iRig (with DIY impedance matching cable?)

[davidallancole]

FB=Ferrite Bead

[PRR]

> I drawn the iRig circuit

Thank you for your very sharp eyes.

> WTF are the components marked as FB? Circuit breakers?

Ferrite Beads to reduce radio frequency signals (an iPhone is full of digital and radio signals).

> Can someone read the code for Q1? Is it a JFET, MOSFET or what?

If R5 is missing, then it has to be a JFET.

> The capacitors C1 and C2 form a band pass filter. Can someone specify those using filters theory?

Not without values.

And I bet it turns out that C1 cuts subsonics, C2 cuts supersonics. Technically it may be a 50Hz-15KHz bandpass, but that is such a wide band that we can usually consider it as a hi-pass and a low-pass which hardly affect each other.

> C3 is also meant to filter high freq signal.

Yes.

> For its acurate calculation we need to know what semiconductor Q1 is.

Maybe not. At higher voltages all the semiconductor devices are "high" impedance, circuit impedance is dominated by resistors. This FET is working at quite low voltages, so may not be very-high impedance, but not much less than the R4 VR1 iPhone resistor network.



R2 can't be 39 ohms. Could it be 39K?

R1 should be a resistor and probably larger than R1 because guitar signals are bigger than the working voltages we can get from an iPhone jack. The maximum clean JFET gate signal is probably 0.1V-0.3V, guitars often exceed 0.1V and can go as high as 0.5V.

[rafaeldmachado]

I couldn't find the Q1 datasheet for the code "G2P 2". (See http://www.qsl.net/dl7avf/smdcode/cg.html#TOC)

It would make sense if the marking were G27, that stands for the 2SK3749: [url]http://www.datasheetcatalog.org/datasheet2/e/0lx3acz5ocwjtl3flkiphx0ss43y.pdf[/

mat, could you please double-check this for me?

[rafaeldmachado]

Thank you davidallancole and PRR. Could think of ferrite beads as I never heard of them. LOL. thanks for explaining!

PRR, thank you for the invaluable comments.

I'll try to do some simulation with these value when in home (unfortunately only the next week).

[PRR]

> 2SK3749

Unlikely. That needs a positive gate voltage to conduct current. If R5 is not used then there is no way to get positve gate bias.

I simmed it with an old-standard JFET, and it barely worked. These "electret mike inputs" are a few K ohms to a few volts DC supply. In an iPhone the 3V battery power is most likely because it is ample for electret mikes. If we also assume that R4=4K7 and about 1K under the source, then there is less than 2V for the FET. My nominal 2V-8V JFET gets "stuck ON", can't pull down and will hardly swing up. Which just means I should get some new (model) FETs... all the $1 electret mikes have a 0.2V-1V threshold JFET, and this is probably an electret mike head-amp plus a resistor-divider to drop the 500mV of gitar to the 100mV that these little amps (and the iPhone mike input) can swallow.

It clear

[rafaeldmachado]

Maybe G2P is in fact 62P : the JFET J201, which the main characteristic is its low voltage polarization needs.

[mat]

I couldn't find the Q1 datasheet for the code "G2P 2". (See http://www.qsl.net/dl7avf/smdcode/cg.html#TOC)

It would make sense if the marking were G27, that stands for the 2SK3749: [url]http://www.datasheetcatalog.org/datasheet2/e/0lx3acz5ocwjtl3flkiphx0ss43y.pdf[/

mat, could you please double-check this for me?

I can check that tomorrow. I could also measure across the components if need be. I have meter to measure caps..

[PRR]

> JFET J201

Yes. That works beautifully.

> The variable resistance needs to be

There is a lot of variation from one FET to another, from one J201 to the next.

If you have the contract to make this iRig, every iRig must hit the Specified Gain; i.e. all iRigs are interchangeable, they all have the same gain (actually loss).

The Gain Spec may have been picked for the Average User.

Are you "average"?

In a DIY product, you could try 1K to 10K at "VR1" to get a good signal without overload.

Actually, I think VR1 should be omitted and R1 should be increased until overload never happens. That leads to the highest input impedance and (within reason) the best noise figure.

[rafaeldmachado]

Good news you have PRR.

I could'n hold myself and installed a PSPICE software in the company's laptop. The circuit has worked fine. I also tried omitting VR1 and increased a little bit the R4 to enhance available voltage for JFET. Also I made some essays in R1, and I had some doubts about the inputs to be used. What is the PSPICE guitar model you use, say, for a Strato and a Les Paul?

Just to be sure of no overload biasing I reduced the gain a little bit by changing R3.

[Skruffyhound]

This is going well 

[PRR]

> I had some doubts about the inputs to be used. What is the PSPICE guitar model you use, say, for a Strato and a Les Paul?

Every guitar AND guitarist is different; all should work well.

My observations of commercial and happy DIY amps suggest a "clean" input (iRig has distortion in software) should take at least 0.5V peak and preferably 1.0V peak without major wave-damage, should present well over 100K input impedance and often over 470K.

I'm guessing the "microphone" input of the iPhone can take a LOUD voice, which may be near 0.1V peak.

Here's values to give 1V to 0.1V with ~~250K input and <2% sweet THD.

[rafaeldmachado]

> I had some doubts about the inputs to be used. What is the PSPICE guitar model you use, say, for a Strato and a Les Paul?

Every guitar AND guitarist is different; all should work well.

My observations of commercial and happy DIY amps suggest a "clean" input (iRig has distortion in software) should take at least 0.5V peak and preferably 1.0V peak without major wave-damage, should present well over 100K input impedance and often over 470K.

I'm guessing the "microphone" input of the iPhone can take a LOUD voice, which may be near 0.1V peak.

Here's values to give 1V to 0.1V with ~~250K input and <2% sweet THD.

Shouldn't you put an output resistance or something (L and C) after "stimulus". I mean, they could affect the freq response and even the total gain of the circuit.

Just for being more linear in lower frequencies (for playing a 5-string bass through, for example) I would suggest we rise C1 a little bit from 20nF to 120nF.

Someone willing to give it a try in a real prototype?

[Skruffyhound]

Point me to the schem and parts values you have settled on and I'll breadboard it and test with my iPhone. Good work

[PRR]

Prepare your iPhone cable.

Measure voltage on MIC port. ***

Add (my) R4 4K7. Measure voltage on MIC port. ***

Add J1 = J201 (do not substitute recklessly), R3 = 1K to 1K5, R2 = 33K to 47K, C3 about 500pFd, C2 about 100pFd.

Measure voltage on MIC port. ***

Do whatever is done so you can hear the iPhone's MIC signal. (This may force you to figure and wire a headphone jack.) Put your finger on Q1 Gate. The buzz from room power wiring should be very audible.

Add R1 as 330K or 470K. C1 0.02u-0.1u is not strictly needed for guitar, but is a wise idea if other (leakier) sources may be used. Plug turned-down guitar. Strum, raise volume, listen. If it seems really too-soft at normal guitar volume setting (allowing for whatEVER gain control is possible inside the iPhone), remove R4 and reduce R1 toward 100K. ***

*** Record and report these values and voltages.

[Skruffyhound]

Ok, I'm quietly on the way with this.
I have 2.69 V on the mic port.
I have breadboarded the whole circuit, but am wondering about

room power wiring

I was thinking we would power off the iPhone, but since that pin is also the input I'm finding that confusing, perhaps you could clarify.
Are we using an external battery or is the mic port also the power source?

This is what we've got. I have the output from the phone running out of right through my breakout box and into a headphone amp and this works fine with music played from the phone.
I've inserted guitar at stimulus, mic port at "out". Would someone be so kind as to draw with a crayon what's connected where in the rest of the circuit so I can slap myself hard on the forehead tomorrow.
Paul, after I'm connected right I'll remove some components and give you those voltages.

[PRR]

"room power wiring"

This is a quick functionality test.

I assume you work in a room, not out in a field. If you have a high-gain audio amplifier, and touch the input, it BUZZZZ!es loudly.

You ever touch your cord-tip to see if your guitar amp is alive?

[Processaurus]

Thanks very much to everyone involved in this, this could be a handy travel accessory, and not just for iphones/ipod touch's.  Someone might have said it already, but the circuit could be applied to laptop mic inputs as well, for running software like Guitar Rig, etc.  Not sure if those computer mic inputs are typically stereo (tip and ring, on the 3.5mm phone plug), but adapting it would just be a matter of getting the plug wiring right.

Here's a 4 conductor 3.5mm plug that'd work for the iphone at Mouser:
http://www.mouser.com/ProductDetail/Kobiconn/171-7435-EX/?qs=sGAEpiMZZMv0W4pxf2HiV7X8OZ%252bjzZHiSUEC9jNX59s%3d

[rafaeldmachado]

Ok, I'm quietly on the way with this.
I have 2.69 V on the mic port.
I have breadboarded the whole circuit, but am wondering about

room power wiring

I was thinking we would power off the iPhone, but since that pin is also the input I'm finding that confusing, perhaps you could clarify.
Are we using an external battery or is the mic port also the power source?

This is what we've got. I have the output from the phone running out of right through my breakout box and into a headphone amp and this works fine with music played from the phone.
I've inserted guitar at stimulus, mic port at "out". Would someone be so kind as to draw with a crayon what's connected where in the rest of the circuit so I can slap myself hard on the forehead tomorrow.
Paul, after I'm connected right I'll remove some components and give you those voltages.

Skruffyhound, the power for this circuit comes from iPhone itself. It is the way electret mics work, they need to be polarized before any sound pressure can be measured (It is even modeled as a FET and a bunch of caps and resists).

We are using the same principle here: instead of polarizing the mic, the voltage coming from iPhone polarizes our JFET, which is then excited by the guitar signal (instead of air pressure).

So, (sorry, i forgot my crayons...) the circuit is connected to the iPhone at one side and to the guitar at the other. If you noticed in my last circuit schem drawing, there is a dashed box. This box is our circuit, the rest are the iPhone and the guitar.

Another thing, just in case you need, J201 = MMBFJ201.

[Skruffyhound]

Well, some news. I've got processed signal coming through, and it's not bad. Some pops on certain settings and a bit of noise, not much on delay, but the amp adds a bit. I am running on a 3G and I actually had to update iTunes to get an OS update to get the app over the weekend, so I'm a bit out of date. Changing the settings from "ultra low" to "low" latency helped a bit, so it's not just the circuit.
1.33V on the mic port with the full circuit (including Raphael's 100k in front of C1) my values are all the same as Raphaels except C1=22n and C3=470p.
I'll try to give a more full report in the next couple of days. You are still very welcome to get your crayons out and mark where the off circuit connections should be so I know that I haven't made some idiot mistake. I'm better with pedals

[PRR]

> 1.33V on the mic port

Take out R4 4K7. I still think that's mis-guided.

If strong guitar overloads the input, increase R1.