Mill Bypass 2 problems

[audioguy]

OK I REALLY apprecite everyones help in these matters. Im still having a heck of a time with this MBP2.
I've built a board exactly as shown here: http://www.geofex.com/Article_Folders/Millenium/mill2bd.gif
then I hooked up the battery and the DPDT switch just to test its switching before I went any further, and I get the light on constant. Ive swapped every component, and rearranged the MOSFET, but still I get the solid light. It was suggested that I remove the MOSFET all together but I didnt try that because isnt the G of the mosfet the - of my LED, and what would I plug my switch into?
It looks exactly like the diagram, using an NTE519, NTE177 for the diodes, and an NTE490 for my mosfet. I thought maybe it had something to do with using an Xwing switch so I made sure that my switch wiring looked like this- http://www.geofex.com/FX_images/stompsw.gif
Now my question also has to be, how many times can I change these things around before its all fried? The parts arent expensive, but a pain in the butt to get- outside of mail ordering.

Thaks again to everyone thats helped out!

Audioguy

[R.G.]

It was suggested that I remove the MOSFET all together but I didnt try that because isnt the G of the mosfet the - of my LED, and what would I plug my switch into?

I suggested that because it's a great way to determine whether the PCB all by itself is allowing the LED to light (it shouldn't) or whether something about the MOSFET was causing the LED to stay on. It was a real, valid diagnostic test. Before we can get down to testing the switching, we have to find out some way to make the LED go off.

I've built that board a number of times, and given a few hundred away like party favors, so the board is well tested. If I had to guess, you may have a problem of
(a) the high- and low-leakage diodes are not doing something right
(b) the pinout of the MOSFET is even wierder than normal, it being an NTE.
(c) PCB short, perhaps almost invisibly small.

You can replace things several times without frying them, with the possible exception of the MOSFET.

If it were me, I would remove everything from the board except the battery leads (+9 and ground), LED and current limiting resistor, and power it up from 9V. LED should be out. If it's not, the PCB has a short.
If the LED is out, use a bit of wire to short the MOSFET drain and source pads together. Does the LED come on when the wire is in place? If yes, then the MOSFET at least has some chance of working.
Then place the MOSFET and high leakage diode on the board. Notice that if you get the diode polarity backwards (that is, bar/stripe end pointing to the MOSFET gate) then the MOSFET will be turned on 100% of the time by the forward biased diode. The bar/stripe of the high leakage diode MUST be connected to the +9V line, the unstripe/unbar end to the gate.

The LED should now be ON, as the leakage through the diode turns the MOSFET on. Short the gate of the MOSFET to ground with a 100K to 1M resistor. LED should go out. Remove the resistor - LED should come on. The Mill2 is now working. The low leakage diode is a protective device. It's there to prevent gate failure under certain (rare) circumstances.

I keep swearing to myself to rewrite all this on the web page. I really have to do it.

[sir_modulus]

I had the same problems! Try using some standard Mil B values that have been proven to work. for \
diode
normal: 1N914, 1N4148
Low: any 1N400X (X should any # from 1-7)

FET: BS170 (note i used a MPF102 and a J201(common) and both are wring fets. Check that the Fet you use is a N-channel FET.

Now, don't take it that i'm calling you dumb, it's just that these are easy mistakes to make.

LM 250+

[Davide]

My led was always off .  Removing the low leakage diode it works great .
(i'll use a npn junction later, i need all my 2n3904 at this moment .. anyway i tested it with the bjt junction and it's all ok.. ). The guilty diode was a 1N4007.

[audioguy]

I keep swearing to myself to rewrite all this on the web page. I really have to do it.

No need to rewrite- Just copy and paste from my posts!
To tell you the truth, I tihnk Im just going to try some 3pdt switches. Doesnt that bypass with out needing the millBP?

[R.G.]

OK, there are some lessons here. From this and other builders, the 1N400x series of diodes are NOT suitable for the Mill2. Apparently a significant fraction are leaky enough to prevent the thing from functioning. The correct diode for that is the CB junction of a transistor it appears.

To tell you the truth, I tihnk Im just going to try some 3pdt switches. Doesnt that bypass with out needing the millBP?

I distributed the first 3PDT switches to the effects trade, years ago. They were so hard to find and expensive that I spent a couple of years looking for some way around them. The Millenium 1 and Millenium 2 are the result.

What's changed since then is that 3PDTs have been copied by Asian electronics makers and have become cheaper. The first 3PDTs I bought to resell were $20.37 each, and I resold them for $23.00. That's come down some.

If you've read "The Techology of Bypasses" at GEO, you know that you can get "bypass" with an SPDT, bypass with indicator LED with a DPDT, true bypass with a DPDT, and true bypass with indicator with either a 3PDT or a DPDT and the Millenium circuits. Use whichever works for you.

Now that you have conquered the Millenium 2, you're unlikely to have more problems with it.

[cd]

I still have a lot more confidence in a Carling DPDT with a Millenium than any 3PDT.

[audioguy]

Really? Why is that? Being new, these types of opinions are good to hear.

[Davide]

And I think Millenium is.. COOLER than a mechanical switch with led and resistor

[TheBigMan]

I still have a lot more confidence in a Carling DPDT with a Millenium than any 3PDT.

I don't.  I've seen WAY more reports of bad Carlings recently than bad 3PDTs.  I recall ZVex reporting a very high failure rate on the bench, never mind in the stores, a few weeks ago.

[audioguy]

Thats good info to know- Anyone have any other input?

[sir_modulus]

R.G. is right again (as usual) I never tried building one till the other day, and voila, my 1N4003 didn't work. I had four other ones, and only 2 worked. Anyone know a better low-leakage diode?

[R.G.]

OK - let me be very clear here.

The low leakage diode is only there for protection of the MOSFET in the exteremely unlikely case that you get static discharges into the Mill 2 circuit.

The low leakage diode can be left out and the Mill 2 works fine.

You can't tell what is a low leakage diode or a non low leakage diode without looking at specs *except* if you USE THE COLLECTOR-BASE JUNCTION OF AN ORDINARY NPN BIPOLAR TRANSISTOR WHICH IS MADE TO BE LOW LEAKAGE.  Just cut the emitter lead off.
[/list:u][/u][/b]

Get yourself a 2N3904, or 2N4125, or 2N(NPN anything) or a BC549 or BC(NPN anything) or 2SC(NPN anything) and cut the emitter off. You now have an excellent low leakage diode.

[sir_modulus]

Hey thanks R.G. seriousley though, this and EA tremelo both need stickies to answer these questions!

[primalphunk]

If possible some of this should be moved to the faq here.  Am I right?

BTW, thanx RG for your post some time ago about There Are No Electrons.  I'm actually starting to understand a tiny bit more about the theory behind some of these projects.  The projects I do are looking a little less like a dot-to-dot jumble to me and frankly that's more than just a little scary.  Long live little greenies!