Autopilot and two HSI sources, switchable?

P

peter-h

Line Up and Wait
Joined
Sep 24, 2007
Messages
613
Location
UK
Display Name
Display name:
peter-h
Everybody here (UK) tells me this is illegal.

I would like to fit a Sandel 3500 EHSI on the LHS and the old King 525 one (or maybe a second 3500) on the RHS, with a switch on the KFC225 autopilot selecting which HSI is to be used to drive the AP.

The reason is just for the obvious redundancy.

The aircraft is US registered.
 
I wonder if the two units provide absolutely identical output?

I've seen similar autopilots set up with a NAV1/NAV2 switch for two identical CDIs, but these were CDIs and not HSIs so all they output was a L/R deviation.

I'd expect it would be "easier" with two Sandels.
 
Peter,

I don't think this would be a good idea. It could be very confusing, especially with HSI 2 not in the pilot's primary view. It is not clear if the additional switching and annunciation would not reduce reliability rather than add to it. What Nav source is HSI #1 hooked to and what Nav source is HSI #2 hooked to? If I recall, you have a Nav #1 and a KLN90B GPS for the #1 Nav system with an annunciator control unit and probably a Nav #2 hooked to a secondary CDI.
 
EHSI #1 would be driven from

- NAV1 (KX155A, VOR/LOC/GS)
- NAV2 (KX165A, VOR/LOC/GS)
- GPS (KLN94)

EHSI #2 would be driven from the same sources.

Currently I have the standard "NAV/GPS" switch which selects the source for the 525 HSI: NAV1/GPS.

There is no NAV2 option on the current HSI (NAV2 drives a CDI, and an RMI). AIUI, a 3500 can accept just about all the signals available and one selects the required one on the 3500 itself. So there is nothing gained in installing a 3500 and not feeding it absolutely everything available.

My panel is here

http://www.peter2000.co.uk/aviation/crete2/tb-panel-big.jpg

It is almost totally factory standard.

The other reason for operating the AP from HSI #2 would be to enable the RHS pilot to do all the flying from his seat.

I would be interested in knowing the legalities of this. I assume that one would need a clear annunciation of the current AP source (HSI1/HSI2). That could be done with an illuminated switch - same as my current NAV/GPS selector switch.

If HSI #2 doesn't drive anything, its heading bug will be virtually useless - except for manual flight.

In terms of wiring, I can see one would need a multipole relay to switch all the signals over. One can get very high reliability sealed relays (not cheap but nothing in this game is).

Incidentally, a friend has a nearly new KI-825 EHSI, on the shelf. It came out of a Jetprop. This has a silly price tag but he might sell it cheaply. Is it any good? I have flown with it; it is very sunlight-readable.

I need to understand the issues because UK avionics shops are well known for delivering a "90% solution", knowing the customer will still pay up because the alternative is to rip it all out.
 
We have many PA46 aircraft with dual HSI installation. One Jetprop has 2 Sandel units installed. It seems there was a switch installed because a ship left one day that the long time owner couldn't get the heading to respond, we found the HSI switch in #2. I'll do some asking. As I write this I remember the local shop doing this with a Sandel/ KI525 arrangement.

Kevin
 
Peter,

Assuming all the signals are going to the EHSI #1 and it selects the display source, how do you plan on providing the resolver information to the selected Nav source. Then how will you do the same with two resolver sources, one from EHSI #1 and the other from EHSI #2?
 
John,

That terminology is beyond my understanding of how these things work, but clearly inputs to the HSIs can be shared (if perhaps appropriately buffered).

The same set of inputs will go to both HSIs. On each you can select which you want to see. I can't see a problem with that.

The issue might be with feeding the autopilot from one HSI or the other.

I don't know what form the course pointer and heading bug go to the autopilot. It might be a quadrature sinewave, and that can be switched. The deviation bar is a voltage, I believe, so likewise.

Currently I have a KI-229 RMI whose VOR needle is switchable NAV1/NAV2. The wiring wasn't trivial but the dealer did it (on the new airplane). There is a relay somewhere.

Are you saying it cannot be done?

Peter
 
In terms of electronics, it's nearly trivial.

Whether (1) it is routinely done, or (2) you can find someone to actually do it for you - those are the real questions. I'm going behind the woodshed to ask an avionics guy.
 
peter-h said:
John,

That terminology is beyond my understanding of how these things work, but clearly inputs to the HSIs can be shared (if perhaps appropriately buffered).

The same set of inputs will go to both HSIs. On each you can select which you want to see. I can't see a problem with that.

The issue might be with feeding the autopilot from one HSI or the other.

I don't know what form the course pointer and heading bug go to the autopilot. It might be a quadrature sinewave, and that can be switched. The deviation bar is a voltage, I believe, so likewise.

Currently I have a KI-229 RMI whose VOR needle is switchable NAV1/NAV2. The wiring wasn't trivial but the dealer did it (on the new airplane). There is a relay somewhere.

Are you saying it cannot be done?

Peter
Click to expand...

It is doable. It would be legal unless someone could show me regulatory docs to the contrary.

Placing a mechanical switch or relay in the middle of this would be introducing a single point of failure while theoretically attempting to provide redundancy.

And to explain what he's talking about with "resolvers", avionics and these displays originated in the analog days. Your KX 170's and KN 75 nav receivers communicated with a bundle of wires. Current on a given combination of wires would select a given frequency. As in.. these three wires select 118, 119, 120 mhz, then these 4 or 5 wires would be used to select .2 .4 .6 .8 .0 mhz..then one final wire would be for the .05 mhz...

Well.. with the CDI's, sending that output from a round bezel ring with an infinitely adjustable number of positions requires a "transformer" and voltage. Think of a signal that looks like a sine wave, with 0 and 180 degrees being the highs and lows, and 90 and 270 degrees being the 0 points (example only, not completely accurate, I dont have a manual in front of me). As you twist the CDI ring, the voltage on the "output" (resolver) varies like a sine wave. The AP uses that resolver output as well as other signals to navigate. If thats what a quadrature sine wave is, then bingo you understand it.. Its been two decades since I schooled in physics and math.. and I dont use it frequently.
 
Lunch today with my avionics guy (who owns the longest-operating avionics shop in the USA).

He says, the 3500 can be put in KI525 mode, and the autopilot input run through a switching relay. Bottom line: do-able, and not difficult.
 
Placing a mechanical switch or relay in the middle of this would be introducing a single point of failure while theoretically attempting to provide redundancy.
Click to expand...
That's very true but... you can buy mil-spec switches for $30 (the kind that you get on a $3M turboprop :) ) and you can buy high-rel sealed gas-filled multipole relays.

Also, if the the relay fails, it will prob99 jam in one position or the other, so if you need AP functionality you use the appropriate HSI. Every airplane already has a NAV/GPS switch on its AP input so there is a single point of failure already.

And, in theory anyway, one can fly without an AP. I have done that enough times, when one of the KFC225 servos went up in smoke :) Got one of those last week actually (the KS270C strain gauge torque sensor failed, as usual).

I don't think the wiring would be totally trivial because one would need one switch+relay to select HSI1/HSI2 and then another switch+relay to select NAV/GPS from the previous relay's outputs.

And then if one day I replace the KLN94 with a Garmin GNSx30, the GPS side of that wiring would be re-done because the Garmin talks directly to the AP via ARINC, AIUI.

the 3500 can be put in KI525 mode
Click to expand...
What does that mean?

AIUI, both the KI-825 and the Sandel 3500 can be used to replace the KI-525, in a very basic mode where you get just the KI-525 functionality. This can be achieved with just a cable adaptor and no other wiring.

In practice one would not do that, because one of the attractions of the 3500 (forget the 825 - silly price) is that it provides a backup display device for my MFD (GPS track, WX500 stormscope, TCAS if I ever buy it). But those are all inputs to the HSIs which can feed both HSIs at the same time.
 
Last edited:
Peter,

The resolver is the analog course datum signals that define the position of the EHSI course pointer to the Nav or GPS. The signals are used by a Nav receiver to determine left-right and to-from by comparing the reference signal of the resolver to the received radial from the Nav receiver. For a given Nav receiver there can only be one resolver input. The resolver signals are also used by the KLN90B when it is in OBS mode to determine the selected course.

So, lets assume you have two EHSI's and both of them select Nav 1. Which course pointer EHSI #1, or EHSI #2 defines the selected course to the Nav 1 receiver? It might be possible to install separate VOR localizer converters and have the course pointers be independent of one another for the Nav radios, but not so for the KLN90B.

From a system point of view, a Nav radio system consists of three main components. A Nav receiver, a VOR Localizer converter and a CDI with a resolver. The Nav receiver tunes the VOR or localizer frequency and provides two main signal outputs, composite VOR (sometimes called video) and ILS Energize, a digital true false signal that when true indicates that a localizer frequency is tuned. The composite VOR is an available output on most Nav receivers, it contains the radial of the VOR signal being received or the localizer signal. These signals are provided to the VOR localizer converter along with the resolver outputs from the CDI and determine the position of the left-right, to-from, and flag on the CDI. ILS Energize is routed to the autopilot along with the CDI left-right, to-from, Flags, and ILS energize. If there is a separate GS receiver, it generates the Up-down and GS valid signals that go to both the CDI and the autopilot.

In the case of a KX165, it has the VOR Localizer built in and is designed to interface to a HSI. It also has a composite signal output available. The KX155 only has a composite signal output and requires an external VOR localizer converter. The KI209 is a CDI that is often used with the KX155 as it has a built in VOR Localizer converter. The KLN90B does not use a VOR localizer converter and is wired directly to the HSI and autopilot (usually thru a relay that is controlled by an push button on a Annunciator Control Unit (ACU) such as a MD41.

To get the two EHSI's to be independent of one another, you need each to have its own VOR Localizer converters. The KX165 has a built in one that can go to EHSI #1 for Nav 1. For EHSI #2, you would probably need to add a VOR Localizer converter and wire it between the KX165 composite output and EHSI#2. For the KX155, you would need to add two VOR localizer converters and wire them both to the composite output of Nav 2 (the KX155) and to their respective EHSI inputs.

The KLN90B would be more problematic, as it can be only attached to a single EHSI and ACU at a time. The outputs can be paralleled, but the inputs can't be.

The switching on the autopilot and the annunciation would also be problematic. The autopilot would need to be able to determine which source, Nav1, Nav2, and GPS, as well as which EHSI.

If you made your project less ambitious and kept the KCS55A as a backup that was dedicated to the Nav2 and the EHSI was dedicated to the Nav1 and GPS, I think your system would be much more reliable and understandable.
 
Last edited:
SCCutler said:
Lunch today with my avionics guy (who owns the longest-operating avionics shop in the USA).

He says, the 3500 can be put in KI525 mode, and the autopilot input run through a switching relay. Bottom line: do-able, and not difficult.
Click to expand...

Not so sure!
 
John,

Many thanks for your detailed reply. I didn't realise that the course pointer goes back to the NAV radio and then the nav radio emits the deviation bar value. I thought that the deviation bar was done in the display device (the CDI or the HSI).

Incidentally my radios are KX155A and KX165A and while I gather these have (or can have) VOR/LOC converters, I still have a KN72 under the seat. One avionics man was puzzled by this but the explanation may be that Socata changed from 2xKX155 to 2xKX155A very shortly before I bought my airplane. I later changed the bottom 155A to a 165A, mainly to get 8.33.

Clearly I need to think about this a bit more. I don't really want a load of old 1970s-style boxes behind the panel. But at the same time I don't want the RH HSI to not have GPS NAV capability, because that is what one is using nearly all the time.

The GPS is a KLN94.

I know you will say I should rip all this 1990s stuff out and install a GNS530W :) The reason I am not ready for this is that I will then need an MFD to do the other stuff I need, and given the dire state of avionics over here I am not ready for anything that major. I can do trips like this without the slightest issue so why spend 20k and have 8 weeks+ of downtime...

Peter
 
Peter,

I enjoyed reading your trip report. I do have one question, what was the power setting (MP and RPM) when you had difficulty climbing to FL180? I know that in my Bonanza, to get to that altitude, I am wide open throttle and maximum RPM of 2700. I can't reach the service ceiling with less than a full maximum power setting.
 
Anything over about 17000ft needs "best power" which is max RPM (2575) and 100F ROP.

Below 17k, one has various RPM options. I tend to fly at 2400 since the engine is smoothest there.

The very best economy (best MPG) setting is at 2200rpm but one cannot get much above about 14k on that. Given nice weather, I stop climb at 10k/11k and 2200rpm and peak EGT.

The MP is what I can get :)

I am sure the problem was caused by sinking air; a mountain wave. We had a strong tailwind and even though there was no turbulence, it would not have taken much terrain to make some of that airflow go up or down.
 
You must log in or register to reply here.
Back
Top