Menu Sign In Contact FAQ
Banner
Welcome to our forums

Glideslope problem - any ideas?

This one is bizzare…

During a recent ILS it was found that the GS was not intercepted by the autopilot; there was no GS indication and both NAV1 & NAV2 GS are flagged.

Using a test generator and standing next to the antennae finds LOC1 & LOC2 fine and GS1 & GS2 both totally dead.

The 1st suspect is the unreliable GS antenna in the TB20 tail. The cable from this antenna runs, via a BNC connector joint, to a DM H24-1 diplexer which splits it into the two GS signals for the KX165A (NAV1) and KX155A (NAV2) radios.

The analog GS outputs from the radios (6 wires each) go to a SN3500 EHSI (NAV1) and a KI204 (NAV2) respectively. Also the two GS outputs go to a relay which selects GS1/GS2 for the RHS SN3500 EHSI.

Injecting an RF signal into the BNC joint makes both glideslopes work. However the only test generator available is a single power one (no attenuator).

However, connecting a separate GS dipole antenna into the diplexer input, or directly into either of its two outputs (i.e. directly into the radios), and using the signal generator right next to the antenna, does nothing. So only a “strong” signal works, via a direct cable connection.

Also, temporarily using a CI507 diplexer (0.5db insertion loss) to split the output of the tail mounted NAV antenna into a NAV and a GS signal, and using the GS signal to feed the GS cable after that BNC joint, doesn’t work either. That is bizzare. Apparently it is common practice to do this in GA; a separate GS antenna is not universal. However the CI 509 diplexer was not tested on a flight to an ILS airport; it was again tested with the same old fixed-power tester.

All I can imagine is that the signal generator is strong enough to work via a direct cable connection but too weak to work via the antenna. But it works with other aircraft.

Another explanation is that both KX radios have their GS inputs partially blown up. How could that happen?

I can’t believe the problem is at the analog GS signals, because a strong enough RF signal makes it work. It has to be on the RF side.

Obviously, I need to get somebody with a proper IFR4000 test set, and work backwards from the diplexer outputs, with the correct signal level set (one just has to know what that is). I have the usual problem of not being allowed to do any work in any hangar where I am based, so the only option may be to fly to an avionics shop, and the only options are the usual flight + train + hotel stuff. Currently, whenever it starts to rain, we have to close up the aircraft and hide. Property owners run the world, as usual…

But I can get somebody with an IFR4000…

Last Edited by Peter at 14 Aug 14:06
Administrator
Shoreham EGKA, United Kingdom

Apparently it is common practice to do this in GA; a separate GS antenna is not universal

No. I originally had a separate GS antenna styled as a winglet located on top of the cabin. That installation suffered a bad case of propeller modulation causing one to fly too low (i.e. actively dangerous), because the 80Hz from the 2400RPM propeller (2400/60*2 due to a 2 blade prop) was not sufficiently discriminated from the 90Hz “too high” signal. But the top rated EASA 145 company in the region didn’t believe me until I instrumented my KN75 and flew a test approach logging the relevant signals with different propeller RPMs…

So now I’m back to a combined NAV/GS antenna in the tail. It’s not great (spec’d at VSWR 1:3 or so for GS), but it works ok and without appreciable propeller modulation.

that both KX radios have their GS inputs partially blown up

Do you have the schematic of the RF front end of your GS receiver?

The LNA transistor is usually quite exposed, it needs to have low noise and high gain which is kind of the opposite of robust. So an ESD event could take out the LNA, which would cause the receiver to still work with very strong signals that would leak past the LNA stage, but not with lower level signals.

LSZK, Switzerland

From the 165A MM:

The BF996 is indeed totally unprotected.

Administrator
Shoreham EGKA, United Kingdom

That seems a rather typical dual gate MOSFET amplifier. At least the BF996 is a relatively common device. And there is quite a bit of band filtering before it, so at least only the energy within roughly the glideslope frequency band passes into the gate.

If you can easily open the box, you could measure the operating point (DC) voltages at source, drain and the two gates to get an idea whether the transistor still works. In case of an ESD event the channel oxide will most likely break down, i.e. you’d get a short from gate 1 to source or drain.

Last Edited by tomjnx at 14 Aug 14:26
LSZK, Switzerland

Thinking about this, I see two explanations:

  • both radios have got their glideslope receivers blown up, and in a similar manner
  • there is a break in the tail mounted GS antenna, and the signal generator used for the ground testing is duff (very low power)

I would think the 64nH inductor across the input ought to provide a reasonable level of static discharge protection.

It would be trivial to pop another KX155A or 165A in one of the two slots, to eliminate the first possibility, but I don’t have a spare.

Last Edited by Peter at 14 Aug 17:54
Administrator
Shoreham EGKA, United Kingdom

Be carefull not to blow things with testing. Your signal generator has quite a high output. An attenuator is a MUST for direct connection testing.

Test equipment is the way to go. You might be able to troubleshoot the system with you signal generator and your spectrum analyzer. Be sure to know what you are doing with respect to power levels. A 100 dB step attenuator might be a good addition to your test equipment.

JP-Avionics
EHMZ

I would think the 64nH inductor across the input ought to provide a reasonable level of static discharge protection.

The input passband filtering certainly protects against simple discharges, not however against one that “hits” the glideslope frequency band

Out of interest, how much power does your signal generator output (at least nominally)?

LSZK, Switzerland

Didn’t check schematic, BF996 can handle quite some power (from your tester) on the input, wouldn’t cause a problem. I do like to use attenuators as standard anyway.

Are you sure your coax cables are fine? Never seen ESD damage to any GS receiver, quite uncommon I guess. Coax cable problems are quite common, though very strange to have the same problem with both coax cables. Could GS2 have been inoperative before (unnoticed / unused) as you would GS1 as primary?

JP-Avionics
EHMZ

The sig gen is an old bit of kit (not mine) which is routinely used for both direct connection and via a whip antenna at close range. I don’t know the dBm output. Yeah – I know I should but the number of people in the UK who do avionics who know what a dBm is could be counted on 5 fingers.

I always fly an ILS with the ILS set up on NAV1 & NAV2 concurrently. So I see the ILS on 2 x EHSI and the KI204, concurrently.

Every ILS worked fine until a mysterious failure to intercept the LOC at LKPR 2 weeks ago. It intercepted it and then disengaged halfway through the turn. I flew it manually, and think the GS was OK then. The next ILS (EGMD, I think) worked fine on LOC but didn’t ever see the GS.

I am hoping to get an IFR4000 tester tomorrow.

Administrator
Shoreham EGKA, United Kingdom

Verdict:

Both radios have a blown up glideslope receiver.

All else is OK. There was something wrong with the old tester which has almost no output when used with an antenna but is OK with a direct cable connection.

The GS works with a direct connection down to -25dBm which is way to insensitive; the KX165A spec is -87dBm. It works with an antenna (at about 5m) down to about +10dBm.

The LOC works with a direct connection down to about -110dBm which is probably about 5 electrons per second It works with an antenna (at about 5m) down to about -60dBm. So that looks OK.

I can get them tested and fixed locally. Those MOSFETs are only 41p each…

We have no way of testing the tail mounted antennae but the VOR/LOC one is obviously OK.

Administrator
Shoreham EGKA, United Kingdom
20 Posts
Sign in to add your message

Back to Top