A friend asked me to look at his Yaesu FT-7800 dual band FM transceiver (ham radio) which stopped transmitting on 2 meters. I put together this video showing the troubleshooting procedure I took to uncover the problem, which appears to be a pair of PIN diodes in the final filter stage. One person commented in the video that these PIN diodes might be obsolete. (Torex XB15A709A0HR). Checking Torex's website confirms that these diodes are discontinued. I did find some from a Hong Kong supplier. However, I'll look for some local (US) sources before ordering from overseas. The datasheet for the original is a bit sparse, so not a lot of data to go on for looking for a suitable replacement. Anyway, here's the "Part 1" video, which shows the troubleshooting process I went through. Hopefully, I'll be able to locate replacement PIN-diodes and finish this repair...
Here's the link to my "repair log" video:
Interesting video, as always. I have that radio's successor, the FT-7900. For curiosity's sake, I did a bit of searching and found that the corresponding diodes are L709CER in the 7900. Yaesu probably ran into the same problem about the original diodes being unavailable and found the substitute. I'm not sure if those are direct drop-in replacements, or if they made some changes around the bias network or something, but if your searches for the original part comes up empty, that might be something to pursue.
Nice video. I might suggest having a chat with your friend about the circumstances surrounding the death of the PIN diodes. When they die, it can be due to a loose/arching connection in the antenna system. If the failure happened while using his regular antenna, then the repair may not last for long, unless any antenna problems gets fixed before putting the radio back in service.
What kind of probe are you using in the video? Not sure I'd dare throwing 100Vpp @ 144 MHz at even a passive probe...
watched this this morning
nice vid, i have to do something similar
with an old hf kenwood of a friends that
has no output
Nice video. I might suggest having a chat with your friend about the circumstances surrounding the death of the PIN diodes. When they die, it can be due to a loose/arching connection in the antenna system. If the failure happened while using his regular antenna, then the repair may not last for long, unless any antenna problems gets fixed before putting the radio back in service.
What kind of probe are you using in the video? Not sure I'd dare throwing 100Vpp @ 144 MHz at even a passive probe...
He was just operating normally. I'll have him check for an intermittent in the antenna. I'm suspecting the failure was due to the fact that these diodes were in parallel with no provision to evenly split the current. One probably took most current, failed, and left the other to die. Just my opinion.
The probe was a Tek P6139A. Rated for 400Vrms at low frequencies, deranged to 50Vrms at VHF and low UHF, so it was OK.
To speed up troubleshooting you could have kept an eye on the current draw at the PSU, I am sure it would have given you a better idea which end of the chain to work from.
Every radio I have fixed with similar faults have drawn significant more power on TX than RX indicating the circuitry past the mixer and pre / driver was working, so you would have worked back from the antenna, not ideal for a video but in a repair business it could mean 2 radios in the same time of 1, doubling the $
The probe was a Tek P6139A. Rated for 400Vrms at low frequencies, deranged to 50Vrms at VHF and low UHF, so it was OK.
Thanks! Was looking at new Tek probes earlier, and couldn't immediately spot the voltage vs. frequency derating curves. It didn't occur to me to search for discontinued models...
Unfortunately it won't compensate 20pF, so that model is of little use to me. Tek I probe interface, 8pF load cap and 500MHz BW would otherwise have been nice to have.
The probe was a Tek P6139A. Rated for 400Vrms at low frequencies, deranged to 50Vrms at VHF and low UHF, so it was OK.
Looks more like the P6139B to me. Rated up to 300 Vrms for low frequencies, though still ~50 V for higher frequencies up to 500 MHz.
Thanks! Was looking at new Tek probes earlier, and couldn't immediately spot the voltage vs. frequency derating curves. It didn't occur to me to search for discontinued models...
You can usually find the derating curve in the user manual, which is available for download at tek.com after registration.
Unfortunately it won't compensate 20pF, so that model is of little use to me. Tek I probe interface, 8pF load cap and 500MHz BW would otherwise have been nice to have.
A larger compensation range would translate to a higher input capacitance. The 8 pF capacitance is already causing unacceptable circuit loading for many applications, which is why people typically reach for FET probes beyond a few hundred MHz. Note that the compensation trimmer cap is not known for particularly tight tolerances, so you may get lucky. I've successfully compensated P6137A probes (400 MHz, designed for the 2465B) to 20 pF scopes with room to spare, even though the compensation range only goes up to 18 pF or so.
The probe was a Tek P6139A. Rated for 400Vrms at low frequencies, deranged to 50Vrms at VHF and low UHF, so it was OK.
Looks more like the P6139B to me. Rated up to 300 Vrms for low frequencies, though still ~50 V for higher frequencies up to 500 MHz.
Yeah, you're right, it was the P6139B. Note that the P5050B is nearly identical but designed to compensate up to 22pF.
The P5050B probe is virtually identical, 500MHz BW, and will compensate to 22pF. Same datasheet as the P6139B.
http://www.tek.com/sites/tek.com/files/media/media/resources/P6139B-P5050B-500MHz-Passive-Probe-Datasheet-2_1.pdf
Looks very nice. There also appear to be other viable options for me among the P613x series, if I don't mind buying used probes. The trick would of course be finding some in good condition.
Currently I am slowly 'rebuilding' my home lab, trying to fill in the major gaps first. To that end I have decided that my old Tek scope will keep serving, yet my collection of probes is mostly junk in dire need of replacement. At the moment 500MHz BW would be mad overkill for my needs, yet it would be nice to have the readout pin and the low(er) input capacitance.
A larger compensation range would translate to a higher input capacitance. The 8 pF capacitance is already causing unacceptable circuit loading for many applications, which is why people typically reach for FET probes beyond a few hundred MHz. Note that the compensation trimmer cap is not known for particularly tight tolerances, so you may get lucky. I've successfully compensated P6137A probes (400 MHz, designed for the 2465B) to 20 pF scopes with room to spare, even though the compensation range only goes up to 18 pF or so.
Thanks for the compensation tip, will definitely keep that in mind.
I'm aware of the issues surrounding the probe input capacitance for passive probes, and I am just looking at my options for generally poking around in non-critical circuits. However if possible I wouldn't mind having 8pF in place of the more customary 12-13pF input capacitance. For the more critical situations I am working on a simple, active probe DIY design, where the main focus is a low input capacitance.
Always enjoy your videos, and if you are starting to video all repairs from now on even better.
That diode looks like a typical MELF packaged PIN diode and the datasheet suggests it is quite similar to PIN diodes from Macom or Microsemi.
A couple of examples of diodes that are in a similar ballpark are the MA4P7101-1072 from Macom or the SM0511 from Microsemi. These diodes might prove to have a slightly higher ESR at 45mA bias but probably nothing significant. I've got loads of MELF PIN diodes here because I've designed lots of TR switches in recent years.
The slight worry with that PIN switch design in the Yaesu radio is that the diodes look to be biased at about 45mA each and the Tx power is around 50W. The Irms for 50W is 1A in a 50R environment so if you allow for a highish antenna ISWR then the worst case scenario would be an ISWR 'standing' peak right at the diodes. If there was 0.8A of RF in each diode and the ESR of the diode was 0.6R then each diode will have Pdiss of about 400mW (plus maybe 50mW Pdiss due to the bias). There are two of them side by side and I doubt the PCB will be designed to conduct this combined heat away efficiently. So this might be a candidate for the cause of failure.
If you do get some diodes and repair the radio successfully I would suggest making sure the aerial has a low VSWR and the ideal thing to do would be to look at the diodes with a thermal camera on Tx (into the aerial) and trim the feeder length to place an ISWR minimum at the diodes. This would make them run a LOT cooler if (for example) the ISWR measured 1.5:1 and you compared an ISWR maximum against a minimum on the thermal camera for different feeder lengths. I've got a lot of experience of doing tricks like this
Whatever diode you choose, it has to have low ESR at 45mA bias (eg 0.5 ohm) to minimise i^2 *R losses and it needs to have a similar capacitance (eg 1.2pF per diode) because the PIN switch will be designed with these two parameters in mind. i.e. the 1.2pF per diode will be absorbed into the overall network design and putting in diode with a lower capacitance might make it perform worse.
One thing you don't need is a diode with a high reverse voltage rating because this diode isn't used in a switch that needs a high reverse bias. So a 50V or 100V diode will be fine.
The MA4P7101-1072 from Macom or the SM0511 from Microsemi are a good starting point and you could email them both to see what better alternatives they can offer. As you work for Tek I would expect you could blag a load of free samples from both of them
The other potential cause of PIN diode failure would be an intermittent bias failure. So it would be worth checking the quality of all soldered joints right back through that TX bias path to the VHF 8V TX switching transistor(s) to make sure one of the joints hasn't gone bad over time. If this happens then the two PIN diodes will toast very quickly.
I guess if this radio is a USA model then it covers 144-148MHz and it probably isn't worth trimming the coax for a low ISWR peak at the diodes because the band is so wide. i.e. it won't be possible to avoid an ISWR peak at the diodes at some point in the band assuming a reasonable feeder length is used.
Thanks for the good comments, G0HZU. I've already asked the owner of the rig to carefully check his antenna. As for replacements, I decided to use the devices that Yaesu is currently using in the successor rig FT-7900R. That RG has an identical RF PA and filter bank, except they changed to a different PIN diode pair. Hopefully the replacements will arrive this week. I'll be sure to double check the switching/bias path for intermittents.
For those that watched Part 1, here is the Part 2 showing the rest of the story with the repair of the Yaesu FT-7800 Dual Band FM Transceiver (ham radio).