Update on the Fender ...

[Arfa Daily]

So, the nonsense bias board arrived. It was in a properly printed Fender
cardboard-topped 'accessory' bag, as you might find hanging on pegboard in
any music shop. "Genuine Fender Parts" it proudly announces. "PCB ASSY
AUTOMATIC BIAS 4 TUBES S - S"

I'm not sure what that tells us exactly, but if Fender have them all bagged
up ready to hang up in dealers ... ??

Anyways, put it in and fired it up. It does a lot of LED flashing and bias
trimming between the 4 tubes, but seems to eventually settle on a
'solution'. Out of the four (brand-new looking) Sovtek output tubes, three
ended up with almost -ve 51 volts on their grids, and the other with just
under -ve 52 volts. For all of this effort though, I'm not sure that there
would have been any practical difference if all four tubes were to have had
the same bias on them. I don't know if these Sovteks were bought as a
matched quad - we didn't fit them - but clearly they aren't, quite. I don't
know how much 'slack' this board is able to take up, but I would imagine
quite a bit.

But the thing that's vexing me a bit now, is whether to recommend that the
owner replaces all four tubes now for a properly matched quad. There's no
signs of any flashovers or other nasties going on, but I still have that
niggling doubt that something might have twatted that board, and that one of
the outputs clearly doesn't quite match the other three. Thing is, it's
already going to have cost them a bit in bench time, even if we let the
board go in for no cost, and to then add a set of output valves that it
looks like have already been replaced not long ago ... :-\

Arfa

[Gareth Magennis]

Anyways, put it in and fired it up. It does a lot of LED flashing and bias
trimming between the 4 tubes, but seems to eventually settle on a
'solution'. Out of the four (brand-new looking) Sovtek output tubes, three
ended up with almost -ve 51 volts on their grids, and the other with just
under -ve 52 volts. For all of this effort though, I'm not sure that there
would have been any practical difference if all four tubes were to have had
the same bias on them. I don't know if these Sovteks were bought as a
matched quad - we didn't fit them - but clearly they aren't, quite. I don't
know how much 'slack' this board is able to take up, but I would imagine
quite a bit.

But the thing that's vexing me a bit now, is whether to recommend that the
owner replaces all four tubes now for a properly matched quad. There's no
signs of any flashovers or other nasties going on, but I still have that
niggling doubt that something might have twatted that board, and that one of
the outputs clearly doesn't quite match the other three. Thing is, it's
already going to have cost them a bit in bench time, even if we let the
board go in for no cost, and to then add a set of output valves that it
looks like have already been replaced not long ago ... :-\

Arfa

Why would you need a properly matched quad in an amp that measures each
tube's bias point and matches it to the rest?
Surely that's just like having 4 separate bias pots onboard, something very
few amps have.
If one tube is down, swap its position and see if the amp compensates
accordingly. So either tube or amp problem.

Plus the bias current is actually measured and set per tube in the amp, at
the voltages that the tubes are actually running at.
Not some arbitrary set of parameters remotely measured in a tube tester that
the tube supplier claims to have carried out to "match" them.




Gareth.

[Arfa Daily]

"Gareth Magennis" <gareth....@ntlworld.com> wrote in message
news:Kxvaw.742333$3F6.1...@fx02.am4...

I hear what you are saying Gareth, and I take your point, but you are
missing the point that I am making. What is concerning me is that the
original board has failed. I cannot know whether this is a 'chance' failure,
or whether it's been knocked out by - for instance - an intermittent
internal short on one of the valves, bearing in mind that it is connected
directly to all four grids and all four cathodes. The fact that one of the
valves clearly has rather different characteristics from the other three,
just rang slight alarm bells in this regard, particularly as the valves all
look almost new, so were *probably* fitted as a matched quad, although
again, I cannot actually know this, as we didn't fit them.

You of course make a valid point that the whole purpose of this automatic
bias / balance board is so that it can match anything from 2 x 6L6 + 2 x
EL34, to a set of four 19 set PA valves, and it is of course doing its job
here in correcting for the odd man out - if that is in fact *all* that it is
...

Arfa

[David Platt]

In article <iIxaw.793388$Au1.4...@fx36.am4>,

Arfa Daily <arfa....@ntlworld.com> wrote:

>I hear what you are saying Gareth, and I take your point, but you are
>missing the point that I am making. What is concerning me is that the
>original board has failed. I cannot know whether this is a 'chance' failure,
>or whether it's been knocked out by - for instance - an intermittent
>internal short on one of the valves, bearing in mind that it is connected
>directly to all four grids and all four cathodes. The fact that one of the
>valves clearly has rather different characteristics from the other three,
>just rang slight alarm bells in this regard, particularly as the valves all
>look almost new, so were *probably* fitted as a matched quad, although
>again, I cannot actually know this, as we didn't fit them.
>
>You of course make a valid point that the whole purpose of this automatic
>bias / balance board is so that it can match anything from 2 x 6L6 + 2 x
>EL34, to a set of four 19 set PA valves, and it is of course doing its job
>here in correcting for the odd man out - if that is in fact *all* that it is
>...

As far as checking the Fender itself for hidden damage - I think the
"swap the tubes around and track the bias changes" idea is a good
one. With three rotations, you could test each tube in each socket
position. See if the bias for each individual tube "follows" the tube
pretty exactly, or if one or more sockets have a (forgive me) biased
bias. This would give you some idea as to whether anything in one of
the board sections was "cooked" by a tube failure (maybe a resistor
overheated and has drifted?)

As to the tubes themselves... best you can do is report your findings
to the owner, make your best recommendation, and ask whether the owner
wants you to install a fresh set that you've personally matched.

What does a good transconductance tester tell you about the tubes?

[N_Cook]

Useless bloody thunderbird, I remember reading the original of this
story but cannot now find it or could you at least repeat the model number.
What was the failure in the board, could you tell whether over-current
or over-voltage, any deliberate track thinnings for tell-tales.

[Gareth Magennis]

"Arfa Daily" <arfa....@ntlworld.com> wrote in message
news:iIxaw.793388$Au1.4...@fx36.am4...

Ah, sorry I did miss the point.

I would tend to assume that these were not a matched quad, as people who
match and sell valves generally like to make this known by marking the
valves with either a sticker or sharpie on the glass.
Although its more difficult with EL84's because they don't have a nice base
for the sticker and the glass is small to mark well in this manner.

I buy my valves 50 at a time direct from JJ Tesla in Slovakia. I've built a
tester with variable regulated HT and constant current sources to match the
valves at a known voltage and current point.
A quick check of the EL84 box shows about a 10% variation between lowest and
highest.
It looks like the readings are probably within a Gaussian distribution, or
whatever the correct Statistical term is involved here, i.e. the majority
are clumped in the middle with increasingly fewer being lower or higher than
the mid point.

Your readings seem to indicate around a 2% difference between the "bad one"
and the others. I'd say this might be pretty typical of a non matched set?

But who really knows, as you are saying!



Cheers,



Gareth.

[Phil Allison]

Arfa Daily wrote:

> So, the nonsense bias board arrived.

** The "nonsense" is all coming from you - fuckhead.

>
> Anyways, put it in and fired it up It does a lot of LED flashing and bias

> trimming between the 4 tubes, but seems to eventually settle on a
> 'solution'. Out of the four (brand-new looking) Sovtek output tubes, three
> ended up with almost -ve 51 volts on their grids, and the other with just
> under -ve 52 volts. For all of this effort though, I'm not sure that there
> would have been any practical difference if all four tubes were to have had
> the same bias on them. I don't know if these Sovteks were bought as a
> matched quad - we didn't fit them - but clearly they aren't, quite.

** One valve, 1 volt out is about as near to matched as any 4 output valves I have ever seen or ever need to be.

Jesus Christ you are a neurotic fuck.

> I don't
> know how much 'slack' this board is able to take up, but I would imagine
> quite a bit.

** The 4 bias currents are IDENTICAL !!

You stupid twat.

> owner replaces all four tubes now for a properly matched quad.

** FFS - there is nothing improper about the set you have.

There's no
> signs of any flashovers or other nasties going on, but I still have that
> niggling doubt that something might have twatted that board,

** Fender have already told you the boards fail all on their own.

> and that one of
> the outputs clearly doesn't quite match the other three.

** Twaddle.

> Thing is, it's
> already going to have cost them a bit in bench time, even if we let the
> board go in for no cost, and to then add a set of output valves that it
> looks like have already been replaced not long ago ... :-\

** If it ain't broke, don't fix it.

Only crooks do that.


... Phil

[Arfa Daily]

>
>
> Ah, sorry I did miss the point.
>
> I would tend to assume that these were not a matched quad, as people who
> match and sell valves generally like to make this known by marking the
> valves with either a sticker or sharpie on the glass.
> Although its more difficult with EL84's because they don't have a nice
> base for the sticker and the glass is small to mark well in this manner.
>
> I buy my valves 50 at a time direct from JJ Tesla in Slovakia. I've built
> a tester with variable regulated HT and constant current sources to match
> the valves at a known voltage and current point.
> A quick check of the EL84 box shows about a 10% variation between lowest
> and highest.
> It looks like the readings are probably within a Gaussian distribution, or
> whatever the correct Statistical term is involved here, i.e. the majority
> are clumped in the middle with increasingly fewer being lower or higher
> than the mid point.
>
> Your readings seem to indicate around a 2% difference between the "bad
> one" and the others. I'd say this might be pretty typical of a non
> matched set?
>
> But who really knows, as you are saying!
>
>
>
> Cheers,
>
>
>
> Gareth.
>

In reply to everyone, the bias did move with the valve - it was one of the
first things that I checked, so the board was 'doing its thing' and
balancing this valve into the other three. Information that I have since
found out today, indicates that one valve was changed on its own a few
months back, so I guess that's probably the one that is different, so I
think I'm going to add this new knowledge to what the man at Fender said
about them having had board problems, and that the new board was a
pre-packaged spare, and declare it "good to go" !

Arfa

[Arfa Daily]

"Phil Allison" <palli...@gmail.com> wrote in message
news:62358070-9217-4fe4...@googlegroups.com...

> Arfa Daily wrote:
>
>> So, the nonsense bias board arrived.
>
> ** The "nonsense" is all coming from you - fuckhead.

The board is a nonsense piece of design. It's not needed, and only adds
another layer of complexity into an item that is needed to be robust and
reliable. For most purposes in this type of equipment, fixed bias at the
simplest or single level-pot and sometimes a balance pot is all that is
required. Owners should not be arsing about changing valves in the first
place, and particularly should not be mixing types.

>
>>
>> Anyways, put it in and fired it up It does a lot of LED flashing and bias
>> trimming between the 4 tubes, but seems to eventually settle on a
>> 'solution'. Out of the four (brand-new looking) Sovtek output tubes,
>> three
>> ended up with almost -ve 51 volts on their grids, and the other with just
>> under -ve 52 volts. For all of this effort though, I'm not sure that
>> there
>> would have been any practical difference if all four tubes were to have
>> had
>> the same bias on them. I don't know if these Sovteks were bought as a
>> matched quad - we didn't fit them - but clearly they aren't, quite.
>
> ** One valve, 1 volt out is about as near to matched as any 4 output
> valves I have ever seen or ever need to be.

You miss the point as always by running your stupid mouth before engaging
that pathetic walnut of a brain. Fucking dope

>
> Jesus Christ you are a neurotic fuck.

Bwaaaahhhaaahhhaaaa !!!!! That from YOU ??!!! Neurotic ? Fuckhead ?
That would be YOU with bells on ...

>
>
>> I don't
>> know how much 'slack' this board is able to take up, but I would imagine
>> quite a bit.
>
> ** The 4 bias currents are IDENTICAL !!
>
> You stupid twat.

Philip. Mirror ...

>
>
>
>> owner replaces all four tubes now for a properly matched quad.
>
>
> ** FFS - there is nothing improper about the set you have.

I didn't say that there WAS. I was just examining the possibility that there
MIGHT be. But then you don't actually read and understand anything before
you run that dopey gob of yours, do you ?

>
>
> There's no
>> signs of any flashovers or other nasties going on, but I still have that
>> niggling doubt that something might have twatted that board,
>
>
> ** Fender have already told you the boards fail all on their own.

So they did. But that doesn't mean that they don't also fail as a result of
something else as well. There was no serial number identification offered by
Fender in regards of this bad batch. Just a statement that they had had a
bad batch "around that time" when told its age. Therefore, this one may or
may not have come from that batch. I can't say for sure, and YOU most
certainly can't ...

>
>
>
>> and that one of
>> the outputs clearly doesn't quite match the other three.
>
> ** Twaddle.

Bollocks. If the board arrives at a solution that sets the bias near as a
gnats cock identical on three of the valves, and sets the fourth a volt or
so different, then that 'odd one' clearly doesn't QUITE (look the word up if
you're not sure) match the other three

>
>
>> Thing is, it's
>> already going to have cost them a bit in bench time, even if we let the
>> board go in for no cost, and to then add a set of output valves that it
>> looks like have already been replaced not long ago ... :-\
>
> ** If it ain't broke, don't fix it.
>
> Only crooks do that.

And you would know, of course ... Now go take your meds, and shut the fuck
up. Twat.

Arfa
>
>
> ... Phil
>
>

[Arfa Daily]

>>
>> Arfa
>
> Useless bloody thunderbird, I remember reading the original of this story
> but cannot now find it or could you at least repeat the model number.
> What was the failure in the board, could you tell whether over-current or
> over-voltage, any deliberate track thinnings for tell-tales.
>

It's a Supersonic 100 head. The board failed to produce any 'controlled'
bias to the valves, and all four grids were at -89 volts. It looks as though
it works by measuring the voltage across four low value Rs in the cathodes
by using A-D inputs on the microcontroller on the board. The bias is then
produced by feeding PWM signals from four port pins to driver transistors
followed by filter caps. In the case of this board, the supply volts, reset
pin etc were all at correct levels, but the micro did not produce any
output. You could not even measure any clock at the xtal pins. No evidence
of any cause was visible. Board was pristine. Double sided surface mount,
and all examined under strong magnification, but no signs of any damaged
components.

Arfa

[Tom Miller]

"Arfa Daily" <arfa....@ntlworld.com> wrote in message

news:nZSaw.1124894$Fy6.1...@fx25.am4...

Any numbers on the cpu?

[Arfa Daily]

"Tom Miller" <tmille...@verizon.net> wrote in message
news:m4gvbm$ghe$1...@dont-email.me...

STM32F
101C6

This appears to agree with what it says on the schematic

Arfa

[Phil Allison]

Arfa Daily wrote:

> "Phil Allison" :

> >
> >> So, the nonsense bias board arrived.
> >
> > ** The "nonsense" is all coming from you - fuckhead.
>
>
> The board is a nonsense piece of design.

** No it is not.

> > ** One valve, 1 volt out is about as near to matched as any 4 output
> > valves I have ever seen or ever need to be.
>
>
> You miss the point

** There was no point get - you lunatic.

Jesus Christ you are a neurotic fuck.

> >

> > ** The 4 bias currents are IDENTICAL !!
> >
> > You stupid twat.
>
> Philip.

** You completely missed the point, as usual.

> > ** FFS - there is nothing improper about the set you have.
>
>
> I didn't say that there WAS.

** You did, three or four times.

> > ** Fender have already told you the boards fail all on their own.
>
>
> So they did.

** Believe them.

> But that doesn't mean that they don't also fail as a result of
> something else as well.

** Sheer, paranoid fear of the new.

If the design was that fragile, there would be constant PCB failure.

You will find there is a very high resistance between the output valve grids and the bias PCB circuitry and the cathode resistors are protected from over-current too.

What a dickhead.


... Phil

[Tom Miller]

"Arfa Daily" <arfa....@ntlworld.com> wrote in message

news:DvTaw.655114$J62.2...@fx03.am4...

Put this whole number in a search to get the datasheet.

STM32F101C6

It is an ARM Cortex 32 bit CPU with 32k bytes of flash. I don't know how
successful you would be copying the firmware over to a new chip..

[Arfa Daily]

"Phil Allison" <palli...@gmail.com> wrote in message

news:605af8cc-4dd2-4781...@googlegroups.com...

> Arfa Daily wrote:
>
>> "Phil Allison" :
>> >
>> >> So, the nonsense bias board arrived.
>> >
>> > ** The "nonsense" is all coming from you - fuckhead.
>>
>>
>> The board is a nonsense piece of design.
>
> ** No it is not.

So, would that be more or less so than your opinion of the HK board ?
And I quote from your post -

> "That H&K model even allows the user to change from EL34s to 6L6s.
> However, the system cannot cope with faulty valves and may even be damaged
> by them OR fail totally like you found. Also, it is sometimes possible to
> put an octal valve in the socket out of key or having broken the spigot
> install it any position you like. Results are often dramatic.
>
> IME having guitar players fool around putting valves in their amps is full
> of traps & hazards - owners are bound to fall into one sooner or later. "

>

>
>
>> > ** One valve, 1 volt out is about as near to matched as any 4 output
>> > valves I have ever seen or ever need to be.
>>
>>
>> You miss the point
>
> ** There was no point get - you lunatic.

Not in your wobble-eyed head ...

>
> Jesus Christ you are a neurotic fuck.
>
>> >
>> > ** The 4 bias currents are IDENTICAL !!
>> >
>> > You stupid twat.
>>
>> Philip.
>
> ** You completely missed the point, as usual.
>
>
>
>> > ** FFS - there is nothing improper about the set you have.
>>
>>
>> I didn't say that there WAS.
>
> ** You did, three or four times.
>

Err, no, I didn't ...

>
>
>> > ** Fender have already told you the boards fail all on their own.
>>
>>
>> So they did.
>
> ** Believe them.


>
>> But that doesn't mean that they don't also fail as a result of
>> something else as well.

As I said ...

>
> ** Sheer, paranoid fear of the new.

Oh fuck off back under your stone you obnoxious arsehole. I've been in this
game for 45 years. Nothing has come along in that time to make me paranoid,
and this isn't going to be the first

>
> If the design was that fragile, there would be constant PCB failure.
>
> You will find there is a very high resistance between the output valve
> grids and the bias PCB circuitry and the cathode resistors are protected
> from over-current too.

There is a couple of hundred k. Hardly a "very high resistance".


>
> What a dickhead.

You know, you really are a sad fool, Philip. Just about any thread that
anyone starts that runs for longer than a couple of days, sooner or later,
there you are running your dumb - and usually foul - gob off. The trouble
with you is that you have this pathetic need to try to sound superior to
everyone, when in truth, you are often at least misguided, if not downright
wrong. Just because something is your opinion, it doesn't necessarily make
you right and everyone else wrong. Your main failing is that you do not read
things properly, and then go off on one of your ill-conceived rants, based
on what you *thought* you read. When your error is pointed out, you then
start madly snipping in an attempt to cover up what *you* said in reply to
what the poster *actually* said, and then start leveling a tirade of abuse ,
peppered with expletives, capital letters, exclamation marks and asterisks,
at anyone who dares to challenge you. Some of the more 'delicate' posters
cave in at this, but I don't know why you keep trying it with me. You've
tried this crap with me on many occasions, and I would have thought that you
would have worked out by now that I am not intimidated by you. But then
again, maybe your brain is either so faulty, so drug-rotted, or so
alcohol-soaked that you can't remember from one day to the next who you have
already unsuccessfully had a go at.

Go get some help. Seriously.

And, unless anyone else has got any sensible comments or questions, I'm done
with this now. The amp is fixed. I'm reasonably happy that there are no
further issues, and it will be returned to its owner today.

Arfa

>
>
> ... Phil
>
>
>
>
>
>

[Phil Allison]

Arfa Daily wrote:

> >> >
> >> >> So, the nonsense bias board arrived.
> >> >
> >> > ** The "nonsense" is all coming from you - fuckhead.
> >>
> >>
> >> The board is a nonsense piece of design.
> >
> > ** No it is not.

>
> So, would that be more or less so than your opinion of the HK board

** The HK auto bias works fine and is very useful - as I explained in the earlier thread.
All water off a fuckwit's back to you, I see.

> >
> >> > ** FFS - there is nothing improper about the set you have.
> >>
> >>
> >> I didn't say that there WAS.
> >
> > ** You did, three or four times.
> >
>
> Err, no, I didn't ...

** A damn liar and a fuckwit.

Typical pommy troll.

> > ** Sheer, paranoid fear of the new.
>
> Oh fuck off back under your stone you obnoxious arsehole. I've been in this
> game for 45 years. Nothing has come along in that time to make me paranoid,
> and this isn't going to be the first

** Stupidity and ASD are all you need to become paranoid.

> >
> > If the design was that fragile, there would be constant PCB failure.
> >
> > You will find there is a very high resistance between the output valve
> > grids and the bias PCB circuitry and the cathode resistors are protected
> > from over-current too.
>
> There is a couple of hundred k. Hardly a "very high resistance".

** How many mA does that limit the worst case current to ?

What a fuckwit, pommy dickhead.

Not even honest enough to use his real name.




.... Phil

[Arfa Daily]

"Phil Allison" <palli...@gmail.com> wrote in message

news:b39f33cd-156a-427c...@googlegroups.com...

snip snip rant rant. Ever the predictable sheep-shagging loony-toon ...

Arfa

[josephkk]

Believable. I wonder if the crystal itself died. Might check that
failure mode out if you are interested.

?-)

[Arfa Daily]

"josephkk" <joseph_...@sbcglobal.net> wrote in message
news:tpdt6a1gotjmitoef...@4ax.com...

I did consider the xtal itself being at fault, but as Fender were sending a
new board out to us, I couldn't be bothered at the time to check. I still
have the board, so might stick some power on it and see if I can prove
anything.

Arfa

[Jeroni Paul]

I also bet the crystal or its associated caps (if any). Another bet would be some residual resine or PCB protector turning conductive with age.