Dr. Boogey mystery

Subzero · December 14, 2007, 05:59:09 AM

I have a friend that works at a pcb manufacturer, who made me a Dr. Boogey PCB using gauss markov's design (actually using the project cad file converted to a gereber file etc.)

Anyway, after populating everything a very strange thing happened that even after checking everything like 4 times and wiring everything twice to make sure there's no mistake, it still continues.

The voltage biasing range of Q4 & Q5 is betwin 8.85 - 9.25 approximatly and the sound is VERY weak altough with a lot of gain. The treble and presence controlls have a lot of effect while the bass and mid controlls have little if any effect.
When I take Q5 out of it's socket, the sound level is restored and has lots of gain but the tone stack has even lass efect then before, the bass and mid controlls do almost nothing.
I take Q4 out as well and the sound becomes even stronger. I put Q4 back in, and Q5 reversed, the sound is just a bit more quiet but basically the level seems ok, again with the eq doing almost nothing.
Note the in any of the situations above, the circuite is VERY quiet, no squeeling, watefall or anything else I've read about, and I did'nt use any grounded wires or anything like that, the gain and volume controlls seem to work ok, I'm able to bias all the other fets with no problem and I've tried several differant fets, I've replaced the trims 3 times thinking the trims might be faulty, but nothing seems to work.

Any one has any idea what's going on

Zben3129 · December 14, 2007, 07:16:37 AM

Hmmmm, I may know the problem.

With this PCB layout, was it made with the correct orientation? Because most of the layouts we use are designed to be printed to pnp and transferred. Therefore, you may need to flip them to get them to come out correctly. Check to see if the words on the trace side are the correct way.

Zach

dschwartz · December 14, 2007, 07:37:18 AM

check your tonestack pots and wiring..pots are kind of tricky..

maybe you have the fets reversed ..did you measure the voltages with the fets connected?

try to bias the thing by ear, starting from the last fet to the first..gaussmarkov´s layout is correct and ultra-verified, i´ve made a few of them and they all work at first power up...

Auke Haarsma · December 14, 2007, 08:05:24 AM

I built one last night with GMs layout. Fired up right away. Could bias Q4 and Q5 correctly.

Subzero · December 14, 2007, 08:43:30 AM

Thnx for the replies guys but ...

The PCB was created using the CAD file (which is intended to be used by the pcb manufacturer and it was created buy a full blown pcb manufacturing plant, with solder masks, silc screen printing and everything else ...) it is double layered and it's oriented correctly.
The wiring of the pots was done twice and was checked time and time again, it all seems correct.
All the fets are correctly oriented and were measured for voltage and verified to be recieving it.
I biased by ear as well .
The main mystery is why does this work with Q5 removed from the circuit completly and with with the Q5 in place makes almost no sound at all ...

BTW, Here's a piece of info you might find usefull, if any one is going to try and have some one manufacture this pcb, the pots are not spaced correctly and you can't use pcb mounted pots. They are spaced at approx. 1.6mm center to center, and they need to be spaced 1.8mm center to center (which also fits nicely into a 1590BB sideways ...), or more ...

John Lyons · December 14, 2007, 09:03:15 AM

Several people have used the Dr boogie layout by Gaussmarkov with total success so you must have introduced a problem.
Taking out any of the FETs will stop the signal. If you are able to take out a FET and still get sound then something is wrong.

Which layout did you use specifically? Can you link to it?
It sounds like you have a layout with the pots mounted on the board. I'm not sure I've seen that one before...

If you used the parts layout below as an actual PCB it will not work as the pots are representations only for layout purposes.
The red lines are just to show where the connections are. The pots are not to scale and are not meant to be.
Sorry for the huge image!

John

Subzero · December 14, 2007, 09:27:27 AM

Hi John

I used the cad file of this:

to create a gerber file which was then used to create a dual layer dual sided pcb.
The pots were not spaced correctly so I cut this part off and wired the pots with wires.
The red connections are on the second layer of the PCB. Maybe I should use jumpers for all the spots that are jumpered to the ground ? but then again, if these spots were not actually connected to the ground the circuite should not work at all now should it

dschwartz · December 14, 2007, 09:41:46 AM

Quote from: John Lyons on December 14, 2007, 09:03:15 AM
Several people have used the Dr boogie layout by Gaussmarkov with total success so you must have introduced a problem.
Taking out any of the FETs will stop the signal. If you are able to take out a FET and still get sound then something is wrong.

exactly john.. if taking out Q5 and still sounding, that´s screaming SOLDER BRIDGE!!!!

I also insist on the pot issue..measure them while turning..i have had very often problems with pots

..i allways press the pot contacts with a plier now...

Subzero · December 14, 2007, 10:00:10 AM

I've just noticed that my version is a little differant and older then the one you posted john. especially around C17 and C21. Could it be something wrong there?
dschwartz, if it works without Q5 and does not work with Q5 in, where could there be a solder bridge ? I checked everything with a magnifying glass but everything looks good ...
Re. pots, should I attach the DMM to lugs 1 & 3 or 2 & 3 ??

Thnx

guitarman89 · December 14, 2007, 10:10:04 AM

the very problem of this kind of circuit is pot and their wiring
Check every pot. One by one. You have to test its total resistance (lug 1 and 3) and its variation (lug 1 & 2 or 2&3)

dschwartz · December 14, 2007, 10:13:01 AM

yes.. is a PITA to check pots.. allways check ém before soldering, you´ll save a lot of pain....

John Lyons · December 14, 2007, 12:20:17 PM

The red lines are jumpers so there is no need for a second side to the board...
Are the holes plated through to the second side? If there is continuity between the jumpered holes through to the second side then no need for jumpers.
Trace the board layout you used against the one I posted and see if there are any differences. The pots are shown with the shaft facing down (just in case) .

Once again. If any of the FETs are pulled out you should get no signal. The FETs are like tubes and are in the signal path as amplifiers. If one is missing there is no signal path and no sound just like an amp. If you get sound with a FET taken out then there is a problem somewhere.
I don't think it is the layout although I don't know if the layouts are the same as I have not checked.

John

gaussmarkov · December 14, 2007, 07:33:23 PM

Quote from: Subzero on December 14, 2007, 10:00:10 AM
I've just noticed that my version is a little differant and older then the one you posted john. especially around C17 and C21. Could it be something wrong there?

hi. i posted a slight revision this morning because someone asked me to switch the pads so that positive cap terminals are square. trim1 is just move slightly, that's all. it's the same layout. the date on it is slightly more recent because i apparently moved the trimmer sometime after releasing the one everyone has been using and that's what was in the file when i opened it this morning.

thanks everyone!

all the best, gm

Subzero · December 15, 2007, 02:07:40 AM

Thnx for the help guys.
I still have'nt found anything wrong. What really kills me is that I have a signal with a missing fet. IAny body ever encounter a bad socket ? I really can't think of anything else.

BTW GM, Could you update the project cad file? The area around C17 & C21 is a bit less clutered and as I said before, if you make the pot spacing exactly 1.8mm center to center it would anable the use of pcb mounted pots with a perfect fit for a 1590BB sideways.

Zben3129 · December 15, 2007, 10:49:23 AM

Hmmm even if it is not a solder bridge, you still have a short somewhere. If it is in a box, take it out for now. If you are sure there are no bridges, then check it up against the layout, and check it exactly, don't skip parts you "know" are right. If that is not the problem, then I cannot think of anything else to do besides taking your dmm on continuity and testing all the components. Possible that you could have a bad resistor or cap somewhere thats shorted, however this is unlikely

Good Luck

gaussmarkov · December 15, 2007, 11:07:22 AM

Quote from: Subzero on December 15, 2007, 02:07:40 AM
BTW GM, Could you update the project cad file? The area around C17 & C21 is a bit less clutered and as I said before, if you make the pot spacing exactly 1.8mm center to center it would anable the use of pcb mounted pots with a perfect fit for a 1590BB sideways.

sure, this is a good idea. but 1.8mm center to center? right now i think they are 200 mils apart, which is about 5mm. but you said 1.6mm previously, which i don't understand.

btw, i did update the CAD file when i updated the layout.

Subzero · December 15, 2007, 01:52:50 PM

oooops ... 18mm lol not 1.8 ... my bad

...

gaussmarkov · December 15, 2007, 02:06:05 PM

Quote from: Subzero on December 15, 2007, 01:52:50 PM
oooops ... 18mm lol not 1.8 ... my bad ...

hmm. i am still confused. bear with me, please.

if you go to taiwan alpha's website and look at their specs for 16mm metal shaft pots, then you will see every pc mount pot (and every solder lug pot) has 5mm centers. now i remember going there originally when i made up the 16mm package.

so what kind of pot are you talking about?

cheers, gm

Subzero · December 15, 2007, 04:12:44 PM

I'm talking about the Taiwan Alpha 16mm pots. If the place them next to each other (actually touching each other...) the distance betwin the ceters of their shafts will be exactly 16mm because half of each 16mm pot is 8mm and you have 2 "halves" of a pot betwin the centers.
In the layout, the spacing is 14mm-14.5mm and because of that, you can't use pcb mounted pots because of that. 18mm spacing will let you mount the pots and fit them nicely in a 1590BB box sidways.

Subzero · December 15, 2007, 04:50:03 PM

YESSSSSSSSSSSSSSSSSS !!! ... MYSTERY SOLVED !!!

For some reason, the bridge near C19 and C20, altough bridged on the component side, for some reason was not bridged on the solder side. I threw in a jumper and it fired up immidiatly

I think the mistake was done at the manufacturer but it might be worth double checking it just in case GM

News:

Dr. Boogey mystery