New module at yusynth.net : Quadrature VC-LFO

oculus · Joined: Oct 30, 2011 Posts: 35 Location: Iceland, Reykjavik

hey yves
first of all your site has been a great help and i love some of the circuits
that iv´e built from your page like the steiner filter.

im really interested in this module but would like to add sync input (retrigger)
ive used this circuit (attached) that tim parkhurst came up with for his 8k lfo for a few different lfos with success,
i was wondering would i only have to put the circuit on C6
or also on C7 on the quad lfo ?
thanks
Fridfinnur

Tim parkhurst 8k lfo retrigger.jpg

Description:

Filesize:

30.06 KB

Viewed:

16398 Time(s)

yusynth · Joined: Nov 24, 2005 Posts: 1314 Location: France

Synching this lFO is not easy indeed. To stop the oscillator it is sufficient to act on a single integrator only, since it's the chaining of both integrator that makes it possible to oscillate.

Still I am not sure that you won't have problem to restart the oscillation since in its normal configuration the circuit enters slowly into oscillation when the CV is low. I am afraid there will be lag with a synch circuit.

Well in my opinion (though I have never tried) is this VC-LFO is not suited for synching application.

wackelpeter · Posted: Thu Sep 25, 2014 7:42 am Post subject:

Hi Folks,

i have just finished this my next yusynth module and finally got it working after some Trouble Shooting (as always i forgot a resistor on my stripboard and interchanged pin 8&9 on one of my tl074's

In the description it is mentioned that the filters slowest rate is 30 seconds per cycle.. mine won't do that... at slightly lower than 1Hz it stops oscillating and also the width of one Phase seems to be longer than the others... at least the led is longer on...
judging from the scope at higher rate everything seems fine and the width of each Phase Looks equal...
i must admit that i haven't matched both caps the resistors are in the 1% area... is this my fault?
and also i noticed that at higher rate the Output is +/-8Vpp and when the frequency get slower increases up to +/-12Vpp... can i reduce it simply with dropping other zener diodes in?

yusynth · Joined: Nov 24, 2005 Posts: 1314 Location: France

In the BOM it is said C6,C7
10nF matched to 1%, polyester

A good match of these two caps is very important for the LFO to oscillate at very low rate. Those I built with matched caps can go down to 30s per cycle without difficulties.

I am also surprised by the output range you get, Z1 and Z2 are there to limit the range under +/-10V.

wackelpeter · Posted: Thu Sep 25, 2014 8:56 am Post subject:

Hi Yves,
thanks for your fast reply.

The caps i used are Wima MKS-2 63V 5% These are Polyester i think... right?
okay as mine are unmatched and you say it's very important for stability at low rates that makes it clearer for me...first i thought it would maybe only have an effect on the Beauty or symetry of the sine waves...

can i use this circuit to match them:
http://www.dartmouth.edu/~aurora/procedures/capacitor.matching.pdf

or is there any other simple way to match them?

regarding the voltage it is as you wrote on your website +/-8V when the oscillator is swinging faster... will check with my magnifier if i used the correct ones ... have taken them out of a bag with a dozen of fresh 5,1 Zeners i got 2 weeks ago from reichelt... the bag is labelled correct of course...

yusynth · Joined: Nov 24, 2005 Posts: 1314 Location: France

Yes its a good way to match caps unless you have a DVM with a good capacimeter function.

The oscillating capacity is insured by the fact that the outputs of the two integrator stages are shifted 90° from one to another. If the caps of the integrators are different then the poles of the integrator will be different introducing a phase shift greater or smaller than the ideal 90°.

wackelpeter · Posted: Fri Sep 26, 2014 10:23 am Post subject:

Back from work where i matched all my 10nF caps with a Benning MM3...

replaced the caps with those matched pairs (10,00 vs 10,02 or 10,16 vs 10,17) which are the closest ones... still not working properly have tried different pairs of those matched caps...
It's slightly better at the low range but still not going below 0,5 Hz...

At some point even in the range of circa 2Hz ist running a few cycles then the cosine Output stays for 1-2 secs longer high or frozen before it switches back to oscillation...
whenever it stops it stopped at between Sin+pi and cosine Output both remaining high...

Ahh have found one troublemaker and pushed him in the bin... it seems that the first tl074 had a Problem with it's first op stage... replaced it with an lf347 and know it goes much lower without stopping the cycle...

okay didn't reach 30secs yet... it runs stable until 12-14 secs per cycle...

any other suggestions or hints how to get it working a bit better in the lower frequency spectrum?

yusynth · Joined: Nov 24, 2005 Posts: 1314 Location: France

No not really more ideas, the quality of component is important (good FET AOPs) an improvement could be using LF444 instead of TL074 for U2. Yet again here at home it works with a plain TL074. What about Q1,Q2, are you using matched BC557C ?

wackelpeter · Posted: Tue Sep 30, 2014 9:23 am Post subject:

Hi Yves, thanks again for your reply... until now i didn't take really much care about matched trannies as all things i've built so far worked really well so far... will give myself a kick and build ian's Transistor matching circuit now... guess that will also show some improvement on the circuits that i thought to be working well already... will also have a look für the lf444 when i order something online next time...

wackelpeter · Posted: Wed Oct 01, 2014 9:46 am Post subject:

hi Yves,
it's me again... built the Transistor matching circuit following Ian's design and matched some trannies... are 1mv difference of Ube enough or must they be closer matched?

Looking at my scope the oscillator didn't stop at it's slowest frequency now ( that was the faulty lt074 chip) but the Level at the different stages sinks when the frequency is slowed down...

at full Speed it's +/- 10v now and the Level then shifts down to quite lower voltages from cycle to cycle... it goes so low that the led's aren't turned on...
when i had it all full Speed and then abruptly turn it to lowest Speed within a few minutes or so the Levels sink and reach only:
0 degree: amplitude swing between 2v to 1,5v
90 deg.: 0,8v to 0,6v
180 deg.: -1,6v to -1,8v
270 deg.: swing very slightly somewhere around -0,65v

also noticed that it Needs some attempts at some time to get over a Point sligthly lower than the Peak before the voltage would go in the direction of 0V/gnd...

can the Led's do cause this? i used red for all of them with the same resistor (the brightness differs a bit, especially at slower rates) i guess that's not the Problem as there is simply not enough voltage at the Input of the buffering Transistors...

hmmm... don't know where to look at besides your Suggestions with an lf444 for U2...
maybe the caps, the lm13700, some of the resistor values??? for the resistors i mostly used 2 resistors in series in lack of 1% tolerance resistors... choosed those that are as with my digital multimeter within that 1% tolerance... for example 67k9 for 68k and None of them more than 0,2k difference...
as i didn't used a etched pcb layout i can imagine that my positioning of my component on my stripboard may be not the best and some of the traces could be placed better... but i don't have a clue if they would affect each other... for some higher frequencies i could imagine some interferences or pops and plops caused by signals with higher level peaks and spikes or capaciator discharging

Anyone with some further suggestions where to look at? Otherwise i fear i must use it as a not so slow LFO... Smile

wackelpeter · Posted: Tue Dec 22, 2015 2:49 pm Post subject:

When you turn pin 5 of the Integrator to gnd then your Quad-LFO is working like charme without stopping soon after you pulled the rate down... Very Happy

got today again over my quad lfo replacing all resistors that should be 1% with a batch i ordered recently and after also changing the bc557 trannies without matching them i noticed after a while that i have forgot this Jumper...

Cycles closely to 30 seconds are now possible without stopping at some point... Smile

beepboop · Joined: Apr 08, 2016 Posts: 4 Location: US

Just about to order some parts for this but the BC557C is obsolete. Is there a specific replacement I should be using? I read about the 2N3906 but for this particular application I'm not sure.

I've had a look on Mouser and have found the 2N3906TA, 2N3906TFR, 2N3906TF and 2N3906TAR, all of which have slightly different specs, none of which exactly match the BC557C, so I'm looking for a little guidance.

LFLab · Posted: Fri Apr 08, 2016 1:29 pm Post subject:

Pinout for the 3906 is different, and it is a device with a lower hfe.
But the bc transistors are pretty common, at least in europe.

beepboop · Joined: Apr 08, 2016 Posts: 4 Location: US

Mouser and Digikey were out but I just found some BC557CG on Arrow. I'm going to buy a bunch. I jumped the gun a little with my post.
Thanks.

LFLab · Posted: Fri Apr 22, 2016 2:07 pm Post subject:

I did a layout for eurorack, using a eurorack PSU connector, board mount potmeters and jacks, and SMT IC's.
But despite the schematic checking out, and no mistakes in placing components, it doesn't oscillate, and my LM13700 is smoking hot Crying or Very sad

This is the layout Yusynth quadrature LFO layout
And this is the schematic Yusynth quadrature LFO schematic

I didn't have local PSU bypassing for the OTA, so I tacked a 100nF cap over the rails close to the IC. The opamps have SMT caps close to them.

What could be causing the OTA to heat up like this?

Note that this is for my use only, not a commercial endeavour.

JRC4558D · Joined: Jun 29, 2014 Posts: 14 Location: London, UK

micromusic · Joined: May 05, 2014 Posts: 58 Location: england

Hi

Have the same problem with my module not oscillating and the 13700 gets very hot
Have swapped out the 74s and the 13700 but still the same
All wiring and components checked and found to be correct
Any ideas anyone

LFLab · Posted: Sun May 08, 2016 12:45 pm Post subject:

Well, still the same problem, I matched a couple of BC557's and got those in, no change. Took the dremel and removed the LM13700, stuck in a new one, still the same. Sigh Rolling Eyes

Edit:
Well, that was quick, but some success. I scratched though the track that feeds the control current from the transistors to the OTA, and hooked the OTA side with a 22k resistor up to the 12v supply. I know get oscillation! It's fixed at about 1kHz. At least I now know the problem is in the control current source. All the outputs work, the LED's light up.

Edit2:
I was double checking the transistor orientation at work, and judging from the datasheet, there is a mistake in the Eagle component!
It is weird though because I have used BCxxx PNP's in my projects before, and never had a problem (but I do have two or three libraries which have this same component, so I might have used another one earlier).
They need to be orientated with the flat side the other way. Can't wait to get home and try!

Yep, transistor orientation is off, and it is in a standard library "transistor.lbr" , I have another library "transistor-pnp.lbr" also a standard library, and that one is correct.

LFLab · Posted: Mon May 09, 2016 9:50 am Post subject:

Sweet oscillating action!
Verified, so glad. Also checked the other layouts I did around the same time, and there was one that needed correction (richardc64's multimode VCA using LM13700).
I'll post some pictures later.

LFLab · Posted: Tue May 10, 2016 10:56 am Post subject:

micromusic · Joined: May 05, 2014 Posts: 58 Location: england

Success!!!!!!!!!! Laughing
Pulled transistors and all LED and jack wiring, checked every solder joint and PCB track, and IC pins and switch and pot wiring
Powered up and the 13700 stayed cold
Matched some 557s with my multimeter on the diode test method
Soldered a matched pair in and powered up all nice and cool
Wired led and jack for 0 degrees powered up and all nice and cool and after about 30 seconds the led started to blink happily
Wired each LED and jack in turn powering up after each one was wired until I had 4 blinking in a nice circle, tested all jacks for voltage, all working and pots doing what they should and the switch reversing the rotation[img][/img]

LFO.jpg

Description:

Filesize:

1.54 MB

Viewed:

197 Time(s)

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

LFLab · Posted: Thu May 19, 2016 12:45 am Post subject:

Very nice!

Regarding the startup behaviour, mine also does take a while to start oscillating.

unrecordings · Joined: May 10, 2017 Posts: 4 Location: Sheffield

Hello all,
I've acquired a partially finished Quadrature LFO and I'm just preparing a panel
(This is a +/-15V build)

Call me daft but I just want one simple bipolar LED instead of four normal LEDs
I'm using MOTM friendly two pin Lumex LEDs. Using ground and any LED output I'm getting a nice bright effect with the installed 1K resistors or clipped on 220R resistors.

I think to hook up a bipolar LED I need to put one wire on the 0˚ output and the other on the 180˚ output - this works but it's very very dim, even with R28 & R30 at 220R

What am I doing wrong ? Would adjusting R24 & R26 remedy things ?
Any advice much appreciated, my fall back position is just a single LED on the 0˚ output...

PHOBoS · Posted: Fri Nov 24, 2017 9:45 am Post subject:

Are you connecting it between the transistor drivers ? because those can only supply a positive voltage.
I think it could work if you connect it directly (with a current limiting resistor) between the outputs of U3c & U3d.

unrecordings · Joined: May 10, 2017 Posts: 4 Location: Sheffield

Yes but the two pin LED basically has two anodes, so in my addled mind when 0˚ goes positive, 180˚ is zero, red lights up (or green depending on which way round you put the LED), when 180˚ goes positive, 0˚ is zero, green lights up

...although after running off and mucking around with two voltage sources, I've come to the conclusion that I fundamentally didn't understand how this LED works.

I'll look at that other method, or I might just go for a single green LED