Another N331JH EFIS update.

Bullet form facts:
***** With the installatoin of the 18 guage autopilot extension wires (previously had 22 guage installed as extensions to BMA supplied wires), autopilot now controlls with sufficient force for roll and pitch.
***** Roll was very jerky and abrupt due to a high gain we set during our previous troubleshooting. Brent and Greg from BMA are tuning it.
***** We do not have the autopilot switch installed as of yet, amd are currently using the button style CB as an autopilot power switch. Now......After we installed the elev. wire to the servo, I neglected to pull the CB before turning on the master, and the EFIS 1 had not finished booting up yet. As mentioned in previous posts, when I turned on the master switch, there was a shrieking sound coming from the elev. servo (aileron was not yet hooked up) and the elevator abruptly went full up. Sooooo as mentioned by Greg, Bob and some very condescending posters here, I shall reiterate, per the instructions....NEVER, NEVER turn on your autopilot controller without the BMA being up and running, and install servo arm stops to prevent ANY over center condition in the case of rod bending, rod end fitting slop, loose servo arm, etc..! ! !
***** Alt. tends to overshoot by 200' or so, but slowly corrects back. Once again, tuning/gain, etc.
*****If hdg sel. was turned more than 30 deg. left, the A/C would turn right to the selected heading, not left. Greg says that the latest update will fix this.
*****Overall, the E-1/autopilot setup is not nearly as simple as having a seperate control head/autopilot, but it appears to be overall a more than acceptable autopilot setup/system.
*****I am not aware of the update we have installed, but it was the one that Bob installed in November. It was supposed to make the refresh rate faster. Is it faster? Yep.... Is it not still jerky? IMHO (anal opinion by some) nope.
***** Jim has been fine tuning his servo stops and end of April they should be ready for those interested.
Glen

hi glen,

I'm in the same boat-- LOVE the EFIS one but can I recommend it? Nope.

Banked twice today and I got the Leans. Stuck. Dynon right on, as always.

Auto pilot-- yep-- those perfect gains settings are really touchy and elusive. When I decide to engage the A/P I make sure I've got a firm grip-- because I don't know what it's going to do.

Sometimes it pulls a 1.5G jerk up or down and then usually works as it should-- nice 300 fPM intercept.

If i documented all the glitches on each flight I'd have approx. 6 anomalies per hr. average.

I designed a really neat panel and the info avail. if amazing. I love the engine page in the E1, works pretty well with MY settings.

I'm looking forward to being able to trust it in some minimal IFR conditions. See more than 8 miles in front of me, add Traffic and WX.

The color is as good as it gets, the GPS screen with the terrain is excellent.

I've said before-- the confidence that the EFIS 1 gives a pilot is unbelievable. It's everything I need, right in front of me.

I just need to get to where-- like other pilots-- it's dialed in right. (and I'm going to start sending out messages for help from others soon.)

My father needs a panel and A/P RIGHT NOW for his HR-II and with the problems I've had, he doesn't know what to do.

FWIW-- btw- if I sound negative, I'm not, it's difficult to communicate without sounding that way.

Anyone that want's to go for ride, let me know.

67hrs. RV-7a flying.

Trying to stay neutral

Ok, without getting into any of this A vs. B.... I do have to wonder why you guys put up with this? Also why you "trust your lives" with it. VFR is one thing, but "marginal IFR". Come on, there is *NO* such thing.... You are either in the clouds, or vis reduced, or you aren't. No gray.

That last thing I'd want is a) an AP that "jumps" when engaged; b) or that does something that isn't expected.

Please don't be a statistic? If it consistantly doesn't work right, don't use it, get it fixed or replaced, and don't every talk yourself into "oh, it will be ok, just for a few minutes while I get thru this layer". That is just darn scary.

Mark, I found it totally believable that you got the leans just from a bad EFIS display, and I'll assume you were in the clear. You mind can play trick on your oh, so fast. If that were me, I'd be screaming at the company that made it, especially if they didn't have an answer, and couldn't fix it.

Avionics is one thing that I'm just not willing to compromise on. I like to live too much and the one thing they teach over and over again..... TRUST your INTSTRUMENTS!.

yep, it's my .02

sometimes it works as advertised.

That's comforting

mark manda said:

sometimes it works as advertised.

Click to expand...

That's not real comforting. For the nearly 3 years I've been involved in the RV world the bad has far outweighed the good when it comes to Blue Mountain. Seems as if there is no medium with this equipment. It either works great and people love it, or people can't get it to work, have trouble with the company etc...

Learned enough to run, not walk away from this one. Good luck to all those that have these. And, as Adam said, there is no such thing as "light IFR."

i'm on the fence-- I've got $20K in this by gawd I'll get it to work or I'll ...

Time is everything

mark manda said:

i'm on the fence-- I've got $20K in this by gawd I'll get it to work or I'll ...

Click to expand...

Careful.... you know what they say about getting what you wish for...

$20K is cheap compared to what *time* you've invested, or will invest.

I think this sums it up the best...

What scares the crude out of me with these guys that do no testing - No quality assurance, revisions of the week, and most importantly they don't value my life as they ask me to be their "test pilot" everyday and I don't even get "hazard pay" from them.

Click to expand...

I honestly can't see why we put up with this. I am trying to be sensative to all of the EFIS guys. Some don't act this way. I don't think Dynon, or GRT do this, but for some reason BMA feel like it's part of being "experimental".

I guess I now see the value between the higher end of the market and the lower. Most importanly - please be careful, you are trusting your life to newly developed, untested code or "tweaks" most of the time.

Reasons for thread.

I started this thread to keep my promise to Greg and Bob at BMA, that I would lighten up and stop hammering their product. They have, in the end, bent over backwards (finally), to get Jim to where he is happy with his system, and to apologize (read excuses?) to me for my previously posted delivery/refund nightmare.

As far as leading edge technology, I believe that in theory, BMA has some of the best and innovative ideas, and, they were the first to push the envelope in a price range we can afford. Now... With that said, I am merely trying to inform here those who are installing,dreaming, etc. about BMA products, specifically the EFIS 1. My preference for this thread, WITHOUT INFRINGING ON FIRST AMMENDMENT RIGHTS OF OTHERS!...... is to propose posts here for updating all of us about their BMA install/flying/etc. experiences. We learned alot during Jim's install from other builders and this forum, and I would like to extend that same kindness to others for their sanity..lol...
**********************************************************
Additional note. I heard that BMA for a small fee will remake the EFIS 1 box so, for example, the plugs can be inserted from the back. This my friends, is worth its weight in gold!!!!!!! Have them do this for your future sanity!
Glen

News flash to Glen and anyone else interested in this thread or BMA products. The new Generation 4 equipment is TOTALLY different in every way, especially the EFIS/One. Installation an order of magnitude easier especially for us RV guys who don't have a lot of real estate. Check out the website and look at the new brochures, especially the new 8" Sport and the engine pod. Oh yeah, did I mention WAAS GPS and 420 degree per sec roll rates?

Bob Northrup
RV-8A
"2 more months!"

Oh no, not the roll rate topic again

n714b said:

News flash to Glen and anyone else interested in this thread or BMA products. The new Generation 4 equipment is TOTALLY different in every way, especially the EFIS/One. Installation an order of magnitude easier especially for us RV guys who don't have a lot of real estate. Check out the website and look at the new brochures, especially the new 8" Sport and the engine pod. Oh yeah, did I mention WAAS GPS and 420 degree per sec roll rates?

Bob Northrup
RV-8A
"2 more months!"

Click to expand...

Bob, I've seen you post on the BMA forums so I'm not sure if you are an employee or just a user... BUT.... People, please don't think that WAAS GPS and 420 degree p/s roll are *good* because they were published in a brochure. There is always more to the story.

First let me say, yes, they have some innovative products. However I have a *major* problem in them *assuming* that I'm willing to alpha/beta test their product for them and be a test pilot for every revision. ALL their users have become their testors and that leads to many revisions and "tweaks" in airplanes that have live people in them when problem is found.

Now to the topic at hand.

First the GPS. *All* the EFIS guys have WAAS GPS and exactly *NONE* of them conform to TSO-145a/146a to be allowed for IFR use (no RAIM, no integrity monitoring)! You can use them all day long for VFR, but if you want to go flying in the clouds, better get something that conforms to TSO-129a (non-waas), or TSO-145a/146a (WAAS). So just having WAAS, if that is something new, then they just played catchup.

Now for the second troubling area. If you take the "solid state gyro" and you sample it at a frequency to allow 420 degrees/sec rates you actually loose accuracy. That is the *last* thing you want an EFIS to do!

The solids state gyro is actually a fixed frequency device. For the sake of argument, let's say it's a 200 cycles per second device. That means that if you are going to do a roll at 420 degrees/sec with a device that can only sample at 200 times a second, then each sample of the gyro is eqivalent to 2.1 degrees. (420/200=2.1)

So, let's look at the EFIS that support 200 degrees/second. That EFIS would provide 1 degree per sample accuracy (200/200=1) and is approx 60% more accurate than the one that does 420 degrees/sec. If you want accuracy, Less is more in this case.

Again, maybe 420d/p/s is ok for an Acro pilot or a VFR pilot, but then why would an Acro guy want to watch his EFIS while he's out performing Acro?

BTW, anyone here do rolls in less than 1 second (360/420=.8), that'd be a pretty violent roll and I'm no expert, but I doubt the RV can do them that quick. The Legacy for example, is only an 80 d/p/s airplane - a roll in 4.5 seconds. Even a device that supports only 200 degrees/sec instead of 420, will allow rolls of 2 seconds or less... Still pretty violent.

Just because these guys throw around these numbers doesn't mean that the numbers are good for you or mean what you think. Please take the time to evaluate your options. Ask questions about testing, certification standards, robustness, how things like EMI, RFI, or lightning might effect the device, etc. But don't just buy because it's *cool* and has all the whiz bang features. Sometimes those features aren't what they are cracked up to be.

aadamson said:

The solids state gyro is actually a fixed frequency device. For the sake of argument, let's say it's a 200 cycles per second device. That means that if you are going to do a roll at 420 degrees/sec with a device that can only sample at 200 times a second, then each sample of the gyro is eqivalent to 2.1 degrees. (420/200=2.1)

So, let's look at the EFIS that support 200 degrees/second. That EFIS would provide 1 degree per sample accuracy (200/200=1) and is approx 60% more accurate than the one that does 420 degrees/sec. If you want accuracy, Less is more in this case.

Click to expand...

You've got to keep something constant in these examples or you prove nothing. In the first you are rolling at 420 degrees/sec and in the second you are only rolling at 200 degrees per second. Your assumed constant sampling rate yields exactly the same accuracy between your examples given equal rolling rates.

Why does roll rate matter anyway when the refresh rate of those screens is probably only 20 to 30 FPS?

Missed the point

mlw450802 said:

You've got to keep something constant in these examples or you prove nothing. In the first you are rolling at 420 degrees/sec and in the second you are only rolling at 200 degrees per second. Your assumed constant sampling rate yields exactly the same accuracy between your examples given equal rolling rates.

Click to expand...

I *did* provide a constant. The "solid state gyro" and it's ability to only sample at a fixed resolution, in this example, 200 times in one second.

The role rates simply show that when sampling for a fast roll rate vs. a slower one that same *fixed* constant of numbers of samples per second, have to grab larger chunks of degrees to keep up with a fast roll rate, there by decreasing the granularity, decreasing the resolution and decreasing the accuracy.

It matters

Jamie said:

Why does roll rate matter anyway when the refresh rate of those screens is probably only 20 to 30 FPS?

Click to expand...

Jamie, Hi, hows building going?

It matters when the "system" is built to support a fast roll rate as a marker for performance, when that faster rate is a tradeoff for accuracy. Yes, I agree, screen refreshes will become a limiting issue.

What I should have said.

mlw450802 said:

You've got to keep something constant in these examples or you prove nothing. In the first you are rolling at 420 degrees/sec and in the second you are only rolling at 200 degrees per second. Your assumed constant sampling rate yields exactly the same accuracy between your examples given equal rolling rates.

Click to expand...

Ah, but it doesn't work that way... If you build the device to support 420 d/s, then the constant on the solid state gyro *and* the internal clocking for system update/refresh (not screen, but internal) will cause that system to *always* sample at 2.1 d/s. Remember it has to provide support for 420 d/s when/if it happened. This occurs no matter what the roll rate is.... Hence the problem.

Obviously this is a complicated issues and it's done in the digital domain and there are all kinds of "filters" etc that get applied, I've simplified it to remove all the overhead of those other items.

aadamson said:

Ah, but it doesn't work that way... If you build the device to support 420 d/s, then the constant on the solid state gyro *and* the internal clocking for system update/refresh (not screen, but internal) will cause that system to *always* sample at 2.1 d/s. Remember it has to provide support for 420 d/s when/if it happened. This occurs no matter what the roll rate is.... Hence the problem.

Obviously this is a complicated issues and it's done in the digital domain and there are all kinds of "filters" etc that get applied, I've simplified it to remove all the overhead of those other items.

Click to expand...

I was simply trying to point out something stinky in Denmark.
Logically following your example above, if I am not rolling, my sample rate will be zero. I can't believe that is the case.

Ok...

mlw450802 said:

I was simply trying to point out something stinky in Denmark.
Logically following your example above, if I am not rolling, my sample rate will be zero. I can't believe that is the case.

Click to expand...

Well, we certainly are talking past one another. So I give up... believe what you want, mathematically the system that support 420 d/s will be less accurate than the system that supports 200 d/s.

Oh, btw, no, your sample rate won't be zero, your ROLL RATE will be zero. Your sample rate will still be 2.1 degrees/sec in the device that supports 420 and 1 degree/sec in the device that supports 200. YA JUST WONT BE ROLLING. In fact, this makes the point exactly, very small movements (I'd like to see you keep things *perfectly* at zero roll), will show up on the 200 d/s system and they *wont* on the 420 d/s system.

Thanks for the discussion,
Alan

aadamson said:

Ah, but it doesn't work that way... If you build the device to support 420 d/s, then the constant on the solid state gyro *and* the internal clocking for system update/refresh (not screen, but internal) will cause that system to *always* sample at 2.1 d/s. Remember it has to provide support for 420 d/s when/if it happened. This occurs no matter what the roll rate is.... Hence the problem.

Obviously this is a complicated issues and it's done in the digital domain and there are all kinds of "filters" etc that get applied, I've simplified it to remove all the overhead of those other items.

Click to expand...

Perhaps you meant to say that the screen updates only every 2.1 degrees, regardless of the roll rate?
Still hard to believe that a constant bank of 2 degrees would never be detected.

-mike

aadamson said:

Well, we certainly are talking past one another. So I give up... believe what you want, mathematically the system that support 420 d/s will be less accurate than the system that supports 200 d/s.

Oh, btw, no, your sample rate won't be zero, your ROLL RATE will be zero. Your sample rate will still be 2.1 degrees/sec in the device that supports 420 and 1 degree/sec in the device that supports 200. YA JUST WONT BE ROLLING. In fact, this makes the point exactly, very small movements (I'd like to see you keep things *perfectly* at zero roll), will show up on the 200 d/s system and they *wont* on the 420 d/s system.

Thanks for the discussion,
Alan

Click to expand...

Sorry to frustrate you Alan.
Units are really important when you start a proof that depends on dimensions and communicating that to people whose livelihood comprises the accurate use of dimensions will easily confused them when the units make no sense (me, for example). I believe I finally see your point that the screen will update every 2.1 degrees not every 2.1 degrees/second unless of course you are rolling at exactly 2.1 degrees per second. (edited note: degrees per second is not a sampling rate, it's a roll rate. sampling rate is samples per second. Hence my confusion)

-mike

Where did you get that 200 rate? Did you pull it out of...? What if it's 210?? Then math comes out to a really nice round number.. Do you know for a fact that BMA's unit is this 200-rate thing?

All of this has nothing to do with "sampling rate" or anything else. I think we're talking about the gyro output and the thought that to get up to 420 degree per second you have to somehow loose some precision. Assuming everything else is kept constant, and assuming that the old gyro went from "stop to stop" (i.e. it was using all of it's dynamic range) that may or may not be true. HOWEVER, typically, these things are analog devices at their core. What determines the high end of what they can output has to do with how the unit's output is scaled (i.e. the DYNAMIC RANGE is predetermined at the factory, and everything is scaled to that). The signal to noise ratio (SNR) determines the "step size", so to speak...in other words, the PRECISION.

Why are you so sure that they haven't simply improved the SNR? If you increase the SNR, then you can either increase the dynamic range or increase the precision (or do a little of both). I don't know anything about what Blue Mountain is doing so I can't speak to that.

Radomir said:

Where did you get that 200 rate? Did you pull it out of...? What if it's 210?? Then math comes out to a really nice round number.. Do you know for a fact that BMA's unit is this 200-rate thing?

Click to expand...

I think Alan was just using it as an example rate. It does seem a bit slow given the rates that data acquisition systems run these days. If it's close to that number then his arguments are valid. If it's 10 times that number, it doesn't invalidate his argument, it just makes it a little specious since a refresh interval of .21 degrees versus .1 degrees of roll would be a bit picky in my opinion.
-mike

I think Alan is talking quantization error.

As jcoloccia says, everything is a matter of S:N ratio...

Alan, in response to your post (#23), let me begin by saying that you're obviously a well educated, intelligent man. I've been reading your posts for a long time now, both here and on the BMA Forum and while most of your comments are right on, you are woefully uninformed when it comes to the design and operational capabilities of the BMA product line. Yes, I'm one of the company techno-geeks, but I'm also an RV repeat offender and soon to be proud papa of a really cool -8A.
Sadly, there are MANY builders and readers of these forums who take a lot of what they read as gospel and draw conclusions from the opinions of others and not facts.
I respectfully invite you to come to our booth at Sun and Fun (D16/17), spend 15 minutes talking design philosophy with Greg, then feel free to comment to your hearts content. Maybe then you'll have more time to finish that beautiful Chelton equipped Lancair!

Bob Northrup
80578
RV-8A
Almost there!

Perhaps

n714b said:

Alan, in response to your post (#23), let me begin by saying that you're obviously a well educated, intelligent man. I've been reading your posts for a long time now, both here and on the BMA Forum and while most of your comments are right on, you are woefully uninformed when it comes to the design and operational capabilities of the BMA product line. Yes, I'm one of the company techno-geeks, but I'm also an RV repeat offender and soon to be proud papa of a really cool -8A.
Sadly, there are MANY builders and readers of these forums who take a lot of what they read as gospel and draw conclusions from the opinions of others and not facts.
I respectfully invite you to come to our booth at Sun and Fun (D16/17), spend 15 minutes talking design philosophy with Greg, then feel free to comment to your hearts content. Maybe then you'll have more time to finish that beautiful Chelton equipped Lancair!

Bob Northrup
80578
RV-8A
Almost there!

Click to expand...

Bob, thanks for the comments... Perhaps I'll do that... Of course, I'll be at the Chelton booth most of the time tho ... I never implied that I knew the BMA technology, heavens I certainly never designed it nor had anytihng to do with its design.

In fact (and this is a little know secret). My panel is also designed for a BMA lite to replace the electronic attitude indicator if I so desire. The advantages of having a backup HSI is what primarily drives this requirement. I'll post a picture later, my wife is waiting to drag me away from my epoxy messes.

See ya at SNF and yes, I do have some issues with the BMA technology, primarily the testing, but there are design issues that I struggle with as well.

aadamson said:

My panel is also designed for a BMA lite to replace the electronic attitude indicator if I so desire. The advantages of having a backup HSI is what primarily drives this requirement.

Click to expand...

Funny. I'm probably a year away from worrying about it, but I was thinking along the same lines that the lite is great bang for the buck exactly because it has an HSI in it (and you can use it as a dedicated HSI the rest of the time).

Here ya go

jcoloccia said:

Funny. I'm probably a year away from worrying about it, but I was thinking along the same lines that the lite is great bang for the buck exactly because it has an HSI in it (and you can use it as a dedicated HSI the rest of the time).

Click to expand...

And done by Lancair Avionics no less ... (if you click the picture below, you can get that *much* larger version)


The picture above is about 2 revs prior to final. I moved the 3 steam gauges to be *centered* in the panel screws. I also moved the Chelton screens as far up as I could get them to give me the clearance for the BMA Lite. Proof will be in the pudding as they say, but I tried real hard. While the AI that will populate that opening will be the "Sportys" electric AI in the first version. I figured that at some point, I *may* want an HSI backup - Just in case.

We'll see, but I did provision for it.

ps. I learned an interesting trick to posting...to do the nested picture. I first clicked the "add a pictures" icon in the message compose window, then I pasted the smaller pictures link into it and entered it in the message, then I cut all of that text and instead, added a hyperlink, but the the "prompt" screen, I used the