CE-2 Debugging, Can you help me?

[CriminalK]

First, I'd like to introduce myself because I am new to this forum, so hello everyone! I'm a student from Holland, I'm currently completing my last year at High School this year (hopefully). I'm not really new to electronics, because this year I've made a Theremin (from Moog's design), and before that I've done some projects too; though I'm absolutely no expert. Therefore I'm asking you for help.

Recently, I started building the Corrral Chorus from the Tonepad website: http://www.tonepad.com/project.asp?id=53. I ordered parts from a local guitar supply shop, except the MN3007 and MN3101. Those, I've ordered from a webseller in Hong Kong. I etched the board myself and soldered components etc. I did all the offboard wiring, but when I tested it, the effect signal was the same as the clean signal. I spended some days debugging at school, but no success yet. After a while I deleted all the unnecessary parts from the offboard wiring (like LED, DC-bus etc), I thought maybe that was causing it. As I expected, It didn't, so I kept on debugging. I removed 1 solder bridge already, but still no success. I've run the board for solder bridges since then for over 10 times, but I haven't found others. I checked the polarity of the capacitors, the IC's and transistors, but they seem to be OK.

With a multimeter, I've measured the voltages of the transistors and IC's. I measured the battery before measuring the voltages, and afterwards. It's really weird that the battery voltage dropped about half a volt in the 10 minutes I was measuring the pins. So I checked the ground & 9v copper tracks for bridges again, no success unfortunately.
The voltages I've measured from the JRC4558 and TL072 (I've substituted the TL022 with a TL072, but that shouldn't be a problem), seem normal to me, but the voltages of the MN3101 and MN3007 are rather strange. First I measured some pins of the MN3007 and MN3101 and they were 0V, but according to André's information, they shouldn't be 0V. So I replaced the MN3101 and then, only the grounded pins were 0V, but still no effect. Later on, I also replaced the MN3007, but still no effect.

I've read the other CE-2 chorus debugging reports on this forum, and I found out something about bad MN3101-chips. They said that it merely occurred with Hong Kong-chips. As I've said, I've also bought my BBD drives from Hong Kong, so I thought that maybe that was the problem. I bended pin 8 of MN3101, but there was no difference, still the effect signal was the same as the clean signal.

I was planning to make a CE-2B (the bass-guitar version), but I was suspicious and therefore I didn't install any mods yet.

(I'm charging my camera's battery so I can show you guys some pics now)
Pin Voltages:

Code Select Expand


Battery: 9.00V

IC's:
IC 1: JRC4558
1) 4.58V 8) 9.00V
2) 4.57V 7) 4.59V
3) 4.56V 6) 4.58V
4) 0.00V 5) 4.58V

IC 2: TL072
1) VAR   8) 8.91V
2) ~4.5V 7) VAR
3) VAR   6) 4.65V
4) 0.00V 5) ~4.5V

IC 3: MN3007
1) 7.11V 8) 6.83V
2) 1.53V 7) 6.83V
3) 3.78V 6) 0.98V
4) 3.91V 5) 0.00V

IC 4: MN3101
1) 7.10V 8) 3.91V
2) 0.97V 7) 0.65V
3) 0.00V 6) 2.72V
4) 1.53V 5) 1.68V

Transistors:
Q1 2N5088
1. 8.68V
2. 3.91V
3. 3.48V

Q2 2N5088
1. 8.65V
2. 4.25V
3. 3.71V

Q3 2N5088
1. 8.64V
2. 3.98V
3. 3.53V

Q4 2N5088
1. 6.98V
2. VAR -> About 4V
3. 0.17V

Q5 2N5088
1. 0.09V
2. 0.64V
3. 0.00V





(I like wagamama cookbook btw)



(As you might think, I'm not a super soldering hero)


I hope that anyone can help me, already thanks for reading my topic!

Edwin

[Mark Hammer]

Hi Edwin,

First, unless you failed many years and you are 25, I am VERY impressed with your handiwork.  Very ambitious and accompished for a "high school student".

Second, your excellent pictures show a great deal of flux around the soldered areas.  I would suggest cleaning the flux off the board so that you can do a better visual inspection for solder bridges or cracks or cold joints.  There are other ways, but personally I find methyl hydrate (paint thinner) VERY useful for this.  A simple cotton-tipped applicator dipped in methyl hydrate will allow you to easily dissolve and wipe off the flux.  Again, the objective is not to make a beautiful board, but rather to be able to inspect it and say "I *know* it is not this that causes the problem".

Third, it is probably mentioned in the threads you looked at but bears mentioning again: delay chips need to see a specific bias (DC) voltage at their input pin or else no signal will pass through the chip.  Imagine a dwarf trying to cross a river by walking.  The river is probably too deep and they will drown before they get to the other side.  If they stand on stilts, however, they may be able to keep their head above water enough to successfully cross.  The DC voltage is a bit like stilts.  And just like stilts, if they are too short, the dwarf will be unsuccessful, and if they re too tall, the dwarf may fall over and also be unsuccessful.  The DC bias must be set (via the trimpot) neither too high or too low). 

It is possible that a wrong bias setting is your problem.

[CriminalK]

Thanks for your fast reply!

Thanks for the tip! I will clean my board as soon as possible.

I've tested the trimpot on several positions, but I never thought that could be the problem, so I wil be testing that soon. I figured out that the trimpot can vary the voltages of the 4558 opamp, darn, but I never thought that it would work that way!

(By the way, I'm 17)

Edwin

[Andre]

Edwin,

Your measured voltages show that the voltage on pins 1 of both the mn3007 as the mn3101 has dropped almost 2 Volts, where is should only drop maybe 0.1 or 0.2 Volts.
I guess this causes your problems.
When looking at the copperside of your PCB. I see some shiny stuff between the Collector of Q4 and one of the pins of C22 right next to Q4.
To me it looks like a short between those pins, but i might be wrong because the pic is not too hires.
But if this is a short, it would explain the voltage drop and the fact that the mn3101 oscillator is not working.

If I'm wrong about the above you could also check if R52 is really 33 Ohm.

André

[alfafalfa]

I ordered parts from a local guitar supply shop, except the MN3007 and MN3101.

Edwin , I can't give you any help on the voltage drop. I am building the chorus myself but have only got halfway.
There are people here on this forum who can help you better.
But I was wondering why you bought those chips in Hong Kong,  they're also sold in Holland : Newtone-online.nl has them too.

Just wanted to let you know this.

Regards, Alf from Holland

[MikeH]

I just had to debug one of these myself; my problem was a solder bridge.  It's a very compact layout and there are soooo many chances for bridges.  Use a continuity meter (you probably already have one on your DMM) and check every trace and make sure it's not connected to any of the others.  If you clean off the flux as Mark Hammer suggested, this job will be much easier.  Print out a copy of the layout, and use it for reference.  After you check a trace with all the others around it, highlight it, and move on to the next.  Sooner or later you'll find it...

[Mark Hammer]

Yeah.  Once you know what you're doing, leaving flux on does not present a problem.  But when you are in unfamiliar territory, I find it helpful to actually be able to see the copper pattern and solder joints clearly.

[signalpaths]

I sometimes use an old toothbrush with strong bristles to clean off flux.  What doesn't come off with the brush I carefully use a pick or a exacto razor to get.  Then another time with the toothbrush and that usually gets it. 

I dont know if this will help you in your situation but I bought a broken cE2 off Ebay.  It made sound but no effect.  So I opened it up and fooled with that trimpot in there and bingo I had a functioning ce2 for low$.

Hope this helps

ERic

[CriminalK]

I've tested the trimpot several times now, but it doesn't make any difference. I've measured pin 1 from MN3101 and MN3007 while turning the trimpot, the voltage changes, but the highest voltage is still about 1V lower than the battery + voltage.

R52 really is 33 Ohm, I've measured it and it says 33 Ohm, so that's not the problem. I haven't replaced it yet, so maybe it's broken or something. In that area, I've checked for bridges over and over again, but I haven't found any. Then I started looking for bridges anywhere else on the board, but after 4 hours I didn't found any.

I've removed the flux and I'll post some pictures in a moment.

Could it be that a part is broken? Like a resistor or something?

Edwin

[Mark Hammer]

BBD-based effects are not impossible, but they do present more possibilities for where things can go wrong, which is why we tend not to recommend them as first or even 3rd projects.  If they work, great, but if they don't it is a bit like the difference between searching for your lost dog in the backyard vs a 300 hectare forest.

Looking at the PDF for the Corrrral, I see that the bias voltage is inserted not at pin 3 of the MN3007 (where the input signal arrives), but a little earlier in the circuit.  Specifically, it goes to the biasing point in the 3-pole lowpass filter built around Q2 just before the BBD.  because there are no capacitors between that point and the BBD itself, the DC bias-voltage can arrive without problem.

But is it?  You need to check pin 3 of the MN3007 and adjust the trimpot.  If the DC voltage does not change at that point (and no audio input is needed for this test), then something in between the trimpot and the BBD is getting in the way.  The most likely cause would be that you either have the wrong transistor or have inserted the transistor the wrong way.  There are many many transistors that can substitute for the 2SC1815s originally used (including the 2N5088), but they do not always share the exact same pinout.  Note that you would also not hear anything if Q3 was the wrong one or inserted incorrectly.

[CriminalK]

Pooh,

No luck yet I guess... I don't think that Q3 is inverted, because I've got sound, but no effect. If I turn one of the pots, I hear a little bit of ticking through my amplifier. The frequency of that ticking can be variated by turning the pot (one of those B-100K), but it's very slightly and you can't really hear it.

I've read some other posts and they say it shouldn't be a problem to substitute the transistors by 2N5088's. In the pictures I've seen with those 2N5088, all transistors are placed exactly like I did on my board. But I will check tomorrow if the pin 3 voltage of the MN3007 will change if I turn the pot.

My battery's voltage dropped significantly while I was measuring the pins. So I was wondering if you know what could cause that.

Like I've promised, I would post a picture of the board with the flux removed:

[Mark Hammer]

2N5088s work fine.  I just wasn't sure if you had used something else that was functionally 5088-equivalent but not pin for pin.

Your comments indicate the LFO is working, which is good.

The comment that the battery dropped significantly is NOT a good sign.  In two ways.  First, the effect should not be consuming that much battery juice.  Second, the bias trimpot is set based upon the battery voltage.  This is, in fact, why Matsushita and Boss switched to the MN3207.  That chip works well at 5vdc.  That means that you can run the 9v of a battery through a 5v regulator to power the MN3207 until the battery hits around 7v (regulators need 2v more at their input than they provide at their output).  Since the bias for the 3207 is set based on a stable 5v, rather than a changing battery voltage, whatever you set the bias trimpot to remains valid for as long as you have enough battery current to power the entire effect.  In circuits using the MN3007 and a bias voltage derived from the battery directly, as the battery ages the bias becomes inaccurate and sound quality starts to suffer.  If one uses a wallwart, that is not a problem, but if one depends on a battery for power, the MN3207-based circuits present a more reliably-performing alternative.

[CriminalK]

Hi,

I tested pin 3, and the voltage changes when I turn the pot, so I guess there's no problem in that part of the circuit then. I've rechecked all parts including capacitors and resistors for values, and the values are what they should be. Maybe one of those parts could be broken or something.

[CriminalK]

BUMP,

Wow.. I haven't been on this forums for about 1 year now Probably because I graduated from high school last summer and went on a long vacation ^^. But anyways, about 4 weeks ago I re-found the spirit of making guitar effects and I decided to build some easy distortion pedal (this chorus thing has really set me up: after 1 month of debugging I threw it in a box never to be seen again). I agreed with a friend, who also wants to build a couple of effects, to buy all our supplies at once. Ohh well, because he is a bit lazy we came to order the parts just the day before yesterday and because I couldn't wait, I fooled around a bit with the chorus I abandoned many months ago.
I found my fathers old Lambda Digital Multimeter (guess it's from the 70's of 80's :p) and went on measuring... again. Of course I didn't find any solder bridges or defective copper tracks and the voltages for most of the pins were pretty normal; except for the MN3101 and MN3007.
As for the 2 MN3101's I own (both Hong Kong shipment), one of them really messes up the input voltage for both MN3007 and MN3101 at pin 1. I read this post and it gave me the idea to check the voltage without the chips inserted. As my walwart's got a bit of a high voltage (9.26V) the input voltage turned out to be about 9.25 V which I think is pretty normal. The other MN3101 didn't do anything with this voltage, but has an output voltage at pin 8 of about 14/15 of the input voltage. I read somewhere that this was normal to bad Hong Kong chips and the solution could be to bend pin 8. I already tried once and again, it didn't work out.
Furthermore, one of the MN3007 IC's seemed to be broken because, in contrast to the other I have, it lowers the input voltage on pin 1 about 2 V.
Conclusion? Both MN3101 could be bad and I didn't trust the MN3007's either.
Solution? I start all over.

My shipment, from a trustable reseller, arrived today (with new MN3101) and I inserted the MN3101 into the socket on my old PCB, just an experiment. It didn't work, no matter how i turned the pots and trimpot. When I was beginning to give up hope, I replaced the first MN3007 with the other MN3007. Guess what?

I guess I do have a CE-2 clone at the moment . I still hear just a little bit of ticking but I haven't put it into the case either. I guess it'll work out and with Matsumin-san's excellent (Japanese) description on how to remove noise I don't think that it will be a problem. I think there could be slightly more depth, but I can tweak the resistors a bit. I do have to install the bass mods as I imagined it to be a Bass Chorus . I only have to figure out what I will do with the spare parts, because I've bought all transistors, resistors, capacitors etc (except for MN3101 and MN3007) to build another  .

Well, all you guys who helped me about a year ago, thanks you very much! I'll post the building reports of this CE-2, Big Muff, Zevex Booster and Phase90 that my friend and I are going to build.


P.S. You guys there thinking to buy MN3007 and MN3101 from Hong Kong, please don't. If you find a trustworthy reseller, fine, but don't use this ebay-crap.