fet Matching

[spoontex]

Hi,

I need to build this circuit for matching some fet's.

It's mandatory the use of 5m resistor? I only have 4m7. It would be fine?

Thanks.

[GibsonGM]

I believe you will be fine with 4.7M    All readings will be 'off' by the same amount - if any.

[FSFX]

Hi,

I need to build this circuit for matching some fet's.

It's mandatory the use of 5m resistor? I only have 4m7. It would be fine?

Thanks.

There is a 10k potentiometer there. The 5M resistor is completely pointless. You can get rid of it.

[antonis]

The 5M resistor is completely pointless. You can get rid of it.

I presume it's there to "simulate" Gate leakage current under real-world bias conditions..

P.S.
I'd suggest R.G.'s more accurate measurement circuit..
http://geofex.com/Article_Folders/fetmatch/fetmatch.htm

[FSFX]

I presume it's there to "simulate" Gate leakage current under real-world bias conditions..

But that is irrelevant as the circuit is measuring gate-source voltage. Any leakage current has no effect on the accuracy of this type of measurement.

Also I don't agree with your suggestion for a better tester. In fact a better matching tester would be to build something like double the original posted circuit but with a common G-S voltage connection and run it in a bridge configuration with the two JFETs to compare the tracking of I/Vgs over a range of Vgs values.

Otherwise, do the right thing and get a Peak DCA75 and plot the characteristics using its PC based curve tracing feature.

[GibsonGM]

Most in the DIY world don't have $350 to buy the Peak DCA75, LOL.   

[FSFX]

Most in the DIY world don't have $350 to buy the Peak DCA75, LOL.

Mine cost about £120 but that is nothing compared to a second hand Tektronix Curve Tracer.

Anyway there are quite a few simple ways of testing JFETs.

[GibsonGM]

Yes, there are several schemes for how to do it, I think that's what the OP is trying to do 

[spoontex]

I'm trying to match some fet's with the attach shematic on first post. BF245A, and I get very high values, comparing VGS 0V to -0,250V. I don't think it is normal. I stop on 0,750v.

[FSFX]

I'm trying to match some fet's with the attach shematic on first post. BF245A, and I get very high values, comparing VGS 0V to 0,250V. I don't think it is normal. I stop on 0,750v.

I would expect a typical pinch off voltage, i.e. a Vgs(off) for a BF 245A to be about -1.7 volts.

The datasheet actually says -0.5v to -8.0v but that probably covers the A, B and C versions and you really can't go by the wild values in the datasheet. That is why I test every JFET I buy and use.

[spoontex]

I'm trying to match some fet's with the attach shematic on first post. BF245A, and I get very high values, comparing VGS 0V to 0,250V. I don't think it is normal. I stop on 0,750v.

I would expect a typical pinch off voltage, i.e. a Vgs(off) for a BF 245A to be about -1.7 volts.

The datasheet actually says -0.5v to -8.0v but that probably covers the A, B and C versions and you really can't go by the wild values in the datasheet. That is why I test every JFET I buy and use.

My fet's are bad? Because If I compare with this table form masonaudio, he also uses bf245a and his values are more logical to me...

[digi2t]

Most in the DIY world don't have $350 to buy the Peak DCA75, LOL.

Mine cost about £120 but that is nothing compared to a second hand Tektronix Curve Tracer.

Anyway there are quite a few simple ways of testing JFETs.

Stompville JFET Matcher has done no wrong by me. The price didn't hurt either.

https://stompville.co.uk/cart/jfet-matcher-kit

[spoontex]

Most in the DIY world don't have $350 to buy the Peak DCA75, LOL.

Mine cost about £120 but that is nothing compared to a second hand Tektronix Curve Tracer.

Anyway there are quite a few simple ways of testing JFETs.

Stompville JFET Matcher has done no wrong by me. The price didn't hurt either.

https://stompville.co.uk/cart/jfet-matcher-kit

I built the schematic that you post. I need to match two fet's for a 1176 compressor build. It's enough the simple method? or I need to trace curves and all stuff that explain here:

https://www.axtsystems.com/joomla31/index.php/projects/11-cat-1176ln-audio-compressor/4-art-1176-fet-matching

[PRR]

> two fet's for a 1176 compressor build.

You need to plot and match the WHOLE curve. Otherwise the stereo image shifts left and right while compressing.

[spoontex]

> two fet's for a 1176 compressor build.

You need to plot and match the WHOLE curve. Otherwise the stereo image shifts left and right while compressing.

I try with this method, but the values are very high... 0V to -0,250V VGS almost 9V of difference.

https://www.axtsystems.com/joomla31/index.php/projects/11-cat-1176ln-audio-compressor/4-art-1176-fet-matching

[ElectricDruid]

> two fet's for a 1176 compressor build.

You need to plot and match the WHOLE curve. Otherwise the stereo image shifts left and right while compressing.

The article says that you need to match the FETs because one is used for the metering and the other for the gain reduction. No mention of stereo? Is this a mono version of a stereo studio original or something?

[spoontex]

> two fet's for a 1176 compressor build.

You need to plot and match the WHOLE curve. Otherwise the stereo image shifts left and right while compressing.

The article says that you need to match the FETs because one is used for the metering and the other for the gain reduction. No mention of stereo? Is this a mono version of a stereo studio original or something?

I will build stereo version. For stereo version you need to match four fet's. Anyway I think that this bf245a are fake... what do you think?

[Clint Eastwood]

Your measurements do not make much sense to me. Fake is a possibility, or you did something wrong. I say use the simple test circuit for Idss and Vgsoff. 
My experience with Jfets tells me they are like eggs, there are small ones and big ones, but the shape is always basically the same. So if you measure the same Idss, you can bet they have very similar Vgsoff, indeed they will have the same curve. Idss is the simplest to measure for a BF245a, just connect gate and source to negative of a 9volt battery, and connect the drain to positive through an amp meter. Voila.

[FSFX]

So if you measure the same Idss, you can bet they have very similar Vgsoff, indeed they will have the same curve.

That is not actually true. Well maybe it is OK to assume that as long as you qualify that statement by adding 'if they are the same type of JFET from the same manufacturer and the same production run'.

The relationship between Vgs(off) and Idss varies considerably depending on the geometry of the JFET and the doping of the semiconductor material.

The transfer characteristic for normal operation is defined by Vgs(off) and Idss which determine the two end points of the forward transfer conductance curve used for typical operation.

This curve can be derived from these two end-point parameters using the following equation:   Id = Idss(1-Vgs/Vgs(off))^2

[spoontex]

This is the seller:

https://www.ebay.es/itm/304157747953?var=603269161743

The fet matching circuit works with a genuine j201, so I absolute convinced that this bf245a are fake.