All transistor Rat

[soggybag]

As a study problem I had the idea to make an all transistor Rat. Here is what I came up with. Looks like it's in the ballpark but has a few questions.

- I need to get about x2000 gain from Q1 and Q2. My idea was to get x10 from Q1 and x1 to x200 from Q2.
- The filter low pass filters R8/C4 and R9/C5 seem like they provide a lot of Rat's sonic flavor. I'm sure if what I have is the same. Seem like the GAIN pot will affect the filter in my design.

[PRR]

> ....get about x2000 gain from Q1 and Q2.

Then that is ALL your distortion. Q2 can't make even 3V, 3V/2000= 1.5mV, a VERY tiny guitar signal, about 40dB down on an average good pluck.

What does the rest of it do?

> Seem like the GAIN pot will affect the filter in my design.

Probably. But is that a flaw or a feature?

I wonder if Q1 will bias happy with its Gate near +4V.

[imJonWain]

You could also make a simple discreet opamp out of 3x trans (Boss style) and put that in place of the lm308.

[idy]

Gain pot affecting tone is a desirable feature of rat or mxr distortion + or.... High gain means lots of clipping, sharp edges, but there is a hump as you push gain: less bass woof, less treble screech.

Many would consider this interactivity handy, even essential. Electrosmash website has a nice collection of freq response vs. gain graphs.

[Scruffie]

You could also make a simple discreet opamp out of 3x trans (Boss style) and put that in place of the lm308.

Did this many years back, not quite how I'd have gone about it now, but perhaps it'll inspire.

[antonis]

Ahhh... Mitchell again..
(always 2 or 3 projects going on simultaneously..)

1. I presume you are familiar with single BJT CE amp (like Q2 stage) maximum gain in conjunction with power supply margins..
For your info, a directly grounded Emitter (in the mean of no R8/R9) can't exhibit gain more than 20 X PS voltage (ideally..!!)
(actually much less due to next stage impedance Collector loading..)

For particular Q2 stage, R10 severely dominates R11 so gain can't be higher than X10(*)..!!
(far away from X200)

2. As Paul pointed above, Q1 bias point should be OK for a Mosfet but not so good for a JFET..
About 4V on Gate call for 5.5V (say) on Source, hence 5.5mA channel current..
I'm pretty sure that Drain can't handle 55V drop..
(minor note is R5/R6 high values)

(*) brute analysis:
For Q2 Collector sitting at 4.5V, quiescent current should be  450μΑ, hence intrinsic Emitter resistor about 55R..
The above resistor is added in series with R8 & R9 equivalent parallel resistance (43R)..
Q2 stage gain is [R11 // (R10+diode dynamic resistance) / (55 + 43)]..
(less than 1000 divided by less than 100..)

[soggybag]

Thanks Paul, you always cut out the heart of the matter in the fewest words. My takeaway is that I don't need to match the gain the of the original circuit I just need enough gain to get those diodes to clip, and Q1 is not biased correctly.

Thanks Scruffy, looks like you did exactly what imJonWain was describing, nice work! If I understand this correctly, it's a discreet op-amp Rat with 18v option. Have you tested this?

Thanks Atonis your insights are always illuminating! I'll read this a few more times. Seems like I need a little work on Q1 and Q2. I have a couple more projects in the works! I'm trying to discussing one per week!

[Scruffie]

Thanks Scruffy, looks like you did exactly what imJonWain was describing, nice work! If I understand this correctly, it's a discreet op-amp Rat with 18v option. Have you tested this?

Almost, it's Joe... Davidson's? Diode compression discrete op amp.

Yeah, I tested it, might even have an old PCB you can have if I didn't toss them in my previous move, nudge me with a PM on Monday to search if you're interested, has 3-way diode switching and I think I even put DIP switching for the diode in the op amp and it fit in a 1590B.

[soggybag]

Thanks Scruffy, looks like you did exactly what imJonWain was describing, nice work! If I understand this correctly, it's a discreet op-amp Rat with 18v option. Have you tested this?

Almost, it's Joe... Davidson's? Diode compression discrete op amp.

Yeah, I tested it, might even have an old PCB you can have if I didn't toss them in my previous move, nudge me with a PM on Monday to search if you're interested, has 3-way diode switching and I think I even put DIP switching for the diode in the op amp and it fit in a 1590B.

Awesome! I remember when the diode compression op-amp was a big discussion here! I never tried it. I have enough trouble getting regular op-amps to cooperate!

[aron]

I really liked the diode compression op amp! I tried it in several pedals.

[puretube]

Wasn`t the virtue of the RAT the limited slew-rate of the opamp it used?

[antonis]

Seems like I need a little work on Q1 and Q2.

If you are fine with about 40dB gain, just implement Distort-o-Matic XI input stage configuration..
If you replace Q1 JFET with a 2N5088/89 BJT (and lower R1 value) overal gain is raised at about 52dB with corresponding input impedance issue, of course..



For more gain, you can increase R5 value (or even delete it - which isn't recomended for stability issues..)

[Vivek]

Wasn`t the virtue of the RAT the limited slew-rate of the opamp it used?

I feel it goes this way

Rat has a peculiar sound

It's Opamp has low slew rate

Some guy who wanted to publish an analysis of the Rat chanced upon this quirk

And attributed more to this quirk than it rightfully deserved

What if I say almost every distortion pedal is slew rate limited to some extent

TL072 has slew rate of 13-V/µs

Suppose we have a distortion stage with gain 800, and feed it with 1vp guitar signal of 3khz, that demands a slew rate of 15.1V/μS. Any higher gain or frequency will demand even more slew rate. If the Opamp is not capable of slewing at that rate, the output will be slew rate limited to some extent.

[Ben N]

Why not a single MPSA13?

[Fancy Lime]

Intersting project. If more gain is needed from transistors, bootstrapping comes to mind. I tried something similar, albeit geared towards medium gains here:
https://www.diystompboxes.com/smfforum/index.php?topic=123610.0#msg1169798
By using normal antiparallel Si diode clippers instead of the Zeners, tacking on a rat-style tone control, and removing some bass from the gain stage, this concept may be worth a try for a "discreet Rat".

Andy

[GFR]

From the datasheet of the LM301A "With proper compensation, summing amplifier slew rates to 10 V/µs can be obtained.", but later on the same data sheet it says SR typical 0.25 V/µs. The datasheet of the TL071 says "16 V/µs typical". For comparison, a 741 has 0.5 typical, 0.25 is really low.

The compensation cap is external for the LM301A, not built-in like in TL07x. Also the inputs are bipolar rather than FET.

I don't if and how much all this affects the sound of the RAT.

I've found this graph with a rule of thumb about distortion in opamps:

https://eng.libretexts.org/Bookshelves/Electrical_Engineering/Electronics/Operational_Amplifiers_and_Linear_Integrated_Circuits_-_Theory_and_Application_(Fiore)/05%3A_Practical_Limitations_of_Op_Amp_Circuits/5.04%3A_Slew_Rate_and_Power_Bandwidth

I remember when I first tried a RAT in the 80's, when using a power supply instead of a battery, it was much quieter (in terms of rejecting hum from the power supply) than Boss pedals with much less gain, that was a great feature.

[Vivek]

I've found this graph with a rule of thumb about distortion in opamps:

https://eng.libretexts.org/Bookshelves/Electrical_Engineering/Electronics/Operational_Amplifiers_and_Linear_Integrated_Circuits_-_Theory_and_Application_(Fiore)/05%3A_Practical_Limitations_of_Op_Amp_Circuits/5.04%3A_Slew_Rate_and_Power_Bandwidth

The graph above makes some assumptions that are mentioned elsewhere in the article

A) The power supply is +/- 15V

B) Therefore the maximum unclipped Vp that we can expect is 13V (assuming 2V dropout from rail)

C) The slew rate is 0.5V/uS  and expected output is 12Vp <<<<<<<<< These are the important specs that give the 12v/6631Hz data point


If any of above change, the graph will change.


Hence that graph is not a rule of thumb for all Opamps and all circuit configurations.

[antonis]

If more gain is needed from transistors, bootstrapping comes to mind.

With all the respect Andy, bootstrapping has nothing to do with any stage gain..
(at least, in the way gain is determined..)

[Fancy Lime]

If more gain is needed from transistors, bootstrapping comes to mind.

With all the respect Andy, bootstrapping has nothing to do with any stage gain..
(at least, in the way gain is determined..)

Why not?

[antonis]

Why not?

'Cause bootstrapping deals with input impedance (voltage dividing BEFORE gain stage) where gain deals with signal amplification AFTER any previous attenuation..

Do you presume that Q1 stage in 250 biturbo should have lower voltage gain without R3 bootstrapped..??

P.S.
Are we talking seriously or what..??