Date: Sun, 19 Jan 97 17:13:22 NZT From: zl4til%zl3ur.bbs@zl3rr To: all@sthlnk Subject: Tiny-2 MK2 defect (1 of 3) Path: ZL3UR!ZL4TAY!ZL4TIL >>>>From the internet Hi Folks. During the repair of a number of PacComm Tiny-2 TNCs that were damaged by lightning, a manufacturing/design defect on the Tiny-2 MK2 model was discovered. It would appear that EVERY PacComm Tiny-2 MK2 ever made has an electrolytic capacitor installed backwards. C23 is installed with the positive terminal connected to pin 6 on the MAX232. However pin 6 on the MAX232 is the NEGATIVE supply filter and output terminal with a typical voltage of -10v. If this capacitor fails, the charge pump voltage generation circuitry is loaded down and your serial port fails or locks up (or both). The only reason I can see that it has not been noticed before this is that PacComm have been generally using 25 volt electrolytics in this spot and they have been able to tolerate the reverse polarity under most normal circumstances. The TNCs I was working on ranged from early production units with PCB ver 1.7 (SN 12,xxx) to pretty recent boards with ver 1.81 (SN 18,6xx) which had been stressed with lightning and it was during the troubleshooting procedure that my assistant started to ask why the board and the schematic did not match. The schematic is correct and the board copper traces are fine if the capacitor is reversed (unless you consider that a square pad denotes the positive terminal). The only mistake the designer made was on the component layout screen printing where the polarity marking was reversed. Of course the board stuffers simply follow the board markings when the components are installed. PacComm has been already advised by e-mail to their trouble reporting address. I am attaching a copy of the confirmation message received from their automatic server. At the time this message is being composed, I have not received any response from PacComm ( I allowed 2 working days before posting this notice). Recommended action: This defect poses only a potential failure problem. If your TNC is working, you do not need to immediately change the capacitor. Schedule a change for the next time you have the unit open. For network/node operators: Examine your own node stacks when next servicing the node stack and replace C23 with a new unit (no telling what damage the existing unit has suffered with continued application of reversed voltage - although the units we switched around (we did not have replacements readily available) seemed to work reasonably well). The replacement must be installed with the positive terminal in the hole nearest C33 (negative lead in the hole marked + ). A dipped tantalum cap might be a better replacement. Note: Some of the problems that I reported some years ago regarding low drive and lockup of the serial ports with certain brands of serial port ICs may have in fact been aggravated by this problem. I know of at least two stacks recently that have suffered similar problems with lockups. Let's not dump on PacComm on this issue. Even the best manufacturers can have errors slip thru their design and testing procedure. We don't want another AEA - NetRom compatible TNCs are getting hard enough to find and we don't want a monopoly by one remaining manufacturer. What we should expect from a responsible company is correction of their current production ASAP and a timely advisory notice to owners outlining the problem and how it can be corrected. Any Comments??? Burt>>> ****************************************************** (response from PacComm trouble report server) Subject: Receipt confirmation of Hardware Report 1071 Date: Sun,12 Jan 97 11:00:47 EST From: reporter@paccomm.com To: burt@rocler.qc.ca Number: PR-1071 ____________________________________________________________________ SUBMITTER: Name: Burt Lang Callsign: VE2BMQ Phone: (514)825-2597 FAX: REPLY E-MAIL: burt@rocler.qc.ca Product: Tiny-2 Mk-2 Release: rev 1.7 & 1.81 Date: SYNOPSIS: Manufacturing/design defect in Tiny-2 MK2 DESCRIPTION: You have a serious problem Electrolytic capacitor C23 is installed with incorrect polarity on all Tiny-2 MK2 boards I have examined. The error is in the component marking screen printing on the board which obviously the installers used when orienting the parts on the board. If you examine the board very carefully, you will find that the positive terminal of C23 goes to pin 6 on the MAX232 chip. However pin 6 on the MAX232 is the NEGATIVE polarity power supply filter point and output terminal and requires connection to the negative terminal of the capacitor. The schematic is correct as is the copper traces on the board. The problems arise when this capacitor starts to leak excessively and loads down the power supply circuitry in the MAX232. The serial port will then fail and may also lock up. How come this was not noticed before this? Well probably because you have been using 25 volt units (instead of the 10v units specified in the schematic) and they tolerated reverse polarity at least long enough for the units to past inspection and during normal operation in the field. Users probably attributed occasional failures to other causes. Recommended correction (on units already out in the field): Replace this capacitor with a new one installed with the polarity reversed to that shown on the board screening. I do not expect you to offer to correct them but I would expect you to issue an advisory notice to owners to warn them of the problem. I would also expect you to correct your production on future units asap. Again I must emphasize that you generally make good products but bugs can be hidden in even the best of equipment. HOW TO REPEAT PROBLEM: Thank you for your interest in PacComm products. PacComm ***************************************** -- End -- Date: Sun, 19 Jan 97 17:25:26 NZT From: zl4til%zl3ur.bbs@zl3rr To: all@sthlnk Subject: RE: Tiny-2 MK2 defect (Part 2 of 3) Path: ZL3UR!ZL4TAY!ZL4TIL >>>>>From the internet Capacitor C 23 Question Let me tell you more about which models are affected and what should be done about it. Amateur radio models affected are the TINY-2 MK-2, TNC-NB96, and IPR-NB96 which all incorporate the TINY-2 circuit board. The TINY-2 pre-MK-2 version is not affected. Those models use the MAX-231 RS-232 chip. The mismarked capacitor crept into the design when the MK-2 changes were made, which included adding the capability to use the MAX-232 RS-232 chip. (The circuit board accepts either the 232 or 231, but all have been built with the 232). Another way to identify the boards with the error is to examine the circuit board revision number. Revisions 1.7, 1.8, and 1.81 contain the error. Revisions 1.6 and prior use the MAX-231, and Revisions 1.9, 1.91 and 1.92 have corrected the problem. Some of the final production run of 1.81 boards may have the capacitor correctly installed (the reverse of the board markings) as the problem was identified late in the 1.81 production run as work was done on rev. 1.9. Visual inspection will confirm if this is the case. This capacitor problem is of the nature of 'you ought to fix it the next time it is in for a major tune-up' and not of the 'park it immediately and tow it to the garage' type. In production we usually install capacitors with a significantly higher working voltage than what is applied to them and they have generally worked just fine in these mismarked boards, even though reversed in polarity. Nothing is going to be damaged if the capacitor is left in backwards, except maybe the capacitor itself. There are no reports of any capacitor explosions or other bad effects. Repair instructions: Measure the voltage on pin 6 of U2 (MAX232). It should be in the range of 8-10 VDC. Remove C23 and reinstall with the negative lead in the + marked hole. Measure the voltage on pin 6 again. If still within the 8-10 VDC range then C23 was not damaged. If the pin 6 voltage is out of range either before or after reversing C23, the capacitor should be replaced. Gwyn Reedy, W1BEL PacComm Packet Radio Systems gwyn@paccomm.com http://www.paccomm.com Date: Sun, 19 Jan 97 17:37:36 NZT From: zl4til%zl3ur.bbs@zl3rr To: all@sthlnk Subject: RE: Tiny-2 MK2 defect (Part 3 of 3) Path: ZL3UR!ZL4TAY!ZL4TIL >>>>>From the internet Capacitor C 23 Question Gwyn Reedy wrote: > > The report you received is correct. Let me tell you more about which models > are affected and what should be done about it. > > Amateur radio models affected are the TINY-2 MK-2, TNC-NB96, and IPR-NB96 > which all incorporate the TINY-2 circuit board. > > Repair instructions: Measure the voltage on pin 6 of U2 (MAX232). It should > be in the range of 8-10 VDC. Remove C23 and reinstall with the negative > lead in the + marked hole. Measure the voltage on pin 6 again. If still > within the 8-10 VDC range then C23 was not damaged. > > If the pin 6 voltage is out of range either before or after reversing C23, > the capacitor should be replaced. > > Gwyn Reedy, W1BEL > PacComm Packet Radio Systems > > gwyn@paccomm.com > http://www.paccomm.com > > Gwyn's test for the bad capacitor is reasonable ONLY IF THE SERIAL PORT IS ADEQUITELY LOADED. If you run the test without anything connected to the serial port, the voltage may well look fine even with a bad capacitor (I can attest to that from personal tests). Before running the voltage test, connect at least a terminal to the serial port. If you are using the TNC in a multi-node nodestack, then you should load it with a 5 or 6 port diode matrix fully loaded with TNCs. Or an equivalent static load if someone can figure out what that load should be. Burt VE2BMQ