In late November 2004, a serious software flaw was discovered in the microcomputer code that runs the PicoDopp DF. The flaw causes the reported DF bearing to contain unpredictable errors. The flaw can causes DF bearing errors as great as 90 degrees, but no greater.

The flaw was the result of a software change that was completed on 26 June 2004. Units shipped after that date ( and before the flaw was discovered ) contain the flaw, and require the installation of a new microcomputer chip, to fix the problem.

The problem was not discovered for some time because the bench tests performed on each PicoDopp DF units ( prior to shipment ) were performed only at 4 cardinal bearings, ( north, east, south and west ) and the problem is only obvious for intermediate bearings. ( bearings between the cardinals )

No-one reported any difficulties with the PicoDopp until early November, and some time was then wasted investigating all the "usual" causes for trouble. ( improper assembly, no DF experience, etc. )

To date, ( 20 Nov 2004 ) only one person has come forward with reports of trouble. Certainly there are others, but some people probably suspect they have done something wrong when assembling the DF unit, and others may have not yet completed the assembly process, and are therefore not aware anything is wrong.

If you purchased a PicoDopp DF unit that was shipped between 26 June 2004 and 20 November 2004, you are affected by this notice. A few people in this category might not be affected, if they were provided with DF units that were assembled before 26 June. ( pre-existing inventory, built before 26 June )

The microcomputers are not "tagged" with a revision number, and the boards are not serialized. For most people, the first hint that anything is wrong will occur when someone actually tries to use the DF in a vehicle.... the DF bearings will not "flow smoothly" around the display, if the vehicle is rotated slowly through a 360 degree circle. Instead, the DF bearings will tend to "dwell" at particular bearings for some time, and then "snap" to a different bearing, as the 360 degree turn progresses.

The problem can be confirmed ( on a test bench ) using the method shown below. A multi-turn trimpot is required. ( 1K to 10K ohms )

Obviously, I ( = Bob Simmons ) will make every effort to correct this problem, as quickly as possible, for everyone. There are three possible ways to fix the problem, and everyone can select one of the three.

Option 1 : The PicoDopp DF MAIN board can be returned to me. I will install a new microcomputer, re-test and return the board via Priority Post, free of charge. Other shipping methods ( if desired ) will require payment for the shipping. Depending on the level of response to this notification, it might take as much as 1 week of time ( after the board arrives here ) for a unit to be repaired, re-tested and re-shipped.

Option 2 : I can provide a replacement microcomputer "chip" via Priority Post, free of charge. You can then install the new microcomputer yourself. Other shipping methods ( if desired ) will require payment for the shipping. if you feel your soldering skills are not adequate for this task, you are encouraged to select another option. The old microcomputer must be ( carefully ) removed from the MAIN board, and the new one must be installed. This is probably the fastest way to effect repairs, I can ship a new microcomputer almost immediately.

Option 3 : I can provide a replacement PicoDopp MAIN board via Priority Post, to replace the MAIN board that was originally provided. There will be no charge for this, but everyone is requested to return the original MAIN board for re-work, as a 1-for-1 exchange.

Some negotiation will be required for this option, since I do not presently have sufficient boards in stock to offer this option to everyone. I presently have 16 boards, ( on 20 November 2004 ) and more will not be available until late December. My records indicate about 30 people are affected by this recall.

If over ½ of the people demand an exchange board, someone will have to wait. ( sounds like the flu vaccine, no ? ) I will try to be fair to everyone... I presently plan to provide 8 units in the 1^st week, 4 in the 2^nd week and 4 in the 3^rd week, so people that respond slowly to this notification won’t be penalized for being "tardy". If everyone ( or almost everyone ) submits their requests quickly, and it looks like I will have enough boards to meet the demands for an exchange, I will accelerate the disbursal of the exchange boards.

To be as fair as possible to everyone, priority will be given to some customers, based on a "point" system I have contrived, using the following rules :

Each thursday, I will tally up the points for all customers that have requested this option, and provide new MAIN boards for those folks that have the highest number of points. The replacement units will ship the following day, along with an e-mail notification of shipment. If desired, you can keep the original MAIN board until the arrival of the replacement board, but ( as indicated above ) there is a 10 point bonus if the original board is returned, first.

Repeating, I presently plan to disburse 8 boards in the first week, 4 in the second week and 4 in the third week. If it becomes obvious that I have enough boards to cover all the requests for an exchange, the disbursal will be accelerated.

I want to extend my apologies to anyone who has suffered or struggled with a PicoDopp as a result of this error, which is clearly my fault. It is now apparent that more rigorous testing would have revealed this problem, prior to shipment of any units. No doubt, some people reading this have spent considerable effort trying to solve a problem that really was not caused by anything they did ( or didn’t ) do properly.

Unfortunately, the lack of feedback from PicoDopp users aggravated the problem, and the first indication of trouble ( feedback from a customer ) did not occur until early November. The problem was finally identified as a software problem on thursday, 18 November. Friday was spent identifying exactly when the flaw occurred, and identifying who bought units after that time. Replacement chips were also ordered on friday. Everyone affected by this flaw was contacted on saturday by e-mail... none of the e-mails "bounced". More PC boards have also been ordered ( in case demand for an exchange board is high ) but they probably won’t arrive until mid December. This webpage was created on sunday, and a second e-mail ( containing a link to this page ) was then sent to everyone affected. Basically, I am trying to correct this situation as quickly as I possibly can...

Again, I regret the trouble this has caused for anyone who purchased a Picodopp. I will do everything I possibly can to rectify the situation as quickly as possible, for everyone involved. Obviously, the testing procedures will be changed to prevent anything like this from ever happening again. There are no more changes planned for the PicoDopp, ( code or hardware ) since it already does about as much as one chip can possibly do. ( no more "meddling" with the design )