Problem with millenium 2

[juancra]

Hi All!!

I've been working on a Millenium 2 bypass to put into a working Fender Blender clone i've built a time ago from topopiccione (http://topopiccione.atspace.com/PJ06FenderBlender.html ) and (as usual for me) it didn't work.
I've followed the article from www.pisotones.com (http://www.pisotones.com/Articulos/truebp.htm). First I had the problem
of the LED-staying-always-on, which I solved fliping LED's contacts and cutting GND wire from the millenium pcb. Now it seems to be
ok, but it is working the opposite way it should be !. the LED turns on when the effect is bypassed and goes out when the effect is
engaged. I tried flipping the BS170 , but the LED didn't show up when i did it. Also I can't get the effect working, there is no sound when it is on, but I ve got clean guitar when bypassed. I also added a pulldown resistor (1M) to the end of the effect, because it ended in a capacitor
and the article said the bypass won't go if the circuit didn't finish in a pull down resistor.

What can I do?


Hope you all understand me, apologyze for my poor english!

[miqbal]

You can always count on the function in this forum, try "Millenium 2" keyword. Mr. R.G Keen had some in-depth answers on some Mill 2 threads.

From me, try to build a working Mill 2 first. You can use bread board for ease of use, so you can try on different BS170, or different diode. I personally made my own a separated small Mill 2 PCB for trial.

Usually, the main issue is the BS170 and the low leakage diode. I got my first Mill2 after my forth trial of BS170, the first 3 was broken, damn!!
Another tips, Mill 2 should be able to work WITHOUT the low leakage diode, and the LED should light OFF if you connect the GATE of BS170 directly to ground.

If everyhing's seems OK, you may then connect your Mill2 to whatever circuit you have.

Good luck! (I know how frustrating it is )

[juancra]

thanks for the answer my friend! I'll see what I can do

[juancra]

Please, anyone can help me out?

I've re-drawn the pcb for Mill2 board and substituted the low leakage with the B-C junction of an 2n3904 but it doesn't move. First i mounted a battery and let Control pad unused. Then something weird happen. if I switch contacts of the battery ( + of the clip with negative of the battery) the LED turns on and BS170 gets VERY hot, so I though that the BS was incorrectly placed.  After this I connected the PCB to the effect and reverse-pinout the BS, but nothing came along. I know there isn't a wiring issue because i triple checked it.

This thing is driving me crazy!

[miqbal]

You have to replace that BS170 with a brand new one. MOSFET is highly sensitive, even to a low static charge.

I use this layout. The MOSFET (Q1) is BS170, the low leakage diode (Q2) is 2N3904, and high leakage diode (D1) is1N4148. I even socketed Q1 and D1.


http://www.aronnelson.com/DIYFiles/up/MILL2.pdf

[juancra]

hi!

Did you use any "out of the box" 1n4148? I read that some problems may be related to the fact that modern 1n14148 haven't so much high leakage, so the led won't light. They said only old gold-doped works. is that true?

[miqbal]

They said only old gold-doped works. is that true?

IMHO, I have to say NO. I use a very ordinary 1N4148. It's very cheap:  only Rp 500 for 5 pieces in my country, it's about US$0.04:)
Virtually, any hi-speed switching diode should work: 1N914, 1SS133, ECG519, etc...
As I said, Millenium2 should work without the low leakage diode, so you can try without it first.
How is your Mill2? Still not working?

[R.G.]

Note: almost all signal switching diodes were made with gold doping (whether they said that or not) at one time. It was the only way to get the fast recombination of carriers in the junction to quench conduction at turn off fast enough. Modern techniques have enabled junctions fast enough without gold doping. A "1N4148" is "any diode which meets the voltage, current, and speed specs on this piece of paper labeled 1N4148". So any manufacturer who makes a diode fitting those specs can call it a 4148, whether it's gold doped or not.

Leakage is (one of the) enemies in semiconductors. Manufacturers continuously try to get leakage lower. There have been cases of people having to use more than one 1N4148 to get leakage high enough. It's reasonable to suppose that they were using newer, lower leakage diodes that also met the 1N4148 specification. On the specification, leakage is not specified as "at least X", it's specified as "guaranteed lower than X".

So any high speed switching diode has the potential to work, especially if it's an older one. The newer it is, the more chance you get one of the improved ones that don't leak as much as we'd like for this (mis)application of the diode.

If you're worried about getting enough leakage, use two - or three, four, etc.

And yes, you're correct, the low leakage diode is only there to protect the MOSFET against static. Looks like this particular one may have needed it    even though in general, no, it's not needed to make the Millenium operate. One advantage to the Millenium C is that the CMOS chips generally have static charge protection inside the chip.

[juancra]

Hello RG and migbal!

Thank you both for comming to save me ! ... My millenium 2 isn't working yet. I've separated it from the rest of the circuit and connected a battery into +9v and GND slots and left CNT line unused. The only way that the LED turns on is if I reverse position the battery (the positive connector with the negative of the battery). Then it turns on and BS170 gets very hot. I've tried to put a 1m Resistor in the CNT line and touched the GND with the unsoldered lead to see if it works as RG posted this as a way to test it.

I couldnt found any BS170 at the store so I bought BS107 instead. May this MOSFET also work in the circuit?

This is driving me nuts!

[R.G.]

It may. Typing "BS107 datasheet" into google brings up an assortment, which leads to the pinout. With the flat face toward you and the legs down, the pins are drain, gate, source from left to right. This is the same pinout as the BS170. It will conduct up to 200ma, should be enough.

If you're in a place that's cold this winter, the humidity is probably very low indoors, and you could kill them with a single touch with static electricity. Wear cotton clothes, take your shoes and socks off (no, I'm not kidding), and work in your bathroom with hot water spraying in the shower/bathtub to wet the air down.

If you are SURE of the pinout and wiring and the LED will not come on with the gate/control line open, then the 1N4148 (or other high leakage diode) is not leaking enough. Temporarily parallel it with a 1M resistor for testing purposes. Now does the LED come on? Does it go out when you short the control line to ground? If so, the diodes are not leaking enough and you need to parallel more or get others that are leakier. If no, you have either killed the MOSFET accidentally or wired things wrong.

[juancra]

thanks a  lot RG, I will try what you posted tonight