Replaced capacitor across power-pole connector?

Jerry - NG6R

My much-loved Flex 6000, purchased in March 2018 and not covered by an extended warranty, developed a power line short (actually 1.8 ohms between power-pole jacks) during normal operation. Was originally thought to be an external power supply issue—Astron 35M showed 1.5V and 1A just after failure—since other PS-attached units went out at the same time; but subsequent testing showed that the “crowbar-ing” culprit was the Flex radio. BTW, the power supply came through the direct short unscathed—PS crowbar circuit did what it was designed to do.

 

Prior community forum posts have attributed this same failure mode to a shorted 10uF electrolytic capacitor across the power-pole connector. Has anyone with this radio (beyond the warranty period of course) ventured deep enough into the hardware to get to this connector and/or locate the one (or more?) electrolytic caps across the DC power jack? If so, do you have any lessons learned that you can share? Access to this area from below looks impractical or at least difficult due to the way the various aluminum shield panels are secured. Access from above appears to be the successive removal of at least two stacked PCBs, but might be the best access approach. Suggestions would be appreciated from anyone who has ventured into the region of the power-pole connector and (likely-nearby) caps.  A photo would be terrific.

Jerry - NG6R

Access to shorted rear power-pole connector (integral with the power amplifier PCB) was easily gained from the top of the unit by removing two PCBs stacked right above it on vertical standoffs. Lots of photos taken. Disconnected the two inter-board ribbon cables to electrically isolate this board, extracted the intact 30A fuse and made measurements to demonstrate that the power-pole connector itself is OK as is the 10uF (C48) ceramic SMD capacitor directly across it, which had been the likely root cause of similar shorts in some previously reported cases.  This board houses the two PA FETs and the two driver/amplifier FETs. Drain terminals were carefully lifted on all four active devices and tested to determine that all four devices are good, yet the ~2-ohm DC power line short persists, possibly within some shunt SMD component somewhere in the PA DC conditioning circuit. I’ve taken troubleshooting to a reasonable stopping point on this mostly SMD board and have determined that the next step is to purchase and replace the PA board, assuming FRS is willing to sell one outright. Will contact Flex and open a Help Desk ticket. Thanks to the responders who expressed an interest in helping to resolve this issue.

Jerry - NG6R

Last (and hopefully final) chapter! I opened a HelpDesk ticket, described the short across the DC power connector and described the troubleshooting tests already performed (i.e., isolated short to PA board, confirmed C48 OK, confirmed all four PA board FETs OK). Since there are no published schematics and the PA board is largely comprised of surface-mount devices, I didn’t know what additional tests to perform to locate the shunt device(s) that was/were causing the short. I expressed my resignation that a new replacement PA board might be needed and asked for next steps to purchase one.

 

Flex HelpDesk representative Ken NM9P responded back quickly and identified a few more components to test before giving up entirely on the PA board. His service technician suggested testing Diode D3 (which, after lifting one side of this SMT device, tested OK) and test a series of four 10uF 25V ceramic capacitors, C62, C63, C64 and C65. I found that two of these capacitors—C63 and C64—were shorted. Ken indicated that replacement parts generally are sourced from Mouser or DigiKey. I ordered several replacement SMT caps from Mouser (ceramic, AVX manufacturer, 10uF, 25V, X7R dielectric, 10% tolerance, 1206 package code) with fast shipping. I installed them, reassembled the three removed PCBs and returned the radio to operation yesterday. So far, the radio is working OK.

 

Ken acknowledged that it would be very difficult to speculate what caused the capacitors to fail, offered to run a bench diagnostic and issued an RMA that’s good for 3 months. I’ll watch its operation and performance closely over the next several weeks and months. Hopefully, I won’t have to take advantage of the RMA, but am very appreciative of its availability.  A big THANK YOU to Flex for their cooperation and support while troubleshooting this anomaly. 

Jerry - NG6R

Make that 6600.

Jerry

VE7ATJ_Don

Hi Jerry.... this is probably not the comment you were looking for, but it might be best to submit a help ticket and get the staff @ Flex to give you a hand. I know you're out of warranty and maybe wanting to do a bit of DIY, but as you described in your post, getting to the culprit might be a tad more difficult than expected, and may have unintended consequences.

Generally, I've heard really good comments about the responsiveness of the Flex staff and pretty good turnaround in the repairs department.

Ronald Panetta

Jerry, can't help you on the repair but can describe my experience and provide a word of caution on powering a Flex. I had a $600 out of warranty repair for my 6600. I can't prove it was the capacitor but theorize it was. In my case the board was charred so bad it required replacement. I too have an Astron 35M which tripped on overcurrent protection. The bad news is I also had a 55Ah battery fed through a power gate in the event of loss if commercial power. That battery was fused at 30A via standard ATC fuses. When the Astron tripped, the battery kicked in and I found that thermal fuses like the ATC fuses are very slow. It never blew, the fusible link deformed but not before the 6600 let the smoke out.

I encourage everyone to power their Flex with a power supply with electronic overcurrent protection and discourage use if a battery. I have since disconnected my battery.

I do encourage you to contact Flex as your repair will likely be cheaper. In my case 2 boards were replaced. Flex staff were very good in communication and recommendations on the repair although it still cost me $600. If as anyone is interested in the write up I did on the failure, please email me. Email good in QRZ.

Ron, WB2WGH