Trio Kenwood TR-9000 TX issue

[GSBX1220]

Hi folks

Small issue with my Trio TR9000 multi mode rig.

When the set warms up either from being only in RX or after a small period (varies) of TX the power drops not gradually, but like a switch from full 10 w to 7w output. This is across all modes.

My thought is as it's common to all modes an issue with the final board mounted at the rear. One thought was the TX protection kicking in but the VSWR is 1.2:1.

Any thoughts? Anyone came across this?

Cheers

[Andrew2]

I can't remember if the 9000 uses a discrete PA or a block type. The blocks tended to suffer low output and intermittencies as they got older, and SSB became low and badly distorted.
IIRC, it was a common problem across all makes that used the block PA.

[Radio Wrangler]

Probably a cracked ceramic substrate in a Mitsubishi PA module.

They can sometimes be repaired under a microscope with fine wire bridges (I've done it)

There are deposited resistors and spiral inductors and chip capacitors. The main transistors mount through holes in the ceramic directly onto the copper spreader. There are bias comp diodes as well.

Ceramic is bonded to copper heat spreader plate and the plate gets bolted to the diecast heatsink. The whole structure is worse than a bimetallic strip! The rigs had no temperature warning or shut down so people rag-chewed until "You went very quiet 20 mins into that last over..."

David

[GSBX1220]

Hi.

Thanks guys.

Yes, it's a Mitsubishi M57713.

It has been reported that my SSB audio is slightly distorted. The station I worked likened it go the amp loosing linearity and the peaks of my speech being compressed. Another station that I worked at the same time, but was much weaker with me, said that I started to fade away during my transmission.

My transmissions on FM are spot on as reported by the same stations. One in Dunfermline and the weaker one in Blantyre.

The room where I have my radios gets quite chilly during the night, so it's a repeatable fault when I switch the set on from stone cold. What I also noticed this morning is the TX power flicks between 10w & 7w just before it finally settles at 7w.

I had a look on line on Thursday night for PA modules. It's an obsolete part. Anyone know a supplier? Some of the Chinese sites claim to have them, but wary of fakes or if they will ever arrive!

[Andrew2]

Yes, the old bipolar modules have been obsolete for some years, now replaced by MOSFET types. I use GH Engineering http://www.ghengineering.co.uk/products/Mitsubishi.htm

The M57713 is replaced by the RA30H1317m.

In the past I have successfully replaced faulty old ones with the new type, but I've made my own PCB to achieve it. The good news is that the bias supply can be just a pot, no need for fancy bias regs or hot diodes!
It's worth e-mailing him, as he's a knowledgable bloke and may even have an easy answer.

[Radio Wrangler]

Oddly, there's a TR9000 I repaired in the Dunfermline area, though Phil went SK several years ago. Could be anywhere now.

I dont expect Kenwood dealers have any stocks left, these things were high usage spares.

Richard's suggestion is best, else it's microscope time...

David

[GSBX1220]

Thanks guys.

I have emailed GH engineering and I will see what they say. I'm not sure what extra work or modifications is involved in fitting the MOSFET part. I have quite limited test gear.

[Alf]

Could this type not replace it? Seems similar.

[Radio Wrangler]

Similar.

It's a VHF power module made by Mitsubishi. They made loads of types, mostly for FM transmitters. The TR9000 is a multimode set and needs one capable of linear operation for SSB. They're much less common.

David

[VT FUSE]

Whilst trawling through eBay last night,I did happen to see a PA unit for the Kenwood TR715e 2M Multimode transceiver for sale - is that module suitable despite not being same number? the module in the PA is an M57727.
No personal connection to listing or seller etc...

[GSBX1220]

It's a higher output power. The PA assembly as a whole probably won't work as the wiring harness most likely will not be compatible for starters.

As for just using the module, I don't know. The pin out may be different. I did wonder if a PA module with a higher rated output but with the same pin out would work. I suppose if it was driven to run well under its rated output it may last longer and not killed by the heat cycling that seems to ruin these modules.

[VT FUSE]

Indeed,I was not suggesting that the entire PA board etc would fit, but was thinking that the module itself may be a sub,of course that depends on how much you want to keep the TR- 9000 functional....

[G0HZU_JMR]

Interesting stuff! I've got an old Kenwood TR9130 here but I don't have the microphone for it so I've never really tested it on Tx apart from bridging the pins on the panel socket when on FM to see if any power was generated.

So maybe I need to check mine is OK on SSB! I think the PA module number for the 9130 is the M57727 according to the block diagram. I used to have a decent stash of the various Mitsubishi modules as we used them at work for making quick and cheap test amplifiers. Sadly I can only find a few of them and the only SSB version I can find is the M57762 for 1300MHz (23cms). This was used at work as part of a non linear radar we developed and this one I have here must have been a spare.

They do look to be fairly easy to open up so I'd follow David's suggestion and have a look under a decent microscope. It might be an easy repair! I really need to make or buy a mic for mine and see if it is OK.

[GSBX1220]

I would give the microscope repair a go, and that may be needed. One small stumbling block is that I don't have a microscope

[VT FUSE]

A cheap Chinese USB microscope would give plenty of magnification,I bought one a couple of years ago and found that to implement it properly would entail fitting it to the mechanical column of a disused photographic enlarger minus its light housing,neg carrier and optics,pity I heaved two Durst B&W specimens into the skips around ten years ago.No doubt simpler and smaller schemes would be forthcoming!!

[GSBX1220]

Today I separated the back from the chassis to give access to the module. Plan was to see if putting firm thumb pressure on the module whilst in TX would cause power variation. No difference.

Still no luck on the hunt. Still waiting a reply from GH engineering. I also emailed Castle electronics to see if they have the part. There is also a place called HiFiSSB that do used working parts. They list some stuff for my set, but not this.

How unreliable are these modules?

[Radio Wrangler]

Unreliable enough to have cleaned out pretty much all stocks of spares. You might be very lucky, so you've got to try.

David

[VT FUSE]

Wonder which other rigs used the required Mitsubishi module,it was surely used by other rigs and even other manufacturers? possibly with a different part number.Given this info the net could be cast further.
I too wish you luck with a solution,it may just take a little more deeper digging to find it.
What scope is there for reverse engineering a replacement,probably external to the rig? after all,the drive to the module is ok and many older rigs were used for 2m ssb with external linear amplifiers after all.

[GSBX1220]

Hi

Not sure what other rigs used this part, but apparently Icom used it but under a different part number. I have found one on eBay. https://www.ebay.co.uk/itm/ICOM-SC-1...-/263280064427

Since these are so unreliable, it might be best to see what GH engineering say about fitting a MOSFET unit and how difficult that is.

Does the microscope repair on these modules give bettter reliability than as new?

Using the set to drive another amp was a consideration, but a last last last resort.

[GSBX1220]

Earlier in the week I contacted Castle electronics. Geoff suggested fitting the higher output M57727 module but running it at around 12w. I think that's the route I'll go down as I have found a reasonably priced new module.

Not 100% on how to run it at the lower power. As far as I know, there is a pot that adjusts the drive, but that's all I know at the moment. Anyone have advice on this?

GH engineering got back to me. Fitting MOSFET solution is a no go basically.

It's all good fun!