Trident mod to API 2520 chip info.
Bill Hatcher
I have a friend who ask if I would put out a request for info on
modifying his Trident Series 65 to use the API 2520 chip in the Master
module and in some of the input modules if possible. Anyone out there
ever done this? My friend is very competent with this type of stuff.
Just needs to know what the mod consists of. Please E-mail me direct
and to the group if you like. Regards. Bill Hatcher
Bill Hatcher
Monte P McGuire
In article <32C544...@bmtc.mindspring.com>,
The API 2520 is not a chip, it's an epoxy potted module about 1.2"
square by .6" high. It has the 'standard' pinout for hybrid op amps,
but that pinout is not at all related to the pinout of a standard IC
op amp. The best way to get a 2520 into a circuit is to make a new
circuit board to accomodate the 2520. If you don't want to lay out a
new PC board, you could put the 2520 on a small bit of perfboard and
use jumper wires to attach it to the main PC board. But, if I have to
tell you this, then you won't be able to get it to work anyways (the
magic word is 'stability') so don't bother!
The real question is 'why do you want to do this'??? If you want the
API sound, you'll have to throw in lots and lots of cheap steel cored
output transformers to get the right kind of distortion. The 2520 is
not a horrible op amp, but it's not all that amazing either. It has a
lot of output oomph, but these days, every load you'll find is a
bridging load, so that aspect is irrelevant. The 2520 is similar
noise-wise to a NE5534, which is not a bad chip at all, but the 5534
is probably what's in the mix bus of your console right now, so you'll
notice no improvement noise wise. I'd skip the modification and
simply use an API mike preamp if you want to add some API grunge to
your signal.
IMHO, mix buses are not the place to add a circuit with deliberate
distortion (or 'character' if you want to be kind). I find it's more
interesting to have an exceptionally clean mixer whose inputs can be
'tarnished' individually as desired. That way, you can have some
choice in which signals you're going to step on and which you aren't.
Also keep in mind that a circuit which is quite dirty may still sound
OK when it's handling a single multitracked channel, but if you feed
it a complex mix, it will probably sound pretty horrible since all of
the different elements in the mix will cross modulate each other,
yielding a thick pile of goo. If that's what you want, then why not
use a clean mixer and run the output through a pair of grungy
circuits; it's simpler and you can have some choice in the matter.
Or, if you really want the API sound, why not get an API console?
There should still be a few out there that haven't been chopped into
modules and API is back in business again if you want a brand spankin'
new one. ;-)
Best of luck,
Monte McGuire
mcg...@world.std.com
Dave Baldwin
Monte P McGuire (mcg...@world.std.com) wrote:
: Or, if you really want the API sound, why not get an API console?
: There should still be a few out there that haven't been chopped into
: modules and API is back in business again if you want a brand spankin'
: new one. ;-)
Do you have an address or Web page for API?
--
-=-=-=-=-=-=-=-=-=-=-=- Check out DIBs and TCJ -=-=-=-=-=-=-=-=-=-=-=-=-
Dave Baldwin: dib...@netcom.com | The Computer Journal 1(800)424-8825
DIBs Electronic Design | Home page "http://www.psyber.com/~tcj/"
Voice : (916) 722-3877 | Hands-on hardware and software
TCJ/DIBs BBS: (916) 722-5799 | TCJ/DIBs FAX: (916) 722-7480
-=-=-=-=-=-=- @#$%^&* I can't even quote myself! Oh,well. -=-=-=-=-=-=-
Mark Plancke (Soundtech Recording Studios)
dib...@netcom.com (Dave Baldwin) wrote:
>Monte P McGuire (mcg...@world.std.com) wrote:
>
>: Or, if you really want the API sound, why not get an API console?
>: There should still be a few out there that haven't been chopped into
>: modules and API is back in business again if you want a brand spankin'
>: new one. ;-)
>
>Do you have an address or Web page for API?
>--
Mark Plancke
SOUNDTECH RECORDING STUDIOS
Windsor, Ontario, Canada
http://www.mnsi.net/~soundtch
mailto:soun...@mnsi.net
Kris Kringle
>In article <32C544...@bmtc.mindspring.com>,
>Bill Hatcher <bhat...@bmtc.mindspring.com> wrote:
>>I have a friend who ask if I would put out a request for info on
>>modifying his Trident Series 65 to use the API 2520 chip in the Master
>>module and in some of the input modules if possible.
(...)
>(...) The 2520 is similar
>noise-wise to a NE5534, which is not a bad chip at all, but the 5534
>is probably what's in the mix bus of your console right now, so you'll
>notice no improvement noise wise. (...)
Hi, Monte,
If this is one of the original Trident 65s, all the op-amps there are
TL072 or 74; There were *no* NE5534s or TDA1034s in it.
I happened to work in the second Trid 65 that ever reached the U.S.,
back in 1985, and I must tell you that this was a "beta" board <g>;
the mike preamps got all the cabbie radio talk, the solo switch on the
inputs cut the signal from any busses it was assigned to, the "fx
returns to cue" switch sent the fx returns to cue, but removed them
from the CR mix, the midrange EQ on the subs didn't work after a few
hours of operation (some strange thing was happening in the circuit),
a real mess...
A few months later, Trident fixed the problems (removing the mid EQ
from the subs), added 8 tape returns (only 16 on the 65), some wood
finish, changed the color of the escutcheons to gray (the original
color was yellow, ouch!), added a patchbay and priced it at 3 times
the 65's list price, naming the new board the Trident 24.
Enough said.
--
KK
Monte P McGuire
In article <32ca10b6...@news-s01.ny.us.ibm.net>,
Kris Kringle <bi...@home.com> wrote:
>On Sun, 29 Dec 1996 10:54:18 GMT, mcg...@world.std.com (Monte P
>McGuire) wrote:
>>(...) The 2520 is similar
>>noise-wise to a NE5534, which is not a bad chip at all, but the 5534
>>is probably what's in the mix bus of your console right now, so you'll
>>notice no improvement noise wise. (...)
>
>TL072 or 74; There were *no* NE5534s or TDA1034s in it.
That's pretty odd! The 5534 is _very_ commonly used for virtual earth
summing stages because it has low distortion and nearly ideal noise
properties for the typical source impedance of a V.E. mix bus. The
TL07x should have about 12dB more much noise as compared to a similar
stage using a 5534... Then again, when the 65 was around, most people
mixed onto analog 1/4", which has about 20dB more noise still...
>I happened to work in the second Trid 65 that ever reached the U.S.,
>back in 1985, and I must tell you that this was a "beta" board <g>;
>the mike preamps got all the cabbie radio talk, the solo switch on the
>inputs cut the signal from any busses it was assigned to, the "fx
>returns to cue" switch sent the fx returns to cue, but removed them
>from the CR mix, the midrange EQ on the subs didn't work after a few
>hours of operation (some strange thing was happening in the circuit),
>a real mess...
Yikes! I actually used one in a small studio that had an Otari MX-80
24 track and it seemed to work pretty well. I can't remember swearing
at it all that much, which must have meant it was a nice console
overall! Then again, I don't use solo switches when the master is
printing... ;-)
>A few months later, Trident fixed the problems (removing the mid EQ
>from the subs), added 8 tape returns (only 16 on the 65), some wood
>finish, changed the color of the escutcheons to gray (the original
>color was yellow, ouch!), added a patchbay and priced it at 3 times
>the 65's list price, naming the new board the Trident 24.
Progress...
Regards,
Monte McGuire
mcg...@world.std.com
Monte P McGuire
In article <32ca05e7...@netnews.worldnet.att.net>,
James E. Hammons <JAMES-E...@worldnet.att.net> wrote:
>
>The spec on a Trident 65 is system unity gain noise -75 db max output
>level is +28 db which is a dynamic range of 103db. The op-amps are
>TLO-71's the mic pre is descrete transistor and noise is given for
>200ohms instead of the typical 150 as -125db this would be a few db
>less if measure a the usual 150 ohms and is within a few db of what is
>theoretically possible at room temperature as are all modern well
>designed mixer. the noise level of a mixer is determined almost
>entirely by the first gain stage and TLO-7X series op amps are used in
>many mixers. by the time the level is raised to +4 operating level it
>makes no difference whether the op amp has a noise figure of 1db or
>3db it is swamped out by the amplified noise of the input stage.
With a console having more than a few inputs, a TL071 is far from
optimal as a summing amplifier. But don't just take my word for it,
let's go through the derivation:
What happens when you have 24 inputs assigned to mix? Well, the noise
gain of the summing stage has to be at least 25, which is 28dB. So,
assuming you're using 10K summing resistors, we have a bus impedance
of 400 ohms. This has an inherent noise level of sqrt(4KTRB) = sqrt(4
* 1.38E-23 * 300 kelvin * 400 ohms * 20000Hz) = 364nV = -128dBV. So,
after the gain of the summing bus, the theoretic noise floor of a 10K
summing bus is -100dBV. You can do somewhat better with smaller
summing resistors, but not many consoles do that. 10K is a safe
assumption...
Now, let's add in the contribution of the summing stage op amp. If we
use a TL071, we have an equivalent input noise voltage of 18nV/root Hz
=> 2.55uV => -112dBV. The contribution of the amp's input noise
current is low, but let's calculate it anyway: .01pA/root Hz * 400
ohms => 566pV => -184dBV...
So, adding the three terms together in an RMS fashion: 364nV +rms
2.55uV +rms 566pV = 2.58uV and multiplying by the noise gain of the
stage yields a mix bus output noise level of 64.4uV = -83dBV.
Now, let's try an NE5534 in the same circuit. The amp has a noise
voltage of 4nV/root Hz, which yields 565nV over a 20KHz bandwidth.
The input noise current is around .5pA/root Hz, which is 28.2nV into
400 ohms in a 20KHz bandwidth. So, the sum of the bus noise and the
amp noise is given by: 364nV +rms 565nV +rms 28.2nV = 673nV = -123dBV.
After the gain of the summing stage, the output noise of a 5534
summing stage is -95.5dBV. This is 12.5dB quieter than a TL071 in the
same circuit...
Let's calculate what the noise floor after the mike amp is: assuming
an ideal mike amp, we would have an input noise level equivalent to
the noise of 150 ohms at room temperature, which is sqrt(4 * 1.38E-23
* 300 kelvin * 150 ohms * 20000Hz) = 222nV = -133dBV. Assuming the
mike amp is not perfect, let's handicap it by 5dB and assume a gain of
30dB. This creates a noise floor of -100dBV.
Now, the next trick may be somewhat of a leap, but follow closely: the
-100dBV noise floor of each input is actually the same as the noise
floor of as many inputs noiselessly mixed together as you'd like. The
reason is that the signal in each channel as well as the noise from
each channel are generally uncorelated. This means that as more
inputs are summed, the signal level rises just as fast as the noise
level. So, the S/N ratio remains the same no matter how many inputs
you sum. If the inputs are correlated, then the S/N ratio increases
as more inputs are summed. So, to be safe, let's simply assume the
uncorrelated case.
In summary, this means that inputs processed through non optimal mike
amps yield a noise floor of around -100dBV. A 5534 summing stage will
yield a noise floor of -95dBV. A TL071 summing stage will yield a
noise floor of -83dBV. Do you still think that the summing stage
noise is unimportant and that it's OK to use a TL071???
>The Trident 65 is a quiet mixer and was designed by
>the famed John Oram and it has a excellent reputation for what really
>counts the way it sounds which is musical.
I wouldn't count -75dBV as quiet, especially when other similarly
priced and equipped mixers offer 15-20dB better noise floors. My
little Neotek is far quieter than -75dBV... even using the stock
chips. If you're mixing to 1/4", then perhaps it really doesn't
matter, but that's not the common practice these days. 16 bit noise
floors are the bare minimum and if you plan to do any overall mix
compression, the requirements get even more stringent.
As far as the way the console sounds, it's a matter of preference. I
remember using one and having no real problems with it, but then
again, we were tracking and mixing with analog where you've got a lot
of noise and distortion in the system to start with. You're free to
like the way the console sounds, but don't claim that it has an
insignificant noise contribution in a real world system!!
Regards,
Monte McGuire
mcg...@world.std.com
Kris Kringle
On Tue, 31 Dec 1996 03:02:45 GMT, mcg...@world.std.com (Monte P
McGuire) wrote:
>In article <32ca10b6...@news-s01.ny.us.ibm.net>,
>Kris Kringle <bi...@home.com> wrote:
>>On Sun, 29 Dec 1996 10:54:18 GMT, mcg...@world.std.com (Monte P
>>McGuire) wrote:
>>>(...) The 2520 is similar
>>>noise-wise to a NE5534, which is not a bad chip at all, but the 5534
>>>is probably what's in the mix bus of your console right now, so you'll
>>>notice no improvement noise wise. (...)
>>
>>If this is one of the original Trident 65s, all the op-amps there are
>>TL072 or 74; There were *no* NE5534s or TDA1034s in it.
>
>That's pretty odd! The 5534 is _very_ commonly used for virtual earth
>summing stages because it has low distortion and nearly ideal noise
>properties for the typical source impedance of a V.E. mix bus. The
>TL07x should have about 12dB more much noise as compared to a similar
>stage using a 5534... Then again, when the 65 was around, most people
>mixed onto analog 1/4", which has about 20dB more noise still...
But Trident's choice of opamp was based on cost considerations only;
the 5534s cost a little more, and taking into account the number of
opamps in a board, this made sense for them economically.
Of course I do not agree with them.
Even the balanced outputs of the recording busses used TL07x's.
>Yikes! I actually used one in a small studio that had an Otari MX-80
>24 track and it seemed to work pretty well. I can't remember swearing
>at it all that much, which must have meant it was a nice console
>overall! Then again, I don't use solo switches when the master is
>printing... ;-)
I am sure that you have gone thru the situation when you are cutting
a superb track, and suddenly something sounds wrong in the middle of
it; you then solo the tape return of the track you think has problems,
then solo the buss, then solo the inputs summing into that buss; if
the damn board does a destructive solo instead of PFL, you are in deep
cow dung.
Even during rehearsals, if you soloed any input on the original 65, by
removing the signal form the other busses you would mess the foldback
mix; very annoying for everybody.
>>A few months later, Trident fixed the problems (removing the mid EQ
>>from the subs), added 8 tape returns (only 16 on the 65), some wood
>>finish, changed the color of the escutcheons to gray (the original
>>color was yellow, ouch!), added a patchbay and priced it at 3 times
>>the 65's list price, naming the new board the Trident 24.
>
>Progress...
Yeah... the original Trident 65 sold for $12.000,00 for a 32in/16
returns frame; the Trident 24 was over $30.000,00 for 28ins/24
returns, with patchbay. The input channels were virtually the same,
only with more assignment switches (12 instead of 4); the
output/monitor sections of the 65 had 3-band EQ, and the 24 had only
bass+treble, plus fader reverse.
For $12k, the 65 (if modified the way I did with that first one I
mentioned) was at the time an incredible bargain; the 24 was terribly
overpriced, but it did sell quite well.
__
KK
James E. Hammons
On 29 Dec 1996 04:41:47 GMT, Bill Hatcher <bhat...@bmtc.mindspring.com> wrote:
>I have a friend who ask if I would put out a request for info on
>modifying his Trident Series 65 to use the API 2520 chip in the Master
>module and in some of the input modules if possible. Anyone out there
>ever done this? My friend is very competent with this type of stuff.
>Just needs to know what the mod consists of. Please E-mail me direct
>and to the group if you like. Regards. Bill Hatcher
The spec on a Trident 65 is system unity gain noise -75 db max output level is
+28 db which is a dynamic range of 103db. The op-amps are TLO-71's the mic pre
is descrete transistor and noise is given for 200ohms instead of the typical 150
as -125db this would be a few db less if measure a the usual 150 ohms and is
within a few db of what is theoretically possible at room temperature as are all
modern well designed mixer. the noise level of a mixer is determined almost
entirely by the first gain stage and TLO-7X series op amps are used in many
mixers. by the time the level is raised to +4 operating level it makes no
difference whether the op amp has a noise figure of 1db or 3db it is swamped out
by the amplified noise of the input stage. The main advantage the NE 5534
series has is its ability to drive low impedance loads. The Trident 65 is a
quiet mixer and was designed by the famed John Oram and it has a excellent
reputation for what really counts the way it sounds which is musical.
James E Hammons
James E. Hammons
Nice analysis however you have made a error in assuming that the Trident 65 uses
a low impedances buss this is not true. There are more than one way to skin a
cat and this is not just my opinion I will quote from page 860 Handbook for
Sound Engineers Howard W. Sams an excellent book available from Mix Bookshelf,
The author is Steve Dove.
"Common sense on first glance says to make the mix resistors as low a value as
possible, but this has the downside that too low a value would cause quite large
signal (hence ground) currents to be thundering about. On a less technical and
more economic level, it necessitates somewhat beefier and more serious buffer
amplifiers on each source to feed the busses.
Ordinarily though, the mix resistors are of such a value that, in the context of
a complete mixer , the combined effectively paralleled resistance is well below
the optimum source impedance of the mix-amp device used, so the primary noise
modes are those previously mentioned device vices. This isn't too difficult
with FET ront-end devices, such as the TLO-71 with the high OSI. These devices
have a couple of other major benefits in this application by virtue of their FET
inputs, Input current (hense,input current noise) is extremely low, and being
FET's they don't have the many low=frequency junction and surface noises
inherent to bipolar devices. It seems a paradoxic absurdity to use an ultrahigh
input impedance device for zero impedance mixing, but in many ways they're
better suited than bipolars. On the other hand the intrinsically superior noise
performance of a 5534 class device can pay dividends in this application. Like
so many cases in console design each individual application needs staring at for
its own optimum solution."
I am sure that the $3 price of a NE5534 was not the reason for the choice of
TLO-71's when you consider that every resistor is a 1% metal film It has rail
type professional faders which cost $100 is modular in construction and uses
real VUmeters they were not trying to cut corners.
Sure you could find other solutions however the Trident 65 is more than adequate
for use with ADAT's and compact disks.
The specs are comparable to other mixers of it's size and you are trashing a
fine piece of equipment when I'll bet you don't even have a schematic diagram.
Results are what counts what are the specs on your Neotek and is it comparable
size. Does it output +28db.
Where are your real world sources with more than 105db signal to noise the spec
of a AKG 414 BULS a very quiet mic is 80db. The problem most people have is
lack of mic pads not lack of gain. When you turn up a room mic the meters on a
ADAT clearly show the room noise. What are you recording to with 105db of
dynamic range and what convertors are you using.
By the way using only 20k bandwidth in calculating the noise of your 150ohm
source is to low. Theoretical noise is around 130db and many of to days designs
are in the range of 128-129.
I also own a Soundcraft Spirit Studio the original made by soundcraft which uses
NE5534 bus amps and I can tell you can't tell the noise is about the same so I
do have a real world comparison.
To those who still think they can improve the Trident 65 by substituting op amps
don't do it. It won't work without a complete redesign. The one place where I
did substitute there is a sub board on the master module which on my mixer
contained a TLO-84 put a TLO-74 there.
Best Wishes in the New Year and thanks for taking time for your reply.
James E Hammons
Kris Kringle
*** I took the liberty of creating a new thread, this was getting too
far away from the initial subject ***
KK
On Wed, 01 Jan 1997 17:21:51 GMT, JAMES-E...@worldnet.att.net
(James E. Hammons) wrote:
>On Tue, 31 Dec 1996 13:05:14 GMT, mcg...@world.std.com (Monte P McGuire) wrote:
>
>>In article <32ca05e7...@netnews.worldnet.att.net>,
>>James E. Hammons <JAMES-E...@worldnet.att.net> wrote:
>>>
(snip)
>>>The Trident 65 is a quiet mixer and was designed by
>>>the famed John Oram and it has a excellent reputation for what really
>>>counts the way it sounds which is musical.
>>
>>I wouldn't count -75dBV as quiet, especially when other similarly
>>priced and equipped mixers offer 15-20dB better noise floors.
(snip snip)
>> You're free to
>>like the way the console sounds, but don't claim that it has an
>>insignificant noise contribution in a real world system!!
(snip)
>>Monte McGuire
>>mcg...@world.std.com
>
>I am sure that the $3 price of a NE5534 was not the reason for the choice of
>TLO-71's when you consider that every resistor is a 1% metal film It has rail
>type professional faders which cost $100 is modular in construction and uses
>real VUmeters they were not trying to cut corners.
I have to disagree here.
1) the VU meters used on the Trident 65 were the lowest-cost line from
Selco/Sifam (pretty good for the price, I shall note); the price for
small quantities was $15.95, and I guess this was substantially lower
in industrial quantities;
2) The faders were "real" faders, yes, but FAR from costing $100 each;
they looked like (yes, I do not have a 65 in front of me now to open
and look) ALPS $30 parts, wich also cost substantially less in large
orders;
The list price for a 32 input Trident 65 was about $12.000,00 , so
let's go for some very basic math:
32 input faders + 8 buss faders + 1 master fader @ $100.00 ea =
$4,100.00;
If the VUs were the best Selco has: 16 buss/track VUs+2 master VUs @
$75 ea = $1,350.00
This alone will cost Trident some $ 5,450.00 ; add the cost of the
XLRs, all the pots, knobs, frame, internal connectors, power supply,
labor, etc, and you'll get a manufacturing cost over $8,000.00.
Trident would go broke selling this for $12k.
>To those who still think they can improve the Trident 65 by substituting op amps
>don't do it. It won't work without a complete redesign. The one place where I
>did substitute there is a sub board on the master module which on my mixer
>contained a TLO-84 put a TLO-74 there.
This alone shows how much Trident cared for the low noise of this
board; TL08xs???
>Best Wishes in the New Year
I wish the best for you too!
__
KK
Marc Stager
Monte P McGuire wrote:
> My little Neotek is far quieter than -75dBV... even using the stock
> chips. If you're mixing to 1/4", then perhaps it really doesn't
> matter, but that's not the common practice these days. 16 bit noise
> floors are the bare minimum and if you plan to do any overall mix
> compression, the requirements get even more stringent.
>
> Monte McGuire
> mcg...@world.std.com
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Monte,
This is taking a left turn from the original topic but you
mentioned something that made me curious.
I have a Neotek console which uses TL-074 chips extensively.
It sounds quite good, actually, but if is about 10 years old
and I wonder if there is something that has come along to
improve the noise and distorsion figures. Two channel chips
are no problem but these are quads.
____________________________________________________ MS
David L. Rick
In article <32ca8de5...@netnews.worldnet.att.net>,
JAMES-E...@worldnet.att.net says...
[James is responding to an instructive noise analysis posted by
Monte McGuire.]
>Nice analysis however you have made a error in assuming that the
>Trident 65 uses a low impedances buss this is not true.
<snip>
James then quotes an article by Steve Dove discussing the tradeoffs
between JFET devices and the ever popular NE5534 as a summing amp:
"Ordinarily though, the mix resistors are of such a value that,
in the context of a complete mixer , the combined effectively
paralleled resistance is well below the optimum source impedance
of the mix-amp device used, so the primary noise modes are those
previously mentioned devices."
This argument could be summarized, "If you pick the mix buss impedance
suboptimally, the summing amp noise dominates, so you needn't worry
about the mix buss impedance." To which I respond, "huh?"
If the Trident designers had been willing to pay for a pricier summing
opamp, then lowering the mix buss impedance would have further lowered
the noise floor. But that would have require more power, which would
have meant more expensive power supplies, and, well, you get the picture.
>I also own a Soundcraft Spirit Studio the original made by soundcraft which
>uses NE5534 bus amps and I can tell you can't tell the noise is about the
>same so I do have a real world comparison.
You can't compare these two boards' noise floors without also comparing
their relative headroom. I know a bit about the Spirit board, because
the design was pretty much a direct ripoff of the Seck 1282 and 1882,
one of which I owned until last year. These boards had much more headroom
than competing products, without any sacrifice in noise performance. A look
at the schematic showed a careful mix of 5532s and TL-072s, with the former
used in all the high-gain positions. Shortly after buying the Seck, I
considered the option of changing out the opamps for something more exotic.
After going though a set of noise computations pretty similar to Monte's,
I realized that I couldn't get more than another 3 dB of noise improvement,
because the noise floor would be limited by the (20k?) resistors.
>To those who still think they can improve the Trident 65 by substituting opamps
>don't do it. It won't work without a complete redesign. The one place where I
>did substitute there is a sub board on the master module which on my mixer
>contained a TLO-84 put a TLO-74 there.
This is probably true, for the same reason as on the Seck/Soundcraft.
The good news in all this is that decent-quality mixers are a lot cheaper
than they once were. The original price on the Seck was 1/10 of the
original price on the Trident, and the summing buss design was better!
(The EQ on the Trident was better, though.) Mass-market boards such as
the Mackies are starting to show up with balanced low-Z mixing busses,
which would have been unheard of ten years ago on a budget board.
David L. Rick
Seventh String Recording
dr...@hach.com
Hank Alrich
In article <5ah8vu$2...@natasha.rmii.com>, dr...@hach.com (David L. Rick) wrote:
<snip>
> I know a bit about the Spirit board, because
> the design was pretty much a direct ripoff of the Seck 1282 and 1882,
> one of which I owned until last year.
<snip>
> David L. Rick
> Seventh String Recording
> dr...@hach.com
As a minor point of clarification, Harmon International, parent company of
Soundcraft/Spirit, bought Seck and used the Seck design(s) as a basis for
the Spirit line. So this may not qualify as a "ripoff" in the
generally-accepted sense of the term. They paid their money.
hank
James E. Hammons
Is your name really Kris Kringle or are you afraid to post with your real name?
The 24 is a 65 and inflation took it's toll that's why they priced it at
$30,000. If you look at the cards you will see there are traces for 24 track
assignment. You must be referring to a 4 or 8 buss when you say they had 3 band
eq on the monitors they had no eq on half the monitors ok for 4 or 8 tracks. I
haven't seen one of these but I have the schematics. Later 65's used the same
cards as the 75 but in the smaller sized 65 frame.
The discussion was about the buss amps. Trident used a high impedance buss which
must use a JFET op amp and that is why the NE 5534 was not used. There are many
ways you could improve on their solution however there is no simple way to
change it. Large boards use a balanced buss which gives a 3db advantage for
instance.
On Wed, 01 Jan 1997 17:33:55 GMT, bi...@home.com (Kris Kringle) wrote:
>
>
>*** I took the liberty of creating a new thread, this was getting too
>far away from the initial subject ***
>
>KK
>
>On Wed, 01 Jan 1997 17:21:51 GMT, JAMES-E...@worldnet.att.net
>(James E. Hammons) wrote:
>>On Tue, 31 Dec 1996 13:05:14 GMT, mcg...@world.std.com (Monte P McGuire) wrote:
>>
>>>In article <32ca05e7...@netnews.worldnet.att.net>,
>>>James E. Hammons <JAMES-E...@worldnet.att.net> wrote:
>>>>
>
>(snip)
>
>>>>The Trident 65 is a quiet mixer and was designed by
>>>>the famed John Oram and it has a excellent reputation for what really
>>>>counts the way it sounds which is musical.
>>>
>>>I wouldn't count -75dBV as quiet, especially when other similarly
>>>priced and equipped mixers offer 15-20dB better noise floors.
>
>(snip snip)
>
>>> You're free to
>>>like the way the console sounds, but don't claim that it has an
>>>insignificant noise contribution in a real world system!!
>(snip)
>>>Monte McGuire
>>>mcg...@world.std.com
>>
>
>>I am sure that the $3 price of a NE5534 was not the reason for the choice of
>>TLO-71's when you consider that every resistor is a 1% metal film It has rail
>>type professional faders which cost $100 is modular in construction and uses
>>real VUmeters they were not trying to cut corners.
>
>I have to disagree here.
>
>1) the VU meters used on the Trident 65 were the lowest-cost line from
>Selco/Sifam (pretty good for the price, I shall note); the price for
>small quantities was $15.95, and I guess this was substantially lower
>in industrial quantities;
>
>2) The faders were "real" faders, yes, but FAR from costing $100 each;
>they looked like (yes, I do not have a 65 in front of me now to open
>and look) ALPS $30 parts, wich also cost substantially less in large
>orders;
>
>The list price for a 32 input Trident 65 was about $12.000,00 , so
>let's go for some very basic math:
>
>32 input faders + 8 buss faders + 1 master fader @ $100.00 ea =
>$4,100.00;
>
>If the VUs were the best Selco has: 16 buss/track VUs+2 master VUs @
>$75 ea = $1,350.00
>
>This alone will cost Trident some $ 5,450.00 ; add the cost of the
>XLRs, all the pots, knobs, frame, internal connectors, power supply,
>labor, etc, and you'll get a manufacturing cost over $8,000.00.
>
>Trident would go broke selling this for $12k.
>
>>To those who still think they can improve the Trident 65 by substituting op amps
>>don't do it. It won't work without a complete redesign. The one place where I
>>did substitute there is a sub board on the master module which on my mixer
>>contained a TLO-84 put a TLO-74 there.
>
>This alone shows how much Trident cared for the low noise of this
>board; TL08xs???
>
>>Best Wishes in the New Year
>
Kris Kringle
>Is your name really Kris Kringle or are you afraid to post with your real name?
Does it make any difference? What does it have to do with the subject
in focus?
>The 24 is a 65 and inflation took it's toll that's why they priced it at
>$30,000.
Mr. Hammons, I am comparing the price of a Trident 65 purchased in
February, 1985, against the price for the Trident 24 in October of the
SAME year.
Inflation? I don't think so.
> If you look at the cards you will see there are traces for 24 track
>assignment. You must be referring to a 4 or 8 buss when you say they had 3 band
>eq on the monitors they had no eq on half the monitors ok for 4 or 8 tracks. I
>haven't seen one of these but I have the schematics. Later 65's used the same
>cards as the 75 but in the smaller sized 65 frame.
The *original* Trident 65 I referred to had 8 busses and 16 tape
returns, with 3-band EQ on the 8 busses OR the first 8 tape returns,
switchable.
If Trident later changed this for some reason (cost? the problem with
the mid-EQ simply disappearing after a few hours of operation? major
redesign? who knows?), I do not know.
What I say is what I know; my experience, as I stated clearly in my
first post, was with the very second Trident Series 65 console that
ever reached the shores of the U.S., and it was in February (March
maybe?) of 1985.
The cost difference I've pointed was based on the price asked for the
Trident 24 at the 79th AES convention in NY, in 1985 too.
>The discussion was about the buss amps.
The initial discussion was: Trident mod to API 2520 chip info, started
by Mr. Bill Hatcher; later, another thread started with: Info on
Trident 65-70-75 etc, by Mr. D.L. Morley;
In both threads I gave the very same reply.
Later on, the discussion included deeper technical issues on buss
amps, but the main subject is still "opinions on the Trident 65-70-75
etc" or "is it worth modifying a Trident 65 with the API 2520".
Since the discussion went far from the two original propositions, I
created a new thread under "Trident Series 65 debate"(should have been
"Trident Series 65 discussion", my mistake), so that we all could
write about our personal opinions and leave the original threads
focused on the original questions.
You do not seem to be replying to any of the considerations I made
about the 65, but else centering the discussion on my use of a real or
assumed name, and in blaming the absurd price difference between the
65 and the 24 on an inflation that never happened.
I do reserve myself the right to remain incognito, and I will not
discuss my reasons for it with anyone in any NG.
A final note: I do respect your technical knowledge, I am in no way
trying to diminish you, I am just saying what I know and think about
the Trident Series 65 and 24.
And my sincere wishes of a very happy and peaceful 1997!
__
KK
David L. Rick
In article <walkinay-ya023080...@news.telis.org>,
walk...@telis.org says...
>> the design was pretty much a direct ripoff of the Seck 1282 and 1882,
>> one of which I owned until last year.
>
>Soundcraft/Spirit, bought Seck and used the Seck design(s) as a basis for
>the Spirit line. So this may not qualify as a "ripoff" in the
>generally-accepted sense of the term. They paid their money.
Fair enough. They also eliminated the in-place solo (bad), improved the
case rigidity (good), retained the fragile and acoustically noisy
power supply (bad), moved the EQ corners (good), and raised the price
substantially so that the product wouldn't canabalize their higher-priced
lines so badly. The original Secks can be purchased very cheaply on
the used market, since nobody knows what they are. Hold out for
the 1882; twelve channels are not enough.
In certain respects (but not all) these were better boards than the
Mackie 8-bus which displaced them in the market. I give Mackie the nod
for better EQ, stereo returns, and slightly better mic preamps. I like
the Seck/Spirit's mix buss design and generally more flexible routing.
The Seck and Spirit can be repaired/modified without a great deal of
fuss, something that's difficult on Mackies due to their surface-mount
components. Do not confuse the recording-oriented Secks (8 busses)
with their sound-reinforcement-oriented brethren (4 busses). The
latter are not worth your time.
Monte P McGuire
In article <32ca8de5...@netnews.worldnet.att.net>,
James E. Hammons <JAMES-E...@worldnet.att.net> wrote:
>
>Nice analysis however you have made a error in assuming that the
>Trident 65 uses a low impedances buss this is not true. There are
>more than one way to skin a cat and this is not just my opinion I will
>quote from page 860 Handbook for Sound Engineers Howard W. Sams an
>excellent book available from Mix Bookshelf, The author is Steve Dove.
>
>"Common sense on first glance says to make the mix resistors as low a
>value as possible, but this has the downside that too low a value
>would cause quite large signal (hence ground) currents to be
>thundering about.
Possibly, but if 10K causes problems, then what happens when the input
module drives a 10K channel fader?? Or an equalizer pot? Or an aux
send pot?? The problems that result when signal current is sent to
ground in a console are very serious and there are only two ways to
solve them: use an enormous amount of metal for the ground system (a
la Neotek) so that the ground currents don't matter or run balanced IO
between modules, so that there is no net signal current between
modules (a la Neve). The first approach makes the console heavier and
larger than what we see today and the second approach requires more
circuitry at greater cost. This stuff simply doesn't get done inside
of small consoles like Mackies.
>On a less technical and more economic level, it
>necessitates somewhat beefier and more serious buffer amplifiers on
>each source to feed the busses.
10K is not hard to drive... 2K can be driven by any op amp fairly
well... If we decide to not use 10K summing resistors and go with 20
or 50K, then the noise floor rises even higher. Considering the
Trident spec of -75dBV and not -83dBV (which is what I calculated for
a 10K bus), this is apparently what they did.
>The specs are comparable to other mixers of it's size and you are
>trashing a fine piece of equipment when I'll bet you don't even have a
>schematic diagram.
No, I don't have a schematic, but if you quote an output noise of
-75dBV I don't care how they got there, it's simply far from the best
you can do with components and techniques of equivalent vintage. Low
noise is not always obtained at the expense of low distortion; with
modern op amps that can drive low impedances cleanly, this is simple
to do. Older op amps can be used with discrete buffers for higher
output drive too...
>Results are what counts what are the specs on your
>Neotek and is it comparable size. Does it output +28db.
My console is a 20x8 Series II and it doesn't output +28 because it's
unbalanced throughout; there's no inverted output to bring up the
voltage swing. However, the noise floor is typically at -100dBV with
only a few inputs assigned and it runs off of 18V rails, so it can do
at least +23dBV. However, it is able to drive 200 ohms pretty easily
now that I have upgraded the op amps in the master and channel output
sections...
>Where are your real world sources with more than 105db signal to noise
>the spec of a AKG 414 BULS a very quiet mic is 80db. The problem most
>people have is lack of mic pads not lack of gain. When you turn up a
>room mic the meters on a ADAT clearly show the room noise. What are
>you recording to with 105db of dynamic range and what convertors are
>you using.
A close miked drum gives a very high acoustic output and a very low
total system noise due to the low preamp gain used. I typically use
about 10dB of gain on a Sanken CU-41 on snare drum; this should easily
have over 100dB of S/N... Sure, the room or the converters aren't
that quiet, but I simply want the equipment to be quiet so that I can
ignore its noise contribution. It makes it simple to use compression,
EQ and gainriding if the console isn't the dominant limitation.
>By the way using only 20k bandwidth in calculating the noise of your
>150ohm source is to low. Theoretical noise is around 130db and many
>of to days designs are in the range of 128-129.
I stand by the calculations... 150 ohms at room temperature in a 20KHz
bandwidth gives you precisely a -133dBV noise floor. This is the
absolute best that can be attained, so that's why I handicapped the
results by 5dB; most mike amps can do at least that. As far as using
more than 20KHz bandwidth, that's not the common tradition for noise
calculations on audio gear...
>To those who still think they can improve the Trident 65 by
>substituting op amps don't do it. It won't work without a complete
>redesign. The one place where I did substitute there is a sub board on
>the master module which on my mixer contained a TLO-84 put a TLO-74
>there.
It's not theoretically impossible to upgarde a console with chip
swaps, but you do have to have very good test equipment to make sure
you actually did something positive to the console and to make sure
you didn't break anything. Furthermore, upgrading a console is
hopefully going to change the sound of it and if you're happy with the
sound of the console, then an upgrade could be a step in the wrong
direction.
In my case, the stock Neotek Series II had some really nice features
and a good basic layout, but it was lacking in some performance
aspects. After upgrading it (not just chip swaps) over a period of
years, it is clearly better sounding than it has ever been and I
regret nothing. However, the goal of the Neotek was to be as neutral
as possible, so it was pretty straigthforward to improve upon that.
Some consoles that are designed to have a certain 'sound' are not so
straightforward to upgrade since usually a lot of colorations are
balanced off of each other; removing one coloration can result in
performance that is technically better but perhaps more annoying to
listen to.
>Best Wishes in the New Year and thanks for taking time for your reply.
My pleasure and happy new years to you...
Monte McGuire
mcg...@world.std.com
Monte P McGuire
In article <5ah8vu$2...@natasha.rmii.com>, David L. Rick <dr...@hach.com> wrote:
>
>Mass-market boards such as
>the Mackies are starting to show up with balanced low-Z mixing busses,
>which would have been unheard of ten years ago on a budget board.
I've never seen the schematics for the 8 bus Mackies, just the 1604
and I don't recall seeing balanced buses. Do they use balanced
summing buses in their 8 bus boards?? If so, hats off to Mackie!!
Regards,
Monte McGuire
mcg...@world.std.com
Kris Kringle
On Wed, 01 Jan 1997 17:21:51 GMT, JAMES-E...@worldnet.att.net
(James E. Hammons) wrote:
(big snip)
>I am sure that the $3 price of a NE5534 was not the reason for the choice of
>TLO-71's when you consider that every resistor is a 1% metal film It has rail
>type professional faders which cost $100 is modular in construction and uses
>real VUmeters they were not trying to cut corners.
I have to disagree here.
1) the VU meters used on the Trident 65 were the lowest-cost line from
Selco/Sifam (pretty good for the price, I shall note); the price for
small quantities was $15.95, and I guess this was substantially lower
in industrial quantities;
2) The faders were "real" faders, yes, but FAR from costing $100 each;
they looked like ALPS $30 parts, wich also cost substantially less in
large orders;
The list price for a 32 input Trident 65 was about $12.000,00 , so
let's go for some very basic math:
32 input faders + 8 buss faders + 1 master fader @ $100.00 ea =
$4,100.00;
If the VUs were the best Selco has: 16 buss/track VUs+2 master VUs @
$75 ea = $1,350.00
This alone will cost Trident some $ 5,450.00 ; add the cost of the
XLRs, all the pots, knobs, frame, internal connectors, power supply,
labor, etc, and you'll get a manufacturing cost over $9,000.00.
Trident would go broke selling this for $12k.
>To those who still think they can improve the Trident 65 by substituting op amps
>don't do it. It won't work without a complete redesign. The one place where I
>did substitute there is a sub board on the master module which on my mixer
>contained a TLO-84 put a TLO-74 there.
This alone shows how much Trident cared for the low noise of this
board; TL08xs???
>Best Wishes in the New Year
I wish the best for you too!
__
KK
Monte P McGuire
In article <32CB94...@ix.netcom.com>,
Marc Stager <sso...@ix.netcom.com> wrote:
>This is taking a left turn from the original topic but you
>mentioned something that made me curious.
>
>I have a Neotek console which uses TL-074 chips extensively.
>It sounds quite good, actually, but if is about 10 years old
>and I wonder if there is something that has come along to
>improve the noise and distorsion figures. Two channel chips
>are no problem but these are quads.
The only real problem with the TL074 is that it can't drive low
impedance loads very well. Even though most loads are high Z lately,
better output drive means less distortion at high levels, so it's
worthwhile to update the console.
For a while I was using the AD713 which specs somewhat better THD
wise, but when you actually put the thing on a distortion analyzer and
look at the distortion spectrum, it seems that the TL074 actually has
less distortion (and a more pleasant form of distortion) than the
AD713. I haven't measured any TL074 directly, but I did measure TL072
in an outboard mike amp and found that the TL072 has predominantly 2nd
harmonic and little else, while the AD712 has a fairly rich mixture of
harmonics; not good! These results are not completely scientific
because I didn't measure the AD712 alone, but I think they're still
valid.
So, these days, I'd recommend using the Burr Brown OPA4132. It draws
a lot more supply current, but it can properly drive low Z loads, it
has less voltage noise and it has far better DC specs. I'm using a
combination of 4132s and Everett Carroll's quad hybrids made from
OPA2604 chips in the master section and a few of the channel modules
and it's working nicely. The power supply doesn't seem to complain
about the extra current draw, but you may want to do some calculations
first...
I need to do more real testing of op amps, but for now, the Burr Brown
OPA132/2132/4132 and the OPA604/2604 are my favorite JFET input op
amps. They're fantastic sounding, simple to apply, not too fussy
about layout and they're not all that expensive either. Digikey now
sells them, but if their catalog prices are correct, you may be able
to get better prices from other dealers like Insight Electronics. See
the back of a Burr Brown databook for dealer information and shop
around. If you're filling a console, you can probably get some nice
quantity discounts.
I also made a few other modifications to the Neotek; the best mod was
to replace the discrete FET switches in the control room module with
some mini hermetic relays. They use some rather nasty JFET switches
that, while they're only in the monitor path, really gunk up the sound
of the console. Oddly enough, the relays were smaller than the whole
JFET circuit they replaced and they are simple to retrofit. I was
able to find some surplus, so it didn't cost much either. If you have
a Series II or can send me some schematics, I can guide you through
the mod if you'd like.
I was also able to jumper out an entire quad amp inside of the control
room module as well. Other mods I've done to the console: replaced
the 5534 summing stage with an AD797 and a DC servo (this yields less
distortion and allows me to jumper a few coupling caps), replaced the
5534 channel fader follower with AD797 and a DC servo (also allows me
to jumper a lot of coupling caps and provides better output drive) and
I also beefed up many of the power supply decoupling caps (changed the
stock 10uF to 100uF in the channels and 2200uF in the summing and
master stages). I have a few channels with only polypropylene and
polystyrene coupling caps also; I've been to busy to retrofit the
entire console.
It's a nice console and it now sounds better than it ever did. Too
bad I mix digitally... ;-) Digital mixing still sounds better than
anything else, but it's still nice to have a clean console now and
then. I still use it as a keyboard and monitor stand for my audio
workstation as well as a patchbay and control room monitor system, so
it still has a purpose. Oh, and it looks a lot more 'official' than
just a Macintosh and some converters... ;-)
Happy new year,
Monte McGuire
mcg...@world.std.com