Alllllmost finished with Small Clone build but having a little trouble...

[deadlyshart]

Hi guys, I've been building a Small Clone using the tonepad design found here. I just did a 2nd etching of the PCB because the 1st was a little unreliable, and this 2nd one is perfect. I've assembled it, powered it up, but I'm not getting the effect when I play through it... The signal is making it through, but sounds exactly the same as the true bypass signal. Just to be rigorous, I'm sure I have the stomp switch set to go through the circuit.

So I've done some poking around with my DMM and function generator + scope combo. I think I'm getting the idea of how the circuit works, but I still haven't found the culprit. Two possibly important details/questions, before I say anymore:

1. The 100k trimpot. Before I soldered it in, I set it to be at about the 50kOhm/50kOhm position, just to hedge my bets. I don't totally understand the purpose of it though. From what I can tell, it determines the bias of the signal coming out of IC1A (the one going to the MN3007). Indeed, if I look at the DC bias of pin 3, adjusting the trim pot changes it, but I'll get to that. Is that roughly correct?

2. The 2.2uF tantalum cap. I couldn't find tantalum caps anywhere around me, so I substituted a 2.2uF electrolytic cap in. I think it's working fine as a replacement, because if you look at the schematic, that whole section seems like it's really just for driving the 4047, which I believe is working correctly (shown below). So if the 4047 is working correctly, can I assume that this replacement is fine?

Okay, so let me give some pictures and solid info. Here is the schematic, with one test point I've added:



And here's a little photo of my PCB wired up in the chassis:



The clips to the chassis are grounds for the function generator, scope, DMM. The red clip is a test signal from the fn generator, 1kHz, 0.1V amplitude sine wave. Here's what that looks like, measured right off the input jack:



And here's what the signal looks like on the output jack (after going through the whole circuit):



(Pretty much exactly the same.)

Okay, so the input voltage is ~9.8V. It seems to lose a little bit of voltage over the 47 Ohm resistor (is that right..?), so the positive rail in the circuit is ~9.3V. Measuring the "4.5V" points (for example, pin 5 of IC1) gives ~4.6V, so that seems correct.

So I poked around with the scope to see what was happening to the signal around the 3007, where the action should be happening. All the following scope photos are with the signal AC coupled, so it won't show the DC bias of the signal (some of the signals were much smaller than their DC bias, so I wouldn't have been able to see them if they were DC coupled). Here's the datasheet for the 3007.

First of all, I think the 4047 is working correctly. If I measure pins 2 or 6 on the 3007 (that come straight from the 4047), I get a square wave whose frequency is modulating. If I change the depth knob, it modulates more, and if I change the rate knob, it modulates faster. These pins are the two clock pins for the 3007, so I think that means the 4047 is working correctly. Here's what the signal on pin 6 looks like:



(You can see that it's a little blurry because it was modulating while taking the photo.)

Now I was trying to look at the output of the 3007, so I was checking out the pins that the signal from IC1 was reaching: 1, 3, 7, 8. Pin 3 has a clean looking signal:



However, this signal has a bias that can be adjusted by the 100kOhm trimpot. From looking at the datasheet, it seems like the gate and drain voltages (pins 4, 5) should be -14V with respect to the GND pin (1), and pin 1 is at 8.2V (with respect to the chassis) and pins 4 and 5 are at 0V (also w.r.t the chassis), so they are at -8.2V (w.r.t pin 1). I've read that you can run the 3007 at 9V, so I guess that much is fine.

According to the manual, the input DC bias should be between -5 and -10V. Using the trimpot, I can get the pin 3 DC bias to be 2.2V, for example (wrt the chassis), putting it at -6V (wrt pin 1). However, I'm still not hearing the effect.

Pins 7 and 8 are outputting a signal that's clearly being modulated at the same rate as the clock signal that looks like this (pin 7):




If I look at test point F in the schematic above, it seems like this signal is getting added to the dry signal:



That modulation on it also changes at the same rate as the clock modulation. Is this what should be happening?

Lastly, here's the signal at test point D, the last point in the signal chain before it recombines with the dry signal straight from the op amp (or has it already at this point? I dunno):



So I'm pretty confused at this point. Is there anything obvious I may have missed? Sorry for the long post, I just wanted to include anything that may be useful. Let me know if there's any other info I should include, thank you!

[GibsonGM]

Have you tried following the signal path with an audio probe?

[Jdansti]

In addition to Mike's suggestion, I recommend confirming some basic things first.

1)  It appears you are using a DC power jack that has metal threads and you've got it wired with a negative sleeve and positive tip. Correct?

2) Does your power supply plug have a negative sleeve and positive tip?  This would be the opposite of most (but not all) power supplies used for pedals.

3) You're relying on the metal case to provide ground connections for your in/out jacks.  Check for continuity between your jack sleeves and a ground point on your board. Do this check without any clips attached anywhere on the enclosure or board. Just use your meter probes.

Some of these questions may seem unecessary since you're getting voltage readings on the board and clean signal, but I find it's always a good idea to start at the most basic trouble shooting point and work up from there.

[deadlyshart]

Have you tried following the signal path with an audio probe?

Hi, thanks for the response. I don't have an audio probe... I guess I could just use my function generator and instead of probing with a scope, just probe with something outputting to my amp? (Is that exactly what an audio probe is?)

Do people usually just use a simple sine wave or something more complicated?

[deadlyshart]

In addition to Mike's suggestion, I recommend confirming some basic things first.

1)  It appears you are using a DC power jack that has metal threads and you've got it wired with a negative sleeve and positive tip. Correct?

2) Does your power supply plug have a negative sleeve and positive tip?  This would be the opposite of most (but not all) power supplies used for pedals.

3) You're relying on the metal case to provide ground connections for your in/out jacks.  Check for continuity between your jack sleeves and a ground point on your board. Do this check without any clips attached anywhere on the enclosure or board. Just use your meter probes.

Some of these questions may seem unecessary since you're getting voltage readings on the board and clean signal, but I find it's always a good idea to start at the most basic trouble shooting point and work up from there.

Hey thanks for the suggestions.

1) & 2) I'm pretty sure it's correct. I know the power supplies I use are center positive (had a negative one once, won't make that mistake again!). I'm also sure that the jack/wires to the board are correct; if I measure between the chassis and the V+ from the jack, it has the right polarity.

3) I just checked to make sure, and was confused for a minute, but it's all good. The DC jack is apparently one of those ones that only completes the circuit when something is plugged into it. So it showed no continuity between the jack sleeve and board GND with nothing plugged in, but 0 Ohm when the power is plugged in.

Thanks for the advice!

[Jdansti]

So the power and grounds aren't an issue-that's good. We can probably rule out a defective bypass switch since you're seeing the signal at various points on the board, although if it we me, I'd temporarily connect a couple of alligator jumpers between the common poles and board poles of the bypass switch to create a positive connection from the jacks to the board, and unsolder one side of the little black jumper on the switch.  This would allow you to check the switch without having to desolder the whole thing. See if that makes a difference.

If it still doesn't chorus, it seems that you're signal is bypassing the part of the circuit that would create the chorus effect. Check the polarity of all of your electrolytic caps, especially the 1uf connected to pin 1 of IC1a.

If the caps look good, make sure your depth pot is working properly.

If all of that looks good, I'd do as Mike suggested and use an audio probe to follow the signal starting at the input jack and trace it through the modulation section of the circuit.

[deadlyshart]

So the power and grounds aren't an issue-that's good. We can probably rule out a defective bypass switch since you're seeing the signal at various points on the board, although if it we me, I'd temporarily connect a couple of alligator jumpers between the common poles and board poles of the bypass switch to create a positive connection from the jacks to the board, and unsolder one side of the little black jumper on the switch.  This would allow you to check the switch without having to desolder the whole thing. See if that makes a difference.

If it still doesn't chorus, it seems that you're signal is bypassing the part of the circuit that would create the chorus effect. Check the polarity of all of your electrolytic caps, especially the 1uf connected to pin 1 of IC1a.

If the caps look good, make sure your depth pot is working properly.

If all of that looks good, I'd do as Mike suggested and use an audio probe to follow the signal starting at the input jack and trace it through the modulation section of the circuit.

Hi, thanks for the advice.

The thing is, the signal is definitely getting into part of the chorus effect circuit. Pin 3 of the 3007 shows the signal, so it's at least getting that far...

Just to be sure, I checked the polarity of all the electrolytic caps, and they're all correct.

I tried with an audio probe (simple one made by this method http://diy-fever.com/misc/audio-probe/ ), but to be honest it doesn't tell me anything useful that the scope didn't... I'm just using the same 1kHz signal, and it sounds good up until pin 3 of the 3007, and then sounds bad after that (not really like the chorus effect).

I'm kind of clueless at this point... my best guess is as follows. The datasheet says that the 3007 needs a minimum of -14V for the gate/drain supply, but obviously most people are running a 9V board, so it works that way most of the time. But maybe my chip can't run at that low a voltage for whatever reason, so it's not working...?

Depending on what the 4047 should be doing, I think it (and the depth/rate pots) are working... changing the depth/rate pots changes the frequency range/rate of the square wave coming off the 4047 pins 10 and 11. Does anyone know if that's what I should get out of those pins?

thanks!

[Jdansti]

Check out these voltage readings on a working Small Clone:

http://www.diystompboxes.com/smfforum/index.php?topic=111493.msg1026084#msg1026084

[deadlyshart]

Check out these voltage readings on a working Small Clone:

http://www.diystompboxes.com/smfforum/index.php?topic=111493.msg1026084#msg1026084

Thanks for the link. I had seen that thread before but hadn't thought to actually do it, but I went through and did it.

The mystery thickens... almost everything on the board matches his values, except a couple things.

One thing that I think may be very important is that it seems like way too much voltage is being lost across the 47 Ohm resistor (where you input the raw +9V). For example, my power supply is 9.6V, but measuring right after the 47 Ohm gives (it seems to change from measurement to measurement...) a range of 9V down to even 7.8V, once. The point is that's way too much I'd guess, right?

So, my 4558, LM358, and 4047 chips all have the exact same readings as him (counting for a slight difference in our supply voltages). I believe he actually has a typo, because his pin 9 of the 4047 says 4V, but pins 7,8,9 are all connected to ground. Indeed, his pins 7,8 are 0V, and my 7,8,9 pins are all 0V, so I think that's just an innocent typo.

Another detail is that some of the values on his pins are dependent on the setting of the 100k trimpot. So once I set that, all the values (except the typo one, but I assume it's a typo) on the 4047 are correct.

So that really leaves just the 3007 and the 3 transistors, and this is where stuff gets weird. I'm sure the transistors are the correct ones and installed in the right directions.

Q3 is correct, very close to his. Q1 is also affected by the tuning of the trimpot, but tuning it correctly for the readings on the other chips made Q1 have the correct values.

Here's where stuff gets interesting/possibly illuminating. So, at this point Q2 had incorrect values on its B and E: .3 and 0V, respectively, and the 3007 chip had the value .35V for pins 7 and 8, so that's wrong.

On a whim (...and because it's easier to pop a new chip in a socket than unsolder a transistor), I switched the 3007 out for another 3007. Now, pins 7 and 8 of the 3007 were pretty high, 6-7V (kept switching), and Q2 had more reasonable B, E values (6.5V, 5.9V, good B-E difference). However, it's still not working (clean sound at output).

Okay, I just switched out Q2 just to be sure, with a fresh one. Same deal.

Honestly, I'm wondering if these 3007 chips just suck or something...

oh I forgot to mention: the 3007 chips seem to get a little hot! Not at all like, burning hot, but noticeably warm.

Any ideas from what I'm seeing though? thank you a lot!!

[deadlyshart]

A few mentions of faulty MN3007 chips...

http://tagboardeffects.blogspot.com/2013/02/zombie-chorus.html

If you're getting a dry signal coming through but no chorus effect, then you've most likely got a faulty MN3007. I bought 10 of them and 3 didn't work. I recommend testing all the ones you've got and bin the faulty ones.

(Exactly what's happening with me!)

https://www.muffwiggler.com/forum/viewtopic.php?t=134781&sid=a91f9da904f9c44353ac80e6bdef700c

edit: okay, tried all 5 of my 3007 chips. Same result with each, so either the whole batch is bad (not impossible), or something else is wrong. Hmmmm...

[Jdansti]

There have been reports of fake 3007's out there. Where did you buy yours?

[deadlyshart]

There have been reports of fake 3007's out there. Where did you buy yours?

Just one of those random ebay sellers...

Now that I look at the ebay listing (here http://www.ebay.com/itm/172018158257 ) it actually seems even cheaper than normal, so I guess that makes it more likely that they're bad...

The thing is that it seems like all the MN3007's I find online have the "M" symbol (or is it a capital sigma?) on them, as far as I can tell. I saw on some forum post where the guy was saying he only uses ones from a certain company (they had three triangles on it, IIRC. edit: it's this one I think: http://www.ebay.com/itm/MN3007-BBD-analog-delay-chip-TESTED-FOR-PEDALS-sent-from-USA-fast-FAST-/142207187419?hash=item211c3579db:g:3LoAAOSwcUBYJCLx but it's $10 for one! As opposed to about a buck each, so that may explain something...).

So there are these on amazon, with prime, which makes me think they're more reliable: https://www.amazon.com/MN3007-bucket-brigade-delay-socket/dp/B01MFC9Y51/ref=sr_1_1?ie=UTF8&qid=1483319857&sr=8-1&keywords=mn3007

but at the same time, they have the same markings as all the others you see everywhere.

[Jdansti]

Being an out of production chip, $2 each seems low. Small Bear sells them for $11 each, but SB is a trusted seller and you can be confident that it's the real thing.

http://smallbear-electronics.mybigcommerce.com/ic-mn3007/

[bloxstompboxes]

I have bought a lot on ebay for next to nothing and they have, so far, all been fine. From what I have heard, if the seller cares about there customers and score, they will refund or replace them if the are, in fact, bad. I have only bought from the ones with crazy tens of thousands of good ratings though.

[deadlyshart]

Yeah, the guy I bought it from has 24k sells with 99.5% feedback, so he's probably fairly legit...


arrghhh, what could be the problem, then? This is driving me nuts, I'm so close...

[Jdansti]

Can you post a layout or link for the PCB?  It doesn't hurt to have a second set of eyes review part placement.

[deadlyshart]

Sure, thanks. The PCB layout was in the link in my OP but here's a direct image:



An image of the top of the board with components installed is in my OP too.

Here's the PCB, etched (but no shot of the backside with components in, sorry):



You may notice a part where the toner transfer missed on the left hand side; that's the ground bar so is actually essential, but I've connected it with a jumper as you can see in the guts shot in the first post.

Let me know if you see anything worrying! thanks for the help.

[Jdansti]

Sorry, I missed the links in the OP. I'll take a look at it tonight.

[analogmike]

I agree with the 3007 being the problem. I have bought hundreds of chips from sellers with fine feedback that were fakes. Getting very tough to find real chips in China/Hong Kong/Taiwan. Do you have a pedal you can test one in? Good luck!

[Jdansti]

Sorry, but things got super busy for me. A quick check of IC, electrolytic cap, diode and transistor polarity looks good. The only thing that looks iffy is the damaged ceramic cap near the output, but it's most likely ok. The individual values of the two sets of parallel caps should summed to achieve the desired value. You probably did it that way, but just mentioning it for completeness.

If I can get some time tomorrow, I'll check the resistor values.

Regarding the MN3007, maybe someone who has a known working one can give you a good deal on it or swap it for something.