I am working on a power board with MPSA06 and 56 transistors that handle
48V. The TO-92A footprint in my PCB software has the pins in line and with
about 0.2mm clearance. It's a bit of a pain, due to the rest of the boards
clearance being >0.5mm easily, so it adds expense when I get them made.
This is no big deal, but when I calculate the rough clearance necessary I
get about 0.2mm minimum, which is a bit close for comfort for my liking with
the 48V difference between pins. I expect the boards to deal with
condensation etc, and need to last for many years.
I was thinking of changing the PCB pattern to put the middle pin back a bit,
like the "bent pin" (can't remember the proper name) versions. Then it will
keep it all at least 0.5mm apart.
What does anyone else think? Am I fussing over nothing?
I suppose I'm just interested in whether anyone else may have the same
concerns. The MPSA06 is rated for up to 80V and the clearance between pins
can be as low as 0.9mm. When you add the annular ring of pads, it does not
leave a lot of clearance.
> I am working on a power board with MPSA06 and 56 transistors that handle
> 48V. The TO-92A footprint in my PCB software has the pins in line and with
> about 0.2mm clearance. It's a bit of a pain, due to the rest of the boards
> clearance being >0.5mm easily, so it adds expense when I get them made.
> This is no big deal, but when I calculate the rough clearance necessary I
> get about 0.2mm minimum, which is a bit close for comfort for my liking
> with
> the 48V difference between pins. I expect the boards to deal with
> condensation etc, and need to last for many years.
> I was thinking of changing the PCB pattern to put the middle pin back a
> bit,
> like the "bent pin" (can't remember the proper name) versions. Then it
> will
> keep it all at least 0.5mm apart.
> What does anyone else think? Am I fussing over nothing?
>
> I suppose I'm just interested in whether anyone else may have the same
> concerns. The MPSA06 is rated for up to 80V and the clearance between pins
> can be as low as 0.9mm. When you add the annular ring of pads, it does not
> leave a lot of clearance.
Okay, to reply to my own post - I have added a large pitch alternative
pattern in the library. I increased the (thinnest part of)annular ring from
0.05mm (!! a lot of PCB houses don't do this low AFAIK) to 0.15mm and
increased the pin spacing from 1.2mm to 1.5mm, giving a 0.5mm clearance. The
maximum pin spacing according to the datasheet is 1.4mm, so I don't think
the fit will be too bad, and I don't intend to have the transistors sitting
on the PCB anyway.
Any comments on all this would be welcomed.
On 18/07/2010 20:24, Oli Glaser wrote:
>
>> I was thinking of changing the PCB pattern to put the middle pin back a
>> bit,
>> like the "bent pin" (can't remember the proper name) versions. Then it
>> will
>> keep it all at least 0.5mm apart.
>> What does anyone else think? Am I fussing over nothing?
triangle layout better and more clearance.
> On 18/07/2010 20:24, Oli Glaser wrote:
>>
>>> I was thinking of changing the PCB pattern to put the middle pin back a
>>> bit,
>>> like the "bent pin" (can't remember the proper name) versions. Then it
>>> will
>>> keep it all at least 0.5mm apart.
>>> What does anyone else think? Am I fussing over nothing?
> triangle layout better and more clearance.
That's what I meant when I said move the middle pin back, to make a triangle
layout (unless you are meaning something different)
However, for now I just moved the pins apart a little (as stated in my
second mail) and it looks reasonable I think, shouldn't place too much
stress on the pins.
I'm more interested in thoughts of sub 0.2mm clearance for ~50V differences
for long term problem free operation.
Oli Glaser wrote:
> The TO-92A footprint in my PCB software has the pins in
> line and with about 0.2mm clearance.
OK, so don't do that.
First, TO-92 is silly in a new design. There are far more SOT23, SOT89, etc
transistors available because there is little demand for the archaic TO92.
You might as well use a 12AU7A, since its pins are spaced farther apart.
Second, you don't have to make the footprint match exactly how the leads
come out of the case of a TO92, and in fact most aren't done that way.
Often you will see them in the common triangle configuration, whether the
transistor has them that way or in line. It's normal to spread the leads a
little when installing TO92s.
> It's a bit of a pain, due to the
> rest of the boards clearance being >0.5mm easily, so it adds expense
> when I get them made.
If you're worried about the extra expense, what the heck are you using
ancient thru hole parts for!!?
> Oli Glaser wrote:
>> The TO-92A footprint in my PCB software has the pins in
>> line and with about 0.2mm clearance.
>
> OK, so don't do that.
>
> First, TO-92 is silly in a new design. There are far more SOT23, SOT89,
> etc
> transistors available because there is little demand for the archaic TO92.
> You might as well use a 12AU7A, since its pins are spaced farther apart.
>
> Second, you don't have to make the footprint match exactly how the leads
> come out of the case of a TO92, and in fact most aren't done that way.
> Often you will see them in the common triangle configuration, whether the
> transistor has them that way or in line. It's normal to spread the leads
> a
> little when installing TO92s.
>
>> It's a bit of a pain, due to the
>> rest of the boards clearance being >0.5mm easily, so it adds expense
>> when I get them made.
>
> If you're worried about the extra expense, what the heck are you using
> ancient thru hole parts for!!?
Yes it would be silly in a new design but I'm using *ancient* thru hole
parts because I am making a repro version of an *ancient* 70's preamp, with
all discrete through hole components, so maybe you can let me off that one..
:-)
I'm not really worried about the extra expense, just noting the fact that my
PCB house has a step up in price when clearances get below 0.3mm and for
smaller annular rings. Yes you're right though it would be a little out of
proportion to worry about an extra £1 per board in these
circumstances(although it all adds up in the end).
I know a lot of PCB design software does footprints slightly differently.
With Diptrace the TO-92 pattern was a triangle but with Altium it's like
this.
No big deal as I have altered it - I was mainly interested in the creepage
and if anyone has experience of failure due to inadequate clearance.
I generally try and over engineer stuff as much as possible as allowed by
budget/time etc. Seems to me logical to if you have the board space etc, as
it costs nothing (stops me worrying too, as I tend to a little too much
maybe about such matters)
>--------------------------------------------------
>From: "Olin Lathrop" <@spam@olin_piclistKILLspamembedinc.com>
>Sent: Sunday, July 18, 2010 11:21 PM
>To: "Microcontroller discussion list - Public." <KILLspampiclistKILLspammit.edu>
>Subject: Re: [EE] Creepage clearance
>
> > Oli Glaser wrote:
> >> The TO-92A footprint in my PCB software has the pins in
> >> line and with about 0.2mm clearance.
> >
> > OK, so don't do that.
> >
> > First, TO-92 is silly in a new design. There are far more SOT23, SOT89,
> > etc
> > transistors available because there is little demand for the archaic TO92.
> > You might as well use a 12AU7A, since its pins are spaced farther apart.
> >
> > Second, you don't have to make the footprint match exactly how the leads
> > come out of the case of a TO92, and in fact most aren't done that way.
> > Often you will see them in the common triangle configuration, whether the
> > transistor has them that way or in line. It's normal to spread the leads
> > a
> > little when installing TO92s.
> >
> >> It's a bit of a pain, due to the
> >> rest of the boards clearance being >0.5mm easily, so it adds expense
> >> when I get them made.
> >
> > If you're worried about the extra expense, what the heck are you using
> > ancient thru hole parts for!!?
>
>Yes it would be silly in a new design but I'm using *ancient* thru hole
>parts because I am making a repro version of an *ancient* 70's preamp, with
>all discrete through hole components, so maybe you can let me off that one..
>:-)
>
>I'm not really worried about the extra expense, just noting the fact that my
>PCB house has a step up in price when clearances get below 0.3mm and for
>smaller annular rings. Yes you're right though it would be a little out of
>proportion to worry about an extra £1 per board in these
>circumstances(although it all adds up in the end).
>
>I know a lot of PCB design software does footprints slightly differently.
>With Diptrace the TO-92 pattern was a triangle but with Altium it's like
>this.
Altium has a bunch of different TO-92 footprints, some are as you describe,
some are spaced out more, some are in a triangle. I don't see the one have
used most (a triangle with the center lead out in front). I don't like the
spread-pins-in-a-line variant so much because vibration in the wrong direction
can easily fatigue the leads. Check the TO-92W footprint and the TO-92_variant
footprint.
Anyway, should take you almost no time** to make a new footprint _exactly_ the
way you want it. For example, for easy hand assembly, I prefer relatively big holes,
depending on whether the transistor is pre-formed or not.
** admittedly more if you want to create a new solid model of the case + leads
and tie the 3D body in). Attached is representation of a similar item-- an image
of a "test point" 3D model that I created recently. The footprint is just one
through-hole pad, but getting the swept path of the flattened wire in the 3D model
to look right took a bit of time.
> Altium has a bunch of different TO-92 footprints, some are as you
> describe,
> some are spaced out more, some are in a triangle. I don't see the one have
> used most (a triangle with the center lead out in front). I don't like the
> spread-pins-in-a-line variant so much because vibration in the wrong
> direction
> can easily fatigue the leads. Check the TO-92W footprint and the
> TO-92_variant
> footprint.
Oh dear, my brain must be really fried today, must try to get some more
sleep. Not been using Altium long, but why I didn't think of doing a search
on TO-92 footprints using the convenient "search" tool, I'll never know.
Just did one and there they all were, as you say :-) The variant version is
what I was looking for, and the W version is pretty much the same as the
custom one I made earlier (for no reason apparently :-) )
I was looking under the footprints for the MPSA06 model which only has the
in line version. It would be good if different versions of the footprint
were included for choice - I think I will recompile the library with more
choices (W and variant) added as I have been doing for other stuff.
> Anyway, should take you almost no time** to make a new footprint _exactly_
> the
> way you want it. For example, for easy hand
> assembly, I prefer relatively big holes,
> depending on whether the transistor is pre-formed or not.
Yes, making the assembly as easy as possible is another reason for wanting
reasonable clearance. Bridges are less likely and more visible when they do
occur.
> ** admittedly more if you want to create a new solid model of the case +
> leads
> and tie the 3D body in). Attached is
> representation of a similar item-- an image
> of a "test point" 3D model that I created recently. The footprint is just
> one
> through-hole pad, but getting the swept path of
> the flattened wire in the 3D model
> to look right took a bit of time.
Looks like a nicely done model - I like the 3D models but as you say, they
do take more time to create/position. Some component manufacturers provide
STEP models, which makes life easier though.
Oli Glaser wrote:
> Yes it would be silly in a new design but I'm using *ancient* thru
> hole parts because I am making a repro version of an *ancient* 70's
> preamp, with all discrete through hole components, so maybe you can
> let me off that one.. :-)
I still don't get it. You are here asking about footprints, so obviously
you are in a position to redesign the board. If this is really supposed to
be a 1970s preamp, maybe you should look more seriously at the 12AU7-A I
mentioned instead of them newfangled wussy transistor thingies.
Is this amp supposed to look like a 1970s amp on the inside, be a similar
circuit, or act like one on the outside?
>I still don't get it. You are here asking about footprints, so obviously
>you are in a position to redesign the board.
Not all products are suited for high-volume SMT production. A small
manufacturer will often choose to stay with through-hole components
simply because they are easily stuffed into PCBs by hand.
I suspect that the original OP plans to sell these things in
singles. There is NO need to design the board for SMT devices in that case.
Also consider that he has a significant number of through-hole
components that can't be SMT. These include the connectors, switches, pots.
Its worth the effort to design for mixed through-hole and SMT devices
when you are manufacturing thousands or more of something.
In this case, however, going to SMT will most likely mean that the
boards cost MORE to produce than staying all through-hole.
FWIW: I spent a significant portion of my previous life doing work on
Neve (and others) consoles (mid-'70s through late-'80s). Well-built
consoles have a VERY long lifetime - some of those old consoles are
still in use today.
They just sound "great". I can't quantify WHY they sound so good but
the difference is obvious when you are working with original source
material (live musicians making wonderful sounds in the
studio). That's really the only way that you can evaluate such
things - everything starts going downhill (quality-wise) after the
source material hits the recording medium.
Those old consoles are also EASY to repair and maintain. All
through-hole parts means that you can read component numbers or
values - and you can still get those parts as you need them.
I'm glad that someone is trying to follow in those footsteps and am
looking forward to hearing a unit sometime.
Dwayne Reid wrote:
> Also consider that he has a significant number of through-hole
> components that can't be SMT. These include the connectors,
> switches, pots.
>
> Its worth the effort to design for mixed through-hole and SMT devices
> when you are manufacturing thousands or more of something.
But it's also worth it for singles. One advantage to SMD is that parts can
be placed by machine. For small volumes like only 10s at a time, it doesn't
make sense to use pick and place machines. However, even if all the parts
will be populated manually, SMD is still cheaper since it takes less time
per part.
> In this case, however, going to SMT will most likely mean that the
> boards cost MORE to produce than staying all through-hole.
Quite unlikely. Ask a local manufacturer what they prefer, even for manual
work.
On Mon, 19 Jul 2010 11:58:05 -0400, "Olin Lathrop" said:
> Dwayne Reid wrote:
> > Also consider that he has a significant number of through-hole
> > components that can't be SMT. These include the connectors,
> > switches, pots.
> >
> > Its worth the effort to design for mixed through-hole and SMT devices
> > when you are manufacturing thousands or more of something.
>
> But it's also worth it for singles. One advantage to SMD is that parts
> can
> be placed by machine.
I think the OP said that this product was a replica and being somewhat
of a "fashion item" his sales would be hurt if he used SMT components.
At 09:58 AM 7/19/2010, Olin Lathrop wrote:
>For small volumes like only 10s at a time, it doesn't make sense to
>use pick and place machines. However, even if all the parts will be
>populated manually, SMD is still cheaper since it takes less time per part.
I have to disagree with you here.
We are a small manufacturer who does primarily through-hole
production. Our SMT production is perhaps 2 or 3 percent of the total.
I can flat out guarantee that placing through-hole parts into boards
costs us much less than placing SMT parts. Part of that is component
prep - we have automated machinery that cuts and bends component
leads as necessary. It takes only a few minutes to turn a full reel
of 5000 resistors into parts that are cut to the right length and
bent - ready to place into the board.
The boards are all stuffed, then run through a wave-solder
machine. Fast, easy, reliable.
Contrast that with SMT devices. First, you have to get the parts off
the tape into bins - we use circular aluminum containers from Lee
Valley Tools. This isn't particularly time-consuming - but it does
take longer than running that full reel of resistors through the
lead-forming machine.
Then you have to grab each SMT device (we use tweezers) and place it
on the board. Again, we've timed it. It takes longer to do that
with SMT than with through-hole parts.
What I haven't talked about is how you solder the SMT devices. We
currently don't use a reflow oven, so what we do is to flood the bare
board with Kester AZ2331 water-soluble flux. As we place each
component, we touch one end with the soldering iron. The solder
plating on the bare board grabs onto the component until it can be
properly soldered.
Admittedly, this is least efficient method possible. Much better
would be a solder-paste stencil and reflow oven. I hope to be using
that shortly.
The way I see it, the only way for my company to be cost-effective
with SMT is to go whole-hog and invest in a full SMT production line:
stencil printer, pick-and-place, conveyor-belt reflow oven. That is
in the plans - but not until we move to a larger building. We just
don't have the space.
Now back to Oli's situation. He is a very low-volume manufacturer -
much smaller than us, for example. He is going to most likely build
his boards in batches of 10 pieces or so, at least to start.
Through-hole is EXTREMELY cost effective in that situation.
> However, for now I just moved the pins apart a little (as
> stated in my second mail) and it looks reasonable I think,
> shouldn't place too much stress on the pins. I'm more
> interested in thoughts of sub 0.2mm clearance for ~50V
> differences for long term problem free operation.
Consider turning the transistor up down and bending the middle leg one
side and other legs another side. You'll have a lot of clearance and
you'd glue the transistor to PCB, - better would be not to allow
vibrating anything on the pre-amp.
> Is this amp supposed to look like a 1970s amp on the inside, be a similar
> circuit, or act like one on the outside?
All three to a certain extent. Most of the circuit is set in stone for this
version.
Bob has a good point in that using SMD stuff would be out of place and
possibly put people off(whether there is any logic in that is irrelevant if
they are not going to buy it)
As does Dwayne - the production volume will be very small to start
with(10s), and probably won't speed up *too* much anyway, for various
reasons (cost, competition, market) Also as mentioned a lot of the old
through hole modules lasted well and were repairable easily. The aim here is
to build it like a tank, and make use of space rather than try to keep
everything as compact as possible.
Speed of production is certainly not an issue here, as it would be a nice
problem to have if they were being sold as fast as we could make them :-)
Compared to other projects I'm involved in that use SMD stuff exclusively
(and production of will probably be done elsewhere), this is an entirely
different beast so I thought it should be approached in a different manner.
On Mon, Jul 19, 2010 at 1:24 PM, Oli Glaser <RemoveMEoli.glaserspam_OUTKILLspamtalktalk.net> wrote:
>
>
> --------------------------------------------------
> From: "Olin Lathrop" <RemoveMEolin_piclistTakeThisOuTspamembedinc.com>
> Sent: Monday, July 19, 2010 1:14 PM
> To: "Microcontroller discussion list - Public." <EraseMEpiclistspamspamBeGonemit.edu>
> Subject: Re: [EE] Creepage clearance
>
>> Is this amp supposed to look like a 1970s amp on the inside, be a similar
>> circuit, or act like one on the outside?
>
> All three to a certain extent. Most of the circuit is set in stone for this
> version.
>
I hope you are doing the PCB layout with tape on mylar film.
>On Mon, Jul 19, 2010 at 1:24 PM, Oli Glaser <oli.glaserSTOPspamspam_OUTtalktalk.net> wrote:
> >
> >
> > --------------------------------------------------
> > From: "Olin Lathrop" <spamBeGoneolin_piclistSTOPspamEraseMEembedinc.com>
> > Sent: Monday, July 19, 2010 1:14 PM
> > To: "Microcontroller discussion list - Public." <KILLspampiclistspamBeGonemit.edu>
> > Subject: Re: [EE] Creepage clearance
> >
> >> Is this amp supposed to look like a 1970s amp on the inside, be a similar
> >> circuit, or act like one on the outside?
> >
> > All three to a certain extent. Most of the circuit is set in stone for this
> > version.
> >
>
>I hope you are doing the PCB layout with tape on mylar film.
>
>Regards,
>Mark
>markrages@gmail
Once when I visited one of my local PCB makers (almost 10 years ago)
I asked about
the digitizer they used to use to program hole locations. He said they
hadn't had to use it in years (and it was pretty dusty looking and sitting all
lonely in a corner). I bet few places even have the ability to deal
with tape layouts/
film anymore. My (former) local film service bureau doesn't do any
silver based
work at all.
OTOH, to use too many TLAs and FLAs, IIRC, MOMA in NYC has a
single-sided PCB hanging
on a gallery wall.
Oli Glaser wrote:
>> Is this amp supposed to look like a 1970s amp on the inside, be a
>> similar circuit, or act like one on the outside?
>
> All three to a certain extent. Most of the circuit is set in stone
> for this version.
That seems a bit strange to me then. The 1970s was when transistors were
replacing tubes in many new designs. Real audiophools insisted on tubes
because they supposedly sounded better. This was probably helped along by
early transistor amps often not being designed with enough headroom.
Transistor amps were the cheap ones more people could afford. Tube amps
were the phancy audiophool ones.
You are clearly selling to a nostalgic crowd, but I'm surprised anyone would
be nostalgic about transistor amps from the 1970s. I can see some folks
being nostalgic about tube amps. Maybe there's room for a few 12AU7-A and
12AX7 yet, even if they don't actually do anything but glow convincingly.
> On Mon, Jul 19, 2010 at 1:24 PM, Oli Glaser <.....oli.glaserspam_OUTtalktalk.net>
> wrote:
>>
>>
>> --------------------------------------------------
>> From: "Olin Lathrop" <TakeThisOuTolin_piclist.....TakeThisOuTembedinc.com>
>> Sent: Monday, July 19, 2010 1:14 PM
>> To: "Microcontroller discussion list - Public." <TakeThisOuTpiclistKILLspamspammit.edu>
>> Subject: Re: [EE] Creepage clearance
>>
>>> Is this amp supposed to look like a 1970s amp on the inside, be a
>>> similar
>>> circuit, or act like one on the outside?
>>
>> All three to a certain extent. Most of the circuit is set in stone for
>> this
>> version.
>>
>
> I hope you are doing the PCB layout with tape on mylar film.
I was waiting for that reply from Olin :-)
That's why I said "to a certain extent", the PCB is our design - a 4 layer
board with ground and power planes.
We were never trying to recreate stuff exactly down to the the finest
detail, just provide a reasonably priced, high quality, solid preamp based
on a very popular circuit.
Oli Glaser wrote:
>> I hope you are doing the PCB layout with tape on mylar film.
>
> I was waiting for that reply from Olin :-)
When I saw Mark's comment, I was going to add that you should also be doing
the schematic on clay tablets with chisels, or at least on D size vellum
with pencil and eraser, but decided not to.
> Oli Glaser wrote:
>>> Is this amp supposed to look like a 1970s amp on the inside, be a
>>> similar circuit, or act like one on the outside?
>>
>> All three to a certain extent. Most of the circuit is set in stone
>> for this version.
>
> That seems a bit strange to me then. The 1970s was when transistors were
> replacing tubes in many new designs. Real audiophools insisted on tubes
> because they supposedly sounded better. This was probably helped along by
> early transistor amps often not being designed with enough headroom.
> Transistor amps were the cheap ones more people could afford. Tube amps
> were the phancy audiophool ones.
>
> You are clearly selling to a nostalgic crowd, but I'm surprised anyone
> would
> be nostalgic about transistor amps from the 1970s. I can see some folks
> being nostalgic about tube amps. Maybe there's room for a few 12AU7-A and
> 12AX7 yet, even if they don't actually do anything but glow convincingly.
I had similar thoughts quite some time ago, when starting to use this kind
of stuff as a musician. There is certainly a lot of snake oil sold in the
recording business, and a lot of the nostalgia is certainly not based on any
sound logic in my experience.
However with these particular amps, as Dwayne said, the sound has a real
distinct quality and they just simply "sound great".
The input/output transformers probably add a lot more colour to the sound
than the transistors also.
A lot of stuff in the recording business is not designed for low distortion
and transparency as with Hi-fi stuff, hence the reason tube amps and preamps
like this are still popular, people like the colour and warmth they add. I
have always used valve amps for my guitar, and I still use a reel to reel in
my studio a lot of the time...
People use all sorts of weird and wonderful equipment people use to create
certain styles of sound.
> Oli Glaser wrote:
>>> I hope you are doing the PCB layout with tape on mylar film.
>>
>> I was waiting for that reply from Olin :-)
>
> When I saw Mark's comment, I was going to add that you should also be
> doing
> the schematic on clay tablets with chisels, or at least on D size vellum
> with pencil and eraser, but decided not to.
:-) I like the chisel idea, maybe that's how the Romans made their PCBs...
>That seems a bit strange to me then. The 1970s was when transistors were
>replacing tubes in many new designs.
I suspect the preamp design Oli is using dates from the late '70s or
early '80s.
Rupert Neve was a real pioneer in designing audio recording
consoles. This is not your average consumer electronics - this was
state-of-the-art electronics designed to record live audio as
faithfully as possible. He succeeded <grin>.
Quite simply, his consoles sounded better than anything else
available at the time. This is genuine A-B testing: listen to what
is coming out of the speakers in the control room, then walk into the
studio where the musicians are playing and listen to the original.
You paid a hefty price for that quality. This is NOT something that
average consumer could ever hope to afford. I recall that one studio
I was doing work for upgraded their older Neve console to the latest
and greatest - the trade-in price was something like a quarter
million bucks (in early '80s dollars). Serious coin.
I started working with recording and radio-broadcast consoles in the
mid '70s. Some of the stuff I had to repair just didn't sound as
good as other stuff. That's just the way it was: some stuff was OK,
some stuff was good, some stuff was great. Neve was the best I ever
worked with.
To this day, I was never able to explain why a Neve console sounded
better than something like a McCurdy or Gates or Western Electric
console. From a measurement perspective, they all tested about the
same (VERY good) in terms of noise floor, THD, IMD, frequency
response. But some consoles just sounded 'better'.
Some time after that, rock and roll consoles (Soundcraft, etc) became
available. Some of those sounded pretty good - but never as good as the Neve.
I guess the best way to put it is: if you haven't actually worked
with that level of equipment, it can be hard to understand why people
are so passionate about the gear they use. True - there is a LOT of
snake-oil out there. An awful lot. But some gear is genuinely good.
Did you ever do a fully blind test? (Have a number of amplifiers which
were fed to you at random and see if you could pick out the Neve just
based on the sound)?
> At 02:45 PM 7/19/2010, Olin Lathrop wrote:
>
>>That seems a bit strange to me then. The 1970s was when transistors were
>>replacing tubes in many new designs.
>
> I suspect the preamp design Oli is using dates from the late '70s or
> early '80s.
>
> Rupert Neve was a real pioneer in designing audio recording
> consoles. This is not your average consumer electronics - this was
> state-of-the-art electronics designed to record live audio as
> faithfully as possible. He succeeded <grin>.
>
> Quite simply, his consoles sounded better than anything else
> available at the time. This is genuine A-B testing: listen to what
> is coming out of the speakers in the control room, then walk into the
> studio where the musicians are playing and listen to the original.
>
> You paid a hefty price for that quality. This is NOT something that
> average consumer could ever hope to afford. I recall that one studio
> I was doing work for upgraded their older Neve console to the latest
> and greatest - the trade-in price was something like a quarter
> million bucks (in early '80s dollars). Serious coin.
>
> I started working with recording and radio-broadcast consoles in the
> mid '70s. Some of the stuff I had to repair just didn't sound as
> good as other stuff. That's just the way it was: some stuff was OK,
> some stuff was good, some stuff was great. Neve was the best I ever
> worked with.
>
> To this day, I was never able to explain why a Neve console sounded
> better than something like a McCurdy or Gates or Western Electric
> console. From a measurement perspective, they all tested about the
> same (VERY good) in terms of noise floor, THD, IMD, frequency
> response. But some consoles just sounded 'better'.
>
> Some time after that, rock and roll consoles (Soundcraft, etc) became
> available. Some of those sounded pretty good - but never as good as the Neve.
>
> I guess the best way to put it is: if you haven't actually worked
> with that level of equipment, it can be hard to understand why people
> are so passionate about the gear they use. True - there is a LOT of
> snake-oil out there. An awful lot. But some gear is genuinely good.
>
> dwayne
>
> --
> Dwayne Reid <@spam@dwaynerRemoveMEEraseMEplanet.eon.net>
> Trinity Electronics Systems Ltd Edmonton, AB, CANADA
> (780) 489-3199 voice (780) 487-6397 fax
> http://www.trinity-electronics.com
> Custom Electronics Design and Manufacturing
>
At 10:37 PM 7/19/2010, Sean Breheny wrote:
>Did you ever do a fully blind test? (Have a number of amplifiers which
>were fed to you at random and see if you could pick out the Neve just
>based on the sound)?
Not with the Neve consoles and not in that era. Generally speaking,
a recording studio has ONE console. Recall the price tag I mentioned earlier.
Obviously, large facilities in certain US cities have multiple
recording studios in one building. But the studios I did work for
had only one studio in the building - and one audio console.
I also think that you may be missing what I said: I was comparing the
live sound that I would hear in the studio where the musicians are
playing, then go into the control room and listen to what was coming
out of the speakers. We had some truly talented recording engineers
in our city and some of the two-mixes I heard were simply stunning.
Live audio in the recording studio is as good as it gets. Its all
downhill after that - no matter how much time I spent aligning the
tape decks (Ampex, then Otari 24 track machines in the studio that I
spent the most time in), what came off of the tape NEVER sounded as
good as what I heard live. The multi-track audio then gets mixed
down to stereo (more tape), then sent to the record pressing
plant. What you as the consumer heard from radio or from a record
was NEVER as good as what I would hear in the control room as the
song was being recorded.
The hallmark of the Neve preamps and recording consoles was their
transparency. The idea was to pass the incoming audio as accurately
as possible, without introducing artifacts.
Things are somewhat different these days. Now you can purchase
outboard preamps to bring the mic audio up to line level and so
bypass whatever preamp is inside the console that you are
using. Some of those external preamps are really good, some look
stunning but don't sound good at all. Snake oil, in other words.
I'm not involved with recording studios much these days and don't get
to see all the latest toys. Unfortunately.
Yes, I understood what you said, and I agree that it shows that the
Neve units were really good, but it doesn't prove that they are better
than the other units beyond the possibility that you were
psychologically affected by the foreknowledge of the brand. The
question is whether two units which show the same electrical test
results (as I believe you stated) can really sound significantly
different to human ears and the only way I can see to resolve that is
the old "blind taste test" to see if you can distinguish one from the
other without knowing in advance which kind is being used.
> At 10:37 PM 7/19/2010, Sean Breheny wrote:
>>Did you ever do a fully blind test? (Have a number of amplifiers which
>>were fed to you at random and see if you could pick out the Neve just
>>based on the sound)?
>
> Not with the Neve consoles and not in that era. Generally speaking,
> a recording studio has ONE console. Recall the price tag I mentioned earlier.
>
> Obviously, large facilities in certain US cities have multiple
> recording studios in one building. But the studios I did work for
> had only one studio in the building - and one audio console.
>
> I also think that you may be missing what I said: I was comparing the
> live sound that I would hear in the studio where the musicians are
> playing, then go into the control room and listen to what was coming
> out of the speakers. We had some truly talented recording engineers
> in our city and some of the two-mixes I heard were simply stunning.
>
> Live audio in the recording studio is as good as it gets. Its all
> downhill after that - no matter how much time I spent aligning the
> tape decks (Ampex, then Otari 24 track machines in the studio that I
> spent the most time in), what came off of the tape NEVER sounded as
> good as what I heard live. The multi-track audio then gets mixed
> down to stereo (more tape), then sent to the record pressing
> plant. What you as the consumer heard from radio or from a record
> was NEVER as good as what I would hear in the control room as the
> song was being recorded.
>
> The hallmark of the Neve preamps and recording consoles was their
> transparency. The idea was to pass the incoming audio as accurately
> as possible, without introducing artifacts.
>
> Things are somewhat different these days. Now you can purchase
> outboard preamps to bring the mic audio up to line level and so
> bypass whatever preamp is inside the console that you are
> using. Some of those external preamps are really good, some look
> stunning but don't sound good at all. Snake oil, in other words.
>
> I'm not involved with recording studios much these days and don't get
> to see all the latest toys. Unfortunately.
>
> dwayne
>
> --
> Dwayne Reid <spamBeGonedwaynerEraseMEplanet.eon.net>
> Trinity Electronics Systems Ltd Edmonton, AB, CANADA
> (780) 489-3199 voice (780) 487-6397 fax
> http://www.trinity-electronics.com
> Custom Electronics Design and Manufacturing
>
> I have added a large pitch alternative
> pattern in the library. I increased the (thinnest part of)annular
> ring from
> 0.05mm (!! a lot of PCB houses don't do this low AFAIK) to 0.15mm and
> increased the pin spacing from 1.2mm to 1.5mm, giving a 0.5mm
> clearance. The
> maximum pin spacing according to the datasheet is 1.4mm
You might want to check your actual transistors. Many of the TO-92
like packages I've been seeing have the leads pre-formed to an in-line
footprint with 2.5mm pin spacing (which leaves plenty of room for LOTS
of clearance.) I don't think I've ever seen a PCB footprint that
matches the actual lead-spacing AT the TO92 case...
Em 21/7/2010 04:29, William "Chops" Westfield escreveu:
> You might want to check your actual transistors. Many of the TO-92
> like packages I've been seeing have the leads pre-formed to an in-line
> footprint with 2.5mm pin spacing (which leaves plenty of room for LOTS
> of clearance.) I don't think I've ever seen a PCB footprint that
> matches the actual lead-spacing AT the TO92 case...
I have seen one, once. The transistors were inserted all the way, their
bodies standing flush with the board. The holes and the pads were really
small.
Isaac
__________________________________________________
Fale com seus amigos de graça com o novo Yahoo! Messenger http://br.messenger.yahoo.com/
talktalk.net>
