Millenium Bypass - Please help!

[zpyder]

Hey there.  So I'm baffled as to why my millenium bypass circuit is not working.  I'm building it on a protoboard using 2N5485 and 1N914 both bought from Small Bear Electronics.  I've tried both the ProCo Rat style (http://www.muzique.com/lab/true_bypass.htm) and the tonepad style (http://www.tonepad.com/project.asp?id=35 Type 3) to no avail.  With both the best I can get is an LED that is always on regardless of switch position.  I'm using a DPDT.  I have checked and double checked both schematics to my protoboard and can't figure it out.  The tonepad version does not make sense to me as to exactly how it works (1N914 reverse-biased across the transistor and +9v?) but the Rat version does make sense (except why two diodes instead of just one?)

I would greatly appreciate any advice someone could give me.  Someoen must have gone through this before.

A couple leadering questions that may help:

I am trying to do JUST a millenium bypass - ie the "effect" that is utilized when switched on is nothing more than a straight connection.  It is not connected to +9v.  Should this be a problem??

Though I have tried all combinations of transistor leads, I still don't REALLY know for sure which is the gate, drain, or source.  Any good way to test this with a multimeter/battery?

THANKS!!!!

[Mark F]

Hi. Welcome to the Forum. You should visit GEOFEX.com and read all about the Millenium and all of its variations. I would suggest you read all of RG's stuff. Hope that helps and Good Luck!

[brett]

Hi
The JFET milleniums are nice robust circuits, so there's no problem with the design.
You should check that the pinout is as you expect.  From memory, most of the 2N548X series make them DrainSourceGate when looking at the "flat" front. (But don't trust my memory - I don't).
Also, check your 1N4148 - they are fairly easily damaged.  If it short-circuits, it may stay on all the time, which is your symptom.

In the Millenium 2, a MOSFET is used, and the gate is held at at a voltage that is determined by the leakage through a high leakage device (1N4148/1N914) and a low leakage device (BE-junction or a 1N400X).

[R.G.]

The Millenium Bypass is a sensitive detector of the difference between open circuits (that is, several inches of air) and a 1M or lower resistance to ground.

The first thing to do is to get the Millenium bypass working by building just the Millenium and making sure that it works by touching a 100K to 1M resistor between ground and the control line. When that works, you can put it into a circuit and be sure that the Millenium works.

The way the Millenium 1 works is this:
- a JFET is connected between +9V and the LED. JFETs are normally ON, in that they conduct unless you pull the gate lower than the source. So if you leave the JFET and LED alone, the LED comes on because the JFET comes on. This may happen quite slowly, as the gate charges up by leakage through the gate/channel junction.

- The reverse biased diode between the gate and +9V makes the LED go on faster. Diodes leak a little, and the leakage is quite constant over a wide range of reverse votlages, so the reverse biased diode is effectively a constant current source, but a very tiny one. Typical values of leakage are 5 to 25 NANO-amperes. A nano-amp is one one-thousandth of microamp, or a millionth of a milliampere.

- The diode/JFET/LED circuit comes on much faster than just the JFET. And "on" is the ordinary state of the Millenium when nothing is connected to the gate of the JFET. If you connect a resistor of less than 1M ohm between gate and ground, the resistor sucks the gate voltage to ground essentially instantly. This forces the JFET to turn off. In the Millenium 1, the "off" may not be perfect, and there may be a dim glow from the LED. A parallel resistor or a series diode on the LED forces it to be fully, truly off. The slight imperfection results from the fact that some JFETs can be slightly on if there is as much as 5 to 10 Volts negative on the gate, and so the "off" may not be perfect. The parallel resistor or series diode clean this up.

I am trying to do JUST a millenium bypass - ie the "effect" that is utilized when switched on is nothing more than a straight connection.  It is not connected to +9v.  Should this be a problem??

It may be if your straight connection has no resistor to ground. Hook a 1M or less resistor from the signal line to ground, then try it.

Though I have tried all combinations of transistor leads, I still don't REALLY know for sure which is the gate, drain, or source.  Any good way to test this with a multimeter/battery?

Yes. A JFET is conceptually a sideways diode. Imagine very wide diode with one connection (the gate) to the anode/arrow side, and two connections to opposite ends of the cathode/band side. These are the drain and source. The drain/source path conducts through the diode "sideways" if you don't do anything to the gate, and the channel resistance is just the resistance of the semiconductor material. When you reverse bias the anode/gate, the electrical field inside the diode sucks the charge carriers out of a region near the gate. This effectively narrows the amount of the channel area that can carry electricity, just like squeezing a garden hose cuts down on the amount of water than can get through.

If you connect up your meter on ohms, diode setting position if it has one, then for an N-channel JFET, you will see a "diode" kind of indication between the gate(anode) of the JFET and both the drain and source, because both the drain and source are the "anode" of the JFET. In most JFETs, the drain and source are interchangeable, so distinguishing between D and S is of secondary importance.

However, there is another, and simpler way to find the pinout. Google for your JFETs part number and "datasheet" and go look.

[pupina]

Dear R.G., i have the same problem...my millenium bypass 2, built whit the BS170 mosfet and the 1N4007 as low leakage diode stay always on...i have to use that circuit in a Bad Stone clone by Topopiccione project. That's happened for the first time in my Little DIY experience.
I'm going crazy 'cause i cannot able to turn off the led...only when i measure the current whit my DMM from "change" lug of the switch and ground it work...
What i have to do whit it?
Please help
Carlo

[R.G.]

The Millenium 2 is a different circuit. It works by using a MOSFET which is switched off by the resistor to ground.

The pull up on the gate of the MOSFET is the same: a reverse biased diode acting as a constant current source. There are a few problems which people always seem to have with the Mill2. Most common is that they get the pinout of the MOSFET wrong.

MOSFETs are different from JFETs as used in the Millenium 1. The MOSFET in the Millenium 2 does NOT have interchangeable drain and source terminals. If you get the MOSFET drain and source mixed up, there is a body diode in the MOSFET which will let the LED stay on all the time. But if that were the case, it would not turn off when you used your multimeter.

The other thing they do wrong a lot on the Millenium 2 is to get the diodes mixed up on the input. The high leakage diode must go between the +9V (for negative ground circuits and N-channel MOSFETs) and the gate of the MOSFET. The low leakage diode must go between the gate and ground.

The low leakage diode is only there as a static electricity protector. If you are having problems with the circuit working, remove the low leakage diode until you get it to work properly, then put the protection diode back in.

I would be very surprised if what is wrong with your Millenium 2 is not one of:
- MOSFET installed with the pins connected improperly. You simply HAVE to check the pinout, and for the BS170, there are TWO pinouts, depending on which manufacturer made the chip; so you MUST look up the datasheet for the one you have, or try it both ways.
- high and/or low leakage diodes connected incorrectly
- soldering problems.

The Millenium 2 (and 1 for that matter) is a very simple circuit - only four parts, including the LED and resistor. But it relies on some very sensitive actions of the high leakage diode and the FET to work properly, so it is unforgiving of any mistake.

[pupina]

Thanks R.G.
I added a 1M res from out of the circuit to ground and everything was OK!
now the led turn on/off when i switch the pedal, like every time
Ciao

[zpyder]

Yea RG, thanks... I placed a 100 ohm resistor from the signal path to ground and VIOLA!!  my frustation is over (for a while)  Sweeeeet...

thanks again
zpyder