Maestro FSH1 not working

[platinum95]

Hi there.
I've just made the Maestro FSH1 effect pedal from tonepad. (http://www.tonepad.com/project.asp?id=50).
I made my own board and gathered all the components, then soldered them in.
I used the BF245a jfets (which I flipped) and 4558 Op amps.
I used 2n3904 transistors.
I couldnt find a 2m2 pot so I just used a 2M.

Wired it all up, applied power and..... nothing. 3 seconds after the nothing, from my speakers I heard something along the lines of "Zwwwoooooopppppp, zwoooooooooo...".
In other words, there is some kind of wave (sounds like sine) that starts high then rapidly decreases and holds the frequency somewhat lower.
Obviously, something was not right. I took out the op amps and the same thing happened. Then I started poking at the Pots, and found that the sound happens depending on how the Resonance pot is situated. So, say I lay it down flat, I get no wave and I can hear my guitar playing through (though without any effect, it just sounds bypassed). Then if I hold the pot up, i get the "zwooooopp" again, and cant hear my guitar.

If anyone could help me out on this, I would be very grateful.

Thanks.

*edit* Forgot to mention, I'm using the old layout. By the time I realised that it was the old one, I had already etched the board.

[Govmnt_Lacky]

I have built this in the past and let me say.... This is not an easy build. And debugging can be a pain.

I recommend (unfortunately) that you re-etch the current version and start again. No telling what could be wrong but there must be some reason that it was re-done 

The circuit works.... choosing the S & H transistor is teadious.

Good Luck 

[platinum95]

Hey, thanks for your reply.
After a good poke with the multimeter, i found that the jumper that delivers 9v to most of the board was not soldered in on one end. Soldered it in, now the filter's working fine.
As for the sample/hold, I suppose i need to do a bit of tweaking. How do I set the trimpots right?
Thanks again

[hannibal827]

Setting the trimpots right depends on the NPN transistor you're using for noise.  More than one person [myself included] has found that 2N3904s don't do it.

Try whatever kinds of NPN you have available, and search the other discussion threads on this project for ideas.  My FSH-1 has a BC109 in the "noise" socket--that wasn't what I planned on, but it sounded better than the other NPNs I had.  I don't understand this theoretically--it's not a matter of higher gain sounding better, as I tried a darlington and it didn't work out.  It seems to come down to the individual component, which is the "tedious" part of tuning this pedal.

There was a suggestion in some thread or other that grounding the base of the noise transistor [the available schematics and layouts suggest that the base not be connected to anything] makes it so that any NPN will work, but I haven't opened up my pedal and tried this yet.

Good luck!

[Cliff Schecht]

NPN's vary all over the place with their noise, you have to sort through a bunch of them to find the ones with the highest noise floor. What you are doing in this circuit is putting the BJT in reverse breakdown (avalanche) operation (the knee on the left side of a diodes I-V curve). This causes a field to form inside of the PN junction that eventually heats up the electrons enough to break them free and start knocking other electrons free, starting an avalanche effect.

But BJT's aren't designed to operate well in this region (heavy reverse bias) and so they will vary all over the place. If you don't want to sort through a bunch of BJT's finding that magically crappy part, you can replace the BJT with a 6.8V Zener. Zener diodes are specifically designed to operate in reverse breakdown and generate a fantastic amount of noise from part to part. You would have to sort through a big batch to find a few that aren't noisy! I don't know if a Zener is a direct fit in the FSH-1 circuit (I haven't seen the schematic in ages) but my guess is that fitting this into the circuit would require minimal changes.

FWIW PAiA sells a product which requires somebody to sit down and sort through transistors to find the noisy ones (in the 9700 series VCA). I designed a simple noise generator that allowed the use of a Zener or BJT and the Zener diode beat the crap out of the BJT with respect to consistently making a lot of noise.

[platinum95]

So, would I be correct in assuming that the noise transistor is the one where the bass is not connected to anything?

[birt]

you can 'improve' transistors to work in this circuit. if you heat them up hard enough and cool down again they become more noisy. the key is to find the point just before you destroy it.
npn GE's also tend to be noisier. or you can add a wire to the base (and metal shell if your transistor has one) as an antenna to get more noise.

[Keeb]

I built this on vero from the tonepad schematic (there's a layout in the gallery).

I used a 2n3904 and I went and measured it just now:

E  0.95
B -7.19
C -7.79

It works, maybe not as good as I would have hoped but when I found a setting that worked I just left it.
Maybe you can get yours going (just see that if it works) and then tweak from there.

Good luck!

[frequencycentral]

Connect the base of the noise BJT to -ve.

[platinum95]

If i leave the noise transistor as it is, I get a fuzzy noise. If i ground the base of it, I get a high pitched squeel. Which one should i be looking for?

[Govmnt_Lacky]

If you search the forum, there is a really good thread about debugging/troubleshooting this by chromesphere.

EDIT: Sorry... apparently, I started the thread.  Although, there are quite a bit of helpful threads about this build. Also, chromesphere has a good YouTube video with a really good sounding build!

http://www.diystompboxes.com/smfforum/index.php?topic=91637.0

[platinum95]

Im having trouble finding this thread. Anyone have a link handy?

[chromesphere]

Funny i was lurking here and was going to offer a link to that thread but couldnt find it lol. 

Heres the youtube video, probably wont help you fix it, but might give you ideas for mods when you get it going and an example of what it should sound like.  Best of luck!  Its a tricky beast, but worth the tears.

http://www.youtube.com/watch?v=5x_2F6iOYkg&list=UUGhzS1GbX-yxyBrUJtnUMoA&index=44&feature=plcp

Paul

[platinum95]

So istayed up till 4 last night, cramming this into a box. Got it in eventually, and it worked!
Woke up this morning, worked intermittently, checked the voltages on the max1044 and i was only getting -4 out instead of -9.
I don't want to go spending another 5 euro on a max1044 chip.
Paul, I know you mentioned using an lt1054 in one of your videos. I'm having trouble finding that around here, so would the LMC7660 work? I checked the pins and they're all the same. Plus, its like 1.50 for it.

[Govmnt_Lacky]

Paul, I know you mentioned using an lt1054 in one of your videos. I'm having trouble finding that around here, so would the LMC7660 work? I checked the pins and they're all the same. Plus, its like 1.50 for it.

I know this was directed at Paul but, I figured I would chime in to offer some advice 

The problem with the MAX1044 AND the 7660 chips is that they have such a small allowable current draw when used as a voltage doubler (around 10mA if I recall)

The benefit of the LT1054 is that it can go up to 100mA 

Just to be safe... you should try to order the LT1054. You could try a 7660 but you may end up with the same results 

[chromesphere]

Govmnt_Lacky has got it.  And its the reason i went with the LT1054 as i think i mentioned in that video, the max1044 worked for about a day, then it stopped working.  I burnt my finger on it, it was red hot.  Swapped it for the LT1054 and its been sweet ever since!  Futurlec have the LT1054, i think its about 2 bucks.
Good luck
Paul

[platinum95]

Sorry, I didnt mean to direct it at anyone
Right, I'll wait around a while to see if tayda has it. (or if anyone knows whether they do or not, that'd be great) They're currently updating the site so it'll be a day or two till theyre back up and running.
But in the mean time, just for testing purposes, I was thinking of doing the ol' 2 9v batteries in series to get a +9v and a -9v rail. Would that even work? xD

On a completely different note, in the 10 minutes I had the pedal working, it was brilliant, I love the sound off it.
One problem I did have, though, was that when on the filter setting it would only work with the up/down switch on one setting (I'm not sure whether it was up or down) and on the other setting I just got a ticking noise at the same frequency as the lfo from the sample/hold side of the pedal. This is pretty strange as they pretty much work as 2 separate circuits and aren't connected at all.
Any ideas?

[platinum95]

Right, I can't actually find a supplier of the lt1054 that has an acceptable price. The cheapest I can find is ebay, where it's 7 euros for 1.
Are there any other options that i can take?