frequency shifter

Funky40 · Posted: Tue Sep 01, 2009 11:21 am Post subject:

I could patch the FS-1A today for around 40 minutes.
Just tried with a Drumpatch, 808 BD. ---> Frequency shifters are bad things hehe.
The Module is not finished, also have to change some things, but electronics work as it seems to me. Must test with some speaking thru it.
I'll report back.
#2 is nearly ready for final wiring too

ahhh, if changing the Resistor for comp out gain : higher or lower ?
what i can say is: it looks as i also have to add a 50K Resistor into the Feedback path.
thanks so far Jürgen !

Funky40 · Posted: Sat Sep 12, 2009 6:16 am Post subject:

Thank you very much for this wonderful Module Jürgen !
I patched a little bit, mostly noises, and tested my FS-1A on a Recoring with some voice. Seem to work fine.
I'm very proud that i have built successfully the FS-1A i must say .
...and it's a Noise montser Shocked

I mean when patching nonsense/Noise and adding Feedback. Results can be astonishing pleasing.
( sorry, no possibility to record anything )

In the meantime i built a second one.
That one has a oscillation somewhere, and calibrationstep 1 is not working.
I thought it must be the wiring, but after a first checking this looks good to me.
Must check again and post my questions directly when looking.
Need some measurmentpoints to compare my two Modules
Jürgen if you have some hints by now, i'm glad to hear them

jhaible · Posted: Sat Sep 12, 2009 6:57 am Post subject:

Funky40 · Posted: Thu Sep 17, 2009 4:30 pm Post subject:

a crappy Picture from my FS-1A

what i can say so far:
the FS-1A is a amazing Noisemaker Shocked

------> make a Feedbackoutjack, or SUM/Diff outputs. It's a must ! ( for patching nosie )
I have all three but SUM/Diff not wired for now

If you are into patching noise:
i suggest to make two CV inputs WITH Attenuators and Inverters and muteswitches for both, lin FM and exp FM.
Or put two mixers beside the FS-1A . But i say: this is not the same with mixers Wink

Thanks again to Jürgen for the help and the whole Project, to Tobias for the Partkits , and the help from you DIY geeks ! VERY Appreciated !!
I'm very happy and also very proud.
( belive it or not, even with that module that is like a cockpit of a Airplane, my modularsynth can not fly Laughing

)

IMG_0108.JPG

Description:

my FS-1A #1 in Eurorack format. The XLR/Jack and mic Gainstage is not wired till now. The LEDs are

Filesize:

107.79 KB

Viewed:

968 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

vtl5c3 · Posted: Thu Sep 17, 2009 6:49 pm Post subject:

Wow. That's a nice looking panel (for an awesome circuit)!

Funky40 · Posted: Sat Sep 19, 2009 10:32 pm Post subject:

Thanks VTL.

just patched 6 Hours nonstop with the FS-1A, not counting the hours the day.....
---------> build that FS-1A with a Feedback output i say ( or SUM and DIFF outputs ), or both.
I'm VERY impressed with the results from first patchings within the feedbackloop. IMO the FB insert is a must have ( go for a addon module if your FS-1A is allready built )
tested so far: LPG, CGS Grinder, 512stage BBD ( Doepfer) and all together.Grinder and BBD have muteswitch.
it's just WOW, and it "can" be VEEEERY musical. Easyer and better than FB without additional patchings.
..........speaking from patching crap/nonsense/Noise..............but the groovy one that sounds........ Wink

never heard more pleasing noises.

Funky40 · Posted: Mon Sep 21, 2009 3:42 pm Post subject:

Wow, debugged my JH FS-1A #2.

I thought it must be the wiring, and i double checked all the connections going to the Board,
but guess not checked the connections between the Boards.
the interconnections of SigCos and Input where exchanged Embarassed

,
don't know, my Eyes could not see it........it was now a coinsidence while surching for the fault. Debugging was then very easy Laughing

Just finished calibration. gonna recalibrate both units later again.
now, what to take out to get space ?
wow, i'm happy. Very Happy

Each JH FS-1A was near to 400.-CHF / 350.-$
( with B-Funk partkits and supercheapo faceplate, but good apem switches and good jacks )

zthee · Joined: Feb 20, 2008 Posts: 414 Location: Stockholm

Just thought I'd share another clip of my FS1A.

http://www.thehumancomparator.net/tmp/fs1a.mov

I've got some carrier bleedthru - as heard when the frequency is at the top. Don't know why though...

sduck · Posted: Wed Oct 14, 2009 7:19 pm Post subject:

That's great! I loved the grinning zthee at the end.

That carrier bleed is pretty extreme - maybe just some recalibration will cure that? Probably Jurgen will know...

zthee · Joined: Feb 20, 2008 Posts: 414 Location: Stockholm

Thanks! It not me grinning though, but my friend who just helped me calibrate it Very Happy

I always mix up carrier and signal.. The internal VCO is not hear, but it seems like the signal coming into the FS1A is not entirely processed. There seems to be no way to trim it away?

/mr · Posted: Fri Oct 16, 2009 5:19 pm Post subject:

Happy to see that you have a good time... Laughing

Funky40 · Posted: Wed Dec 02, 2009 4:06 pm Post subject:

How to patch the FS-1A that it tracks 1V/oct ?
I mean, whats the basic setting ?

vtl5c3 · Posted: Sat Dec 05, 2009 6:21 pm Post subject: Mic Input Wiring Q

I'm finally getting around to wiring up my first FS. I'm a little confused by the diagram for the Mic Input. Is the shield connected to ground on the TRS Jack and the GND pad on the Mic input? Is this a case where you want the coax shield connected on both ends?

R.

[/i]

jhaible · Posted: Sun Dec 06, 2009 2:18 am Post subject: Re: Mic Input Wiring Q

paulshillito · Joined: Mar 15, 2010 Posts: 5 Location: UK

I have recently built an FS1a but it rapidly developed a strange fault that I cant seem to get to the bottom of.

After the build and setup i noticed that the UP output out was very loud and distorted. I used a sine as an input and on investigation I found the SumOut was just a squarewave output switching from +/- 13.5V all the time instead following the diffOut which goes from a few hundred Mv up +/-12V when it starts to clip depending upon the input. I traced it back to the pin 12 output of U17B LM13600. Only when the input is at a very low level a few hundred mV does the SumOut have a positive sine top and a flat -2V bottom as is was almost half wave rectified, any more and snaps back to a full output squarewave.

I checked the inputs on pin 14 and 15 and these are about +15mV to -20mV (there a little bit of negative DC offset) and clearly shows the two outputs from the 1496's fine and out of phase. There is -12V in to pin 16 and 1 from the compander section and that is working all OK. If someone put a 5V 1KHz sine wave into the input of the FS1a, what would you expect to see on pins 14 and 15 of the U17B LM13600?, that should go a long way to seeing if the problem is with 13600 or before then in the 1496's.

The other half of the LM13600 for DiffOut is working fine, so I have no problems with the DOWN output. The whole thing had been working OK for about 30 minutes or more after which the problem occured, which pretty much rules out and wrong value resistors as it would not have worked at all with that issue.

To see if was the output of the LM13600 or the op-amp that was causing the problem I cut the track that joins them and still had +/-13.5V on the output of the LM13600. So I replaced the LM13600, everything worked OK and the output was normal level again. I ran thru the calibration again then the problem reoccured after about 30 minutes, exactly the same problem of the LM13600 acting like a switching comparitor rather than an amplifier. I have checked all the components, joints and signals in and out, I even replaced the LM13600 with a third one but this time it did not fix the problem. I have adjusted R127 and R133 but to no avail. The oscillator output is balanced and all the other calibrations are as they should be.

I have come to a dead end as to why this half LM13600 would be do this hard switching instead of amplifying. THe only thing I can think of is the phase of the inputs but that would be determined by the design which obviously works. Do you have any idea as to why this maybe occuring??

Regards

Paul

jhaible · Posted: Mon May 10, 2010 7:38 am Post subject:

Hi Paul,

that's strange indeed, and I wouldn't expect the 13600 to be the culprit.

Maybe there is a bad solder joint that goes from connected to unconnected after some time. As you level dramatically increases, I suspect the bad solder joint is on a shunt resistor (OTA input divider maybe), or on a feedback resistor of an opamp. Just a guess, of course.

JH.

paulshillito · Joined: Mar 15, 2010 Posts: 5 Location: UK

Thanks Jurgen,

If there was a faulty joint on one of the 470R shunts then i should be getting a much larger input on either of the 13600 inputs that was affected but they are within a mV of each other whilst the output of the 13600 is clamping hard on before it gets to the opamp, this is most strange.

Is there anything in the phase difference to the inputs that would make the OTA work like a comparitor?, I cant see it but...

Paul

paulshillito · Joined: Mar 15, 2010 Posts: 5 Location: UK

Problem Found !!! at last and with the aid of some freezer spray and a hair dryer it turned out to be a dry joint on a shunt resistor, just as you suggested, you D'man Very Happy

jhaible · Posted: Thu May 13, 2010 9:07 pm Post subject:

vtl5c3 · Posted: Sat Aug 14, 2010 11:33 am Post subject: Finally completed!

I've pretty much completed my first JH FS. Today I did a simple test run to see what it could do to a drum machine. Here's the (raw and unedited) result:

http://soundcloud.com/jetfinger/haible-fs-and-ry30

I had no idea that a frequency shifter could provide so much tonal variety!!! Thank you Juergen, for making such a rare module available to us!

Serenadi · Joined: Jul 03, 2007 Posts: 89 Location: Germany

I just made a mystic experience.

Last days, I built my two sets of pcbs of the fs1a.
I always test pcbs on my workbench before making any wiring or housing. So I did with these four fs1a pcbs. I only wired the "range" pot of the mainboard (exponential control) to make the first adjustments. Everythings works fine with the first one (adjustment steps 1-4).
Then i came to the second one. I discerned a strange behavior of the range pot. It worked like an acceleration pot of the frequency of the QVCO. Turned to the right a bit, the frequency started to increase, even when I stopped the pot movement. The frequency increased until its highest value (ca 25kHz). The same when I turned the pot to the left (only a bit), the frequency decreased to zero. So, adjusting a certain frequency was like turning in the direction of the wanted value a little bit, then stopping the increasing by turning the pot to the opposite direction a bit. It feels like a rubber band.
After a while running, the frequency decreased to zero without any action from me.
Also connecting the "direction" led, the frequency decreased to zero immediately.

Having two identical pcbs, one running perfectly, I thought it would be easy to compare and to find the error.

Wrong thoughts!
I spent several hours searching for this fault, replacing several components, couldn't find anything.

I have to say, I worked mostly at night, with two lights (normal 60W bulbs) over my workbench.
The third evening it wasn't that late, it is summer, I worked without the lights. While experimenting and searching for the fault, it got darker outside, so I switched on the lights.
The frequency of the VCO decreased to zero immediately !!!

I couldn't believe my eyes.
You know, there isn't any opto-electronic component on that board.
I tested again and again - switching on the lights, the VCO stopped working, switching of the lights, the VCO starts running. The rubber - behavior of the range pot was still there, of course. I moved a little black piece of foam over the pcb to find the place, it was light dependent.

To make it not that long:
The diodes D1 and D2 (after U3) were those two behaving improperly.
They are housed in glass, but I never heard of light dependance of 1N4148, but you can notice it also when measuring the forward voltage of a glassy diode.
I replaced them with other 1N4148, even with 1N914 (also glass housing), but no improvement. By the way, I noticed the same light dependance behavior at the first pcb also, but not so dramatically. It changes the frequency just 20% when switching on the lights.

I finally replaced the two diodes with 1N4001, the only ones I had with black plastic housing.
And voila: Everything works fine now, even the rubber band behavior of the range pot has nearly disappeared.
But not completely, in the upper range (ca. 5 kHz), the frequency still tends to rise - and this is on both pcbs.

So, the circuit seems to be very sensitive to very little changes in voltage at this point (U3). I really would like to read about your experiences, or a little comment of Jürgen. Thanks.

Cheers,
Bernd

jhaible · Posted: Sat Aug 21, 2010 7:27 am Post subject:

Interesting.
So, is an ordinary enclosure (for the whole Device) good enough to prevent this effect?
Otherwise, switching to diodes in plastic package, should be an easy, and unexpected improovement.
Did anybody else notice something like that?

JH.

Serenadi · Joined: Jul 03, 2007 Posts: 89 Location: Germany

jhaible · Posted: Mon Aug 23, 2010 2:24 pm Post subject: