Monday 1 July 2013

EHX Pulsar

Here's the original '79 two knob version of the EHX Tremolo....



Here's a modded layout that may cure the reported LFO clock bleed. Note that the semi-transparent 47µ cap is there just as an example. You should probably try different values there to find out what works the best. Also the 470R resistor should be tweaked for perfect value. Both of these are there to tame the LFO clock from bleeding back to mixer opamp supply.


42 comments:

  1. I swear sometimes you guys must be freakin' mind readers, I've got a day off today and planned to start on a trem/phaser/delay monster pedal that an old friend has asked about. Looks like I've found the Tremolo section! Just going to take the hound out for a walk (not a euphemism) and then get on it.

    ReplyDelete
  2. Ok, the layout is definitely good, everything functions except that the pots work backwards at the moment. The only problem I'm having, and it's probably an issue with my build rather than the circuit, is that I'm getting some very noticeable clock noise coming through the signal. It's really loud when the power is switched on and then dies down after a few seconds but it's still loud enough to be heard when playing, there's no problem in bypass. Anyone got any ideas on how to kill the noise? It's a decent sounding trem otherwise so it'd be good to get it sorted as I can't put it into a pedal for my mate like this.

    ReplyDelete
    Replies
    1. Hi Rick! I'd use a TL061 instead of the TL071, and would try too an input resistor (about 100r). I will not hurt and may help to reduce the tick sound.

      BR

      Delete
    2. Thanks Javi, I'll switch out the opamp to see if that helps. The input doesn't seem to be the problem but I'll still give the resistor a go, I'm pretty sure it's comming from the LFO as it's perfectly in sync with the sweep of the tremolo. I'll see if the new opamp helps and let you know.

      Delete
    3. Transistors make up the LFO, so i would probably try swapping that right hand side link with 100R-220R to see if that would stop the tick bleed from getting into the opamp (A diode could also be worth a try in that spot.). You could also try putting extra filter cap (47µ) from that strip to ground. That could help too.

      FSB discussion about this circuit had some doubts about LFO tick bleed, so other people have had the same issues with builds from the same schematic.. Let us know how it goes...
      +m

      Delete
    4. Thanks Miro, the resistor definitely helped but it's still got a slight clock noise when the intensity is turned up fully. It's usable now but I wouldn't necessarily recommend it unless you were prepared to put up with it. I'll try the diode in the morning and let you know if that sorts it out.

      Delete
    5. Did you put on that extra filter cap? What value you had for the resistor? I'm thinking that upping the value to ~470R could sort it without killing the LFO. I also think that the type of transistors is a factor..

      If you can crack it, i'll add "fixed" layout..

      I think we can call this verified for now, even though the LFO clock bleed is still a problem.
      +m

      Delete
    6. Ok, after a little bit of time with it this morning I think a 47uf electrolytic cap in the middle of the board from the ground row to the row below (like you suggested) and a 470r in place of the far right jumper just about sorts it. At least with those changes you have to have your ear right next to the amp to hear the clock noise, if you were using a noise gate it'd be cut out and even when recording it's not noticeable when you're playing. Thanks for the suggestions Miro.

      Delete
    7. I used 2n5088s by the way but I'm thinking something with less gain may help too, also the pulse is fairly aggressive even on the standard setting so the change might help smooth things out a little there too.

      Delete
    8. Hmm. Cool. I thought those could help a little. 10µ could be even more effective instead of 47µ. If you used sockets, then 2N3904 would be my next target to try. Does 820R or 1K resistor kill the LFO? 'Cause we'll still want the LFO to work, but we don't want it to tick back to the +supply pin of the mixer IC...
      +m

      Delete
  3. I think the loops on the video is from Tortoise....great

    ReplyDelete
  4. Thanks so much!! Love the sound of this one, definitely gonna through on together..Any chance of a schem? I'm interested in making a fabbed PCB

    ReplyDelete
  5. could you combo this with a Chorus and get a univibe/rotary effect?

    ReplyDelete
  6. Replies
    1. Oh. A missing note...
      That's added polarity protection, so 1N5817 or 1N4001 will do just fine.

      If you're building this, note Madferret's comments if you run in to LFO clock bleed noise.
      +m

      Delete
  7. Hey guys,

    I have built this with slight success. I am a couple of issues though. When the toggle switch is in the down position, whenever the Intensity pot is below 75%, i experience a low hum . The tremelo effect is still working, but it is not the guitar coming through, just a sort of watery low sound. Also, the speed knob works, however once it reaches 50%, it starts to slow back down from anything after. Any thoughts? Any help is greatly a

    ReplyDelete
    Replies
    1. it sounds like you have swapped lug 2 with either 1 or 3 on the speed pot. that would explain why it goes up and then down again in the same direction (which is i think what you're saying).

      As for the LFO (low freq oscillator) there are mods posted in OP for how to fix it.

      Delete
  8. I used sockets and apart from the 2n5088's also tried 2n2222a tin cans, 2n3904, and BC547. I found that the 2n2222's have the smoothest tremolo.

    also, I forgot to put the op amp in its socket, only to discover it was missing when the effect had a 50% volume drop. I put it in and found that it had unity. So I began to wonder if there could be a blend knob modded in somewhere to give more or less dry signal?

    ReplyDelete
    Replies
    1. also, this is a great, simple build! great job, guys!

      Delete
  9. This comment has been removed by the author.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
  10. the effect goes with intereferencia radio, but not the guitar sound.

    ReplyDelete
  11. got it working great, but if I switch the trem/chop switch, it's got REALLY loud ticking. I tried different caps/resistors in the slots, and with the switch in the other position it's still super usable, but I basically can't use it with the switch clicked down. Any ideas?

    ReplyDelete
    Replies
    1. You could try to add a small cap in parallel with 47µ to see if that helps. Something like 100p or 220p. Other thing could be tweaking the value of that 470R resistor from supply row to LFO row.
      +m

      Delete
    2. Thanks Miro. I tried upping the resistor to 820r, and then even to 1k (where that basically just killed the entire lfo), and I tried a 10u cap instead of 47, and added a 220pf in parallel, then a 150pf in parallel, and nothing did the trick. basically when I switch the trem/chop switch, it actually sounds like it's tapping ground ie: the tick is REALLY loud compared to the signal/effect, so I'm thinking I've got a problem somewhere else in the line. Here are my voltages using a battery that's reading 7.9V

      Q1 (upper most transistor)
      E: 0
      B: 440-540
      C: .002

      Q2 (transitor directly below Q1)
      E: 0
      B: .470-1.0 pulsing
      C: 0, but pulses as well

      Q3 (transistor on the bottom left hand side of pcb)
      C: .200-.800 pulsing
      B: .100-.400 pulsing
      E: 0v

      IC
      1: .135
      2: 4.04
      3: 4.02
      4: 0
      5: .002
      6: 7.65
      7: 4.04
      8: .135

      Those are all readings with the switch set to the "usable" setting. I switched it to the non usable setting and the readings wouldn't even stay still long enough to get ballpark numbers.

      Delete
    3. Those voltages do not indicate any big problem. which transistors/opamp you are using? Swapping them to some other models could possible be less prone to bleed.

      One more thing to try could be replacing that limiting resistor (the one you tried different values for) with a 1N4148 diode. I have no idea if that will work, but in theory it should prevent the pulse leaking back to opamp supply.
      +m

      Delete
    4. I'm using a TL071, which I've tried a few of to make sure I didn't have a bad chip. And I tried both 5088's and 2222's all with the same result. The diode trick didn't work, which is not to say it wont in someone else's case, but it just didn't work for me. It's pretty sad I can't seem to suss this out, as it sounds incredible in trem mode, and just plain unusable in chop mode!

      I'm close to just ditching the switch and building it up without one, as it sounds great in trem mode anyway!

      Delete
  12. Thx for the schematic, it's work. But mine is louder than unity, I place it just after the OD effect. Is there a way to reduce the volume? Thx

    ReplyDelete
    Replies
    1. and I don't have enough volume. When enabled the effect, the volume is quieter than when the effect is deactivated.

      Delete
    2. I just pulled this old thing back out of the cabinet and am realizing I too have the problem where it's much quieter when engaged. Perhaps all the stuff I added to tame the clock bleed killed it? Any thoughts anyone?

      Delete
  13. a variable resistor on output to tame it, 100k maybe :)

    ReplyDelete
  14. Hello,

    Anyone could tell me what is the difference between the SPEED and RATE for the Tremolo Effect?

    Thank you

    ReplyDelete
  15. hi there. still having problems with lfo bleed. any suggestions? i'm using the moded circuit. thks

    ReplyDelete
  16. Can anyone please help with suggestions about why the LFO section is not working on my build of this? I have double checked everything and tried both 2n5088's and 2n2222's but no luck so far. Thanks!

    ReplyDelete
  17. Awesome effect. Do you think I should try trimpot (1K) in finetuning LFO (not 470ohm fixed resistor)?

    ReplyDelete
  18. How to add mix-pot for mixing dry and wet signal?

    ReplyDelete
  19. I used the values indicated on the updated layout with the extra elec cap and 470r resistor and it’s working great with no bleed or ticking . I socketed the trannies and ended up with a bc549, bc108, and mpsa13. Kinda weird but sounded best to me.I also used b100 pots and swapped wires from lug 1 and lug 3 on both pots. Anyone else find the switch barely makes any change?

    ReplyDelete
  20. How would you go about adding an expression pedal?

    ReplyDelete
  21. To avoid LFO bleeding it is necessary to shield the output jack with aluminium paper or maybe aluminium tape. The output jack acts like an antenna receaving all the strong signals coming from the rate solder points in the pcb. This is the best way to isolate no wanted signals.

    I have modeled the pulsar in LT Spice and it works good with 470 ohm / 1K resistor in the power section.

    I will try to build this one the next week as soon as the components I bought arrives.

    I hope to publish good news later.

    Saludos desde Venezuela

    ReplyDelete
  22. This comment has been removed by the author.

    ReplyDelete
  23. Ready, no weird noises, no LFO, no bleeding... I redesigned the pcb board for these small changes. So far they are very good results.

    ReplyDelete
  24. an enlightening post on "tick-freeing" the pulsar:

    https://www.diystompboxes.com/smfforum/index.php?topic=18058.msg108913#msg108913

    ReplyDelete