Roberts 606MB output pair

[Hermit6345]

There is a Roberts 606MB on my bench. The silicon output pair are playing up so I would like to replace them. According to my service info they should be BC465A and BC464A. I have searched high and low on the web for a source of these but drawn a blank. Can anyone suggest a source or as a last resort, an equivalent for these devices?

Ian

[paulsherwin]

Don't replace like with like as you don't want more Lockfit problems. I used a BC327/BC337 pair which have worked well despite being a bit underspecified. You could use any complementary silicon output pair you have which will physically fit. Obviously it's a bit easier if they have a c-b-e leadout.

You will need to check and adjust the quiescent current afterwards.

[Hermit6345]

Thanks for the tips Paul. Will do as suggested.

Ian

[Hermit6345]

The plot thickens. The service info I have for the R606-MB shows the output pair to be BC465A/464A. However, it turns out that the model I have uses BC368/369 complete with parasitic stoppers. The triangle of holes in the pcb are 90 degrees anti -clockwise compared with the transistor orientation shown in my service info. This allows the BC368/369 to "fit" the holes because the pinout is not the same as the 465/464 or its equivalent. Does anyone have the schematic for the version I have? I suspect the the circuit of my model may be slightly different from that shown in my paperwork.

Ian

[paulsherwin]

I think these will be an old repair. They are later transistors but good subs for the originals apart from the leadout.

[Hermit6345]

That's exactly what we thought. However after very close examination of the soldered joints using a glass, coupled with the fact that the pcb had obviously been made to accommodate the BC368/369 pair, or something else with an e-c-b pinout, we came to the conclusion that this pcb could possibly be slightly different in other areas as well. The amp certainly doesn't work with the devices fitted as per my 606MB documentation but with BC327/337 output pair. Every resistor has been checked and all electrolytics checked by substitution. I am minded to replace the Lockfit devices with BC1x7 type devices, but all of the transistors have already been replaced as per the schematic we have with tested Lockfit devices

[tri-comp]

Possibly the 606 exists in more variations.
My schematic says Ge AC187/AC188 for output pairs.
Have a look.

rgds,

/tri-comp

[Heatercathodeshort]

There are two versions of the R606MB. The first version [service manual dated Jan 1975] uses AC187/AC188. Versions produced after serial number 600001 show BC464A/BC465A as these two scans confirm.
Not one of Roberts best. Sold quite a few but had tuning capacitor/wavechange switch/output transistors/tuning drift and other probems. Quite unbelievable when you look at it's origins. Hope you get it sorted, John.

[paulsherwin]

All the audio lockfits are suspect and I would change them. They are just general purpose NPN/PNP types so replacement is cheap and easy.

Then measure some voltages in the amp. These DC coupled designs are a pig to troubleshoot.

[Hermit6345]

Thanks for the inputs, most useful. I will check the serial number but there is definitely an inductor in series with C44 and the latter is 470uF. This would make it a post 600001 model using BC465/464 pair ( Earlier model no inductor and C31 680uF) So far so good. The parts layout for the later model shows the flat sides of the BC465/464 facing east The arrangement of the mounting holes for the output pair in our board are such that not only would lockfit devices have their flat sides facing north but the BC465/464 pair would not work because the holes are “wired” for a b-c-e device not an e-b-c configuration like the 465/464 pair. The BC368/369 pair matches the b-c-e configuration. I need to look at the serial number of course but could this be some kind of late variant to overcome the problems with the lockfit output devices?. I was unaware of some of the other problems with this set John. Only the fact that I don’t like being beaten keeps this receiver out of the scrap bin. Paul you are right on both counts.

[paulsherwin]

I don't think there were problems with Lockfits during the production run - they seem to have emerged during the last decade as the transistors aged. I wonder if this is actually a prototype or short run model made between the end of Ge and the adoption of the BC465/464. The serial number may provide some info.

[Hermit6345]

Well, the serial number is 695213. Obviously not at the start of the new version production run, but how near the end? Anyone got any ideas? Would that mean that they made at least 95,211 units if they started at 600002 or does the S/N contain some other info?


Ian