Login / Register
 
Best Opamps to replace TL072's in TAC Scorpion-II
New Reply
Subscribe
#91
4th December 2011
Old 4th December 2011
  #91
Lives for gear
 
Joined: Sep 2004
Location: UK
Posts: 6,770

Matt Syson is online now
Hi
Get the master and aux modules recapped and cleaned first.
If your monitors are connected and not humming, don't bother balancing them.
Balancing the main outs could be good but again if you have no problem and your setup is pretty 'static' you probably don't need to do these either.
Auxes may be better balanced as it would save messing about when connecting different gear. Using a SSM2142, DRV134 or THAT1646 chip with the very small handful of components (basically 4 capacitors and a couple of resistors) would add the balance to an output. To be useful as a balanced out it MUST have a 'transformer like' characteristic where it automatically keeps the level out constant even if you short hot OR cold to ground (unbalanced). The above chips do that.
For a few extra output amplifiers you don't need to mess with the supply.
A proper power supply 'upgrade' should mean bigger heatsinks and transformer as other efforts are just 'fiddling around'.
Matt S
__________________
Matt S
www.mseaudio.co.uk
#92
5th December 2011
Old 5th December 2011
  #92
Lives for gear
 
Joined: Aug 2008
Location: Karlsruhe, Germany
Posts: 975

jensenmann is online now
#93
14th December 2012
Old 14th December 2012
  #93
Gear interested
 
Joined: Aug 2009
Posts: 5

heavyllegas is offline
Hi
I recently bought a tac scorpion console used.
And I would like to upgrade.

I wonder where I can buy the opamps:
BurrBrown OPA1642 fet
National LME49720NA

truth there are so many models which are not purchased

I appreciate your help
regards


moises villegas
heavyllegas@hotmail.com
donnie7
Thread Starter
#94
15th December 2012
Old 15th December 2012
  #94
Gear maniac
 
donnie7's Avatar
 
Joined: Sep 2004
Location: Boston, MA
Posts: 156

Thread Starter
donnie7 is offline
moises villegas,
You have a lot of work to do. The upgrade is no small thing. Your capacitors being replaced (recapped) are much more important than replacing the opamps. You can Google whatever you need. Proceed with CAUTION ! Season's Greetings.
#95
15th December 2012
Old 15th December 2012
  #95
Gear maniac
 
Samuelito's Avatar
 
Joined: Jan 2009
Posts: 186

Samuelito is offline
Hi moises,

You can get them either at Digikey or at Mouser, but be aware that you'll need the adapter for the OPA1642, the .1 uf mono ceramic caps and of course a new power supply to make the board run with those IC's.

GRS will help you with this.

Or you can go to AML get the kit for the scorpion, recap both PSU and console and be happy with it.

Maybe you can upgrade U5 with LT1358's and live with same PSU.. Don't know really.

Someone else can answer this??





Sam
#96
21st December 2012
Old 21st December 2012
  #96
Gear interested
 
Joined: Aug 2009
Posts: 5

heavyllegas is offline
Thanks sam

Quote:
Originally Posted by Samuelito View Post
Hi moises,

You can get them either at Digikey or at Mouser, but be aware that you'll need the adapter for the OPA1642, the .1 uf mono ceramic caps and of course a new power supply to make the board run with those IC's.

GRS will help you with this.

Or you can go to AML get the kit for the scorpion, recap both PSU and console and be happy with it.

Maybe you can upgrade U5 with LT1358's and live with same PSU.. Don't know really.

Someone else can answer this??





Sam

Then I change the TL072 serve by OPA1642AID.
because I need other psu?
#97
21st December 2012
Old 21st December 2012
  #97
Gear interested
 
Joined: Aug 2009
Posts: 5

heavyllegas is offline
The truth change the TL072's thought by eq OPA1642
TL072 fader and the LME49720NA

none of this will work?
#98
22nd December 2012
Old 22nd December 2012
  #98
Gear maniac
 
Samuelito's Avatar
 
Joined: Jan 2009
Posts: 186

Samuelito is offline
Hi Moises,

Like Jim Williams said, the bottle neck of the channel srtips is the U1 position (preamp).

I suggest you go first to Audio Maintenance and get the preamp kit for the s1000 modules that comes with a full recap and explanation on how to upgrade the U1 position.
Recap also the s2000 et s3000 modules.
For the master fader (s4000) you can recap it and/or mod it. I sent it to Jim and he has done a great job with mine. Just think how you will be using the board... only for the preamps or are you going to be monitoring and maybe summing, through the master module.

And don't forget to recap your PSU, is the heart of your board.

This will give you time with your hands on the board and you will start understanding what you want and what you don't want to do with this console.

Good modding!!





Sam
#99
23rd December 2012
Old 23rd December 2012
  #99
Gear interested
 
Joined: Aug 2009
Posts: 5

heavyllegas is offline
I want to use the console to mix songs.
then will move forward slowly with AML kits for preamps

I would also like to improve the sound console equalizer

thanks sam
#100
26th December 2012
Old 26th December 2012
  #100
Gear maniac
 
Steele's Avatar
 
Joined: Jul 2008
Location: Dalton, MA 01226
Posts: 153

Steele is offline
Opamps in a Scorpion

Quote:
Originally Posted by Jim Williams View Post
Hi Donnie!

Did you try out the amps I suggested before?

Avoid OP275's in the EQ circuit as they have too much bias current and will make the frequency pots scratch. Use either a high beta opamp like the LT1352 or a fet input opamp like OP2134. OP2604's may require extra psu bypassing and prehaps some local phase margine tweaks to stabilize. I also like AD825's mounted on a Brown Dog adaptor.

I like the LT1358's in the fader amp position of the 4001 inputs. You can remove all the coupling caps in that section if you use this opamp. It's also very good in the stereo sum amp position as noise is reduced and loop gain increases for lower buss distortion and better clarity when bussing a large number of inputs.

Jim Williams
Audio Upgrades
Yea, listen to Jim. I've always upgraded my gear and jim is one of my references. With my Scorpion II I went thru everything. One thing different, I put LME49720's in U5. No more caps.
#101
26th December 2012
Old 26th December 2012
  #101
Gear interested
 
Joined: Aug 2009
Posts: 5

heavyllegas is offline
Thanks for the help

Quote:
Originally Posted by Jim Williams View Post
I believe AML has sub boards for replacing the SSM2015 chip. In recent years new devices have been released that would be a fine addition to the Scorpion.

In the stereo sum stage of the 4000 module I use the National LME49720NA. That is a very low noise opamp with huge open loop gain at 10k hz. That will clean up a stereo mix quite well. Coupling caps can be yanked out, those are precision devices.

In the S1000/1001 inputs, the new BurrBrown OPA1642 fet opamp is a winner in the EQ section, very low bias current (no pot scratch) and very low 5 nv noise. Use LME49720 for the fader amp and remove all the el caps there.

good afternoon

My last question:
-The LME49720 NA in U5 not need upgrade PSU?
-What amp could replace the TL072 (U2, U3, U4). not requiring upgrade psu?
-What brand of OPAmps would have to buy? (texas Instruments or other brand)


Thank you very much.
#102
9th August 2013
Old 9th August 2013
  #102
Gear interested
 
Joined: Jan 2008
Posts: 11

weary is offline
bumping an old thread... re the tac scorpion 1... I'm wanting to change U1 and U5 in the inputs (preamp and fader amp). For U1 I'll either get the langley line in thingy from AML or have someone do another line in thingy for me. For U5 I want to go with the LME49720 or LME49860. Also same LMEs for the master. I had someone recap the master and a couple channels, put the new chips in U5 and master and it's a huge improvement.

But I'm not sure about the power supply? Those of you who have changed many opamps in a scorpion, what did you do with the PSU? I'm going to recap it at least, but I was looking at GRS' website and they don't seem to change the basic PSU components... Will I need to make any changes to this thing or will it handle the whole 32channels? If not, is there an easy way to mod it?

thanks!
#103
9th August 2013
Old 9th August 2013
  #103
Gear interested
 
Joined: Jan 2008
Posts: 11

weary is offline
bumping an old thread... re the tac scorpion 1... I'm wanting to change U1 and U5 in the inputs (preamp and fader amp). For U1 I'll either get the langley line in thingy from AML or have someone do another line in thingy for me. For U5 I want to go with the LME49720 or LME49860. Also same LMEs for the master. I had someone recap the master and a couple channels, put the new chips in U5 and master and it's a huge improvement.

But I'm not sure about the power supply? Those of you who have changed many opamps in a scorpion, what did you do with the PSU? I'm going to recap it at least, but I was looking at GRS' website and they don't seem to change the basic PSU components... Will I need to make any changes to this thing or will it handle the whole 32channels? If not, is there an easy way to mod it?

thanks!
#104
9th August 2013
Old 9th August 2013
  #104
Gear maniac
 
Samuelito's Avatar
 
Joined: Jan 2009
Posts: 186

Samuelito is offline
Hi weary,

In my experience with this board, I only could use 5 channel with LME49720 in position U5 before I started starving the original PSU.
It's not that easy to tame the LME49720.
If you send some audio through the inserts (will go directly to the U5 position) you'll find probably no problems, but when using the line or mic input you'll notice the instability.
Try tracking 2 OHs through the modified strips and you'll see what I'm talking about.

After several years of playing with this board the best advice I could give you is, get the AML kit that comes with a recap and send the master to GRS for an upgrade. You won't waste your money and you'll have straight forward results.
If you still want to upgrade U5 position, try with OPA2134. It's not a drop in replacement for the TL072 but is as close as it can get, or so I read.

Get an oscilloscope and learn how to use it, you'll be needing that.
Good modding and happy reading.

Cheers,


Sam
#105
9th August 2013
Old 9th August 2013
  #105
Lives for gear
 
Joined: Sep 2004
Location: UK
Posts: 6,770

Matt Syson is online now
Hi
The previous post is incorrect in that 'simply fitting some LME49720 overloads the supply' is complete rubbish.
These faster chips need extra 'stabilisation' components to stop them oscillating in this particular circuit. This was not done, or at least not properly and as a result they will oscillate. For 5 of these chips to start to pull the supply down would have them in flames.
A TL072 draws about 4 milliamps and, you must read the data sheet for yourself but the LME part will draw about 10mA or thereabouts, so suggesting that an 'extra' 30 milliamps total will stall the supply is complete nonesense.
If you want to swap chips you MUST understand correctly ALL the issues.
Bandwidth, stability, quiescent current consumption, input bias currents, output drive limitations are among the more important aspects.
Matt S
Quote
2
#106
9th August 2013
Old 9th August 2013
  #106
Gear maniac
 
Samuelito's Avatar
 
Joined: Jan 2009
Posts: 186

Samuelito is offline
Hi Matt,

For sure I had almost no experience when I first swap the TL072 to the LME49720 and they were not properly stabilized, but I can assure you that I could only use 5 strips when I just swapped the TL for the LME.
I did exactly what I was told to do, and the results were as I mention.
That's why went back to the original state.
I think we've been through this and I learned my lesson. I'm just trying to help people not to make the same mistakes I did.
That's why I posted "In my experience".
Best,

Sam
Quote
1
#107
9th August 2013
Old 9th August 2013
  #107
Lives for gear
 
Joined: Aug 2004
Location: Germany
Posts: 3,589

living sounds is offline
Try the OPA2227. Very stable and - used within its limits - transparent chip.
#108
9th August 2013
Old 9th August 2013
  #108
Gear interested
 
Joined: Jan 2008
Posts: 11

weary is offline
Hi,

thanks for the input. I don't plan on tracking with this, just mixing, hence the need for line ins. I read about the SSM chip and the pad for line ins, and I'd like to remove all this and get a real line in. Also, I did a couple channels plus master already and it's a huge improvement, but when using the whole board, the difference is still minimal, my guess is that the other "stock" channels are overwhelming the improvements in the master.

To Matt, I'm pretty sure the LME's have been installed with some ceramic caps near their power pins, is this enough to keep them stable?

The SSM2015 says it uses about 12-16mA, the ne5532 is about the same, I'm looking for a more power economic dual opamp to make a line input out of, so that might counterbalance the power needs of the LME's. Otherwise, it's 32 channels plus groups and master, would mean 40+ opamps changed.. At 6mA more each, that's about a quarter amp more... Doesn't seem like much?

Living sounds, thanks for the tip, I'll look into it, but I've got a source for a bunch of LME's at a good price so I'm leaning towards those..
#109
9th August 2013
Old 9th August 2013
  #109
Lives for gear
 
Joined: Sep 2004
Location: UK
Posts: 6,770

Matt Syson is online now
Hi
The LMEs may need to have caps on the feedback as well as on the supply rails.
The TL0 chips are pretty docile and hardly need any special care to keep them happy, it is just that the performance is only adequate and you have to be careful where you use them. Fitting 'fast' chips will bring out gremlins from the circuit layout.
An extra quarter of an amp would not worry the power supply if it is in decent condition. Is yours the 'heavy duty' version with a fan or the 'lower power' version. Adding a fan if you don't have it would be a good move for longevity.
Matt S
#110
10th August 2013
Old 10th August 2013
  #110
Gear interested
 
Joined: Jan 2008
Posts: 11

weary is offline
thanks for the reply. According to the schems, the TL072's already have caps in the feedback path. Well, most of them anyway. Is that what you mean by adding a cap int he feddback? Or do you think I recalculate the F of the feedback RC filter and up the cap value to lower the frequency? I think that's what increasing the cap would do, at least from what I have read in topics about API preamps anyway.

As far as the PSU is concerned, I have to pull it out of the rack to see if it has a fan. I was planning on recapping it in any case, if there is no fan, I guess it won't be too hard to add one.

One thing I've been wondering is that I've seen the ads from GRS saying they change the rectifiers to fast ones, but I can't find any bridges this size that are fast/ultrafast or whatever. Is it just marketing BS? I did find some schottky rectifers from mouser.com but they are TO220 types and they are singles or duals, not 4-diode bridges. From pictures of their finished PSUs, they look like normal rectifiers to me. So, paying $300 for $30 worth of parts seems like a bit of a rip off to me. And I don't think the duals will work because they are 3 pin ones, with one cathode, so I think they can't be used for the negative rail? Or maybe they can? The singles can be used, obviously, but again, the pictures on GRS's website look like normal bridges to me, not fast/ultrafast or whatever they claim they are.

Bottom line, I can't justify the cost of the AML kits for what they give you (1000+ GBP for the whole board, or more with import taxes!) then add labor cost to that, and I'm weary (hmmm, that's my name, don't wear it out!) of GRS after reading about these fast rectifier bridges. But maybe I'm mistaken?
#111
10th August 2013
Old 10th August 2013
  #111
Lives for gear
 
Joined: Sep 2004
Location: UK
Posts: 6,770

Matt Syson is online now
Hi
You need to investigate WHY the LMEs seem to be misbehaving and sort that aspect out. Raw 'quiescent current draw' is NOT an issue for a handful of chips from the supply that usually attends a Scorp desk.
Fast rectifiers have value 'in isolation' but if the standard types do not make their appearance known at the desk, then there is no real reason to change.
It is like someone down your road having a loud party, if you can't hear it or see it without going over to their place, you can ignore it.
Regulators and power rail caps in the desk eradicate practically all effects from the supply.
Matt S
#112
10th August 2013
Old 10th August 2013
  #112
Gear interested
 
Joined: Jan 2008
Posts: 11

weary is offline
Hi Matt,
Well, I don't have any issue with the LMEs, so far, was your post meant towards Samuelito? I only have 4 in there right now. (2 in the recapped master, and 2 in the recapped channels). I'm pretty sure they have some ceramic caps on the rails, and I upped the main power rail caps on each module from 220uF to 470uF.
I think GRS says that new, low ESR caps can blow old "tired" rectifiers. I believe they use the word tired. Anyway, my point with that is that I can't find a bridge that looks like the ones in their ads and ebay items, that is a fast type. So it seems like false marketing to me, but maybe they don't want to show which type they use?
I am a little worried about power surge at turn on from doubling all the channel power caps.
Back to the LMEs, do you have experience with putting thee (or other, more power hungry chips) in a scorpion, and if so, have you had issues? Do you think the two 10,000uF caps be enough, or would an upgrade be necessary?

I like the fact that (according to Jim williams) these LMEs can be directly coupled.
#113
21st October 2013
Old 21st October 2013
  #113
Gear interested
 
Joined: Jan 2008
Posts: 11

weary is offline
back to the TAC! I am almost done with the recap/mods by now, and want to add a 1646 to the output. Since the 1646 has +6dB gain, to keep overall level the same, I thought of lowering the gain in the output amp. Seems it's set to about 16dB? Would changing the feedback R (and C) present any other problems? (I think that chaning the 68K to 33K would change the output from 16.7 to 10.4dB gain... according to the calculator here:
Convert Voltage Gain to decibels (dB)

Or am I better off with a 1:1 trafo for balancing?

Also, I noticed the feedback C are very small throughout the desk, I calculated the RC filter frequencies and they are very high, like 200-250KHz. If I were worried (or extra cautious) about stability, could I raise those C's to filter more of the highs above, say 100KHz?

For example, in the master/mix amp, I was going to put in 220pF or 330pF for C25 and C26, and C22 for C31 and C32.

So far, we've already redone the PSU (recapped and modded, and put a filter on the fan.. the heatsink had years of dusk caked on it). I didn't change the rectifiers like they do on ebay, I thought of putting in some fast rectifiers but I could only find those in singles/duals and not bridges, so I put it on hold for now. Each module's PSU caps were changed from 220uF to 470uF, and all new chips have 0.1uF caps on their power pins. The PSU seems fine with the new stuff so far.

thanks!
Attached Thumbnails
Best Opamps to replace TL072's in TAC Scorpion-II-s4000_output.jpg  
#114
21st October 2013
Old 21st October 2013
  #114
Gear maniac
 
Joined: Apr 2006
Location: Aveiro, Portugal

Aerophone is offline
Be careful with lowering the rc pole in feedback networks. You will introduce lots of phaseshift in the audio band and will lower bandwidth. As an example my D&R was designed in a way that they could pass a 40khz square wave without overshoot to keep it 100% clean in the audioband.
200-250khz its just about right, although my D&R consistently uses 600khz -3db point to achieve what i described earlier.

Jim Williams recommended 22pf for values up to 20Kohm and 10pf for values between 20k and 30kohm. You can use 8,2pf and 6,8pf onwards.
Seem like good practice to me.

I have successfully used Lm6172's, LME49720's, LT1358's, ADA4610-2's, AD8512's and opa2134's without stability issues. I just added local 0.1uf bypasses from power rails to ground as near as i could from the Opamp power pins.
#115
24th October 2013
Old 24th October 2013
  #115
Gear interested
 
Joined: Jan 2008
Posts: 11

weary is offline
Quote:
Originally Posted by Aerophone View Post
Be careful with lowering the rc pole in feedback networks. You will introduce lots of phaseshift in the audio band and will lower bandwidth. As an example my D&R was designed in a way that they could pass a 40khz square wave without overshoot to keep it 100% clean in the audioband.
200-250khz its just about right, although my D&R consistently uses 600khz -3db point to achieve what i described earlier.

Jim Williams recommended 22pf for values up to 20Kohm and 10pf for values between 20k and 30kohm. You can use 8,2pf and 6,8pf onwards.
Seem like good practice to me.

I have successfully used Lm6172's, LME49720's, LT1358's, ADA4610-2's, AD8512's and opa2134's without stability issues. I just added local 0.1uf bypasses from power rails to ground as near as i could from the Opamp power pins.
hmm, thanks for the tip. I figured it should be OK, 250K sounds awefully high for such an old board. I know API's used something around 80-100KHz, and from what I've read here and elsehwere, the TL072 are very "easy going" ICs, in that they don't oscillate as much as newer faster chips. Anyway that was my line of thought with lowering the LPF. Will let you know how it works out. The ICs you mention, are these specific to the TAC or to your D&R? I'd be interested to see a schemo of the input channel of your board if you have one to post.
thanks!
thanks!
#116
27th October 2013
Old 27th October 2013
  #116
Lives for gear
 
drtechno's Avatar
 
Joined: Aug 2006
Location: Rapid City, SD
Posts: 1,784

drtechno is offline
I would suggest just changing things the same if you like the way it behaves.
otherwise:
there is something to say about op amp feedback bypass cap.

People don't understand "why" it is there in the first place. second they think just simply calculate it. it does't work like that. this is why interfaces suck till they get modded. (wow, I found 2700pf bypassing op amps in motu, what magic I found changing it to the value usually seen with that op amp )

op amp change without changing the cap can give varying results.

so experiment, use: not installed vs. 6pf, 8pf, 10 pf, and/or 22pf.
__________________
and that is why they call me the Mad Scientist.....
#117
27th October 2013
Old 27th October 2013
  #117
Lives for gear
 
drtechno's Avatar
 
Joined: Aug 2006
Location: Rapid City, SD
Posts: 1,784

drtechno is offline
I for got to mention.

there are sever factors that this cap is sized. and some of them I seen is questionable. because it can compromise performance

its because of:

gain vs positive feedback (oscillation) due to too much bias for low +/-Vcc

gain vs loading effect of op amp in/out vs overall noise floor vs impedance differentials.

a lot of mistakes I see lately:

Relying on computer simulation for this value instead of selecting due to physical perimeters not available virtually: 1. stray inductance and capacitance in devices causing paralleled circuit across the whole die, leads and circuit lands. 2. power filtering and the sum of the differential noise and whether or not it was integrated as an internal noise reduction circuit via feedback networks. 3. Differential level error causing bias change 4. the linear response of the circuit or lack there of that over compensate and shifts the response in a logarithmic fashion in the wrong direction.
#118
31st October 2013
Old 31st October 2013
  #118
Gear interested
 
Joined: Jan 2008
Posts: 11

weary is offline
Thanks for the tips. Drtechno, I'd be interested in hearing more about the relations between bias/loading/impedance. If you have any interesting links to share, I'd appreciate it.

We ended up changing the feedback networks to give between 100-130KHz -3dB points, and it's working just fine for now. There is a weird noise in one channel, but I suspect that it is something oddball, as all the rest are fine.

Also, I can confirm, the big TAC PSU (the one with the fan), recapped and modded to remove the regulator sockets(regulators soldered directly to PCB) can handle the extra current from the LME opamps in a 32/8/2 Scorpion. This is the fader amp in every module, 2 opamps in the groups, the auxes, 4-5 in the master channel, etc. Also, the input channels were changed to a line only using the LME chip, so that's 32 more. main filter caps are 15000 uF (originals were 10000), all module filter caps were upped from 220uF to 470 as well.

thanks to all for the help!
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+ 
 
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
BenJah / Low End Theory
31
stuffradio / So much gear, so little time!
0
Jacklynn / Geekslutz forum
17
stickstome2 / High end
0
j xav / So much gear, so little time!
1

Forum Jump

SEO by vBSEO ©2011, Crawlability, Inc.