Reawakening: Timmy Verification Update

[matopotato]

Thought of replying in the thread, but was advised to open a new thread instead.
I have a Timmy on a breadboard and I am experimenting with various values, especially ICs.
MC1458 just edging out NE5532, but differences are not huge IMHO. Have an NJM4559 arriving in the upcoming week I hope.
I was following the layout referenced here https://www.diystompboxes.com/smfforum/index.php?topic=121481.msg1142928#msg1142928
And some kind of mix of schematics (there are a few around) beginning with http://drdfx.hu/en/jimmy/ (although I suspect it might originate from elsewhere). Then moving over to the one made by G. Tranter modified by Rob Strand labelled "Paul C. Timmy ~2014" (https://www.diystompboxes.com/smfforum/index.php?topic=121481.msg1142222#msg1142222)
And it implies a few adjustments.

1. Adding the C4 47p RF (?) cap, is it to remove disturbances (RF meaning Radio Frequencies?)
2. Change R7 1.5k to 1k (Treble R, Resistor Cap), what effect would it have?
3. The 3k3 to 10k pair in the second stage feedback, I understand would yield some more headroom
4. The 2n2 in the second stage feedback, what would it do?
5. Change out the A10k volume pot for an A25k + R10 18k, I read what it will do. As I do not have any A25k right now, is there some other combination of A50k + R or A10k + R that would demonstrated the effect? Or I just have to wait until I get my hands on an A25k.
6. If I change the D1 I had as 1N5817 in series to parallel, should I also change to 1N4001, or will they work the same? I do not plan to use the 100R as per the layout picture, but read that moving from series to parallel might increase the headroom a little bit.
7. These layouts and schematics are for V2, since then there is a V3 and I found a layout on another site, but it seems to be very much different from the V2 in here. Any thoughts on V3 vs V2 or perhaps that would be better as a topic for a different thread.

Thanks for all who contributed to sort out most things in the original Timmy Verification Update thread!

[ElectricDruid]

1. Adding the C4 47p RF (?) cap, is it to remove disturbances (RF meaning Radio Frequencies?)

Yes, exactly. It'll take any radio frequency noise the cables picked up to ground. Best place for it.

2. Change R7 1.5k to 1k (Treble R, Resistor Cap), what effect would it have?

Not much. It changes the maximum cutoff frequency (e.g. the minimum treble cut) of the R7/C8 filter slightly. 1K5 gives 10KHz, 1K gives 16KHz. Up there, you're not really going hear much difference.

3. The 3k3 to 10k pair in the second stage feedback, I understand would yield some more headroom

If both R8 and R9 were 3K3 and they both change, there's no overall gain change. Sorry, I don't know the original schematic that well.

4. The 2n2 in the second stage feedback, what would it do?

Provides a roll-off above 7.2KHz in tandem with R8/10K. Which is another reason why the R7 change is going to make no odds at all.

5. Change out the A10k volume pot for an A25k + R10 18k, I read what it will do. As I do not have any A25k right now, is there some other combination of A50k + R or A10k + R that would demonstrated the effect? Or I just have to wait until I get my hands on an A25k.

The important thing for the 25K+18K combo would be the ratio of the resistors, so if you wanted to use a different pot value, you'd have to scale the R10 appropriately to get the same effect. E.g. for 10K pot, you go down to 18K/2.5 = 7.2K, so 6K8 or 7K5 closest values.

6. If I change the D1 I had as 1N5817 in series to parallel, should I also change to 1N4001, or will they work the same? I do not plan to use the 100R as per the layout picture, but read that moving from series to parallel might increase the headroom a little bit.

I'd go with either the 1N5817 in series *or* the 1N4001 in parallel, but not both. The 100R turns C1 into a proper lowpass filter. In my view, without it, it's a bit pointless.
The series/parallel diode thing is something people argue about. Personally, I can cope with the small volt drop from a series diode in order to have the greater safety of knowing that it won't short out my power supply if I connect it up wrong (which is what the parallel diode does). After all, if I want more headroom, I can just run it at 12V or 18V or whatever.

7. These layouts and schematics are for V2, since then there is a V3 and I found a layout on another site, but it seems to be very much different from the V2 in here. Any thoughts on V3 vs V2 or perhaps that would be better as a topic for a different thread.

Sorry, I don't know - but I'm sure someone else will step in!

[matopotato]

Thanks a lot @ElectricDruid!

Quote
4. The 2n2 in the second stage feedback, what would it do?
Provides a roll-off above 7.2KHz in tandem with R8/10K. Which is another reason why the R7 change is going to make no odds at all.

So is the 2n2 more "useful" in a way?
Trying to read up on roll-off frequency and I am *guessing* the 2n2 will make the treble cut adjust 7.2kHz and above.
And R7 have the same effect but for 10kHz/16kHz and 7.2kHz being lower puts the R7 out of play?

The 100R turns C1 into a proper lowpass filter. In my view, without it, it's a bit pointless.

OK, 100R is back in the game

After all, if I want more headroom, I can just run it at 12V or 18V or whatever.

Might be that JRC4559 has a max of 18V on Vcc+, the caps can take 25V max so it might be ok. Not sure how "appreciated" it is by the IC to run on max, although I suspect there are some margin that will allow it to run on 18V. The rest seems to be fine with 18V as far as I could tell.

This is the V2 I am looking into making (link above, first post):



And this is the V3 I stumbled upon. (http://dirtboxlayouts.blogspot.com/2021/05/paul-cochrane-timmy-v3.html)


They seem to differ a bit.

[FiveseveN]

JRC4559 has a max of 18V on Vcc+

Maximum supply is ±18 V, i.e. 36 V so you're fine.

[ElectricDruid]

So is the 2n2 more "useful" in a way?
Trying to read up on roll-off frequency and I am *guessing* the 2n2 will make the treble cut adjust 7.2kHz and above.
And R7 have the same effect but for 10kHz/16kHz and 7.2kHz being lower puts the R7 out of play?

Yes, exactly.

The 100R turns C1 into a proper lowpass filter. In my view, without it, it's a bit pointless.

OK, 100R is back in the game

Jolly good!

Might be that JRC4559 has a max of 18V on Vcc+, the caps can take 25V max so it might be ok. Not sure how "appreciated" it is by the IC to run on max, although I suspect there are some margin that will allow it to run on 18V. The rest seems to be fine with 18V as far as I could tell.

Like FiveseveN said, typical maximum supply for op-amps is +/-18V. +/-15V (e.g. 30V supply) was standard for audio work. Our stompboxes are generally running on much lower voltages than these things were designed for.

[matopotato]

JRC4559 has a max of 18V on Vcc+

Maximum supply is ±18 V, i.e. 36 V so you're fine.

Thanks, wasn't sure how to interpret the datasheet.

[Rob Strand]

And this is the V3 I stumbled upon. (http://dirtboxlayouts.blogspot.com/2021/05/paul-cochrane-timmy-v3.html)
They seem to differ a bit.

I recently stumbled on some info that said the latest Timmys use the toggle switch to change the gain.  The diode switching has moved back to internal dip switches.    (Using the toggle switch for gain is something which has
been adopted by a few pedals now eg.  King of Tone, Rocket Pedals.)

(IIRC) also that the Bass and Treble pots are now boost, not cut when advanced clockwise.

FWIW, from what I can see that V3 layout doesn't exactly replicate a V3.

This post has a text file attachment summarizing V2 Timmy versions upto and including the 2014 V2.

https://www.diystompboxes.com/smfforum/index.php?topic=121481.msg1142460#msg1142460

[matopotato]

And this is the V3 I stumbled upon. (http://dirtboxlayouts.blogspot.com/2021/05/paul-cochrane-timmy-v3.html)
They seem to differ a bit.

I recently stumbled on some info that said the latest Timmys use the toggle switch to change the gain.  The diode switching has moved back to internal dip switches.    (Using the toggle switch for gain is something which has
been adopted by a few pedals now eg.  King of Tone, Rocket Pedals.)

(IIRC) also that the Bass and Treble pots are now boost, not cut when advanced clockwise.

FWIW, from what I can see that V3 layout doesn't exactly replicate a V3.

This post has a text file attachment summarizing V2 Timmy versions upto and including the 2014 V2.

https://www.diystompboxes.com/smfforum/index.php?topic=121481.msg1142460#msg1142460

Thanks Rob!
I think I will stay with the V2 I am working on then.
Yes, I read your version document. Very useful for my understanding where things are at.
Any thoughts on #2 and #4 as to using 1k if a 2n2 is applied anyway? I am currently thinking of dropping the 2n2 and using the 1k "only", but maybe the other way is smarter?
Or is 1k + 2n2 and the 10k pair so closely linked that it is either all or nothing?

Any idea regarding the 100R in the powerstage vs having it left out? I guess it was not part of the original V2 but would make some sense as per earlier replies.

[Rob Strand]

Any thoughts on #2 and #4 as to using 1k if a 2n2 is applied anyway? I am currently thinking of dropping the 2n2 and using the 1k "only", but maybe the other way is smarter?
Or is 1k + 2n2 and the 10k pair so closely linked that it is either all or nothing?

Points #2 and #4 are straight forward by they are laborious to explain.

2. Change R7 1.5k to 1k (Treble R, Resistor Cap), what effect would it have?

With a simple RC low pass filter like that the differences are quite small.   When the pot is fully open the filter is at its lower frequency.   If the pot value is left at 50k then the lowest frequency is set by the value of the treble cap.   With the treble pot closed the filter cut-off is at the highest frequency and if you have already chosen the cap value the resistor is used to set the the frequency.

The filter cut-off frequency in all cases is f = 1 / (2*pi*R*C)

Another factor for commercial pedals is what it sounds like with the treble control set at noon.   This is largely set by the capacitor and the pot value.  The resistance at noon is around 15% to 20% of the pot, so 7.5k ohm to 10 kohm.  Given people compare and use pedals with the tone/treble at noon it can influence the cap size chosen.   (And in this case the frequency when the pot is fully open is sort of a byproduct and is usually low enough for people's tastes.)

If you look at filter frequencies:
V1  R=1k5 and C=10n so f_upper = 1/(2*pi*RC) = 10.6kHz.
V2  R=1k and C=10n, so f_upper = 15.9kHz
V3_layout  R=2.2k // 50k = 2.1k, C=6.8n, f_upper = 11.1kHz

The feedback cap on the second opamp adds second a but subtle step filter.   It's not so much the specific resistor or cap values but the cut-off frequency both create.
V1 = no filter
V2 = 2.2n and 10k =>  f = 7.2kHz
V3_layout = 6.8n and 3.3k => f = 7.1kHz

If you compare the treble frequencies of V1 and V2, V1 has a lower cut-off.  However V2 adds a second filter on the last stage.  The combination of the two filters make the effective cut-off of V2 similar or lower than V1.   Where the second filter of V2 helps is when you back off the treble on V1 it only has one filter but on V2 the second filter is always present so it removes more nasty rubbish from the high frequencies.    In general the V2 scheme would seem better.

I'm reluctant to compare V3_layout as it doesn't seem to match the V3 PCB's I've seen.  What I will say though is you can see the cut-off of the second filter is at 7.1kHz whereas V2 is at 7.2kHz, almost the same.  You can see that different resistors are used in each case but the caps have been tuned on the V3_layout to keep the same cut-off.   The first filter of the V3_layout is different to V2 so that's where the similarities end (the difference will be small just the same.)

Any idea regarding the 100R in the powerstage vs having it left out? I guess it was not part of the original V2 but would make some sense as per earlier replies.

For most pedals, this resistor is used to filter 50Hz humm from non-regulated supplies.  For regulated supplies it provides only minor results.     What's more the issue is the Timmy hinges on clipping the second opamp.  If you have too much voltage drop on the supply, due to the added resistor, it will change the clipping.    This comes down to very specific things like the opamp, the loading on the opamp.  It's a fine detail which needs to be checked on an oscilloscope.

So I don't know if the 100 ohm is put there out of design habit or as a subtle means of tuning the sound for different opamps.   By the way adding series diodes also impacts on this.

If I had to guess I'd say the original Timmy set-up the gains so the last opamp didn't clip.  So adding resistors or diodes to the supply might upset that.

[FiveseveN]

maybe the other way is smarter?

Is your goal to outsmart Paul C?
The tone (treble cut) control is an RC low-pass filter: https://www.electronics-tutorials.ws/filter/filter_2.html
The role of R7 is to prevent the "R" part from going to zero when you turn the pot all the way down. 1K is 2% of 50K while 1.5K is 3%. You will not notice the 1% difference. You won't even have signal in the 10 KHz+ region and guitar speakers would not be able to reproduce it anyway.
The second op amp stage is a non-inverting amplifier: https://www.electronics-tutorials.ws/opamp/opamp_3.html
Adding C9 also gives it a low-pass function, see "The AC Op-amp Integrator with DC Gain Control" here: https://www.electronics-tutorials.ws/opamp/opamp_6.html The values of C9 and R8 set this filter's corner frequency, as described in the article. At 7.2 KHz I'd say it has more impact on hiss than anything else, but try it out on the breadboard (with the correct values) and see for yourself.
Learn the math, trust the math, it's the only way to make sense of how circuits behave.
PS: well, Rob kinda beat me to it while I was writing

[matopotato]

Any thoughts on #2 and #4 as to using 1k if a 2n2 is applied anyway? I am currently thinking of dropping the 2n2 and using the 1k "only", but maybe the other way is smarter?
Or is 1k + 2n2 and the 10k pair so closely linked that it is either all or nothing?

Points #2 and #4 are straight forward by they are laborious to explain.

And I *really do appreciate* you taking your time to explain this to me. It is a learning experience and I will be reading it through several times until it settles in. Thanks.

The V3 I think I have dropped, and the V3_layout might not be very true to the original, I haven't dug much deeper. There is something with V2 that stuck with me and after breadboarding it, this is the "V" I will build. Perhaps there might be more versions in the future, but for now I will settle with V2. (Famous last...)

And for me it boils down to 1k + 10n and 2n2 + 10k (in the pair) I think. (I was not aware that my first schematics was V1.)
In the original thread I think the V1 layout was sorted out as well to accommodate the V2 differences. Just have to sort out how to add the clipping diodes on the stripboard, but I am quite liberal with expanding the board size (and just did on a red snapper clone to keep components on the board rather than diodes in the DPDT) so that should be fine given the previous discussions.

Any idea regarding the 100R in the powerstage vs having it left out? I guess it was not part of the original V2 but would make some sense as per earlier replies.

For most pedals, this resistor is used to filter 50Hz humm from non-regulated supplies.  For regulated supplies it provides only minor results.     What's more the issue is the Timmy hinges on clipping the second opamp.  If you have too much voltage drop on the supply, due to the added resistor, it will change the clipping.    This comes down to very specific things like the opamp, the loading on the opamp.  It's a fine detail which needs to be checked on an oscilloscope.

So I don't know if the 100 ohm is put there out of design habit or as a subtle means of tuning the sound for different opamps.   By the way adding series diodes also impacts on this.

If I had to guess I'd say the original Timmy set-up the gains so the last opamp didn't clip.  So adding resistors or diodes to the supply might upset that.

OK. @ElectricDruid suggested putting the 100R in to make sense with the cap and a lowpass filter.
No I am leaning more towards setting a 1N4001 (I hope 1N4004S will be equivalent enough. Read that "higher 1N400x works as replacements) in parallel instead of the 1N5817 (currently in series on the layout) and dropping the 100R after all to keep the clipping in the diodes rather than the opamp. I will hopefully be getting 4559 this week to compare with 1458 and then I can settle which final config to go with.
I guess the 100R was added by others in their versions of schematics. Some have it some don't.

Again a huge thanks for your effort and input!

[matopotato]

maybe the other way is smarter?

Is your goal to outsmart Paul C?

Wouldn't dream of it. (Here I would have put a surprised/shocked emoji)
I was a bit puzzled as to why there was a 100R in some schematics and others without, but that is clear to me now. I think. 

The tone (treble cut) control is an RC low-pass filter: https://www.electronics-tutorials.ws/filter/filter_2.html
The role of R7 is to prevent the "R" part from going to zero when you turn the pot all the way down. 1K is 2% of 50K while 1.5K is 3%. You will not notice the 1% difference. You won't even have signal in the 10 KHz+ region and guitar speakers would not be able to reproduce it anyway.
The second op amp stage is a non-inverting amplifier: https://www.electronics-tutorials.ws/opamp/opamp_3.html
Adding C9 also gives it a low-pass function, see "The AC Op-amp Integrator with DC Gain Control" here: https://www.electronics-tutorials.ws/opamp/opamp_6.html The values of C9 and R8 set this filter's corner frequency, as described in the article. At 7.2 KHz I'd say it has more impact on hiss than anything else, but try it out on the breadboard (with the correct values) and see for yourself.
Learn the math, trust the math, it's the only way to make sense of how circuits behave.
PS: well, Rob kinda beat me to it while I was writing

Thanks! I will take up on more reading and digest what is in there.
This whole effects pedal thing/industry/hobby/life is definitely a journey. I guess one just have to keep an open mind and learn and absorb as long as it lasts.

[Rob Strand]

I was a bit puzzled as to why there was a 100R in some schematics and others without, but that is clear to me now. I think. 

Scanning over the schematics.  The schematics which match the originals have a reverse diode whereas the DIY versions add *series* diodes and/or 100 ohm resistors.   So  I guess the added parts come from a habbit of choosing a particular power supply configuration more what is actually done in the pedal.

[matopotato]

The 100R turns C1 into a proper lowpass filter. In my view, without it, it's a bit pointless.

OK, 100R is back in the game

Jolly good!

Sorry to disappoint, but I will likely drop the 100R after all. Authenticity? I don't know, I probably can't hear much difference, but will try to keep in line with the design made for the 4559.

[matopotato]

On a side-note. I tested 1458 vs 4559. I just got to the conclusion that I preferred the 4559 and just wanted to try the 1458 one more time, but accidentally put it in the wrong way.
Effectively connecting gnd to +9v and Vcc to ground instead, and of course the amps backward as well. It buzzed in agony so I removed it quite soon. But now it seems fried.
When the datasheet says +/- 18V that is still when you have gnd to gnd and Vcc to either of 9V?
So there is no "forgiveness" putting it backwards that the span of +/- would mean it could cope with my mistake, right?

[Rob Strand]

On a side-note. I tested 1458 vs 4559. I just got to the conclusion that I preferred the 4559 and just wanted to try the 1458 one more time,

Paul C initially used the JR4559 then used LM1458 for a while in V2 but ditched it and went back to the JR4559.  Perhaps the opamp clips occasionally.   The LM1458 doesn't sound bad clipped so I suspect it's a very subtle difference.

When the datasheet says +/- 18V that is still when you have gnd to gnd and Vcc to either of 9V?
So there is no "forgiveness" putting it backwards that the span of +/- would mean it could cope with my mistake, right?

Definitely no forgiveness.

The +/-VDC spec means +VDC applied to the opamp's +V power pin and -VDC applied to the opamp's -V power pin.  In essence 2*VDC is applied across the power pins.   So from the datasheet's perspective a 9V single supply is like +/-4.5V. (The +/- thing implies a dual supply.)

You need to be careful about single supply opamps because they spec power as +VDC and GND.  The VDC spec is the voltage across the opamp both power-pins.  Some single supply opamps specs will show both a +VDC and ground spec, and +VDC/2 and -VDC/2 spec.

It pays to look at the datasheet and try to decipher the details.

[matopotato]

Thanks Rob!
Another lesson learned.