I just found this and we can talk when we meet at the Ski fly-in. But for everyones info I had the panels glued on and also riveted. There should be no way any fluid should get behind the panels if done properly. Most of my spot corrosion is not where the rivets which is great news but in various places where the panel edge ends.
In any case Our fix was to lift the panel slightly and use sandpaper to get rid of the corrosion and Alodyne and paint. After the paint dried they put ACF 50 under the entire length of the panel. A few days later blew out the ACF 50 residual and then a few days after that seal the trailing edges closed.
Some US pilots have spoken to CAV in the US about this and were told that they had just one TB20/21 kit left in stock, for many years, and that it was me who bought it 
So anyone else who wants to follow this route can expect a significant lead time. However, I can’t see why it should be longer than my wait, which was something like 1.5 years.
I have managed to sell all the old propeller TKS parts but still have the control unit (the little box) left, for someone who is installing the prop only TKS.
Another TB owner has kindly pointed me to a Socata SB from 2002 which attempts to fix this fluid spillage issue
It is a damn pity that either they never told CAV (the STC owner), or perhaps CAV lost track of the notification (they knew nothing about it when I asked).
I guess most installers don’t tell the STC owner of discrepancies in general. They want the job done, get paid, and don’t want to get into certification to cover the mod… And the CAV STC had small discrepancies e.g. the TB GT battery compartment is bigger, so the provided metalwork doesn’t fit, so some “brain work” was needed, done IAW AC43-13. Of course every installer working on a GT would have found this, but clearly none of them told CAV
That Socata mod is poorly executed. They allow the fluid to spill out into the bottom of the filler door, and from there it can drain via the little tube. But if there is any movement or shaking of the airframe, or the fluid shoots out of the vent hole with some speed (which is very obviously does) that will still dump a lot of fluid in the rear hull cavity. Prior to my mod, I found fluid over a large area of the hull cavity, all over where the directional gyro was, and within about a 3ft radius around there. Clearly the fluid shoots out of that vent hole with some speed, most probably during ground movement. Also this SB has a non ice protected drain; this will be ok if fluid spillage indeed occurs only during ground movement.
I did a test to see how well the system removes existing ice.
This accumulated in a few seconds in IMC around -5C
The TKS system did remove it but it took a long time. At max flow (which empties the ~35 litre tank in under an hour) it removed much of it in a few mins but took around 20 mins to do it all.
Aren’t weeping wings anti-ice, i,e, you need them on when entering icing conditions. Once the wing has ice they technically are not de-icing systems.
The DA42 system also has a tendency to struggle to keep the outer leading edge clear in moderate icing, not ideal given that is where the ailerons are.
Yes I am sure this usage mode is not intended
RobertL18C wrote:
Aren’t weeping wings anti-ice, i,e, you need them on when entering icing conditions. Once the wing has ice they technically are not de-icing systems.
At least for the TB20, the POH says that TKS is both anti-ice and de-ice, but that you should preferably use it as anti-ice.
As a summary of this project I typed up the following for another TB owner in the US:
The reasons it is not FIKI on an N-reg is because the FAA requires two alternators, a heated stall warner, and I believe two fluid pumps.
Under other registries it is certified for flight in icing e.g. on the G-reg.
A backup alternator is installable via a Field Approval and it’s been done (B&C) but is a complex field approval project. A heated stall warner is easy. A dual pump is also easy but CAV don’t offer it in the TB kit.
There is no issue with the amount of fluid carried. An SR22 lasts even less time on max flow. Fuel vents don’t need to be heated; they are ice protected using accessories included with the kit. There is an ice light, and zero space is taken up in the luggage compartment (the fluid container is all below the floor).
I installed the system with my A&P/IA.
It is a somewhat irritating job not least because CAV have almost no TB parts left and would have to manufacture them. When I bought my kit, they had not sold any for at least 10 years and a lot of what arrived was 10-20 years old stock. Some was in a poor condition. I believe CAV make enough from some high value kits (Caravan, and military business) and probably the Cirrus business so don’t get too involved in oddball models now.
The installation was straightforward but beyond the average mechanic because the STC drawings have various errors especially for the GT variant which has a different size battery box. The solution is obvious and trivial but only if you spot it before you start cutting. I saw some photos of installations done long ago which showed that the installer cut metal according to the STC and then had to patch it up :) Fairly obviously numerous installers didn’t report this to CAV because they didn’t want to rock the boat; everybody wants to be paid and get on with the next job. Also the TKS panels simply don’t fit and need a lot of shaping using specially made wooden fixtures; this needs to be done just right for obvious aerodynamic reasons. Most installations produce a ~5-7kt speed loss but mine didn’t produce more than about 1kt and I believe the other numbers are due to incorrect fitting of the elevator panels (evident on installations I have seen).
There is a simple but major defect in the design which causes a lot of fluid to be sprayed into the rear hull cavity, out of a vent hole, and there it damages the KG102A directional gyro. CAV developed some mods but they aren’t effective. Socata did a good mod for this but almost nobody knew and CAV never heard of it. I discovered this issue early on and did what turned out to be the same thing as Socata!
Electrically the standard is generally good (very tight to install the control unit unless you relocate the transponder or the ADF receiver) but the STC requires connection direct to the main bus so the control unit is unprotected from starter motor induced transients. I have the schematic and it is clear from this that the absolute maximum input is 35V. I reported this to CAV who said they never had one blow up… The obvious solution is to connect it to one of the avionics buses because they are protected from this.
The fluid level readout varies wildly (by say 30%) with aircraft attitude. I changed the sensor once before realising this is inherent.
Filling it with fluid is tricky because of a constriction at the tank which is much smaller than the hose before it, and this causes the fluid to fly back up all over your face when you get to the full point :) I solved this with a little portable $10 transfer pump and a level sensor which fills it very slowly, allowing the air to escape, and stops the pump when the fluid starts to rise.
CAV supported me very well but ultimately it took about 1.5 years for all the (correct) parts to arrive so you are looking at a long timeframe and some patience.
In Europe, CAV have an agent in Germany; a “difficult” company which is not interested in talking to customers. They get the dealer discount on the parts but since nobody else in Europe gets that discount nobody else is interested in doing the job either, so the aircraft owner has to really drive it and in practice renting a hangar (where work is permitted – uncommon in Europe) and having a freelance A&P is almost the only way. For an EASA-reg you can’t use an A&P; it has to be an EASA Part M company and CAV offer some route to certifying it. Many years ago Air Touring (a UK and German Socata dealer which went bust about 15 years ago) installed quite a few systems.
The system is extremely effective, coping with ice accretion rates in the region of 1 inch (2-3cm) per minute (!!!) which would obviously result in a certain death within 5-10 minutes in an unprotected aircraft. It doesn’t appear to do anything useful on the slow speed setting when even light ice accretes, so it is Max or nothing.
The window spray bar is not needed due to enough fluid coming off the propeller.
