Heads up as per Jurb...hope this helps someone.

Just a few notes on these transistors and the results of some experimenting.
I had a KV27ts32 and a KV32s25 in the shop with the classic HOT shorted and
Q601,602 shorted. I had some 2SC4834M in stock and ordered some plain
2SC4834 from Tritronics and some 2SC4834P from Acme Enterprises (Orlando).
I tried each type in both sets and found that the higher rank (higher
letter, higher gain) transistors ran progressively cooler. The unranked
versions ran almost 20 degrees F hotter than the best, the P rank version.
I matched the transistors from batches of 10 by forward bias voltage on the
junctions then in circuit for nearest to 50% duty cycle. I have found that
swapping pairs or swapping the caps on Q601,602 can often improve the ratio
to closer to 50%.

I don't post the exact temps because I don't have a well calibrated meter,
but I got a good relative indication. I always used the M rank version and
never had any recalls due to the transistors, but was curious after hearing
someone say that they only used the faster versions. Always looking for a
better solution.

Summing up, it appears that the higher rank transistors may be the better
choice. Acme has them at a great price (10 for $15.95) and has free
shipping for the time being on orders over $25. They are a new operation
but Glen, the main contact there seems to know the business well and seems
very interested in winning over the business of shops like us. So far they
seem to be good to work with, though their inventory selection is not up to
other vendors yet.

Also, I had the idea of using my Sencore with the drive signal (H key pulse)
to test the MOVs (VDRs) in these and it seems to work well. The only
problem is that it only goes to 300 vpp so the ones with higher ratings only
get tested to that level, but not seeing any leakage at that level I assume
that they are ok. At least I know that they aren't leaky below 300. The
others can be tested to the correct breakdown voltage.

BTW, the KV32S25 was a bit atypical in that the HOT failure was due to a bad
FBT. The price was pretty steep and the customer did not want to fix it, so
I gambled and tried an aftermarket transformer from HR. It seems fine so
far and at the lower price the customer went for the fix after being given
the option. We'll see if it holds up. I don't like subbing flybacks but
these have had a good reputation with other techs that I talk to.

Leonard G. Caillouet

"Leonard Caillouet" bravely wrote to "All" (07 Nov 03 22:36:30)
--- on the heady topic of "FYI 2SC4834 Sony power supply Q601,Q602"

There might be a design issue with the horizontal driver circuit if the
gain of the hot is that critical. Is there some field service notes
issued about this problem? It just seems a shame to be burning up
expensive transistors when all along it was another problem...

LC Reply-To: "Leonard Caillouet"
LC Xref: aeinews sci.electronics.repair:10648

LC Heads up as per Jurb...hope this helps someone.

LC Just a few notes on these transistors and the results of some
LC experimenting. I had a KV27ts32 and a KV32s25 in the shop with the
LC classic HOT shorted and Q601,602 shorted. I had some 2SC4834M in stock
LC and ordered some plain 2SC4834 from Tritronics and some 2SC4834P from
LC Acme Enterprises (Orlando). I tried each type in both sets and found
LC that the higher rank (higher letter, higher gain) transistors ran
LC progressively cooler. The unranked versions ran almost 20 degrees F
LC hotter than the best, the P rank version. I matched the transistors
LC from batches of 10 by forward bias voltage on the junctions then in
LC circuit for nearest to 50% duty cycle. I have found that swapping
LC pairs or swapping the caps on Q601,602 can often improve the ratio to
LC closer to 50%.

.... Power is obtained by current meeting resistance

We are not talking about the horizontal output here. These are the
switching transistors in the power supply. While in some sets a higher gain
HOT will run a little cooler and more reliably than the original, that is
another discussion altogether.

The problem in the H drive circuit on these sets is nearly always bad solder
connections at the HD transformer causing HOT failures. Anyone servicing
these sets effectively always resolders the HDT and related parts.

Leonard Caillouet

"Asimov" wrote in message
...
"Leonard Caillouet" bravely wrote to "All" (07 Nov 03 22:36:30)
--- on the heady topic of "FYI 2SC4834 Sony power supply Q601,Q602"

There might be a design issue with the horizontal driver circuit if the
gain of the hot is that critical. Is there some field service notes
issued about this problem? It just seems a shame to be burning up
expensive transistors when all along it was another problem...

LC Reply-To: "Leonard Caillouet"
LC Xref: aeinews sci.electronics.repair:10648

LC Heads up as per Jurb...hope this helps someone.

LC Just a few notes on these transistors and the results of some
LC experimenting. I had a KV27ts32 and a KV32s25 in the shop with the
LC classic HOT shorted and Q601,602 shorted. I had some 2SC4834M in
stock
LC and ordered some plain 2SC4834 from Tritronics and some 2SC4834P from
LC Acme Enterprises (Orlando). I tried each type in both sets and found
LC that the higher rank (higher letter, higher gain) transistors ran
LC progressively cooler. The unranked versions ran almost 20 degrees F
LC hotter than the best, the P rank version. I matched the transistors
LC from batches of 10 by forward bias voltage on the junctions then in
LC circuit for nearest to 50% duty cycle. I have found that swapping
LC pairs or swapping the caps on Q601,602 can often improve the ratio to
LC closer to 50%.

... Power is obtained by current meeting resistance