Help Debugging Faulty Component/Stripboard on Tillman Pre-Amp

[lukemtesta]

I need help with some debugging if you could lend a hand.

I used the vero layout of the tillman amplifier http://4.bp.blogspot.com/-W0Xq9uuEy4A/U5Jch8L7Y3I/AAAAAAAAHsM/FHdS0KRV6Ko/s1600/A+Pair+o%27+Preamps.png to create some schematics for breadboarding http://imgur.com/a/jEuVg.

Parsing an input through the system produces nothing at the FET output (DC cap at drain), but bypassing the system (connecting jacks directly) work verifying my jack sockets and connections are fine.

I've noticed an ammeter beeps (looking for shorts) when testing points A and B in the diagram, but if IIRC the impedance across the cap should be oo at DC - Does this imply a short in the FET or breadboard? I have verified 8.2V at the supply rail at A (supplied by a battery). I have also tried biasing the FET at 2.6V and mixing the input with the guitar output with no luck.
Am I along the right lines? Any help would be awesome!

Note: The caps are electrolytes with their negative terminals facing towards ground. IIRC the 2n7000 and caps trace back to my student days, so are not verified working (can't see obvious signs of damages on the caps).

[robthequiet]

Welcome to the madness.

I've noticed an ammeter beeps (looking for shorts) when testing points A and B in the diagram, but if IIRC the impedance across the cap should be oo at DC - Does this imply a short in the FET or breadboard?

I suggest removing power from the circuit and use your multimeter on the Ohms function to measure for continuity there. Because caps charge you may see the numbers moving. If you see a dead short, then we need to check out the wiring, etc.

There is a debugging thread you can visit that might get you going, It may seem a bit tedious but it's a good place to find answers.

In the meantime, maybe you can post an image somewhere and post a link to it so that we can have a look?

[duck_arse]

are you using the 2N7000 mosfet, or the J201 Jfet? the jfet will bias correctly with a gate to ground resistor as shown in your tillman schem, but a mosfet needs a positive bias voltage on its gate to get-up and mambo correctly.

oh, I just saw you have 3k3 for your gate resistor. that should be more like 3M3 in a jfet circuit, to prevent loading the signal source. at the very least, you could change that 3k3 to 3M3, and add another 3M3 between gate and +9V, and it might be close enough to correct bias to get something out of the mosfet.

[edit :] ..... and an input cap, to block the mosfet gate DC bias from the signal source.

[lukemtesta]

sorry about the late reply and thanks for offering help. I'll post up a pic in a bit of the breadboard (always a bit embarrassed about my arrangement for some reason).

Didn't realise the cap would charge during a continuity test, you learn something new I can confirm I am using a 2n7000 but cannot confirm the manufacturer. Most datasheets confirm the 2n7000 is a MOSFET. If the Vgs threshold is set at 3.0V, would the MOSFET 'switch on' when the voltage drops below that threshold? Do I need to change the 3M3 to a pull-up? Have to admit I never really understood how MOSFETs work - Are they essentially voltage switches, where a gate voltage below (thanks for clearing that up) the threshold would force the device to draw power from the rails? My memory of those lectures is poor to say the least.

Sorry I made a mistake in the gate resistor value - I was meant to write 3M but wrote 3k out of habit. I can confirm I am using three 1M ohms resistors in series. I'll pop in a DC cap at the input, any recommended values?

[lukemtesta]

Just cleared up on some notes - Updated myself on Depletion and Enhancement MOSFETs. Looks like this is an enhancement, which means I need to bias the gate at +Vgs - Is it safe to assume I can bias it at Vs/2 to allow room room for the input to mix without clipping? Still, it doesn't really explain why my (luckily almost correct) test of biasing Vgs at ~2.5V failed if a possible min Vgs was .8V.

p.s. loving using hardware again. A career in the dark-side (software) almost lost me all my hardware knowledge!

Resources
http://www.electronics-tutorials.ws/transistor/tran_6.html
https://www.onsemi.com/pub/Collateral/2N7000-D.PDF
http://www.electronics-tutorials.ws/transistor/tran_5.html

[lukemtesta]

Breadboard - http://imgur.com/a/WQzId

I've biased the gate at 2.3V at its producing sounds! I can't tell on the 1W amp aurally whether the tone is changing with the pre-amp, but if I remove the bias the signal level drops considerably, which is enough proof to me its almost there.

I was thinking to adjust the power of the buffer - Does a gate potentiometer make sense here?

Finally last problem i see is it works when not grounding the jack sockets properly, but if I ground them with the circuit the output cuts out, presumably because the MOSFET is switched off. Any ideas?

[robthequiet]

Seems to be going well -- have you tried biasing at 2V?

As for the grounding of the jacks, the negative (sleeve) lug on each jack should meet on the ground rail of the board. The tagboard layout shows only the output ground point, but we may be presuming the the input would be grounded through the enclosure? Try connecting the input jack ground lug to the same trace on the breadboard as the output jack, as opposed to the ground lug from the other jack.

[lukemtesta]

I will try 2V, need a moment to sit down and calculate the resistor values needed.

Originally I had the ground terminals connected to a common ground (yellow and green crocodile clips), but doing so would cause the output to be grounded. Any idea what could be causing that? I don't understand it because the input should be oscillating between the bias and ground.

[robthequiet]

I should clarify: The grounding should eventually go to one place, what we call star grounding, and that is to the metal chassis itself. Then the ground point for all connectors would go to the input jack shield. Here is a link to the Tonepad site that has an article on outboard wiring:

[lukemtesta]

Thanks for the link, but maybe I still don't follow: Is my understanding right... all connectors will eventually be going to the same ground. In a breadboard circuit, that would mean connecting ground (battery negative terminal) to all components as the ground terminal, the input and output jack connectors? That seems to be backed up by the offboard wiring pdf in the link

[robthequiet]

Yes, the ground is the reference for all parts of the circuit, so on your breadboard it's the ground row, usually the blue line.

[lukemtesta]

Cool, yeah so when I originally shared the same ground between the buffer and the input/output jacks, the drain is dragged down to 0v for some reason. If I leave ground floating on the input/output jack, I get a buffered signal from my guitar to the 1W amp. Any ideas what could be going wrong?

[robthequiet]

I think you may have a crossed wire somewhere. Start at the input on the breadboard from the input jack tip lug and trace the circuit to make sure it is not somehow also connected to ground.

Edit: I think it's a simple wiring issue. Another test is to run your clip leads connecting the input and output connectors to make sure they agree, as in a hard bypass of the circuit. Then we've isolated the issue to either the board or the jack wiring.

[lukemtesta]

Thanks I'll give this a try tomorrow when I get a chance.

Edit: I think I have done this 'hard bypass' test before to test my socket connections were right. I could go straight into the 1W amp via two jack sockets (Guitar -> socket -> socket -> Amp) with no noise implying it was grounded okay.

[robthequiet]

Say, correct me if I'm wrong, but do you have the 3M resistor going to the +V of your circuit rather than ground? I am looking at where the battery red lead comes into the breadboard and it seems to be on the same row as where the resistor would go to ground. Could be my eyes.

[lukemtesta]

Yup, moved it to a pull-up

[anotherjim]

Afraid you cannot really "pull-up"  the gate resistor to the + supply. That's pretty much hardwiring the MOSFET into on state. Not very linear.
FETS are awkward in linear mode, hard to give a one size fits all solution.
My preference, although it adds parts and trouble, is a trim pot wired as a potentiometer across the supply providing an adjustable  gate bias voltage. The middle wiper of the pot supplies the gate via the 3M gate resistor. Adjust to taste, but it will be somewhere 4 to 6v at the drain as a guide. Find the loudest/clearest sounding setting by ear. 10k to 100k trimpot value.

You really must have a DC blocking capacitor at the input. The guitar is a ground facing source and will pull your "pull-up" voltage right down toward ground without the cap. Because of the really high input impedance, the input cap can be low value - no need for electrolytic. The usual choice would be a film type cap. 10nF is plenty big enough.

[lukemtesta]

Thanks for the feedback @anotherjim. I mocked up a quick VCO with an arduino to avoid using my guitar as a test generator - http://imgur.com/a/WQzId gate = pin 2 of MOSFET.

If what you are saying is true, the cap is probably why I don't get an output when switching from the VCO (which works) to the guitar - Tad bit of an unfair test with the 2.5V biasing *slaps head*

I'm a tad confused on ground facing source. Is there a resource I could look at?

[anotherjim]

I'm a tad confused on ground facing source. Is there a resource I could look at?

I lapsed into jimspeak there. The real term is "Ground referenced".
In the guitar, the "common" reference is what we call 0v/ground. The signal voltage is generated by the pickups referenced against ground. Being AC, the generated signal is swinging some millivolts positive of ground and also negative of ground.

Ground reference is the standard interconnect between audio devices. That way, we can design with known assumptions, input and outputs of those devices in a way that does not cause one device to alter the internal circuits of another.

Your preamp already has an output capacitor that blocks the operating voltage of the circuit and allows the final output to settle back to ground reference.

There is an easy automatic bias method for MOSFET amplifiers, and that is to connect the gate resistor to the drain resistor/drain connection. Because the drain resistor sees some +supply voltage, the gate voltage will settle to some +value dependent on the qualities of the particular transistor and source and drain resistor values. This is often perfectly adequate, but I don't like it too much - the bias is what the combination of resistors and transistor wants. I don't think that's necessarily the best sounding or most efficient bias.

I recommend a study of the AMZ mosfet. This is a fully featured design, you wouldn't want to do much more with a single transistor boost than this does.
http://www.muzique.com/schem/mosfet.htm

[lukemtesta]

I noticed I was still not biasing the gate properly. I tried popping a pot divider across the gate with a coupler and it worked with the guitar great, thanks for the suggestions - I'll calculate the cut-off and double check i'm not filtering the input later.

I definitely need to address this gate biasing. I'll give the pot and gate-drain tweeks a crack for completion. Haven't come across the gate-drain design so its gotta be given a go. Thanks for the suggestions/help and will keep you updated!