Ten-Tec Paragon 585 Repair.
I've recently bought this American made Ten-Tec Paragon 585 as a non-working transceiver from an auction, intending to give it an overhaul before using it on the station here. Here is some of the story of the repair work.
As can be seen in the photo above this particular one was in excellent exterior condition. The interior condition was excellent too. The fault report from the seller said that there was a PLL fault and it needed re-aligning. PLL faults on these can be fun (not) to diagnose !
The transceiver was placed on the bench for initial testing...........
On getting the unit I powered it up and immediately the display started flashing a PLL Lock failure warning message. Time to get out the scope and start investigating the PLL operation on the major and minor loop boards.
Diagnosing the PLL fault.
The loop boards (81338 , 81337) and reference oscillator board (81336) are inside a shielded box on top of which is Low Level Amplifier / PSU board (81340) shown above. The lid of this is the heatsink for the regulator IC and transistors on this board. By removing the lid securing screws, removing the connectors from the noise blanking controls (connectors R and 67) the control board assembly can be lifted and and turned upside down and placed to the right to allow access to the loop boards. I placed an insulating sheet under the PCB to prevent short-circuits.
There was evidence that some previous repair work had been carried out on these boards, as if an attempt had been made to locate a fault.
The PSU voltages were checked first, these were all OK. The first checks on the reference oscillator board showed that all was OK, the 2.1, MHZ, 42Mhz, 84 Mhz and minor loop mix signals were present and correct. So off to the loop boards. A check at IC U1 pin 7 of the major loop board PLL synthesiser IC showed that it was logic low, the PLL was unlocked.
The reference signal was present at U1 Pin1 but the mix-down input at U1 Pin 12 from the circuit was not. Checks at the mixer IC U3 showed that the minor loop signal was present (pin 10), but no Local Oscillator (LO) signal (pin 1). A check of the VCO / LO circuit showed that there were no VCO outputs from any of the VCO's at any frequency on the display.
There are four VCO's, each spanning part of the frequency range of the transceiver. The appropriate one is switched on by the CPU on the logic board according to the selected frequency band on the display. The four select inputs are on connector 90. One of these should be logic low, switching on the supply to the varicap controlled VCO oscillators (via Q5,6,7 or 8) inside one of the VCO cans. On this transceiver all these control voltages were high, no VCO's were being selected !
Connector 90 was removed from the Major Loop board and each VCO select input was grounded in turn, checking for a VCO output at transistor Q17. Each VCO fired up OK. This meant the problem wasn't on the PLL boards after all.
Attention was then directed to the CPU Logic side if the transceiver (board 81267). On its way back to there, the signals pass thru a small control filter board (81377) on the side of the CPU Logic screening box. All appeared OK here so it was time to take a look at the logic board itself (81267). The command to enable the VCO's come directly from the MC146805E2 CPU IC, U8 via connector 66. The display was all working OK, so were all the controls, so it seemed unlikely that the CPU was at fault, or was it ?
All the outputs concerned came from the same port, Port A. The port is bi-directional and the CPU decides whether to make them inputs or outputs determined by what's in the firmware EPROM program. Port A appeared to be completely dead inside this CPU.
A new MC146805E2 CPU IC, U8 was fitted (see above), and
the fun really started ! On powering up, a row of random digits appeared on the
display. The CPU had been in my static protected parts drawer for a long time,
it could have been bad so the old CPU was refitted and the transceiver powered
up. This time there was nothing on the display, even after removing the battery
and pressing the reset button as described in the manual !
After the initial panic my brain kicked in again. In the past I've had problems with old IC sockets before, the pins tarnish and changing the IC causes a bad connection to occur afterwards. The IC socket was cleaned with non-oil residue cleaner and the new CPU IC fitted again. After power up and going through the reset sequence the Paragon powered up with the VFO-A indicator lit and there was sound from the loudspeaker ! The PLL stage was working once again !
The morals here are that to diagnose PLL lock faults the best approach is to use the right test gear (especially a high frequency oscilloscope and frequency meter) and actively trace the signals through each stage. Also to not forget that on modern transceivers the processor is closely tied in with the whole frequency control system. The PLL fault may not even be in the PLL stages at all !
The Paragon was then put through a series of tests with an RF analyser (including spectrum analyser) to see if it was performing to its specification and the legal requirements before it could be put on the air.
Here many more problems were discovered !
Everything seemed OK until the transmitted RF spectrum was analysed. Intermittently a "wobble" could be seen in the spectrum plot. Tapping around the transceiver soon revealed that the problem lay in the PLL section, either something was microphonic or there was a bad joint. Tapping the major loop board caused up to a 75KHz deviation in the output frequency.
The major loop board (81338) was removed and inspected, many of the soldered joints looked very bad. All the joints were carefully resoldered and the major loop board was re-fitted. The fault had now been cleared, the output was reasonably steady on tapping the board.
Another problem was discovered on these tests, the TX carrier suppression on USB operation was very poor (only 20db). The causes were found to be on the transmit audio / BFO board (81339). The various BFO oscillator trimmers were first adjusted in accordance with the manual using an accurate 8 digit frequency meter. NOTE if later software is fitted the settings differ, use the frequencies given in the firmware notes. There was still a slight problem with carrier suppression however, it was found that the DSB modulator balance adjustments R38, R398 and T1 needed re-adjusting whilst observing the DSB modulator output on a spectrum analyser. This completely cured the carrier suppression problems in USB mode.
All the remaining tests came out OK.
Two modifications were then carried out. One was to allow the TX output power to be reduced to 5 Watts, the other to place a pull-up resistor on the RTTY Mark / Space control input to allow it to be used with my RTTY adaptors open-collector outputs. One good thing about older transceivers like this that use a modular construction and through hole components is that they are relatively easy to modify.
The transceiver was then ready to go on-air !
First QSO's with the Paragon.
The Paragon 585 was installed in the shack on 12/08/2006,
and adaptor leads made up for my PSK / RTTY adaptors. My first QSO with it was
in PSK mode with IW5ELL (thanks Cosimo). The next two QSO's were using RTTY with
LZ80R and EA5FL (Jose). I'd like to especially thank Jose EA5FL for his very
helpful signal report.
Later I had several SSTV QSO's using the Paragon, including a Transatlantic one with AB1AI. Here I was able to see the images I sent on his SSTV Live page http://ab1ai.com/sstv/ (see picture below).
After running it for a week it became apparent that all was not well with the receiver performance, it was very poor. In particular there was a "splatter" problem (sounding like transmission "splatter") up to 6Khz away from any strong signal. The Paragon was put back on the bench for receiver circuit testing.
The problem was soon found to be due to mis-aligned IF stages. In particular the 75Mhz IF stages on the first mixer board (81331) and 9Mhz IF stages on the second mixer board (81332) needed re-adjustment. Also the VXCO frequency range on the PBT (Pass Band Tuning) board (81333) was incorrect. The VXCO frequency and adjustment range was re-adjusted to the settings in the manual. The IF stages were re-aligned by injecting a signal at IF frequency into C19 on the first mixer board (81331). After checking the remaining IF stages, the receiver was returned to the shack and re-tested on air. The "splatter" problem had now gone away.
Weird lock-up fault !
During testing, what appeared to be a CPU fault was seen. On power up the display would be dead (blank) and could only be brought back by pressing the CPU reset button on the side of the Paragon. The cause wasn't the Paragon at all, but my 13.8V bench supply ! On switch on the voltage was slow to rise and was causing the problem to occur. Changing to a different PSU cleared the problem entirely.
Adding a cooling fan and more modifications.
I often work digital modes and SSTV, during which the 100% duty cycle caused the convection cooled PA stages to run rather hot. Whilst on the test bench this time, a cooling fan was added to the PA stage heatsink. The cooling fan was a temperature controlled 12V clip-on type for PC CPU cooling, bought from a recent ham fair (Alvaston rally I think). The plastic clips of the fan body clipped over the heatsink fins nicely and was secured using a few dabs of epoxy glue. A 12V power lead was made up and plugged into the 12V AUX power RCA jack at the back of the Paragon.
A 9V Ni-Cad (actually a Ni-MH equivalent) rechargeable memory backup battery was added by fitting a modification that was standard on later versions. A 2K2 resistor was added across diode D21 on the logic board 81267 (the resistor is shown as R71 on later models). The diode is next to the 9V battery PCB connector on the logic board. The manual says that on these versions an Alkaline battery can still be used...........personally speaking I wouldn't ever try that !
The Paragon was given a thorough testing during the SARTG RTTY contest (19-20/08/2006), it never missed a beat ! The receive performance was astounding, especially when working weak signals close to much stronger ones. When running at 50-60 watts directly into the antenna at 100% duty (RTTY QSO's) the cooling fan did the job well, the PA stages kept cool. It's temperature control reduced the speed (and fan noise) once the heatsink was cold.
The Ten-Tec Paragon is now most definitely working fine and I'm really impressed by the performance of this old transceiver in comparison to more modern models from other manufacturers.
I'd also like to thank Ken M0DQS for his help in repairing this transceiver and the loan of some of his RF test equipment, and my company, Fletcher Moorland Ltd for the after-hours use of my workbench for this project.
I'd also like to thank N4NQY who's Unofficial Ten-Tec Pages contained very useful information and modifications for this (and other) Ten-Tec transceivers.
Now for a well earned beer !