DIY G5 HSI installation

[bradp]

Hi all

About to embark on yet another project.  I’ll make this a running thread.  This will be a G5 HSI install interfaced to a century IIB autopilot. 

Various resistors, capacitors and transformers provisioned from a combination of mouser, Newark, digikey, spruce and one other supplier that I wasn’t previously familiar with.  If anyone wants the list of various items needed for the autopilot interface (the king and century iii interfaces are much simpler than the century IIB) pm me I’ll be happy to provide .  It was about $200 of various parts and wiring in addition to the unit kit itself.  A very big thank you to @LANCECASPER for ordering the G5 for me.  Thanks Lance.  This work is under the supervision of my IA.  

Step 1: determine the optimal location for the GMU11.  After looking at what it would take to fashion a hat shelf / bracket in the empennage sufficiently far away from the trim motor, I thought twice about the empennage as a location even though I had two sets of wires pre run for just this purpose.  Let’s move to where mooney installs their magnetometer in the wing opposite the aileron roll servo.  Cool.   Open the wing and locate the proper spot along the wing rib.  I got a laser level (auto level on a gimbal) for this job ($50 on Amazon and can be used for a number of other projects around the house and even as a steering line for putting the plan in the hangar at night. Specs exceed that called for in the G5 installation manual).  Cool.  The aircraft sits 2.4 degrees nose up on level ground.  So- the magnetometer will sit 2.4-deg up from the level line.  Of note the bottom skin sits 1.6-deg nose down due to its chord at this inspection panel / station so that would have to have been taken into consideration if I wasn’t using an auto-leveling function.  The holes will be drilled in the rib skin on a level line, so the bracket I’ll fashion will need to sit on a slight tilt up to match the angle of the aircraft on the ground.  I had to dig out the trigonometry and determine for a 2.4-degree angle that my forward screw hole would need to be offset 0.178 in down to raise the front of the bracket up.  I figured it would be easier to manipulate the bracket holes than trying to offset holes on a rib skin in tight quarters that I really can’t have error for.  I can always make a new bracket. 

Anyway the guys at the flying club were nice enough to give me a piece of scrap aluminum and they have a full sheet metal shop (minus band saw), so it was pretty easy to make something actually decent.  Couple of additional pieces of scrap for doubler plates and we’re in business.   A little self etching primer and a top coat and I’ll have a somewhat good looking shelf.  (Now I just need to decide whether to blind rivet or river nut plates onto the doublets...). 

Step 2: I am out of circuit breakers.  The installation manual calls for two additional breakers to be installed. That and I have a strikefinder to be installed during this down time so three separate breakers are needed.  I think what we’ll do is what @jetdriven did for extra breakers which was to create a row in the center pedestal of the footwell (Byron I’m thinking the copilot side - still somewhat visible but more out  of the way of control linkages and throttle quadrant cable routes than the pilot side).  Anyway - I’ll run a 14-ga wire from the avionics bus to a new bus bar and will drill 3 holes for new CBS in that pedestal.  Fashioned and drilled out a nice little busy bar from some copper 1/2 in stock today. 

The weekend project will to finish installing the Gmu-11, route the wires (as well as a sync wire for LED anti collision lights - it’s an awful job if you’re not an arachnodactylic contortionist), and fashion a new mounting shelf that will house the GAD-29b 

My other goal is to do nothing too invasive that the plane can’t be put back in service on relatively short notice (ie keep her flying).  Thus I’m not touching the panel until the panel is ready to be worked on.  Not doing a flush mount this go around - too invasive, etc.).   I’ve also hopefully saved some time by prewiring all the 430W connections from my last project so that that tray doesn’t have to be messed with.  

In all I’m anticipating about 20-hours.  Spent 3 on that shelf so far today.  I’ll see how well the bradvionics shop’s quote come in this time.  

 

[BDPetersen]

I will be interested how the magnetic interference test on the GMU-11 goes for you. I had issues with the anti collision light. After tearing my hair out for awhile we discovered that being grounded to the airframe was the problem. Running a ground from the beacon directly to the battery solved the problem. Don’t know if there was a more elegant solution, but that worked and it passed the test,

[N201MKTurbo]

I just mounted it to the wing skin. With the plane leveled it was within spec. If it wasn't, a couple of washers would fix it.

[bradp]

11 hours ago, BDPetersen said:

I will be interested how the magnetic interference test on the GMU-11 goes for you. I had issues with the anti collision light. After tearing my hair out for awhile we discovered that being grounded to the airframe was the problem. Running a ground from the beacon directly to the battery solved the problem. Don’t know if there was a more elegant solution, but that worked and it passed the test,

I have new Orion 600 in there.  I’m hopeful that the interference test will work as 1) it’s not grounded to the wing - I have no idea where though and 2) the voltage drops to ground are not nearly as high as with the strobe box.  So we’ll see....

[bradp]

Got a bit more of the project done today.  Mounted the tray for the GMU11 and ran the wiring through the wing (again not fun).  I had to determine that the wing ribs are parallel to the long axis of the airplane (they are) using a laser site and a measuring tape at the fore and aft wing. Each spot is exactly 105.25 inches from the aircraft centerline.   Cool.   I ran another length of sync wire for the anticollision lights using my adapted mechanics fingers (below)  

Tonigh I’ll finish the harness for the autopilot interface and tomorrow will be the actual install.  Calibration and static check hopefully Wed or early next week.  Right on schedule in terms of estimated time - it will be right around 20 hours. 

[OR75]

is the length of the wire less than 15ft ?
the manual specifies that .... it gets quickly to 15ft  when you run wire from the left of the panel to the right wing and leave some service length 

[bradp]

@OR75 The maximum total length of the CAN bus is 85ft.  3” max from splice to connector.  No information provided regarding max length from an individual box to an individual box  

[bradp]

Got a bit more work done today on the HSI installation.  Found out that I needed minor panel surgery to get the G5’s some vertical clearance from each other.  I needed to fashion two more instrument cover adapters to be able to make an oval instrument hole for the top G5.  When I get some additional motivation, I think I’ll flush mount the things.  

I also addressed my circuit breaker issues and made a mini bus bar for the g5 and cut three holes in the pilot foot well to accommodate them.  

I finished the wing wiring for the magnetometer and completed that CAN bus harness. 

Tonight I’m going to make the most complicated harness... for the GAD29B to century autopilot.  

I’ll probably need another 8 hours total to finish the project. 

-B

[bradp]

Almost there one more afternoon and the thing should be wired in, and hopefully tested / configured.  I have to go to work tomorrow for a couple of days, but have an appointment with the avionics shop on Friday to do the pitot / static.  I also pulled the altimeter and had them bench it as I am do for 93.411 (and got off schedule between the xpdr and altimeter so now they're 7 months apart - oh well).

Today's progress was:

1- Finished the most complicated wiring harness (this included learning all about transformers, resistors, inductors and capacitors and what happens when you start combining them in series or parallel).  This is the harness that interfaces between the Gad29b and the Century IIb autopilot.  I also had to go to the back of the radio coupler and figure out which leads are which on the CD33 connector.  They were looking rather beat up, so rather than just clipping them and resoldering all the leads to the connector, I thought I'd try to "do the right thing" and map and resolder all the leads, including the two new spliced leads from the Gad29b).  This way if I even needed to put things back the way they were before (e.g. to test if something doesn't work right), I can just plug the harness back into the DG and it should be good to go.

2- Finished the panel work.  Made some panel adapters.  One set came for the HSI as it needs to screw in the DG's hole pattern.  The other I made as there isn't quire enough vertical clearance between the units.  So some panel surgery was done per the installation manual.  One of the photos shows it pretty well, but new screw holes are made 3/6 in or so north of the old, and the top of the instrument hole is ovaled out to create clearance for the G5, then sandwich adapter plates are used to mount the G5 to it's mounting adapter. 

3- Re-routed vacuum lines.  I had another question somewhere about how best to do this.  The DG and AI vacuum sources both come off the regulator and fresh air enters through the instrument air filter.  Before, the vacuum gauge read off the back of the DG via a little manifold.  Now, I used a 1/4 to 3/8 barb adapter to measure vacuum at the regulator.  I used another 1/4 to 3/8 barb adapter to connect the vacuum gauge to instrument air - now I don't have to worry about unfiltered air getting into the regulator.

4- Finished plumbing pilot and static lines for the new instrument.

On Thursday I'll need to make another tray for the Gad29b.  I'll need to decide whether to make it out of carbon fiber and put two shelves in the tray (one for flightstream 210 and the other for Gad29b), or fashion a tray out of aluminum.  Then all the clamshells get put together and wires get tidy.  At that point, we should be good to go.  Almost there.

[carusoam]

Nice work, Brad!

You didn’t let any of the internal smoke become external smoke!

Best regards,

-a-

[MIm20c]

Great work @bradp.

Did you have a tutorial on the 660 install posted up on MS?  Just picked up the unit and bare wire mount and I want to do as much as possible for my IA. 

[bradp]

@MIm20c I’ll see if I can dig up some photos.  It’s pretty simple to wire the 660.  I’m happy to help provide some instructions or whatever else.

You just need either a wired RS-232 source from your navigator or some way of syncing up your flight plan (Bluetooth source).  You don’t have to do this but it’s worthwhile. 

You can also wire it for traffic from a GTX and to your audio panel for terrain / traffic / airspace alerts.  

I got the hardwire kit ($60 nice quality) and the Air gizmos ($100 - overpriced).

 The AC charger can not be required to be a data cable - well it can but it’s not worth the effort and I’m a pretty CB as CBs go.  

 

Do you have an Aspen and a 430W?  If WAAS the best method may be to pin the 660 to a free RS-232 (if you have a EDM and an Aspen you may not have a bunch of free ports and the format you’d want is MapMX which is very Garmin specific). With Mapmx you get curved paths depicted on the 660 ... which is nice.

The flat out easiest way would be to verify that the 430 is putting aviation out format to the EDM and splice a couple of wires and “listen in” on that.  Option 3 might be to listen to RS-232 on the aspen.   I haven’t resd the Aspen manual so I don’t know how it uses RS-232...

Feel free to PM or post your current setup and we can help figure out the least painful way to get the 660 hooked up... 

 

[N201MKTurbo]

20 hours ago, bradp said:

Almost there one more afternoon and the thing should be wired in, and hopefully tested / configured.  I have to go to work tomorrow for a couple of days, but have an appointment with the avionics shop on Friday to do the pitot / static.  I also pulled the altimeter and had them bench it as I am do for 93.411 (and got off schedule between the xpdr and altimeter so now they're 7 months apart - oh well).

Today's progress was:

1- Finished the most complicated wiring harness (this included learning all about transformers, resistors, inductors and capacitors and what happens when you start combining them in series or parallel).  This is the harness that interfaces between the Gad29b and the Century IIb autopilot.  I also had to go to the back of the radio coupler and figure out which leads are which on the CD33 connector.  They were looking rather beat up, so rather than just clipping them and resoldering all the leads to the connector, I thought I'd try to "do the right thing" and map and resolder all the leads, including the two new spliced leads from the Gad29b).  This way if I even needed to put things back the way they were before (e.g. to test if something doesn't work right), I can just plug the harness back into the DG and it should be good to go.

2- Finished the panel work.  Made some panel adapters.  One set came for the HSI as it needs to screw in the DG's hole pattern.  The other I made as there isn't quire enough vertical clearance between the units.  So some panel surgery was done per the installation manual.  One of the photos shows it pretty well, but new screw holes are made 3/6 in or so north of the old, and the top of the instrument hole is ovaled out to create clearance for the G5, then sandwich adapter plates are used to mount the G5 to it's mounting adapter. 

3- Re-routed vacuum lines.  I had another question somewhere about how best to do this.  The DG and AI vacuum sources both come off the regulator and fresh air enters through the instrument air filter.  Before, the vacuum gauge read off the back of the DG via a little manifold.  Now, I used a 1/4 to 3/8 barb adapter to measure vacuum at the regulator.  I used another 1/4 to 3/8 barb adapter to connect the vacuum gauge to instrument air - now I don't have to worry about unfiltered air getting into the regulator.

4- Finished plumbing pilot and static lines for the new instrument.

On Thursday I'll need to make another tray for the Gad29b.  I'll need to decide whether to make it out of carbon fiber and put two shelves in the tray (one for flightstream 210 and the other for Gad29b), or fashion a tray out of aluminum.  Then all the clamshells get put together and wires get tidy.  At that point, we should be good to go.  Almost there.

I just cut off the extra legs of the transformers.

[bradp]

3 minutes ago, N201MKTurbo said:

I just cut off the extra legs of the transformers.

I was trying to decide which would be least harmful and just decided to heat shrink them ten times to Tuesday.  I like getting ideas and feedback though - the stuff is fun but it probably takes me twice as long as a semi motivated avionics tech to do this stuff. 

How did you orient your resistors?

[N201MKTurbo]

5 hours ago, bradp said:

I was trying to decide which would be least harmful and just decided to heat shrink them ten times to Tuesday.  I like getting ideas and feedback though - the stuff is fun but it probably takes me twice as long as a semi motivated avionics tech to do this stuff. 

How did you orient your resistors?

I twisted them with the transformer legs and then soldered them together. 

Yours look great! Nothing wrong with what you did.

[bradp]

Additional progress yesterday, and one setback.  

Got the Canbus, power, GAD29B all wired up yesterday.  Also taxied over to our new and beautifully painted compass rose to do the magnetometer interference test and do the compass calibration.  The GMU11 is spot on.  The wing rib location opposite wing of the  I also swung my whiskey compass while I was at it (results were less than impressive for that one). 

I still need to fashion a mounting tray for the GAD29B. I think I found a nice piece of scrap aluminum and I’ve got some nut plates lying around that I can rivet in.  

The autopilot interface seems to be functional.  It may require adjusting the gpss scaling factor. Also I kept thinking I had messed something up horribly.... the autopilot would track the heading bug nicely and when it got there there would be a slight drift always to the left. Wtf.  Oh yeah I was testing it with the engine off and the AI was off and showing a right bank.  I forgot about how the AP actually inputs the AI function.  When I did the compass swing I had the plane parked on a slight incline so- same behavior just a lot less pronounced.  Will need a solid test flight in VMC to put the autopilot though a workout.  

So the bad news is I have a 200 fpm leak in the alternate air static.  Seems like a simple manifold with a selector valve and I can feel the rubber gaskets.  My hunch is it’s the old knurled brass connectors - plastic compression rings look brittle and cracked. I also found and removed a half century worth of teflon tape which was probably doing something.  I’m going to replace the connectors first and then test it with my home brew apparatus of a syringe and a three way stop cock.  I also had a very helpful tech that was willing to run the static check up a few times so I could generally slice and dice where the leak was.  The empennage including the static drain is tight.  So that’s good news.  We ran it up on each side of the alternate valve and found the leak there.  However there was an additional leak (total was ridiculous- like 700 fpm and it was tight last spring).  The additional 500 fpm was shared by an old nylo-seal T connector and my new G5 was a little finicky creating a seal at the unit itself.  In the past I’ve been able to hand tighten without a wrench or a socket - I didn’t want to damage or overtighten but it took an extra turn with a socket (no wrench) to seal it up.  

So if all goes well I’ll have a couple of new brass connector plastic compression rings next week.  If it doesn’t go well I’ll have to dig into the alternate static.  I may end up posting separately about that.  

The empty hole in the panel is where the CDI used to live and where a new to me strike finder is going to live.

Cheers

b 

[carusoam]

Post a pic of the teflon tape use... if you can.

This has been an issue in the past... for Teflon bits showing up where they don’t belong.  Compression fittings certainly don’t need it.

NPT type threads usually use it, but done properly to avoid sending the Teflon into the system.... National Pipe Threads are conic in shape that create and interference fit when used.... the Teflon tape seals the air from escaping down the long tortured path created by the threads...

PP thoughts only, not an instrument guy... I installed a dishwasher today, was a mix of one NPT fitting and a bunch of compression fittings after that.... on the water supply line...

Best regards,

-a-

[bradp]

4 minutes ago, carusoam said:

Post a pic of the teflon tape use... if you can.

This has been an issue in the past... for Teflon bits showing up where they don’t belong.  Compression fittings certainly don’t need it.

NPT type threads usually use it, but done properly to avoid sending the Teflon into the system.... National Pipe Threads are conic in shape that create and interference fit when used.... the Teflon tape seals the air from escaping down the long tortured path created by the threads...

PP thoughts only, not an instrument guy...

Best regards,

-a-

From what I gathered speaking to the avionics tech (who’s an old guy and has been doing this a long time...) he and I agree Teflon tape is pretty awful.  Teflon sealant can be used sparingly on certain fittings like the NPT that Anthony mentioned but have to be careful to not introduce it into the system or contaminate (just like spark plug anti seize).  He personally doesn’t like the brass connectors - a hard metal and soft plastic interface less well than two plastics over time.  

I may go to Stein’s star and look as his quick connects - they’re pretty awesome from what I hear.  

[Browncbr1]

@bradp   I didn’t read every single word in this thread, and you may already know, but wanted to advise you that you will need to hook up the aera660 rs-232 in parallel with the rs-232 that goes from the GTN to the primary G5.    The GTN/GNS navigators only allow ONE output of MapMX protocol.   You cannot get curves and holds to display on the 660 without mapmx.    Garmin says they can’t enable the navigators to output more than one mapmx signal, so any unit that needs to receive mapmx must be pigtailed together inline somewhere or at either end of terminations.  

[bradp]

Thanks @Browncbr1.   Done and done.  It was far more convenient to parallel another RS232 split from the G5 input to the 660 rather than at the tight quarters and overcrowded 430 connector.  In fact I did it last summer when I installed the first G5. Ran a parallel RS232 from the first G5 to a potential future second G5 and a potential Aera 660.  Fortunately both have been installed since   

One piece of general advice I can give for these piecemeal panel upgrades is to run wiring for what you think you’ll do in the future - it can save hours upon hours of future labor.  I have two runs of two conductor shielded wire to the empennage for I’m not sure.  Maybe a GDL-52R is in the future...

Wing wiring.  Hmm.  Well you don’t have to pull the interior for either side for access.  Although the seats do need to be removed.  The panels take a few minutes and although painful the wiring for anything in the wing takes about an hour.  That run though the baggage compartment is definitely more time consuming.  

What I should be wiring now are a couple of runs for a GFC-500 autopilot ;-p. BTW does anyone think the fact that we have control rods will push back the STC for these digital autopilots compared to other brands?

[Browncbr1]

1 hour ago, bradp said:

Thanks @Browncbr1.   Done and done.  It was far more convenient to parallel another RS232 split from the G5 input to the 660 rather than at the tight quarters and overcrowded 430 connector.  In fact I did it last summer when I installed the first G5. Ran a parallel RS232 from the first G5 to a potential future second G5 and a potential Aera 660.  Fortunately both have been installed since   

One piece of general advice I can give for these piecemeal panel upgrades is to run wiring for what you think you’ll do in the future - it can save hours upon hours of future labor.  I have two runs of two conductor shielded wire to the empennage for I’m not sure.  Maybe a GDL-52R is in the future...

Wing wiring.  Hmm.  Well you don’t have to pull the interior for either side for access.  Although the seats do need to be removed.  The panels take a few minutes and although painful the wiring for anything in the wing takes about an hour.  That run though the baggage compartment is definitely more time consuming.  

What I should be wiring now are a couple of runs for a GFC-500 autopilot ;-p. BTW does anyone think the fact that we have control rods will push back the STC for these digital autopilots compared to other brands?

Exactly.  I ran three extra runs of three pole shielded wire to the empennage as well for whatever might happen.    I’m thinking possible GFC as well.

[bradp]

Bit more of the progress today 

1) fixed that static leak today.  That thing is tight.  Redo leak test (91.411) tomorrow.    

2) made a tray for the GAD29b 

3) made a video showing some of the features of the HSI

Enjoy  

 

 

[MIm20c]

Nice video!  I wished Garmin and aspen would make the heading bug a different color like orange or red. 

I’m planning on a 50r (I hoped 39r’s would be coming up on the used or discontinued/discounted market) instead of the 52r. Only hesitation right now is the lack of compatibility with ForeFlight. If FF upgrades the software to receive weather (especially metar/TAF info) I’ll be placing my order. Planning on talking with FF about that this weekend at S&F. 

[bradp]

So passed the static check today.  Good news.

Installed a strike finder today (surprisingly easy) it’s going downstairs and the Nav2 CDI is going upstairs to the old Nav1 spot (the strike finder is long and hits up against the roll cage up high).    

Finished making the tray for the GAD29B and FS210 (pic attached). 

Ran out for time today - had to run to daycare.  The only thing left is to install the tray and screw down three ground wires.  

Total time was more like 25-30 hours of work.  Lots of wing running wires and having to fashion installation trays....

 

 

 

[OR75]

Can you clarify which panel you did install the GMU 11 ?

i m guessing the acces is one of those rectangular shape inspection panel, behind the spar ?