Construction error report: Flatline compressor

Started by David, April 21, 2005, 04:06:24 PM

Previous topic - Next topic

David

CONSTRUCTION ERROR REPORT
=====================

Circuit:                           Flatline compressor
Method:                          PCB
PCB generation method:  toner transfer via laser transparency
Etch method:                  muriatic acid and 40% hydrogen peroxide
Etch errors:                    1 bridge (corrected prior to population)
Circuit changes or substitutions:  
                                    Did not use stereo input jack power trick as illustrated.  I don't care about power consumption.  I wired the battery negative where the input jack sleeve connection is shown.

Circuit status:                 Partially operational.  Functions as distortion with massive output.  LED is not coming on at all. I had a light-tight package, but I opened it up to see what was going on.

What I know:                  LED flashes momentarily as circuit powers up.  9V on pin 8 of op-amp.  Ground on pin 4.  Op-amp 1 (pins 1-3) functional because of distortion effect (I had this happen on breadboard for a different reason).  Oscilloscope shows guitar signal reaching the two diodes that are not connected to pin 7, and also the LED.  Oscilloscope also shows no wave output from pin 7.  DVM reading on pin 7:  about 8.7V (meter set for DC and reading fluctuated).  I do not vouch for the accuracy of this reading.  9V entering bias network.  Sorry, but I couldn't figure out where on the PCB the bias network ended.  It appears that op-amp 2 is not getting 4.5V as specified.  The reading I got was more like 8.7.  I've also tried 2 TL072s.

Questions:

1) Have I given enough information for anyone to help me get a handle on debugging this consarn thing?

2) If voltage readings are required from the op-amp, should I set the DVM for DC or AC?

3) What should my oscilloscope expect to see on pin 7?

Plea:  HELP!!!!!!!

David


James V

Hi there.

I just lurk usually, so I can't answer all of your questions.

If you are measuring the voltage at the op-amp then you are correct in setting your meter to measure DC voltage however.


I'm really posting to make sure you're aware of the hazards that go along with using acidic hydrogen peroxide. It can and does form dangerous organic peroxides when mixed with solvents such as acetone or alcohol, and there are many known examples of accidental explosions arising from this. I worked for a few years in a chemical lab and I knew a guy who had seen such an explosion.

So please, if you must use acidic peroxide as your etchant, be careful!

Of course you probably know all that stuff already, so I sound like a patronising fool :)

David

I was hoping for some information on correcting biasing issues or at least on the correct method of voltage measurement...   :roll:

Sigh...  guess I'd better follow all the rules and post voltages.

David

Something is seriously whacked with this Flatline -- or with my circa-1988 DVM.  With no signal being applied and a good battery in the circuit, these are the voltage readings from the TL072:

1 = 9.1
2 = 9.1
3 = 9.1
4 = 0.0
5 = 9.1
6 = 9.1
7 = 9.1
8 = 9.1

I know bloody well this isn't right!!!!   :evil:  :evil:  :evil:

Now, I had cables connected to the input and output jacks.  My guitar was plugged into the input cable.  My headphone amp was plugged into the output, but wasn't on.  Could this have screwed with the test?

Neither the battery nor the op-amp are getting warm.  I can't believe that I'm getting even distortion with readings like this!!!  I'm getting really frustrated.  WTF???  I have faith in the PCB DESIGN.  I know nothing whatsoever about my execution of said design.  The only other piece of information I can offer is that when I built it, I positioned the bias network capacitors so that their negative terminals would connect to circuit ground rail around the outside of the board.  

Could some PCB expert please interpret what Ma is screaming at me?  Don't make me tangle with that Godforsaken diode bridge on perf!!!

toneman

Hi D,
I guess U mean the Hollis Flatline compresser(?)
I guess U mean the pcb from the GEO site(?)
U provide lots of extraneous info.
don't need 2 know how U etched.
on the info U *do* provide--
first thing i C is pin 5, shouldBe(S/B) @ 4.5V.
take ic out of socket, & measure @ pin 5.
one thing that has bugged me about many of the GEO layouts,
is the schematic is not there also.  got 2 go somplace else.
U *are* using a TL072?  or an 082? or an 062?  al pin compatable.
so, *assuming* U R using the "first" OA as input(pin1=out) and the 2nd as
the rectifier(pin7=out), then pin 5 S/B at PS midpoint.
only other thing i can comment on, is are pots in correct?
first assumption i make with GEO layouts is--they work.
if your circuit doesn't, U undoubtedly have a part wrong,
a solder bridge, or wires m ixed up/reversed.
that's all the help i can give.....
now, if my Oscope probes would reach 2 your house.....
staycompressed
tone
  • SUPPORTER
TONE to the BONE says:  If youTHINK you got a GOOD deal:  you DID!

David

Quote from: tonemanHi D,
I guess U mean the Hollis Flatline compresser(?)
I guess U mean the pcb from the GEO site(?)
U provide lots of extraneous info.
don't need 2 know how U etched.
on the info U *do* provide--
first thing i C is pin 5, shouldBe(S/B) @ 4.5V.
take ic out of socket, & measure @ pin 5.
one thing that has bugged me about many of the GEO layouts,
is the schematic is not there also.  got 2 go somplace else.
U *are* using a TL072?  or an 082? or an 062?  al pin compatable.
so, *assuming* U R using the "first" OA as input(pin1=out) and the 2nd as
the rectifier(pin7=out), then pin 5 S/B at PS midpoint.
only other thing i can comment on, is are pots in correct?
first assumption i make with GEO layouts is--they work.
if your circuit doesn't, U undoubtedly have a part wrong,
a solder bridge, or wires m ixed up/reversed.
that's all the help i can give.....
now, if my Oscope probes would reach 2 your house.....
staycompressed
tone

Yes, it's the Hollis compressor.  Yes, I used the GEO transfer pattern. Yes, I used a TL072.  The pots are connected correctly.  I believe the capacitors are connected correctly.  I know I did something wrong.  I will remove the chip and take measurements again.  It sounds like what you're trying to pinpoint is whether the rectifier is receiving bias voltage.

seanm

I just built the Flatline today. I found it very clean for an opamp circuit, so you probably should not get distortion.

Your first problem is that pin 5 has to be 4.5v. Something is wrong with the voltage divider for Vr.

David

Quote from: seanmI just built the Flatline today. I found it very clean for an opamp circuit, so you probably should not get distortion.

Your first problem is that pin 5 has to be 4.5v. Something is wrong with the voltage divider for Vr.

No kidding?   :roll:   Did you do yours on PCB or perf?  I checked pin 5 and it's 9.1 volts.  I'm having a hard time divining where the voltage divider is.

Does anyone have any ideas how to debug the bias network on a PCB?  I sure can't figure it out!

David

I can't find the bias problem.  I'm going to etch a new Flatline board and populate the bias network first.  If I can't get the PCB to work this time, I'll just have to build it on perf.  Wire-wrap has served me well so far.

seanm

Quote from: David
No kidding?   :roll:   Did you do yours on PCB or perf?  I checked pin 5 and it's 9.1 volts.  I'm having a hard time divining where the voltage divider is.
Solderless breadboard. I always breadboard first. Much easier to make changes (and fix mistakes). Although sometimes you can get hum/RF problems that go away on the perf/PCB.

toneman

the Hollis sch shows 2 10Ks 4 the Vref.
GEOs pcb parts shows R4, R6, R7. & R9 as 10Ks.
from the GEO layout, looks like R6 & R7 form the Vref.
Vref only goes 2 the POSiinput of only 1 of the amps.
with the GEO layout, looks like that wouldB pin 3.
use an ohmmeter to measure ohms without IC in socket.
Vref 2 pin3--zero
Vref 2 pin5--10M
Vref 2 one side of Sustain-zero
Vref 2 pin6, 100K + 10K with sustain max.
note: sustain needs all 3 pins connected.
sustain goes 2 R4, another 10K.
my probes won't stretch that far!
U will have 2 use yours.
stayohmed
tone
  • SUPPORTER
TONE to the BONE says:  If youTHINK you got a GOOD deal:  you DID!

seanm

David, do you have the effect in a box? If not, did you connect the input ring to the output ring? Also, did you plug a mono plug into the inut when you did the tests?

seanm

Hmmm. I also think C5 is wrong. The schematic I have shows 100uF, GEO shows 100nF.

R.G. is using Rev 1, 10-10-2001 and I used Rev 1, 14-10-2001.

This won't affect the biasing though...

David

Quote from: seanmDavid, do you have the effect in a box? If not, did you connect the input ring to the output ring? Also, did you plug a mono plug into the inut when you did the tests?

I breadboarded it fourteen months ago.  That's why I decided to build it.  At that time, I used the schematic on Hollis' site.  I had planned to build on perf, but got real frustrated trying to implement the diode bridge.  This is what pushed me into etching.  The only substantive differences I see between the schematic and the PCB are the deletion of the 100uF cap across the diode bridge and the addition of the 1K resistor on pin 5 and the protection resistor on the input.  I'm quite convinced this is builder error, not design error.

For the PCB, I expected issues, so I'm connecting it to jacks with alligator clips.  As I indicated, I'm not using the stereo input jack trick.  I connected the battery snap negative right to the ground bus on the board.

David

This thing is officially declared Code Blue.  I'm going to make one last try to resurrect it to salvage the diodes.  I built a bias daughterboard to use R.G.'s LM386 trick.  I'm going to try to adapt the circuit to obtain its bias from this instead of the discrete components.  If this doesn't work, it's time for a new board.

If nothing else, maybe I've come up with another tool for the wrapper toolbox.

David

I'm not quite ready to give up on this Flatline just yet.  The bias daughterboard turned out to be more extensive than I first thought.  While I was at it, I set it up to do power supply filtering, and also made provision for R.G.'s polarity protection facility (which I didn't try on this go-round).  I also removed the bias network components on the Flatline board.  After hooking the daughterboard to the Flatline, I have +8.8VDC on pin 8 of the IC socket (V+), 0V on pin 4 (ground) and...   and...  +4.45VDC on pin 5!  We have bias!

Now, the proof is in the hearing.  I hope to test it tonight.  Hmmm... I should probably also do an autopsy on those bias components.  I just CANNOT figure out why I wasn't getting bias voltage.

seanm

Good. Look forward to hearing the results.

Johan Blomdahl's LA-Light Compressor has a really cool way of biasing using an opamp.

David

I have a functioning Flatline now.  It was definitely something squirrely with the bias.  Now for the box...