I have a 1986 hybrid SP11 in for repair for the second time with yet
another
failing noisy j-fet within.
I guess there are approximately 30 fets, maybe 6 different types are
used, some are npn, some are pnp.
The boards, R&C components and tubes are all ok.

I faxed ARC with a description of the problem and their service guy
faxed me back a week later with a hand written note about the j-fets
used in an SP11,
saying they did have some, for usd$13.90 each, but not others,
and no information on the type numbers used.
ARC removed the type numbers from the fets used in this preamp, and the
type numbers
do not appear listed in the manual and schematic, which I have.

Does anyone in the group have a list of equivalent or better
j-fets and mosfets able to be used in the SP11?

Patrick Turner.

"Patrick Turner" wrote in message
...
I have a 1986 hybrid SP11 in for repair for the second time with yet
another
failing noisy j-fet within.
I guess there are approximately 30 fets, maybe 6 different types are
used, some are npn, some are pnp.
The boards, R&C components and tubes are all ok.

I faxed ARC with a description of the problem and their service guy
faxed me back a week later with a hand written note about the j-fets
used in an SP11,
saying they did have some, for usd$13.90 each, but not others,
and no information on the type numbers used.
ARC removed the type numbers from the fets used in this preamp, and the
type numbers
do not appear listed in the manual and schematic, which I have.

Does anyone in the group have a list of equivalent or better
j-fets and mosfets able to be used in the SP11?

Patrick Turner.


I've still got a Mk1 version of the SP11 here. Looking at my schematic
there are also no clues to the fets types apart from three 'lead out'
diagrams.
It does however say the replacement of some fets needs a trim resistor to
be matched to certain fets, though I guess the same will be on the MkII
schematic. If you do find out anything at all about the fets please let me
know. I didn't get mine upgraded to MkII as I compared both and actually
preferred the Mk1 IMO more natural sounding than the MkII.
Best of luck with the repair. Still a very nice & sought after preamp.

Mike

Mike Gilmour wrote:

"Patrick Turner" wrote in message
...
I have a 1986 hybrid SP11 in for repair for the second time with yet
another
failing noisy j-fet within.
I guess there are approximately 30 fets, maybe 6 different types are
used, some are npn, some are pnp.
The boards, R&C components and tubes are all ok.

I faxed ARC with a description of the problem and their service guy
faxed me back a week later with a hand written note about the j-fets
used in an SP11,
saying they did have some, for usd$13.90 each, but not others,
and no information on the type numbers used.
ARC removed the type numbers from the fets used in this preamp, and the
type numbers
do not appear listed in the manual and schematic, which I have.

Does anyone in the group have a list of equivalent or better
j-fets and mosfets able to be used in the SP11?

Patrick Turner.


I've still got a Mk1 version of the SP11 here. Looking at my schematic
there are also no clues to the fets types apart from three 'lead out'
diagrams.
It does however say the replacement of some fets needs a trim resistor to
be matched to certain fets, though I guess the same will be on the MkII
schematic. If you do find out anything at all about the fets please let me
know. I didn't get mine upgraded to MkII as I compared both and actually
preferred the Mk1 IMO more natural sounding than the MkII.
Best of luck with the repair. Still a very nice & sought after preamp.

Mike

A few weeks back one of the pnp source follower j-fets driving the cathode of
the
last gain stage in the phono amp became noisy and since there are a total of 4
gain stages
each with its own lot of NFB and with global NFB for RIAA, its hard to see that
the choice of fet is all that critical where it is.
A BC557 worked fine with no sound change or gain change or DC working change
whatsover.
The sound is determined by the tubes which do most of the gain, and the enormous
amount of
applied NFB.

I might say that the phono amp isn't as quiet as it could be because it isn't a
cascode input;
the very first gain block is an npn j-fet in common source mode, and these are
not much quieter than most tubes, and from what I can see the fet type isn't
something like 2SK147, 2SK369, or 2SK170, which may be a nice choice.

For quiet discrete component amp stages that are a shirtload more simple than
ARC,
go to http://www.vacuumstate.com where Allen Wright has some rather good
simple phono stages using the above listed j-fets in cascode with 6DJ8,
in SE and balanced schematics, with passive RIAA, and no
loops of global NFB.
I use something similar to AW's SE phono stage, and
the snr is at least 20dB better than a 12AX7 unweighted,
and the thd is 0.2% at 10 vrms output,
or if you like, about 0.01% at a volt.

I find that just ONE 2SK369 is all that is needed in a preamp to
allow direct connection of a low output MC.
All following stages are pure tube stages.
But with ARC, there are active j-fets and mosfets in every gain stage;
the amp is riddled with SS devices on which a lot depends; too much imho.

However, the use of the above 2SKXXX listed devices isn't possible in the SP11
except in the input gain stage of the phono amp and 3 in the line stages.
Most of the fets required are pnp types, and I have not searched for suitable
replacement,
and then there is the BS about fitting trimming resistors to suit each
j-fet and that's a PITA.

I myself could never ever be talked into buying an SP11,
on the grounds of its complexity, and that most samples around
may have the same failures as I have witnessed, and the work to
repair these things isn't made any easier by the maker's after sales service.
I suppose they think its ok to not to be too helpful with a model
that is 20 years old, but it did cost a lot when new 20 years ago........



But the next j-fet to fail in the SP11 was one of two which is used as a switch
to bypass the mode switch
in the line stage; this started tweeting for awhile after the amp came on.
A bjt will not do here, and some sort og fet is needed.
Since ARC didn't have the courtesy to tell me what the type numbers of their
fets are
so I could track down the characteristics, I asked my colleague if he knew of
some suitable
fets and there were some similar circuits used in a Mesa amp.
RS components stocked the npn type J112 and pnp type J176 which are commonly
used switching fets
which should replace the existing switching fets, at an average price of usd
$1.50,
so I ordered six of each.


The switching fets are extremely high impedance when biased off by a dc bias
voltage
which is actually switched, and very low impedance when their gate is biased on
by a medium value DC voltage.
In 1986 ARC must have concluded biased fets were more reliable than switch
contacts.
The MC impedance matching switch in the amp I have to repair is also
dodgy, and has become intermittent, so I have bypassed it and used
soldered in resistors. A new switch was quoted at usd $105, plus freight.
Maybe I will strip down the oldun to see if it can be cleaned/fixed, but
its a PITA to get it out of the amp.
Today much switching is done by switching fets arranged in a convenient location

to prevent long signal runs, and the wires to the switch only carry a small DC
voltage
so they can run anywhere without interaction with other signals.
Touch pads are used now instead of lever switches, and its cheaper.

Don't ask me if its sonically purer to run the signal through so many devices
as ARC does; the signal from the phono pick up to line stage output
can have 21 devices to get through.

The switching fets have one other advantage; the bias applied to the gate
is switched on and supplied through some high resistances, with capacitor
bypasses, which slows the switching process down just enough
to prevent ever getting a loud click one might get with a real
metal contact switch.

But I use metal switches, usually NOS wafer types made in the 1950s,
or the Farnell stock rotary wafer switches with the blue plastics and silver
plated
contacts which are the same excellent quality of units made
50 years ago and better than anything made in Taiwan.
Since I bias anything that is switched to stop DC offsets occuring I have
no unpleasant switch clicks.
No damn j-fets in the switching process to go wrong.

Patrick Turner.

Mark Harriss wrote:

I see on the AR website they test and grade
their transistors: possible there are only a few types
matched, graded for noise etc. Still without a
schematic to help indicate fet types it's impossible
to make an educated guess.

jfets are supposed to be notoriously different to each other.

But in a random lot of 2SK369 I tried in my phono stages
I got less than 1 dB gain difference between 10 samples
I bought at aud $1.10 each.
Gm at 5 mA is 40 mA/V, and as such it is great input device.

I may not have to replace more of the fets in this particular amp I have
to repair.
If that is the case, I may be lucky.
But If I do have to replace all the darn fets, I will certainly examine
what may be a suitable type.
There are also protection circuitry using diodes, more fets and
resistors to
prevent the signal fets used from being over voltaged, rather all too
easy
in a box full of tubes with a 300 volt B+ supply.
So not only does the type of fet used matter, but also the suitability
of the existing protection.
J-fets are notoriously fragile devices, often with a maximum safe supply

voltage of only 30v.

Patrick Turner.

"Patrick Turner" wrote in message
...


Mark Harriss wrote:

I see on the AR website they test and grade
their transistors: possible there are only a few types
matched, graded for noise etc. Still without a
schematic to help indicate fet types it's impossible
to make an educated guess.

jfets are supposed to be notoriously different to each other.

But in a random lot of 2SK369 I tried in my phono stages
I got less than 1 dB gain difference between 10 samples
I bought at aud $1.10 each.
Gm at 5 mA is 40 mA/V, and as such it is great input device.

I may not have to replace more of the fets in this particular amp I have
to repair.
If that is the case, I may be lucky.
But If I do have to replace all the darn fets, I will certainly examine
what may be a suitable type.
There are also protection circuitry using diodes, more fets and
resistors to
prevent the signal fets used from being over voltaged, rather all too
easy
in a box full of tubes with a 300 volt B+ supply.
So not only does the type of fet used matter, but also the suitability
of the existing protection.
J-fets are notoriously fragile devices, often with a maximum safe supply

voltage of only 30v.

Patrick Turner.




I'd heard problems with J-fets comes to the SP11 if a tube acts up. I put
those tubes through a tester every 6 months and change any that have
appreciably altered. Agree that ARC overcomplicate their circuit designs
but I've had nothing complain about to date apart from the original price
sticker shock that is, either with performance or reliability over about 20
years of use. I'm touching wood ;-)

Mike

I see on the AR website they test and grade
their transistors: possible there are only a few types
matched, graded for noise etc. Still without a
schematic to help indicate fet types it's impossible
to make an educated guess.