Inductor for AMZ pickup simulator

[bartimaeus]

I want to add a pickup simulator to the input of a lovepedal eternity clone, so I can stick it after buffered pedals and still get a warm tone (otherwise it just sounds like a tubescreamer).

Naturally I thought of the AMZ simulator, which has served me well in the past: http://www.muzique.com/lab/pickups.htm

But to save space, I'd like to use one of those resistor-sized axial inductors. The 42TM019 transformer suggested by AMZ is nice, but too big for this build.

I'd appreciate suggestions for a suitable inductor value. I couldn't find any on the 42TM019 datasheet: https://www.mouser.com/datasheet/2/449/XC-600134-1212477.pdf

I searched through the old threads, but everyone seems to be using the 42TM019 transformer.

[Rob Strand]

AMZ has a small table of inductance values,

http://www.muzique.com/lab/tips.htm

He only gives > 2H since the inductance on this type of coil is quite variable.

Those common axial inductors (RF chokes) probably only get up to the 100's of mH, so 0.1H or 0.2H.
Much smaller values.

Inductors tend to get larger as the inductance goes up.   Transformers tend to use high permeability cores like iron
and have no air gaps in the core.   That will give the highest inductance for a given size but the side effect is
very larger tolerances in the value.   The other way to get more inductance is to put more turns on the coil
but if the coil size is fixed the only way you can get more turns is thinner wire.    There's a limit to how thin
the wire is and those transformers are certainly getting close to that limit.

What you might be able to do is find a slightly smaller size transformer.  The inductance of one of the windings is
likely to be less than the Xicon TM19.   However instead of using one winding you can put the two windings in series
to get the maximum inductance.  If the transformer has equal primary and secondary windings the series connection will
give you four times the inductance of the single winding.    It's a less wasteful connection.   What you have to get right
is the phase of the series connection: one way will give you four times the inductance and the other way will give a very low
inductance, the inductance "cancels" out.

Here's a slightly smaller core (EI14 vs EI19 core), also sold by mouser at some point,
https://www.datasheetarchive.com/pdf/download.php?id=1faa99c8f23d51bc41e5a83ef9787fbaa611fb&type=P&term=EI14

https://www.mouser.com/c/passive-components/audio-transformers-signal-transformers/?q=EI14

Perhaps try, then look at, the 42TL parts,
https://www.mouser.com/c/passive-components/audio-transformers-signal-transformers/?m=Xicon

Unfortunately no info on the inductances.

OK cool found this,
https://www.soundonsound.com/forum/viewtopic.php?f=24&t=25926&view=unread


42TL022 1.5K Ohms .56 Henry
42TL021 4K Ohms 1.5 Henry
42TL018 7K Ohms 2.6 Henry
42TL019 10K Ohms 3.75 Henry
42TL025 17K Ohms 6.4 Henry
42TL017 20K Ohms 7.5 Henry

and

[PRR]

You want the inductance of a pickup.

The pickup is wound of thousands of turns of the *thinnest* commercial (windable) wire on a mostly-iron core. It will be hard to make something smaller.

Yes, a closed core will give more inductance at the same turns, or give the same inductance on a smaller core. You might get half the linear dimension this way.

Pickups are Alnico and plain iron. RadioMetal, Mu-Metal, FINEMET®, VITROPERM®, and others have higher permeability, but low overload level and much higher prices and minimum orders. The "best" may require winding techniques not suitable for the many-many-many turns we need at audio.

BTW: enameled wire comes along in about 1908. Before that we used thread-wrapped, cotton or silk, single or double.

[Rob Strand]

As for what the 42TM019 actually is, you might need to go through some posts,

Here,
https://www.diystompboxes.com/smfforum/index.php?topic=23188.0

Tim Escobedo is saying,

"Mouser's 42TM019 has a primary with inductance of 5H to 6H"

Which places a big range of uncertainty on AMZ's > 2H.


I know this has come-up in the past and I've even posted stuff on it.  I just can't remember it off the top of my head.

I found a few more quotes for the 42TM019 and 42TM018 and they were stating 5H and 4.7H.
So we are talking roughly 5H as a target.



The closest 42TL transformers to get 5H are the 10k input impedance models.   However, if you look at the excel table image you can see on average they are less than 5H for a 10k input impedance.    In order to get 5H you need about  2411 turns.

The turns are listed in the 42TL datasheet,
https://www.mouser.com/c/passive-components/audio-transformers-signal-transformers/?m=Xicon

Not all 10k input impedance transformers have the same turns (or inductance).  According to the excel table closest model to 5H just using the primary is 42TL024 but Mouser don't keep it and it's not in the list.   If however you look at the sum of the turns for the 42TL019 it is very close to the 2411 turns we need.   So if you got a 42TL019 and wired the secondary in series with the primary (and you wired the phase correctly) it would be very close to 5H.   You could also use only the primary of the 42TL019 alone to get a lower inductance option.

The hanging question is going to the smaller 42TL series of any help at all in the first place?
 

[bartimaeus]

Wow, this is incredible help, thank you both!!!

I was hoping to use one of those Tayda inductors: https://www.taydaelectronics.com/inductors.html

But I didn't realize how TINY their inductance is compared to those audio transformers.

And out of the 100,000+ inductors available on Mouser, only ~50 have a value between 1H and 10H. And all of those are large transformers, mostly the Hammond ones you'd use in an amplifier.

Now I understand why AMZ recommends the transformers, I guess it's the closest we can get to a pickup without just using one!

Here's a slightly smaller core (EI14 vs EI19 core), also sold by mouser at some point,
https://www.datasheetarchive.com/pdf/download.php?id=1faa99c8f23d51bc41e5a83ef9787fbaa611fb&type=P&term=EI14

https://www.mouser.com/c/passive-components/audio-transformers-signal-transformers/?q=EI14

Perhaps try, then look at, the 42TL parts,
https://www.mouser.com/c/passive-components/audio-transformers-signal-transformers/?m=Xicon

Looking them over, they're all about 0.5" square.

The 42IF106 is the smallest I think, at about half the size of the 42TM019. But it has only ~150 turns, and lists an inductance of just 680μH!
https://www.mouser.com/datasheet/2/449/XC-600014-1212509.pdf

That's pretty far off from the thousands of windings in a pickup (didn't think of that before, thank you for explaining PRR)!!

The closest 42TL transformers to get 5H are the 10k input impedance models.   However, if you look at the excel table image you can see on average they are less than 5H for a 10k input impedance.    In order to get 5H you need about  2411 turns.

The turns are listed in the 42TL datasheet,
https://www.mouser.com/c/passive-components/audio-transformers-signal-transformers/?m=Xicon

It seems the 42TL019 is about 0.25" smaller than the 42TM019. Not exactly the size-savings I was hoping for, but it'll certainly help!!

And looking at the datasheet (https://www.mouser.com/datasheet/2/449/Yuetone_XC-600127-1212403.pdf), it seems like the 42TL012-RC would give 4200 total if wired in series. Perhaps that's the best available at its size?

[idy]

But I didn't realize how TINY their inductance is compared to those audio transformers.

And out of the 100,000+ inductors available on Mouser, only ~50 have a value between 1H and 10H. And all of those are large transformers, mostly the Hammond ones you'd use in an amplifier.

Now I understand why AMZ recommends the transformers, I guess it's the closest we can get to a pickup without just using one!

Now we see (partly) why inductors, one of the three theoretical passive components (R,C,L) appear in very few of our circuits. They get physically big fast. But what about using an old pickup? Don't most of us have some old pickup lying around...I don't know if you can/should take the magnets out?

[Rob Strand]

And looking at the datasheet (https://www.mouser.com/datasheet/2/449/Yuetone_XC-600127-1212403.pdf), it seems like the 42TL012-RC would give 4200 total if wired in series. Perhaps that's the best available at its size?

It looks good to me.   

Something that hasn't been mentioned is the inductance a transformer not only has a wide tolerance but the inductance depends on the signal level.  Even if you have good equipment to measurement the inductance, if you measured the inductance at different signal levels you will get different inductance value.   If the smaller transformers are likely to be worse for a given level.   It's not something to be overly concerned about for the pickup sim.   It does mean thinking about things too precisely becomes academic - the uncertainties for transformer inductances are large by nature.

If you look at a real  guitar pickup it is much larger than those small transformers.  The magnetic path goes through the air.  What that means is when you make pickups the  inductance is relatively consistent for a given model.  Between models of pickups the inductance varies all over the place - that's part of what makes different pickups different!

[ashcat_lt]

IDK anything about gyrators, but aren't they supposed to be active inductor emulator type things?  Is there a way to implement one to accomplish this?

Honestly, though, it's just an RL lowpass.  An RC lowpass will get you the same thing, and only requires two relatively small and readily available components. 

[Rob Strand]

IDK anything about gyrators, but aren't they supposed to be active inductor emulator type things?  Is there a way to implement one to accomplish this?

In an equalizer with gyrator we arrange the LCR network so the L is grounded and that allows for a gyrator which simulates a grounded inductor.  The circuit for that are relatively simple.   For the pickup sim the both ends of the gyrator are floating so you need what is called a floating gyrator.    There's many circuits for this but the people study these topics in detail have extracted the better ones over time.    The main beef is the circuits are relatively complex and they can also be noisy.

Honestly, though, it's just an RL lowpass.  An RC lowpass will get you the same thing, and only requires two relatively small and readily available components. 

In many cases that's true but the thing the pickup sim is trying to emulate is the interaction between the impedance of the guitar and the input impedance of the circuit.   The circuit could be non-linear so in order to capture the interaction correctly you need an inductor. 

The buffer + pickup sim inductor thing to simulate a pickup has a degree of wrongness to it as well.   For a guitar pickup connected to a circuit the tone of the pickup is *modified* by the loading of the circuit.      When we have a buffer + pickup we have the response of the pickup connected to the buffer and *as well as* the effect of the pickup sim inductor and the circuit.  There's some double counting going on.     The reason it works OK is the dominant effect is the pickup + circuit  and the pickup + buffer part doesn't contribute much.    To make the pickup sim more realistic we can also add capacitors and resistors but in the light of the double counting issue we would be better off *not* adding those parts.  So in the end the simple inductor works fairly well.

[ashcat_lt]

The circuit could be non-linear so in order to capture the interaction correctly you need an inductor. 

People keep saying that, but I'm still not buying.  Plug into a buffered distortion.  Fizz.  Turn down the T pot on your guitar.  Fuzz.  The important part is the filter and anything else is just splitting hairs.

The one thing that doesn't happen here is the way that the V pot on the guitar comes between the pickup and the pedal, which can help to get some treble back as you roll back that knob, and that will work opposite if you use an RC lowpass.  Course, it's just a moot point in this specific case for what should be obvious reasons.

On "double dipping", presumably the cutoff of the filter created by the pickup and a proper high-Z buffer is enough higher than that of the "pickup simulator" that it doesn't much matter.  That is, the simulator will have rolled off the frequencies so far at that point that the little extra bump and then extra steep drop off from the initial filter just won't be much noticeable.

[bartimaeus]

Something that hasn't been mentioned is the inductance a transformer not only has a wide tolerance but the inductance depends on the signal level.  Even if you have good equipment to measurement the inductance, if you measured the inductance at different signal levels you will get different inductance value.

Not sure how large this effect is, but it does seem to be one aspect that a simple RC filter can't replicate?

[Rob Strand]

People keep saying that, but I'm still not buying.  Plug into a buffered distortion.  Fizz.  Turn down the T pot on your guitar.  Fuzz.  The important part is the filter and anything else is just splitting hairs.

The one thing that doesn't happen here is the way that the V pot on the guitar comes between the pickup and the pedal, which can help to get some treble back as you roll back that knob, and that will work opposite if you use an RC lowpass.  Course, it's just a moot point in this specific case for what should be obvious reasons.

On "double dipping", presumably the cutoff of the filter created by the pickup and a proper high-Z buffer is enough higher than that of the "pickup simulator" that it doesn't much matter.  That is, the simulator will have rolled off the frequencies so far at that point that the little extra bump and then extra steep drop off from the initial filter just won't be much noticeable.

I guess the motive for a pickup sim comes about because pedals are buffered bypass.   Going to true bypass limits the need for it in the first place.

The pickup sim certainly has its limitations.  The biggest one to me is the pickup sim only has one inductance value.  Pickups are all over the place.  The pickup sim is a coarse approximation at best and yes it still doesn't do the volume knob thing.

Not sure how large this effect is, but it does seem to be one aspect that a simple RC filter can't replicate?

Well in the light of a single inductor approximating all pickups the tolerance isn't high on the list of concerns.

IMHO, the motivation for adding a pickup sim to emulation the interaction between the guitar and pedal.   In some cases that's a weak effect say on a chorus pedal and there's no real point using a pickup sim.   In some fuzzes the effect is fairly strong.   How closely an RC filter approximates the pickup interaction is up for grabs.   The only way to test the idea is to do an A/B test (guitar vs RC filter) but the results will only be valid for that pedal.   In the case for the inductor at least it starts roughly on the right foot.  At least it will get you fairly close to what goes on with a pickup without thinking about it too much.


As a second comment of inductor vs RC filter.  If the pedal following the inductor/filter has a non-linear impedance the non-linearity will case the filter cut-off to decrease with an inductor and increase with an RC filter.  So if there truly is interaction the inductor does the right thing and the RC filter does the opposite.  There's also differences in the low frequency attenuation.  If there is merely pickup loading due to a low input impedance then the inductor vs RC filter might show less of a difference.   (Neither the inductor or RC filter gets around the double counting issue.)