TS 9 opamps
From jshina–(at)–indspring.com Sun Sep 14 10:34:15 CDT 1997
From: jshina–(at)–indspring.com (John S. Shinal)
Newsgroups: alt.guitar.amps
Subject: Experiments on a TS-9 Reissue (OPAMPS)
Date: Sun, 14 Sep 1997 03:13:28 GMT
Reply-To: jshina–(at)–indspring.com
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
X-Server-Date: 14 Sep 1997 15:13:22 GMT
X-No-Archive: yes
Xref: geraldo.cc.utexas.edu alt.guitar.amps:63642
Experiments on a TS-9 Reissue
What follows are the results I got when doing experimental
mods to my TS-9 reissue.
My objective was to sample different op-amps in the TS-9 and
find which one I liked the best. All descriptions are purely
subjective opinions, but I tried to remain unbiased (pun intended).
Thanks to the many folks who have discussed TS-9 mods over the past
two years, I used many of your suggestions. Honors to R.G. Keen, Jack
Orman, Analog Mike, and Axel Rohde.
All the op amps I tried were simply what I have in my parts
cabinet this weekend. Some of them are old, some are newer. I think
current production op-amps will be identical in a few of the cases.
Equipment :
TS-9 reissue, approx 2 yrs. old.
Fender ’62 reissue Strat with old .011s
a borrowed Fender SF Deluxe (low volume, pure clean)
a 25′ Ernie Ball cable (guitar to TS-9)
a 10′ Ernie Ball cable (TS-9 to amp)
Output level on the TS-9 was set so that the effect produced
the same volume when switched in or out.
Power for the TS-9 was supplied by a Radio Shack ‘wall wart’
with adjustable voltage and 300ma max current. I accidentally left it
set on 6 volts (oof !) but it appears not to have mattered (sounds the
same at 9 volts).
MODS TO THE DEVICE UNDER TEST (DUT) :
1. Installed an AUGAT gold plated machined socket (wire wrap style).
2. Voltage divider at the output of the DUT was originally a
( 470 ohm/100K ohm ). I changed it to ( 470 ohm/10K ohm ).
3. Diode clippers were both changed to 2mm submini LEDs, one red, one
yellow.
The above mods themselves affected the sound of the unit. The
LEDs produced a much smoother clipping, but with significantly less
drive and less high frequency available.
!! THE DRIVE CONTROL WAS SET AT MAXIMUM FOR ALL TESTS !!
For the tests, I swapped the op-amps in the socket, but made
no other changes between test cases.
The original op amp in the DUT was a TA75558P, with no date
code.
For those who don’t know, date codes on integrated circuits
(chips) are frequently in the form –> 9518, which would indicate
1995, 18th week as the manufacture date. There are also usually
letters as well, which generally indicate a factory or production
line.
(Fade in sounds of Mad Scientist’s Laboratory : Bzzt ! Bzzt !)
OP AMPS TESTED IN THE DUT :
1. National Semiconductor LM1458N, date code M8836
Not bad. The low notes ‘gargle’ slightly, but overall this opamp
sounds pretty smooth. There is a good ‘touch’ response : gentle
picking provides clean notes, and hard picking produces distorted
notes. Distortion drive was moderate.
2. Signetics LM358N, date code 8726VF
YUCK ! EECH ! PEEE-YOOO ! Distortion becomes a buzzy fuzz, BUT the
distorted signal is SWAMPED by the clean tone. Distortion becomes
easily heard ONLY as the notes decay. Bass notes gargle badly. Spitty,
non-musical distortion, but not even usable for weird 1960s fuzz
efects. A disaster.
3. National LM358N, date code M9036
Every bit as weird as the Signetics, but also ‘breathes’ like a cheap
compressor. Distorted tone is smoother, but still swamped by the clean
signal.
4. National LM1458N, date code M9012
Nice. Lows are smooth, with no gargling even on strongly picked notes.
Distortion drive medium. Touch response is nice, although a bit
sensitive; requires care in picking or it’ll all sound distorted.
Noticably different touch response than case #1. Overall a pretty good
op-amp in this DUT.
5. Motorola MC1458P1, date code KKAS 9032
Quite good. Distortion drive level fairly high. Good touch response
with a wide dynamic range, it’s easy to pick it clean or dirty. Low
notes quite smooth even during aggresive playing. A very pleasing
choice. This op-amp is a later version of case #6.
6. Motorola MC1458P, date code L8017
Here’s where it gets scary. Notice from the date code that this is a
1980 version of case #5; and there were likely some internal changes.
(Great, a ‘vintage’ op-amp. Groan.)
This op-amp produced an extremely smooth drive. Touch response is
superb. Produces moderate distortion drive, less than #5, but more
than the original TA75558P. This op-amp would NOT gargle no matter how
hard I slammed the strings. A super choice if you have one laying
around.
A FEW CONCLUSIONS :
While the LM358 family is similar to the 1458/1558/4558
family, the TS-9 circuit will need additional changes to allow its
use.
The newer production Motorola parts sound pretty good, and the
Nationals only slightly less so. Both are good choices.
The old Motorola has the best dynamic response in the basic
TS-9 of all the op-amps I tested. I’d be interested to find an old
MOTO databook and compare the internal transistor layout to the later
P1 version tested in case #5.
NOTE : I am certain that with changes to the feedback loop
around the diode clipper, that these minor differences in the op-amps
can be evened out. All it requires is a handful of resistors, and
careful soldering technique so you can try them one at a time without
ruining the PCB from excess heat and abrasion.
The diodes used for the clippers are going to change all of
the above test cases, at least as far as the amount of drive. This is
no doubt due to the recovery characteristics as the diodes switch, and
the response of the various op-amps. The stock 1N914s had MUCH more
clip than the LEDs in there now. I expect to continue swapping
different diodes in and out for quite a while.
I hope you had as much fun reading this as I had doing it.
If I make a breakthrough in diode’s, I’ll post that too.
John S. Shinal
jshina–(at)–indspring.com
WEB : http://homepage.usr.com/j/jsshinal
From mgarvi–(at)–anix.com Wed Sep 17 11:10:10 CDT 1997
From: mgarvi–(at)–anix.com (Mark Garvin)
Newsgroups: alt.guitar.amps
Subject: Re: Experiments on a TS-9 Reissue (OPAMPS)
Date: 17 Sep 1997 08:38:19 -0400
Xref: geraldo.cc.utexas.edu alt.guitar.amps:64040
Hi John,
Haven’t seen you here for a while. Nice post! A couple comments…
> jshina–(at)–indspring.com (John S. Shinal) writes:
>
>Experiments on a TS-9 Reissue
>
> What follows are the results I got when doing experimental
>mods to my TS-9 reissue.
> My objective was to sample different op-amps in the TS-9 and
>find which one I liked the best. All descriptions are purely
>
> OP AMPS TESTED IN THE DUT :
>
>1. National Semiconductor LM1458N, date code M8836
>
>Not bad. The low notes ‘gargle’ slightly, but overall this opamp
>sounds pretty smooth. There is a good ‘touch’ response : gentle
>picking provides clean notes, and hard picking produces distorted
>notes. Distortion drive was moderate.
This is a ‘plain vanilla’ dual 741-type opamp. Expect similar
sound from most manufacturers’ 1458’s and 4558’s.
>2. Signetics LM358N, date code 8726VF
>YUCK ! EECH ! PEEE-YOOO ! Distortion becomes a buzzy fuzz, BUT the
>distorted signal is SWAMPED by the clean tone. Distortion becomes
>easily heard ONLY as the notes decay. Bass notes gargle badly. Spitty,
>non-musical distortion, but not even usable for weird 1960s fuzz
>efects. A disaster.
Correct. 358’s are simply wretched for most applications. Same
for 324’s.
>3. National LM358N, date code M9036
>
>Every bit as weird as the Signetics, but also ‘breathes’ like a cheap
>compressor. Distorted tone is smoother, but still swamped by the clean
>signal.
Ditto.
>4. National LM1458N, date code M9012
>
>Nice. Lows are smooth, with no gargling even on strongly picked notes.
>Distortion drive medium. Touch response is nice, although a bit
>sensitive; requires care in picking or it’ll all sound distorted.
>Noticably different touch response than case #1. Overall a pretty good
>op-amp in this DUT.
I’m surprised at any audible difference from other National 1458’s.
>5. Motorola MC1458P1, date code KKAS 9032
>
>Quite good. Distortion drive level fairly high. Good touch response
>with a wide dynamic range, it’s easy to pick it clean or dirty. Low
>notes quite smooth even during aggresive playing. A very pleasing
>choice. This op-amp is a later version of case #6.
>6. Motorola MC1458P, date code L8017
>…
>This op-amp produced an extremely smooth drive. Touch response is
>superb. Produces moderate distortion drive, less than #5, but more
>than the original TA75558P. This op-amp would NOT gargle no matter how
>hard I slammed the strings. A super choice if you have one laying
>around.
Again, I’m a bit surprised at audible differences between different
‘vintages’ of the Moto’s.
> A FEW CONCLUSIONS :
>
> While the LM358 family is similar to the 1458/1558/4558
>family, the TS-9 circuit will need additional changes to allow its
>use.
Actually, it’s not that close except for pinout, John. The 358 was
designed for single supply operation. Sounds like a good thing, but
it’s not in this case. I suspect serious problems turn up in all
variants of this chip when driving otherwise ‘normal’ opamp loads.
The difference you have noticed may be largely due to the opamps’
output stage circuitry.
> The diodes used for the clippers are going to change all of
>the above test cases, at least as far as the amount of drive. This is
>no doubt due to the recovery characteristics as the diodes switch, and
>the response of the various op-amps. The stock 1N914s had MUCH more
>clip than the LEDs in there now. I expect to continue swapping
>different diodes in and out for quite a while.
Pretty easy to find germaniums, too. Lower breakover voltage than
silicon diodes, so expect more overdrive but less volume. The
opposite of the change from si to LED’s.
There is a less well known type of diode called a ‘Schottky barrier’
diode. They have yet a lower breakover voltage than germaniums.
Check 1N3819 (if I remember correctly) in Digikey and Mouser catalogs.
Regarding the different colored LED’s: it would seem that two different
breakover voltages would be good in that it would create more even
harmonic distortion (asymmetrical output waveform), but that’s not
necessarily so. Try matching two red LED’s and compare the difference.
Other opamps that may be worth trying:
Burr Brown OPA2604 and OPA2134. Both are duals with FET inputs and
low noise. Tough to find. Audiophiles are snatching all the 2604’s.
Texas Instruments TLC2202, TLC2272 and others. These are CMOS opamps
(most FET opamps are JFETs), and may respond very differently in this
circuit. I suspect that they would sound quite good. Not common, though.
NE5532 (various manufacturers). The 5534 was one of the first very low
noise opamps designed to drive 600 ohm loads. The 5532 is a dual
variation (not identical but close). Kinda power-hungry. Possibly
more edge to the sound than other opamps due to the high slew rate and
ability to drive cables, etc. This may be good in this particular
circuit. These are used in med/high Q mixing consoles. Should be easy
to find these days.
LF353 was one of my favorite FET-input opamps for guitar circuits.
I just liked the sound. Should be another easy one to locate.
TL062, TL072, TL082 are similar with differences in power, noise, etc.
These are used in a lot of music/synth gear (esp the 72). Also common.
Thanks again for your post, John. Please let us know if you have a
chance to try any of the other opamps.
MGarvin
