Rat with double opamp

billonious · Post by **billonious** » 20 Dec 2009, 18:21

it is another mod of rat, having an opamp output buffer instead of transistors. Perfboard, schematic and diy-file uploaded

culturejam · Post by **culturejam** » 20 Dec 2009, 19:09

Moved to the appropriate forum.

Nice job!! I've often thought about doing something similar to this. But now I don't have to.

The only thing is that it definitely won't sound quite the same as a Rat with out the externally phase-compensated op amp (LM308, LM301, TL070, CA3130, etc).

DWBH · Post by **DWBH** » 20 Dec 2009, 19:28

Now make that 2nd opamp identical to the one used in the OCD and you got yourself something truly beautiful

billonious · Post by **billonious** » 24 Dec 2009, 08:32

DWBH wrote:Now make that 2nd opamp identical to the one used in the OCD and you got yourself something truly beautiful

I think that 2nd opamp is a middle-pass filter, right?

The TL072 can be replaced with a RC4558 for more vintage (warm) tone, although, the first one does a more modern-thick distortion. These 2 opamps are the best choices for my ears. I tried TL082 that does a "grunge" tone, similar to the 70's fuzz. NE5532, OP275 are terrible for such high gain cirquits. They fit in low gain overdrives only.

Dirk_Hendrik · Post by **Dirk_Hendrik** » 24 Dec 2009, 18:59

how much difference is it really going to make... replacing a FET buffer by a FET opamp buffer?

2% in output level..

And a sonic difference by using a FET opamp in the clipping stage instad of a BJT.

Now what will be the audible difference? The clipping stage or the buffer.

I'll put my money on the clipping stage 'kay?

billonious · Post by **billonious** » 25 Dec 2009, 09:57

Dirk_Hendrik wrote:how much difference is it really going to make... replacing a FET buffer by a FET opamp buffer?

2% in output level..

And a sonic difference by using a FET opamp in the clipping stage instad of a BJT.

Now what will be the audible difference? The clipping stage or the buffer.

I'll put my money on the clipping stage 'kay?

Rat has a "theoritical" maximum gain of 3200 (from gain-pot 150 k and 46 Ohm resistor to ground) making the opamp clipping the signal (when a power chord of 400mV is amplified x 3000) , before it passes from diodes. As a result, the audability of rat depends heavily on the (1st) opamp. Experimentally, TL072 & RC4558 clip gently when other opamps (even with better specifications) don't. It is not accidental the fact that 95% of distortion/overdrive boxes use TL072 & RC4558 (they are not the only cheap opamps around). As another user in this forum has mentioned, companies don't design opamps for overdriving purposes, so we have to try them and keep which it is musical in ears. Theoritically (again), an output buffer has a very limited effect on the signal.

Dirk_Hendrik · Post by **Dirk_Hendrik** » 26 Dec 2009, 17:50

I thank you for your kind and complete answer. Much appreciated.

however,
My questions you've so nicely answered did have a cynical tone since I'm fully aware of the points you mention.

billonious · Post by **billonious** » 17 Sep 2010, 22:01

having improving my design skills I tidied up the perfboard and the schematic and made a back-side layout (traces are less confusing)

back-side perfboard

: rat nude perf.gif (16.93 KiB) Viewed 4281 times

front-side perfboard

schematic

: RATschematic.gif (10.14 KiB) Viewed 4281 times

diy file for modification

rat layout.diy: (12.98 KiB) Downloaded 176 times

stringsthings · Post by **stringsthings** » 19 Sep 2010, 16:00

Dirk_Hendrik wrote: ... a cynical tone ...

that's a clever marketing slogan for a rat !

DahakaNishoba · Post by **DahakaNishoba** » 19 Sep 2010, 16:29

billonious wrote:having improving my design skills I tidied up the perfboard and the schematic and made a back-side layout (traces are less confusing)

Really nice....never built a RAT.....will try this one it when i got the pots i ordered (at least all 3 here are 100kB)
will try with a 082 and NE dual opamps (have a lot of them here)

anyway is this layout verified ? u do actually tried ?

DrNomis · Post by **DrNomis** » 25 Sep 2010, 13:31

billonious wrote:
Dirk_Hendrik wrote:how much difference is it really going to make... replacing a FET buffer by a FET opamp buffer?

2% in output level..

And a sonic difference by using a FET opamp in the clipping stage instad of a BJT.

Now what will be the audible difference? The clipping stage or the buffer.

I'll put my money on the clipping stage 'kay?
Rat has a "theoritical" maximum gain of 3200 (from gain-pot 150 k and 46 Ohm resistor to ground) making the opamp clipping the signal (when a power chord of 400mV is amplified x 3000) , before it passes from diodes. As a result, the audability of rat depends heavily on the (1st) opamp. Experimentally, TL072 & RC4558 clip gently when other opamps (even with better specifications) don't. It is not accidental the fact that 95% of distortion/overdrive boxes use TL072 & RC4558 (they are not the only cheap opamps around). As another user in this forum has mentioned, companies don't design opamps for overdriving purposes, so we have to try them and keep which it is musical in ears. Theoritically (again), an output buffer has a very limited effect on the signal.

Dr Nomis:

All an output buffer does is lower the output impedance of a circuit,my guess is that it will be something like about 100 ohms,so you won't get any loading effects from long instrument cables,the end result should be a brighter sound....

I used to have a Proco Rat 2,the thing had a tendency to self-oscillate at high gain settings,especially when you turned the filter control up,not surprising given the maximum theoretical gain of 3200.....

billonious · Post by **billonious** » 02 Oct 2010, 14:12

DahakaNishoba wrote:
billonious wrote:having improving my design skills I tidied up the perfboard and the schematic and made a back-side layout (traces are less confusing)
Really nice....never built a RAT.....will try this one it when i got the pots i ordered (at least all 3 here are 100kB)
will try with a 082 and NE dual opamps (have a lot of them here)

anyway is this layout verified ? u do actually tried ?

oopss, forgot this thread. There is a mistake at voltage divider. R7 is 47k (not 100k)

billonious · Post by **billonious** » 03 Feb 2011, 17:54

Rat was redesigned to offer:

1. buffer with constant & low output impedance, after what Micester suggested in a post
2. reversed phase of output signal to limit oscillation, as the signal of input cables interacts negatively with the signal of output cables (in the non-inverting buffer version of Rat, I noticed that oscillation levels vary according to the distance of cables)
3. improved supply cirquitry and protecting diodes for opamp input

Looking in datasheets, I ended up with RC4558 from TI, because the low slew-rate (1,7V/uS) and the high input-impedance (5M) make it close to the LM308. I think to use a higher value capasitor (200pF) for the fist opamp's feedback to mimic better LM308' slew-rate.

the perfboard is built and verified. I hope there isn't any typo.

EDIT: perfboard re-uploaded

Greg · Post by **Greg** » 03 Feb 2011, 23:51

culturejam wrote: The only thing is that it definitely won't sound quite the same as a Rat with out the externally phase-compensated op amp (LM308, LM301, TL070, CA3130, etc).

I have to agree with this..
It will probably still sound great, but much of the character of the RAT seems to be tied to Slew Rate Distortion.

AFAIK there are no dual opamps that offer external compensation.. so building a true RAT, there'd be no real advantage or reduced part count using an opamp buffer.

I'm not trying to take away from this design billonious.. it's a cool design..
I've tried the RAT circuit with other opamps and it can still be a great Distortion.

Greg · Post by **Greg** » 04 Feb 2011, 02:23

DrNomis wrote:All an output buffer does is lower the output impedance of a circuit,my guess is that it will be something like about 100 ohms,so you won't get any loading effects from long instrument cables,the end result should be a brighter sound....

I used to have a Proco Rat 2,the thing had a tendency to self-oscillate at high gain settings,especially when you turned the filter control up,not surprising given the maximum theoretical gain of 3200.....

A couple of thoughts...

Lower output impedance isn't all the buffer does... the buffer's input impedance affects the efficiency of the passive filter before it.

The low slew rate of a RAT in standard configuration limits high frequencies. That's why it can normally get away with such high gain without oscillating.
I've never seen a stock one oscillate.. but I haven't had a RAT 2 either.

billonious · Post by **billonious** » 04 Feb 2011, 07:34

Greg_G wrote:
DrNomis wrote:All an output buffer does is lower the output impedance of a circuit,my guess is that it will be something like about 100 ohms,so you won't get any loading effects from long instrument cables,the end result should be a brighter sound....

I used to have a Proco Rat 2,the thing had a tendency to self-oscillate at high gain settings,especially when you turned the filter control up,not surprising given the maximum theoretical gain of 3200.....
A couple of thoughts...

Lower output impedance isn't all the buffer does... the buffer's input impedance affects the efficiency of the passive filter before it.

The low slew rate of a RAT in standard configuration limits high frequencies. That's why it can normally get away with such high gain without oscillating.
I've never seen a stock one oscillate.. but I haven't had a RAT 2 either.

My motive for building this cirquit is "why using the obsolete & expensive 8-pin LM308, since I & many people have stocked 10-30 pieces of various double opamps with better tech that have 2 opamps in 8 pins?" .

About oscillation, this new inverting-output-buffer version deals with oscillations pretty better than previous versions. There is still oscillation when gain & volume & treble are set to maximum but only when so.

Greg · Post by **Greg** » 04 Feb 2011, 08:27

billonious wrote:My motive for building this cirquit is "why using the obsolete & expensive 8-pin LM308, since I & many people have stocked 10-30 pieces of various double opamps with better tech that have 2 opamps in 8 pins?" .

About oscillation, this new inverting-output-buffer version deals with oscillations pretty better than previous versions. There is still oscillation when gain & volume & treble are set to maximum but only when so.

As I said.. I think it's cool, but not exactly like a RAT due to the opamp.

You can readily buy an LM308 for less than a dollar, so to me they're not what I'd call expensive.

billonious · Post by **billonious** » 04 Feb 2011, 17:06

Greg_G wrote: As I said.. I think it's cool, but not exactly like a RAT due to the opamp.
You can readily buy an LM308 for less than a dollar, so to me they're not what I'd call expensive.

it is not exactly like the Rat, but I am trying to get it closer to the original Rat

, succefully I believe

culturejam · Post by **culturejam** » 04 Feb 2011, 17:36

I use LM301, CA3130, or CA3140 for my Rat builds. Sounds the same as LM308 to me (especially the 301), and they are still being made.

billonious · Post by **billonious** » 04 Feb 2011, 19:24

Greg_G wrote: As I said.. I think it's cool, but not exactly like a RAT due to the opamp.

let throw some light

I am writting down the slewrates from some popular double opamps

slewrates (V/microsecond) :

LM308 = 0.15 (ok, it is not double opamp)
tlc2262 = 0.35 - 0.55
RC4558 = 1.1 - 1.7
TLC272 = 2.5 -4
TS272 = 5.5
NE5532 = 9
TL082 = 13
tl072 = 16
OPA2132 = 20
OP275 = 22
OPA2604 = 25

TLC2262 seems to be the closer to LM308 and the best choise regarding low noise (12 nV/√Hz) contrary to the ridiculus hiss of LM308, but it is a bit more pricey (1.30€ at Banzai, 2.28€ at RS)

The old & classic RC 4558 seems to be a good compromise with slew-rate 1.1-1.7, very cheap (0.25€ at Banzai), with noise levels at 8nV/√Hz. A higher value of the feedback capacitor may lower its slew rate even more.

Other opamps have too high values I think, but I have tried TL072 and I liked much.