Building a Tiny FET Impedance Converter

[JHanko]

This really doesn't fit into the stompbox category, unless the barrel of a 1/4" phone plug can be considered a tiny stompbox, but I think this group would understand what I need better than anyone else. I need to make an impedance converter to connect a guitar to a Sony UHF transmitter. Sony already makes a cable for this: http://www.pro-sound.com/p/sony-k1161-guitarmusical-instrument-adaptor-cable-for-wrt-8b-and-wrt-822b-body-pack-transmitters.html , but at $100, it's not going to happen. The transmitter has a balanced input and supplies 5V phantom power through a small 4 pin connector. The connector pinout is shown below. Donald Tillman posted an image and schematic of a preamp cable back in 2001. You can see it here: http://www.till.com/articles/PreampCable/index.html. I wish I could use his design as-is, but a couple of things prevent that. First of all, he designed the circuit with a 9v supply. He mentions that the preamp starts sounding bad when the supply drops down to 6v. I only have 5v to work with. His circuit provides 3dB of gain. I would rather it not provide any gain and have more headroom. I would like it to be as quiet and transparent as possible. Here are my questions:

1) Is there any way to adjust Donald's circuit to run on 5V instead of 9V?
2) His circuit is based on a Siliconix J201 FET, which is obsolete. Is there a better, smaller device better suited?
3) How to alter circuit for no gain?
4) In his circuit he used a 20K resistor on the input to protect from static. Could this be eliminated to save space if I used Neutrik silent plugs that ground the tip when unplugged?
5) Would it be better to just start from scratch instead of altering someone else's design.

I would appreciate any input on this. My soldering skills are up the the task, whether it be through hole or SMD. My circuit design understanding is poor though. That's why I have all the questions. Thanks!

Sony Cable:


Transmitter Connector:


Donald's Circuit:

[PRR]

Welcome!

Consider this:



CORRECTION - input cap 0.01 !!

[JHanko]

Thanks for taking the time to look it over! Would this still be using the J201 or something else? I was hoping to possibly eliminate 1 component, but this adds 3. Could all of this fit in a 1/4" phone plug barrel?

[Groovenut]

I think something like this will fit. .312" x .625"
You could definitely make it smaller if you go 2 sided.

I would suggest something like a MMBFJ201 for the fet. SMD size is 0805. I'm certain you can find al el cap short enough. If not you might want to consider mlcc but there goes all your mojo

[JHanko]

Thank you for the board layout. So, all of the component values that PRR posted would still be correct with the MMBF201?

[JHanko]

I always thought that aluminum electrolytic's were a no-no in the signal chain..

[Groovenut]

Thank you for the board layout. So, all of the component values that PRR posted would still be correct with the MMBFJ201?

I believe so, yes. PRR posted the Vp and IDSS. You can have a look at the data sheet for the MMBF5457 and the MMBFJ201 to check the specs.

[Groovenut]

I always thought that aluminum electrolytic's were a no-no in the signal chain..

For audiophiles yes I suppose. Though you can get audio quality electrolytics if you want. Otherwise what choice do you have in this situation 

[JHanko]

PRR posted the Vp and IDSS. You can have a look at the data sheet for the mmbf5457 and the mmbf201 to check the specs.

I'm sorry, but you lost me there...

[Groovenut]

If you have a look at the spec sheet of the Jfet you're considering using, there will be a spec for V_GS(off) aka V_P and for I_DSS
In PRR's post he listed both specs on the schematic above the fet

[JHanko]

If you have a look at the spec sheet of the Jfet you're considering using, there will be a spec for V_GS(off) aka V_P and for I_DSS
In PRR's post he listed both specs on the schematic above the fet

OK, I got confused as the datasheet had no listing for Vp. I an not versed in this terminology.

Edit: IDSS is in range, Vp possibly not.

[JHanko]

Some more questions: What type of capacitor for the 0.1uF on the input? I may put the output capacitor in the shell of the transmitter plug to reduce the components in the phone plug shell. Is that OK? What wattage for the resistors? If size was not an issue, what type of capacitor would sonically be most preferred? Thanks

[PRR]

> What type of capacitor for the 0.1uF on the input?

MY TYPO! Should be 0.01uFd. (0.005 might be ample.)

Guitar amplifiers are FULL of aluminum electrolytics, as are many-many studio boxes (and probably your wireless rig). Used with care they are inoffensive. I would not worry about the input cap. The output cap *should* be Al electro, because big ceramic and MLCC caps are *worse*, perhaps audibly worse.

[antonis]

You can use a non-polarised 4μ7 electro if you worry about virtual degradation...

[Groovenut]

Some more questions: What type of capacitor for the 0.1uF on the input? I may put the output capacitor in the shell of the transmitter plug shell to reduce the components in the phone plug shell. Is that OK? What wattage for the resistors? If size was not an issue, what type of capacitor would sonically be most preferred? Thanks

I designed the layout for a generic film type cap. To keep the height down you can fold it over to lay flat against the pcb after soldering. Placing the output cap in the shell of the transmitter should be fine. Standard 0805 resistors are 1/8 watt and should be fine for this. Most tend to prefer film caps in the signal path. Wima makes one that's 7.2mm x 7.2mm x 13mm with 5mm lead spacing MKS2B044701K00KSSD

[JHanko]

Just curious: If the input and output are passing the same frequency range, why is there such a huge difference in the capacitor sizes. Is it because the output is passing higher current because of the lower impedance? Why is a capacitor needed on the input? Is the capacitor on the output required because there is DC present there? Is the 20K resistor on the gate necessary? Sorry for all the questions, but I don't want to just build this. I'd like to understand it too.

[Groovenut]

Just curious: If the input and output are passing the same frequency range, why is there such a huge difference in the capacitor sizes. Is it because the output is passing higher current because of the lower impedance? Why is a capacitor needed on the input? Is the capacitor on the output required because there is DC present there? Is the 20K resistor on the gate necessary? Sorry for all the questions, but I don't want to just build this. I'd like to understand it too.

The input impedance is formed by the 3M//3M gate bias resistors. The input cap forms a high pass filter with those resistors. The input cap is needed to block the gate bias dc from whatever is attached to the input. The same goes for the output cap. It's blocking the DC that is on the source from whatever is attached to the output. The reason the output cap is large is mainly a frequency safety feature. We don't really know what impedance will be attached to the output so we have to guess. At a bare minimum you'll have the load of the connecting cable, so you cold figure based on that. However, the fet buffer isn't going to drive a very low load well anyway so a higher estimate is probably acceptable.

The 20k gate resistor is there for very basic ESD protection. I would recommend leaving that in.

[JHanko]

What is the frequency and order of the .01uF/3M/3M filter? I don't know the formula.

[antonis]

What is the frequency and order of the .01uF/3M/3M filter? I don't know the formula.

f = 1/2π*R*C = 0.16/R*C, R in Ohms, C in Farads.. R = 1M5 because is the parallel combination of the 3M resistors..
(which is about 10Hz in this case..)

[Groovenut]

What is the frequency and order of the .01uF/3M/3M filter? I don't know the formula.

It's a first order HPF. The formula is 1÷2PiRC (R in Ohm, C in Farad). So the 3M are in parallel giving 1M5, so 1÷1M5*.01uF*6.28=~10Hz

edit: Antonis was faster on the keyboard than I