Another sitar circuit

[col]

You beat me to it with the 386! I'd been thinking about this all day at work and the main part of the circuit seems to be based on the earphone interaction. Remember that this is from March 1967 and there wasn't the wealth of components that we have today. They were also relatively very expensive. I'd been thinking about using an LM386 or some transistors from the BF or BD series to get the volume up to a level high enough to drive the second speaker or whatever it is.

The mag it came from has some really interesting adverts e.g " The Transistor amplifier you have been looking for!"; 3W at £17.15.0 or with cabinet £18.19.0, that's pounds, shillings and pence for those of you too young to know. That must have been at least a  weeks wage or so back then. I wasn't even 3!

Bargain of the month! Genuine Mullard OC44 and OC45 only 1/11 each with 12 or more post free. I'll bet they're a bargain even now compared to that.

There is also a project for a 'Digital' radiation counter. Digital in the way that it turns numbers around on an old style counter. It states that silicon transistors are ideal but then gives a list of the Germanium types that you can use. There are also details of where to get the Geiger-Muller tubes from. Can you still get them? Would you want to?

I actually bought the mag for an article on a proximity alarm to see if I could utilise it to drive a wah/theremin/tremelo speed etc. But it would probably need the sensitivity setting for each time the effect was used in different ambient lighting conditions. My ideas aren't dead but I need to know a modern equivialent to a selenium cell type B3M from the International Rectifier Co.

One last thought. "This month we introduce 'Hertz' into our pages as the unit of frquency, in place of the long used and familiar 'cycles per second.' The symbol Hz replaces c/s." Yes it really is that old.

[darron]

everyone's being soooo fussy about the speakers. could someone please just try it with anything and post a clip?

[GREEN FUZ]

           these being oc72 pnp transistors, shouldn't this be positive ground? I'm pretty new to this stuff so I dunno...

Think you`re right about that, my..em..layout needs revising. I`m taking a lot of these comments on board particularly regarding what exactly those earpieces constitute. Gonna try speakers and probably some of those earpieces if I can find another one. It`s possible modern components would do a better job of the amplifier part of the circuit but as a tribute to M.P Hamer of Bath I`ll try it his way first. Lucky I just happen to have a small amount of OC76`s which should do the job. Now all I need is the bleedin` caps.

[gez]

I've asked a question about this circuit on the EPE message board.  You'd be surprised, there are members there who remember all this stuff the first time round!

I'll keep you posted if anything interesting crops up...

[gez]

Got a reply to my question:

"Earpieces fom the 60's had impedances of several thousand ohms. Somewhere between 2k and 6k depending on brand if I remember right. You could also get 600ohm headsets"

Pretty much what I thought.  8ohms is a bad idea.

If this were me, I'd re-jig the circuit using transducers.  I accidentally soldered the wrong wires together yesterday.  I wanted to provide a common ground for both devices, but wired the ceramic connections together by mistake, which meant the metal connections acted as a short, hence my need for insultion when I taped the two together.  Did things the correct way today, which meant I could dispense with the insulation, and the signal was much improved.  Pretty cool treble booster: you get a bright twangy sound but minus the earache you get with a lot of boosters.  Only drawback is it's pretty microphonic.

Anyway, if you wired up a 4k7 collector resistor in place of the earphone, the transducers could couple the signal to the amp - use them like an output cap.

[Scotty Did]

everyone's being soooo fussy about the speakers. could someone please just try it with anything and post a clip?

working on it

[col]

I'm not going to be able to get to my computer for a week or so but if this hasn't been solved by the time I'm back I'll get on with this and see if I can help sort it out.

[Scotty Did]

I hooked my my little gem up to the speakers and into my amp and it definatly has an effect.  I'll post sound clips later.


it doesn't sound like a sitar, kinda but not really but the sound might change if I had the transistor amp specified.  also i's definatly not ment for a stage, or live gig type effect.

more of a studio type thing.  so a bypass switch really isn't nessisary.

[gez]

A further reply on the EPE board:

"Gerry, as an oldee here, I can confirm your looking for a DC resistance between 500 Ohm to 5K; common to that period was 1 or 2K types, since these were used for early hearing-aid amplifiers"

So, for those building it, 8 ohms is a bad idea - the last stage will probably bias towards saturation, in which case 8ohms isn't going to limit current much.  Seems my comment about the mojo of early earphones was correct!  

I used to have an early earphone (60s period) courtesy of my parents.  What I remember is that whatever you played through it was tinny and thin, which was why I thought this circuit was ingenious: my first reaction on seeing it was "Cool, a built in hi-pass filter...makes sense!"  Not sure modern speakers/phones would be appropriate, therefore, as their fidelity is too high.  What I can confirm is that you get the hi-pass effect using cheap disc transducers wired up like an output cap.  I should think that arrangement would complement the complete circuit well.

[GREEN FUZ]

Has anyone besides Scotty Did tried this yet? ( where`s that soundclip by the way?). I`ve been struggling with the amplifier section let alone the earpieces. There`s something drastically wrong with my layout. I`ve spent the last two days trying to debug it. Initially I made a PCB and soldered it up with sockets for the trannies. Pop in some Raytheon PNP`s. No sound. Try some AC128`s. That`s better. Getting some sound now. Very low, but....slightly sitarish ? Is that an octave down I detect, with a hint of Ring mod ? Hang on. what the #$@%? The battery isn`t even connected. True story...I could go on.

But I wont.

Finally, after breadboarding and much trial and error I discover that the second resistor is somehow attenuating the signal to the second cap. What is it about my layout that is doing this? By removing that resistor it sounds fine, a sort of trebly boost with a bit of distortion. So where is the missing resistor supposed to be? I have plenty more questions brought up by this circuit but at least it`s working, after a fashion.

In any case, I have something to be going on with. Next step, earpiece section. I`ll keep you posted.

[GREEN FUZ]

Well here we go, for anyone still interested.

http://s21.quicksharing.com/v/9646774/sitar5.wav.html

For the output section I used one of those earpieces from a Science Fair Electronic Project Lab taped to a transducer. It certainly gives it some of that tinny sitar-like resonance. I think the amp should be more of a clean boost as it introduces some distortion which probably isn`t needed. I`m going to tinker with it a bit more, see if it will clean up. As I said the important part seems to be the cheap earpiece and yes it is very microphonic.



For those who care to know it was recorded through the tube preamp patch of a POD straight to a crappy soundcard via the magic of Windows Sound Recorder. No effects. Guitar was a noname Japanese thing. Possibly Guyatone. Dropped D tuning.

[gez]

Can't seem to download the file.  Do you have to register with them or something??

[GREEN FUZ]

No, you don`t need to register. Click on the link then double-click where it says download file. It does seem to take a long time to download. Maybe I need to compress it or convert to MP3 or something.

[gez]

For some weird reason it just won't open, though I managed to save it then play it no problem.

Not bad...sounds like a Ge fuzz played through an earphone.

[GREEN FUZ]

Not bad...sounds like a Ge fuzz played through an earphone.

Hehheh Yeah, I`m sure Ravi Shankar isn`t sweating it. I`d still like to know why R2 in my layout cuts off the signal. Any ideas ?

[Harry]

Hey thanks for sharing that clip.

For those who still can't find it just follow the link and just look for the 'Download File' link, it's not real obvious and it doesn't look like a link but it's there.

[mac]

I guess the point of using earphones is that the freq response is ideal for the task. A couple of small speakers will work but they may add some unwanted freqs.
In many indian instruments there is thin thread that lift up the string a little. When hit, the string gently touches the edge of the bridge and produces the characteristic sound. Some freqs are decreased, mainly the bassy ones. BTW, adjusting the threads is a pain in the... well, you know.

BTW, this is a real sitar

http://s15.quicksharing.com/v/579674/Mathar_Ad_VW_Polo_.mp3.html

Enjoy

mac

[GREEN FUZ]

In fairness I was never expecting this circuit to sound like a sitar. Or even particularly close. It`s an instrument rich in complex overtones. How does one replicate the drone strings for a start. The idea of trying to make a guitar sound like one has been tried many times with varying results. Notably the Danelectro (coral ?) guitar. The best I could have hoped for was something like that. A western approximation. That said, I do believe the earpiece concept with some tweaking can make a fair fist of the job. It just needs a more suitable circuit in front of it. What I don`t know but I`m going to keep trying.

Suggestions as ever, welcome.

[gez]

I`d still like to know why R2 in my layout cuts off the signal. Any ideas ?

With the inclusion of R2, a resistance of 4k7 is now in parallel with the input resistance as seen at the base of TR2, and in turn this combined resistance is in parallel with the previous stage's collector resistor - R1 - pulling it down in value and lowering the gain of the first stage.  So, by removing R2, you bump up the first stage's gain a little.  Perhaps this might not have been necessary had a higher gain trannie been used for TR1?  The reason you can do this is because of the high leakage current of Ge trannies - it's enough to bias the device by itself.

Ah, the joys of Ge transistors!

[gez]

In fairness I was never expecting this circuit to sound like a sitar. Or even particularly close. It`s an instrument rich in complex overtones. How does one replicate the drone strings for a start. The idea of trying to make a guitar sound like one has been tried many times with varying results. Notably the Danelectro (coral ?) guitar. The best I could have hoped for was something like that. A western approximation. That said, I do believe the earpiece concept with some tweaking can make a fair fist of the job. It just needs a more suitable circuit in front of it. What I don`t know but I`m going to keep trying.

Suggestions as ever, welcome.

Can't remember the track - Heart Full of Soul? - but Jeff Beck famously used an old Fuzz box (no doubt a Ge tonebender) to get a similar sound to a Sitar on the guitar because the session musician the Yardbirds employed to play the lick on a real sitar didn't cut it.  If you play guitar loud enough for all those years, it sounds close enough!