Anything with a USB for comms will increase the price substantially due to their ridiculous license fees. Program your chips through a JTAG/J-Link/FDI or whatever the chip likes. And damnn the bootloaders-- you don't need that crap. Anyway, it's easy peasy.

If you were really crafty with the pcb design, you could likely fit the whole package into a short tube, with an ID to allow AA-sized batteries to easily slip in.

I think a pie has far too much processing power and requires far too much battery power for what you want to do, and the resulting package would be too physically large and expensive to be of any value compared to the many complete, turnkey offerings on the market.
#405573
TjW wrote: Wed Nov 14, 2018 2:19 pmFrank's hot-wire bridge was great , but it ate AA batteries like nobody's business.
But it's going to be like rec.aviation.homebuilding used to say: If you want to build, build. If you want to fly, buy.
TjW,

Yeah, and those old-style AA batteries weren't very good back then, either. I agree on both counts.

Actually, I was thinking more along the lines of my ancient Litek (Owens Valley) vario. That unit essentially read the resistance across a pair of diodes in the airstream to and from the air chamber, and it ran like "forever" on a single 9V transistor battery.

As soon as you say programming and vario in the same sentence, though, you have already lost most of the people who would otherwise be a customer. As with any project, you must consider exactly who will buy into it. I do have some "geek" in me, but if anything needs programming to work, it always makes me look elsewhere first for what I need. Now if the hardware can come to me already programmed, that would be realistic.
#405575
I've built a couple. First one used an arduino pro mini, second one was just the atmega328 on a custom pcb (which is what I currently fly with). You want a fairly sensitive barometer; I'm using the ms5611, which works great. I tried cheaper ones but they weren't sensitive enough. An rpi would be total overkill, I think. Currently working on a new version with a bigger display and a 1284p. I do have gps and the 1284p will make adding an sd card for logging much easier. I think cost with display, li battery, gps, barometer, pcb is around $100.

There was another person who did something similar but also had an accelerometer and kalman filter, there was a youtube video of him demonstrating it and I think his code is on github.

The idea of a kit is neat and I've thought about that but I'm pretty sure the overlap of people who 1. Hangglide 2. Want to build their own vario and 3. Are willing to do smt soldering for components without exposed leads is probably pretty close to zero.
#405578
TjW :But it's going to be like rec.aviation.homebuilding used to say: If you want to build, build. If you want to fly, buy.

I don't think that applies to smaller projects. There are plenty of hours in the evening and on unflyable days when having such a project is good fun. I don't think I've ever missed a flying opportunity because of such a project - and I've done a lot!

RM
#405644
Hi Frank, the first vario I had was one of your original ones with sound only. Then I stepped up to the new model with 2 battery sets and 2 up scales and able to have up only sound. Best vario ever! I inherited a Flytec 6030 from a friend and it can't hold a candle to the COLVER for sensitivity and response time. When I fly I'll use the COLVER for my vario while I use the 6030 for the other info. airspeed ect. Yeah it sucks batteries but I always put 2 new sets in every day I flew and never worried about it.