What you probably didn’t know about the Bendix-King KI-525A HSI

The Bendix-King KI-525A HSI and associated KCS-55A compass system was a popular avionics upgrade for light aircraft until the advent of the Sandell electronic HSI and, more recently, glass-cockpit systems from Garmin, Aspen and others. Although it is a mechanical device driven by some electronics, these units can give years of trouble-free service. I know of one component they use that has been reported to fail quite often (a bipolar tantalum capacitor, C110) – when it does, the symptom is that the glideslope pointer fails to appear, or takes a long time to appear, or appears erratically. However, I want to share with you my experience of another glideslope pointer issue with one particular unit, and suggest a test to perform if you have a KI-525A in your light aircraft.

KI-525A HSIThe issue came to my attention when I had a Garmin 430 installed. The KI-525A had been fitted several years previously. Naturally, I had the KI-525A coupled to the 430 (it had previously been coupled to a KX165 nav/comm set).

Some while later, I noticed a problem. After starting the aircraft engine, while waiting for it to warm up I would turn on the avionics and the GNS430 would go through its self tests. One of the tests that the 430 does is to set the HSI deviation to half-left and the glideslope to half-up. What I noticed was that the deviation was always half-left, but the glideslope indicator was nowhere in sight. The strange thing was that if I actually flew an ILS, the glideslope pointer appeared and behaved properly, and in agreement with the no. 2 display driven from the KX165 (which had been relegated to nav/comm 2).

I reported this to the company that had installed the unit and serviced my avionics. They sent the HSI off to an authorized repair facility. It came back “no fault found”. So I had them send it off to a different facility – with the same result. Altogether, I think I had the unit sent away 4 times, and in each case no fault was found. I asked to speak to one of the authorized repair facilities directly, but I was told they didn’t want to speak to me and weren’t interested in trying to fix the problem.

My avionics company suggested scrapping that KI-525A and putting in a previously used one that they had available. They lent it to me to try out, and it worked just fine during the self-test. So there was definitely something wrong with my HSI, despite all the “no fault found” reports.

I was reluctant to pay a significant amount of money for a used HSI that was actually older than mine and an earlier variant. Seeing that licensed avionics technicians had failed me, and being knowledgable in electronics, I decided to track down the problem myself. I obtained a copy of the KI-525 maintenance manual from the internet and studied it. It wasn’t long before I spotted something:

  • The KI-525A runs from two power sources: a regulated 15V supply provided by the KCS-55A compass system, and an unregulated 11-14V supply from the aircraft bus
  • Most of the circuitry in the KI-525A uses the regulated supply
  • Only the NAV flag and the glideslope pointer use the unregulated supply

Could the problem be related to aircraft bus voltage? I flew the aircraft and ran the GNS430 self-test in flight. The HSI responded correctly, the glideslope indicators visible in the half-up position. Then I turned the alternator off and repeated the test. No glideslope indication this time. So the glideslope indicator worked when the alternator was on-line and the engine rpm was high enough to raise the bus voltage to its usual value of 13.6 volts, but not when the engine was idling at 1000 rpm and the bus voltage was 12 volts or less.

Why hadn’t the avionics “experts” picked this up? Well, the KI-525A maintenance manual (revision 7, dated March 2002) includes a Final Test Procedure. This calls for some kit that provides power at both a fixed voltage to simulate the regulated supply, and a variable voltage to simulate the unregulated supply. The NAV flag tests are carried out with the unregulated supply set to 11.2V (assuming the HSI is configured for 14V nominal), and then again at 14V. However, the unregulated supply is left at 14V all the while when testing the glideslope pointer. The glideslope pointer is only tested under conditions that simulate the aircraft battery being on charge.

I regard this as a serious deficiency in the test procedure. The specification in the Pilot’s Guide for the KCS-55A compass system says the system as a whole requires 11.0V to 15.8V, so the glideslope indicator should work down to 11V. Imagine the scenario: you are flying your light single in IFR when your alternator fails. Fortunately, you notice this (in UK registered aircraft there will be a big red flashing light to tell you about it). Even more fortunately, there is an ILS-equipped airport just a few minutes away. The weather there is only just above minimums, but you reckon that if you conserve power, your battery will last long enough to make an ILS approach. So you make for the airport and intercept the localizer… but the glideslope never appears. The best you can do is try a localizer-only approach, if there is sufficient visibility.

What should you do about it? I suggest the following:

  • Test your KI-525A at low battery voltage. If it is driven by a GPS that can output test signals to it, this is easy to do on the ground. If not, you can do it while taxiing at an ILS-equipped airport. Or you could fly an ILS in good VMC and try taking the alternator off-line after you have captured the glideslope. [If you don’t want to do any of these, I guess you could add a placard WARNING: glideslope indication may be unavailable in the event of alternator failure :)]
  • If you ever send your KI-525A away for servicing, insist that when the repair facility runs through the final test procedure, they do the glideslope pointer tests at 11V as well as at 14V.

If you do have a KI-525A with a glideslope pointer that doesn’t work at low voltage, can it be fixed? For the glideslope to appear, the current in the glideslope coil must produce sufficient force in combination with a permanent magnet to overcome the glideslope pointer counterweights. The troubleshooting chart mentions “magnet adjustment” and “pointer magnet” in the glideslope section, so unless there is a problem with the coil itself or the drive electronics, that’s the likely area for the repair facility to look at.

Are there many KI-525A units out there that don’t work at low voltage, or is my experience a one-off? My KI-525A had previously been sent for repair due to a glideslope issue, so maybe it worked at low voltage before that. Is there a more recent version of the manual than revision 7, with an updated final test procedure? I’d love to know the answers to these questions. If you do run a low voltage test on your KI-525A, please post your results in a comment to this post.

This entry was posted in Light aircraft avionics and tagged , , . Bookmark the permalink.

30 Responses to What you probably didn’t know about the Bendix-King KI-525A HSI

  1. Graham says:

    We have a similar fault, one KI-525A has a GS pointer that does not come into view, the other one works fine, aircraft was operating from ground power, so supply voltage is not an issue. Both are -07 versions & both work fine on the bench at both 28v & 20v DC (28v aircraft) & also with 15v unregulated supply turned down to 10v DC, so I do not think our problem is caused by low supply voltage.
    The GS pointer can also stick on the out of view stop, you could try running a piece of clean paper between the counter weight & stop screw while gently holding the bar against the stop.

    • davidcrocker says:

      From research and personal experience, it appears that the usual cause of the GS pointer not coming into view or being slow coming into view at normal voltage is leakage in the bipolar tantalum capacitor C110. I don’t think bipolar tantalum capacitors are manufactured any more, however it is possible to get a modern metallized film capacitor that fits in the space available.

      • Hans Dau says:

        Hello together

        We have the same problem here with one indicator for a Beech Travel Air, regitered in Germany. GS pointer is not visable during GNS selftest. First we tried to increas the AC voltage, but didn’t help. We losen the intrument from panel and shaked it mor or less smothly by pitching it up and down, till the pointer was visable. After that the pionter started corect work (max up and down and center). After activation of the GS flag (visable on GNS means for KI-525 pointer is fully not visable) pointer raised in parkpostion. After that, it was not possible to activate the pointer again, without shaking. After opening the housing we figured out that is not “gluing” problem of the leaver or so. I agree with David, it must something electronical, maybe the C110. I sent the unit to the shop today and will see what they will find out. I put the blog conversation with the unit, to help the shop duriing searching the mistake. Thanks for that information here. Greeting from Germany, Hans

      • davidcrocker says:

        If the problem only occurs when the bus voltage is low, then it’s probably the “magnet adjustment” problem that I talked about in my blog entry. Otherwise, it’s probably C110, which is a known failure point. The reliable cure is to replace the bipolar tantalum capacitor by a metallized film one, but I don’t know whether that’s approved.

  2. Graham says:

    The OEM capacitor is still available but it is $110. I had already replaced C110 as someone had fitted a bogus bipolar aluminium. I agree this is the most common fault on these.

  3. dbiern6900@aol.com says:

    I am having a similar problem with the localizer needle . it doesnt make much difference whether I have GPS selected or VOR/LOC selected. The needle indicates only two needle widths left or right of course.

  4. epicchetti says:

    I have a similar issue, what is the value of the capacitior and where is it located within the unit?

    • davidcrocker says:

      It’s a rectangular black capacitor (the only one AFAIR) on the electronics board, not far from the connectors at the end. The value in my unit was 2.2uF but I think some revisions of the board may use a different value. If you need more detail I can find the manual.

  5. sunarya says:

    please help ..
    I had a compass system kcs – 55a . at the time of the test heading flags ok but when the flux valve is in turn followed indicator HSI only 3 degrees per minute at the time of the slave , while at free ok …
    is such that its workings ?
    thank you …

  6. Ray Thompson says:

    Hi David,
    A very interesting post, which I found accidentally.
    I live in Australia and have a Cessna 402C with a KI525A fitted with dual GNS430W’s.
    My problem was that the Nav Flag on the KI525A would not disappear at startup.
    I believe this is driven by the same power as the ILS needle and has the same magnet system.
    When a loan KI525A was fitted it worked fine.
    I had this overhauled initially and then sent to 3 different workshop for fault finding – The magnet adjustment was done ( presumably correctly) – still no joy.
    I was told the unregulated power that drives the flag solenoid comes from the KG102A – so this was checked and new capacitors fitted – they thought that maybe there was not enough power available for this particular KI525A – did not fix the problem – another KG102A was fitted – still did not fixed the problem !!!!!
    With the charges for the overhaul, the fault finding and the removal and installation each time, this cost me a small fortune – $10,000. (Still haunts me)
    I eventually purchased another KI525A to solve the problem.
    I feel the problem was the magnet adjustment ????
    I kept the old KI525A on the shelf for sometime as I did not want to throw it away – earlier this year I “thew” it in the rubbish – i now feel somewhat better !!!!!

    • davidcrocker says:

      Hi Ray, I’m sorry to hear about your expensive experience with a KI525A. I don’t believe there is any magnet adjustment available for the NAV flag as I think it’s an integrated meter movement rather than the separate components that make up the GS pointer. The troubleshooting chart in Fig. 5-3 of the manual doesn’t mention magnet adjustment for the NAV flag, instead it suggests looking at 14 electronic components if the trip threshold wrong, and 5 components plus the meter if it is OK. I can see why your maintenance people suspected the power supply, because the NAV flag is the only component besides the GS pointer that is driven from the unregulated supply.

  7. FRANCISCO says:


  8. Jim Adams says:

    Having a similar problem. So far has defied trouble shooting. GS indicator became erratic on the ILS and then stopped responding to the 530 self test. Had the HSI removed and a new reconditioned unit installed. The second unit seemed to work for a few days and then also became erratic but was responding to the self test. Had both the first and second units sent out for testing – both were reported to pass all tests and show no problems. Began to think the 530 had a problem and sent that to Garmin. Also had it upgraded to WAAS. Just installed it – no glideslope response on the HSI, although the lateral indicator responds perfectly to the self tests. I was thinking it must be a connector or wire harness but could it be the bus voltage?

    • davidcrocker says:

      If it’s a KI-525A HSI and the GS pointer doesn’t respond even when the bus voltage is normal (i.e. battery on charge), then most likely capacitor C110 has come to the end of its life. The best solution technically is to replace it with a metallized film capacitor. However, if you have tried two units and they both pass all the prescribed tests, then I agree that it could be a problem with low bus voltage – although an intermittent connection in the wiring harness is also a likely cause.

      I once had an intermittant problem with one of the output signals from a Garmin 430. We tested the 430 on a bench and found we could make the problem come and go by waggling the connector. I forget which signal it was, however I traced the signal path from the connector on the PCB as far as I could, and re-soldered the joints on that trace. That cured it. I guess there must have been a dry joint somewhere.

      • Jim Adams says:

        Thanks David. Here’s the latest.

        – Have tried two GPSs in the rack. The HSI behaves the same with either – all functions work fine except for the GS pointer.
        – Have tried two HSIs, both tested by an avionics shop and supposedly with no issues detected.
        – HSI 1 seemed to work a bit better when I put the battery charger on and the system voltage increased to 27 v and above. But while the GS pointer finally appeared it was erratic and unreliable when using the self-test functions on the 530 (Max down, full down, center, full up, max up). I noticed that if I rapped on the HSI, the GS pointer would appear and generally responded well to the 530 self test signals, but if the signal was terminated (and the pointer disappearing up) then it would no longer respond when the signal was reinitiated. The avionics shop assures me that this HSI has been cleaned and rebuilt and there is nothing wrong with the GS pointer mechanism – as regards sticking.
        – HSI 2 also works perfectly except for the GS pointer. Sometimes it appears, sometimes it doesn’t. It doesn’t seem to be sensitive to rapping on the box. When it is responding (visible) I noticed that it twitches every 3 seconds.
        – Both HSIs: I’ve excercized the P1 and P2 connectors and wiring harness and don’t see any indication of an intermittant connection.
        – Voltages: I tested the 15 v unregulated, 5 vdc, 15vdc regulated positve, and 15vdc regulated negative circuits at the P1 connector. The results were respectively, 12.7 v, 5.4 v, 15.45 v, 15.75 v. At the time, the system voltage was 28 v.

        Is it possible the unregulated voltage circuit at 12.7 v is the problem?



  9. davidcrocker says:

    Jim, 12.7V is a little low for the unregulated supply. In an aircraft with a 12V electrical system, about 13.6V would be more usual. The test procedure for the GS pointer is done at 14V.

    As your aircraft has a 28V electrical system but the unregulated supply is 12.7V, I guess you have the 12V version of the KI525 and you are using a DC-DC converter to derive the 12.7V from 28V. In that case, there may be a voltage adjustment on the converter that you can turn up a little. The alternative it to ask your avionics shop to do the GS test at 12V and perform the magnet adjustment to get reliable operation at that voltage.

    • Jim Adams says:

      Davic, just wanted to give you an update. Looks like it was the voltage supply in the gyro. I have a 24V plane. We switched out the gyro and everything is now working great.

  10. Gary Hall says:

    Bang on with that David, I test these in the field and the Glideslope pointers don’t operate when the battery voltage is low on around half the units under test, generally a fist to the instrument panel usually overcomes the magnetic friction to enable a successful test. The Achilles heal of a generally reliable indicator.

  11. Tarmour says:

    Any updates on a replacement C110 ? Mr. Bob Bramble of King fame for many years is looking at it now. He says the capacitor is acting up and quotes $260 for the part alone. Also the…. I think he said photocell is having an issue. My issue is no flag and then it drops all the way to middle or just above middle. Surely there is a replacement. $260 for a capacitor is crazy. If anyone needs the maintenance manual: http://www.mikeg.net/hobbies/aviation/avionics/maintenance/king%20ki-525a%20maintenance.pdf

    • davidcrocker says:

      I am not qualified as an avionics engineer, but speaking as an electronics engineer, the bipolar tantalum capacitor was never a good choice anyway. The correct engineering solution is to replace it by a modern metallised film capacitor. Whether you do so without getting approval is up to you. I just happen to have a few suitable capacitors in stock that fit in the available space.

  12. Silentflier says:

    Hi David, During my time at CSE (Oxford) I must have done dozens of KCS55’s and the one real ‘problem’ with them (apart from the plastic gears stripping,which was fixed by Mod. 5) was getting the G/S pointer to ‘drop in’ properly. The torque produced by the magnet/coil assembly is really quite weak and any friction or stickiness in the pivot really affects how well the G/S arm drops in.
    The magnet adjustment is difficult to do accurately (hold the magnet and the pointer and do up the allen screw at the same time. Three hands anybody?) but with a little perseverance it can be done. It usually took a couple of attempts and that does take time. A lot of shops are obviously trying to cut service time to a minimum to move onto the next job…

    A couple of other things I saw which may or may not have had an effect but seemed to help (straw clutching times!) were:-
    1. When the GS VALID signal is lost the pointer zips up out of view very quickly and slaps against the edge of the casting on the main frame of the instrument. With older units that have used the G/S a lot this causes a dent in the top of the G/S arm (there is no ‘pad’, it’s metal to metal contact). This can form a burr which possibly – just possibly – could be enough ‘drag’ to hold the pointer up. I used to polish off the burr on the pointer arm, touch up the arm with a spot of paint and things seemed to wok better.
    2. (and this was really scraping the barrel time) At the place where the GS arm hits the main frame the surface (anodising?) of the aluminium main frame also wears away. When you got this bare metal to bare metal contact it seemed as though the pointer was more reluctant to drop in. We discussed this a lot and really odd theories about changes to the magnetic field or earth current loops were the best (or weirdest?) we could come up with. Whatever it was, once we had coated the main frame mark with a touch of epoxy and put some hard setting lacquer on the G/S arm mark and let it set for 24 hours things seemed to work better.
    If anyone has a really stubborn G/S arm problem and it has to go into the shop I’d always ask them to polish up and coat the marks where the G/S arm hits the main frame while it is in the shop. We might have just been imagining it but it always seemed to work better afterwards…

  13. Benjamin says:

    Hello there, there is a manual revision 8 that has an updated test procedure that should catch this problem.

  14. Kami says:


    Do you still have any replacement capacitors available (110 replacement)? Also can you forward pictures of the board showing the cap?


    • davidcrocker says:

      The only capacitors I have are not PMA-approved. You can find PMA-approved capacitors (the original – and unreliable – bipolar tantalum ones) by searching for the part number 096-01193-0000.

  15. Kami says:

    I am inquiring about the non Pma caps. How can I get set from you? Any info about replacing them would help as well. Thanks

  16. davidcrocker says:

    I am on vacation but I will email you when I return a few days from now.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s