15 malfunctioning MN3007?!

[GiovannyS10]

Hey guys,
Several months ago i started a project to build a dual-effect pedal for a friend. One of these effects is a John Hollis' super flanger. I did my own layout using this schematic below:
https://s14.postimg.org/gmvpbkyqp/ultraflanger3.jpg
I bought a MN3007 in Brazilian market and when it came, not worked. The chip started to get warm... I said it to the seller and he sent me more 4. No one worked. A friend bought more 10 for me on eBay and sent me. No one worked! I started to think maybe i did something wrong on my layout, although i already verified it and a friend too comparing to the original schematic. So i built the circuit on breadboard. Worked at the same way than in the PCB - No sound! When remove the MN3007, clean sound (opamps were working). So, i started to think maybe the Ultra Flanger Schematic was wrong, so i built on breadboard a Zombie Chorus for test all my 15 MN 3007. No one worked! Most get warm and the sound continue clean ( on Zombie chorus), no effect, some get very hot, i burned my finger when i tried to remove one from bb. And some do not get warm but don't work... 15 ICs, not one, two or five... Fifteen!
Already it happened with you?
I used this Zombie schematic: http://www.hollis.co.uk/john/zombie.jpg
It getting hot really seems we have a short inside it, but i can't believe how bad-luck i am in getting 15 malfunctioning MN3007 from different sellers...

[DFX-PCBS]

China fault, counterfeiting communists

[deadastronaut]

^ yep....been there....fkng annoying..

[Rob Strand]

If you can't send them back then you could try a few things:

1)  Verify the DC voltages in the circuit.  Perhaps with the MN3007 removed.

2)  Try adjusting the "Vref" voltage where the 10k and 12k join.  (bottom left http://www.hollis.co.uk/john/zombie.jpg)
     Maybe  try (10k, 10k), (10k, 8k2), (10k, 15k)

   You could put a pot there.

3) Add R70, R71, C35. As shown on

http://moosapotamus.net/files/ADA-MN3007-rev5-Jan2010-Documentation-rev20120225.pdf

I'd measure the power supply current before you start with the MN3007 in place, the do a mod
and see if the current reduces at all.  Try to work out.

Try at least two chips for each mod.

If it's just fake junk you are wasting your time, however, you have nothing to lose.

[reddesert]

There are some ICs that are long out of production that are often reported as fakes. The MN3xxx BBD chips, and the CA3080, seem to be especially prone to faking. With such chips I wouldn't consider buying on ebay, only from a reputable dealer like smallbear (I get that shipping makes SB less attractive outside the US). If a price is too good to be true, it probably is.

[Juan Wayne]

Pretty much why I don't bother with MN30xx anymore. I'll get MN32xx from a reputable seller or stick to V32xx. If I can't get a decent sound out of those, the circuit is not worth the trouble for me, or the stupid prices you pay for MN30XX.

[Mark Hammer]

Is it possible that what was sold to you as an MN3007 was actually an MN3207?

The 30xx and 32xx series have different Vdd and Vgg from each other.

[vigilante397]

I ordered a batch of MN3007 from eBay a while back and got lucky to have 2 functioning chips out of the batch of 20. I'll add a +1 for Smallbear, but I agree shipping can be tough outside the US.

[GiovannyS10]

Hey guys!
I don't want to believe i have 15 ICs in such condition.
All trash!

It's really MN3007 and no 3207.
Thank you for your suggestion Rob Strand. I will try it soon, but i am almost sure it will not help.

[12Bass]

Can you provide a clearer photo?  It is difficult to make out the markings.  A macro photo would be ideal.

[Rob Strand]

Is it possible that what was sold to you as an MN3007 was actually an MN3207?
The 30xx and 32xx series have different Vdd and Vgg from each other.

Despite what is written on the package that could well be the case.  Worth investigating IMHO.
Like it could be an NE555 which doesn't work - who knows.

[amptramp]

We had a problem with this in the aviation industry starting in the 1980's.  One delivery we got was 54LS74's.  The device said 54LS74.  The paperwork said 54LS74.  But they were 16-pin devices.  A 54LS74 is a 14-pin device.  It would have been interesting to see what it actually was.  BTW the 54LS74 is the military temperature range version of the 74LS74.  Counterfeiters are everywhere.

[Rob Strand]

One delivery we got was 54LS74's.  The device said 54LS74.  The paperwork said 54LS74.  But they were 16-pin devices. 

IMHO it was probably an 54LS174.   It's a D flip-flop with a similar number.  Once it gets into the system it's hard to get it out.   
http://circuits.datasheetdir.com/104/DM74ALS174-pinout.jpg

When I was a kid I remember repairing one of those Tennis games that plugged into the TV (from Tandy/RadioShack).  It was a very much like this one,
http://www.eurozahler.de/data/bilder/172217/Tandy_60-9001_TV_Scoreboard_172217_1.JPG

All the chips had the numbers removed.   The faulty chip was some uncommon CMOS chip.  Somehow I workout the faulty chip and back-engineered what type it was.   Don't underestimate the determination of kids with time on their hands!

[GiovannyS10]

Hey guys,
I tried all, no work! It isn't a 3207 or a TL072 for example. I don't know what is it and i am so frustrated to continue trying. Maybe i can use a 3207 in this project by just adjusting the power supply?

[Rob Strand]

I tried all, no work! It isn't a 3207 or a TL072 for example. I don't know what is it and i am so frustrated to continue trying. Maybe i can use a 3207 in this project by just adjusting the power supply?

That's annoying!

Obviously you can modify it to work with a MN3207.   However it won't be a matter of swapping the pin numbers.  It will need a complete redesign.  It also means making the design a bit more complex to cope with the different supply voltages.  This goes against John Hollis's minimal design concept.

- Change connections to suit MN3207 pinout
- Add 5V regulator.
- Add coupling caps to input and output of MN3207
- Add biasing circuit for MN3207
- Power the 4046 from 5V
- The current design uses a quad opamp from LFO and the Audio and is powered from 9V
  It is probably easier to use two dual opamps.  Power the audio opamp from 9V and the LFO from 5V.
  Otherwise you have to translate the LFO from 9V and the 5V domain of the 4046 input.
  (The battery voltage will vary as it flattens making the translation less than ideal.
   In other words the behaviour changes a bit with battery voltage.) 
- Adjust the LFO part values to suit 5V
- Add separate bias network for LFO.
- Tweak the 4046 part values to suit 5V
(- you might need to supply a different voltage to the Vgg pin.)

Need to check the 4046 will be fast enough at 5V, I suspect so.

So there's many details.
Use the Boss BF-2 as a reference.

[pinkjimiphoton]

i totally feel your pain. i gotta flintlock flanger i've been having issues with for months. it may be something buggy in your layout... if the chip is getting warm, that suggests a shorted power supply to me...
can ya post us voltages? in my limited experience, if ya get wildly different voltages  between different chips, they may not all be bad. also check your voltages with the chips pulled and see what they become once you put the load back on. sometimes that helps narrow stuff down too.

[DrAlx]

3207 max supply is 10V.  Circuit doesn't need full redesign. Just need to handle the connections to the BBD pins since 3207 runs on positive supply.  Set BBD bias by a trimpot divider instead of the 10k/12k divider.

[Rob Strand]

3207 max supply is 10V.  Circuit doesn't need full redesign.

I had to look it up max is 10V and absolute max is 11V
An AC adaptor could go over that.
Running chips at max. is asking for trouble.
For a DIY it's OK to try your luck but it's not a good policy in general.

[Edit: maybe a simple regulator for the delay chip and 4046 like this:
http://www.madbeanpedals.com/projects/PorkBarrel/PorkBarrel_2015.pdf
]

[GiovannyS10]

Hey guys, thank you for all your reply.
Strand, i was thinking about add a 5V regulator to turn on the 3207. I checked the 3207 Electric Mistress version and i noticed some different than the original 3007 version, but the thing that i don't know is why should i wire the 4046 to 5V? Why can't it continue as it is now? Im sorry i didn't got it.

[Rob Strand]

but the thing that i don't know is why should i wire the 4046 to 5V? Why can't it continue as it is now? Im sorry i didn't got it.

The 4046 clock signals drive the MN3207.  The supply voltage of the MN3207 determines the input clock voltage and the 4046 clock level needs to match it.    So that constrains the 4046 power supply to match the MN3207 power supply.

When the 4046 power supply is changed  the 4046 VCO input voltage changes.  The follow-on effect is the LFO output swing needs to be reduced.   One option is to power the LFO from the 5V, like I mentioned above, and as Boss do.   It's basically a redesign.   Keep in mind the original circuit has a lot of DC coupling between the BBD and analog signals and that complicates the redesign.

While redesign is technically the best solution, what about this:

In an attempt to avoid redesign we can take a different approach.   The main problem with the MN3207 is the high supply voltage. If we limit to the voltage to 8V or 9V, using a regulator, that problem goes away.   So I propose the simplest fix for a MN3207 is to power the *whole circuit* from a 9V regulator.   This avoids changing a lot the circuit.  You will need to change the MN3007 pin out and perhaps tweak the biasing network.   

It would be better if the regulator was a low-dropout type.   However, even if the regulator was not low drop-out the circuit would run at 7.5V with a new battery but it will still work.     Here the regulator is not regulating but  is avoiding overvoltage of the MN3207.  This is a small price to pay compared to a redesign.  Note also the John Hollis design allows the voltage of the whole circuit to change when the battery goes flat.

[Edit: when you use low drop-out regulators (LDO)  make sure you use *both* input and output capacitors; as shown in the datasheet for that brand regulator.  Perhaps use 8V if you can get it.]