Pages

Thursday, 19 June 2014

Mutron III

A few people have requested this, and I hope they all don't mind making links :o)  LED/LDR layouts are such a pain!

A very well regarded envelope filter that many think is the best sounding one of them all with this being based on RG's Neutron project files due to the unobtainium status of some of the original parts.  I have modified this slightly from the original because I didn't like the idea of the Gain pot and first opamp stage being permanently in the signal path of your effect chain, and there's really no reason why it should be now DPDT and 3PDT stomps are available so cheaply.  So I have made this one true bypass so just use your preferred method.

Some people have used two LEDs or vactrols to make this but I have opted for a single LED with the LDRs on either side, so you can just bend them to face the LED as you see in some Univibe type effects.  To get the best out of it the LED and LDRs need to be in darkness which obviously isn't a problem when the effect is boxed, but for testing before being boxed you will want to put a cover of some sort like a small purpose made cardboard surround to keep out the unwanted light.

For LDRs Javi suggested the GL5539 which is available cheaply and in quantity on eBay and is 40-100k light, 5M dark resistance.  But I'd suggest socketing them, and probably the LED as well so you can try a few in there and stick with the combination which sounds best for you.

The trimmer was included simply to allow you to fine tune the brightness of the LED which should allow you to tweak the best setting out of it.

The effect needs a +/-9V supply and so I have shown the usual voltage inverter board.  You can add 5 rows to the main board if you prefer but I thought this would make more sense and allow people to get it in a 125B pretty comfortably with the supply board stuck on an inner wall somewhere out of the way.  You can also omit that entirely and use a +/- 9V supply if you have one, or two batteries if you prefer.

The original included a rotary switch for the Mode selection, but there are only 3 selections and so this is easily accomplished with a DPDT on/on/on toggle which again saves lots of space.  If you go with my choice here make sure the contacts change as shown in the diagram.  Some of these switches have the opposite contacts in the middle position, and so if you get one like that you can still use it but you'll need to swap the poles.  So lug 1 becomes 4, 2 becomes 5, 3 becomes 6 and vice versa.  Then it will work fine.  And don't forget the link you need between lugs 2 and 4.

I couldn't find any marketing blurb for a 1972 effect (the year I was born) so for fun here's the patent blurb instead:

A sound effects generator for modifying a sound generating signal is comprised of a voltage-controlled filter having a variable peak response frequency for transmitting at a maximum amplitude those frequency components included in the sound generating signal that correspond to the instantaneous filter peak response frequency. The peak response frequency is determinable by a control voltage applied to the voltage-controlled filter; such a control voltage being proportional to the envelope of the sound generating signal as extracted by a control voltage generator. The filter peak response frequency is thus varied in accordance with the varying envelope of the sound generating signal.

Eh?  Start again please but slower.

And I have to say it really does sound fantastic in this vid.







And as per Javi's request this is as per the original with a 4 pole 3 way rotary switch for Mode, and with a simplified bypass that leaves the first gain stage in circuit.  Just a DPDT stomp is required for bypass, or SPDT if you don't want LED indication.  I left the charge pump board separate though because I just can't face adding any more links to it! :o)



95 comments:

  1. Fuuuuuu

    Ok. This just climbed to the top of my list. Just got an LDR assortment from MKLEC too so perfect timing

    Any suggestions on the 150K log?

    ReplyDelete
    Replies
    1. I'd probably just use 100K but you can try a 500K linear with a 220K resistor soldered across lugs 1 and 2. I'm not sure what that will do to the taper though.

      Delete
    2. If I remember right, you could also do a fancy trick with a dual B100k to emmulate a A200k. I'll see if I can find the link in the morning ;)

      Delete
  2. Built this one a year ago from a PCB bought from GGG.
    Great pedal.

    ReplyDelete
  3. Ah nice, and should be totally doable in a 1590B enclosure :)
    I have an MXR Envelope filter (from tonepad) which is absolutely awesome, I'm curious to compare both

    ReplyDelete
    Replies
    1. Yes I reckon someone could squeeze it in, and the awesome build of the month award will go to the first person who posts a pic :o)

      Delete
  4. Hi! Just two notes, as I don't think I'm going to build another one in the near future:

    1. I had problems with the 15uf cap, it should be no polarized. Not sure if this is mandatory, as I had problems too with the 1M reversed pot. Had to change both, and the effect came to life.

    2. Following Harald's directions, I used a multiturn 250k trimpot, so you can play with different LDR models. After several tries, the GL5539 with a 5mm diffused led were the choosen ones, but this is not a rule of thumb, try as many as you can (or have patience to :P)

    J.

    ReplyDelete
    Replies
    1. Another note: for the 150k pot, I used a 250k with a 360k resistor between lugs 1-3.
      Just like to add, that this effect was one of the most difficult ones that I've ever built, so take it easy when dealing with it.

      J.

      Delete
    2. 250K? Is the trimmer on Harald's for the LED or to tweak the LDR resistance? You're not likely to want any more than 10K for the LED or it will be too dim, in fact 10K is probably a bit high.

      Delete
    3. Is for the leds too. This effect has a 18v potential, and with HB leds 10k is too little to get the leds not lit in Up position. Harald had the same problem when dealing with handmade vactrols, and started with a 5kb trim, and finished with a 247k one :)

      Will have to try finally this board, to see how it works :P Have no On-On-On switch, so will have to do it wit an (original) rotary :P

      Delete
    4. The pedal has an 18V swing, but the LED is between +9V and ground. In fact not even 9V because the anode is connected to the output of one of the opamps. I'll have to experiment too to try to understand it, even with super bright LEDs I've never used anything like 250K, I think the most I've ever used is about 22K and I couldn't imagine wanting the resistance to be so high that the LED would be very dim, if you wanted it dim why use super bright LEDs? Strange indeed

      Delete
    5. No luck until now. Maybe is my wiring as all works ok... except the envelope :P I get clean signal but no filter.

      Gain works as it gets really overdriven sound at max (as it should). The led lits up as you play. In a dark room (just the led of the amp and under a card box, sometimes whn you hit hard, you set some envelope.

      Anyway, I have to take a closer look tomorrow.

      PS I think Range 4 and 5 are reversed, can you check it, please?

      Delete
    6. Yes Range 4 and 5 are reverse when compared to the schematic, but I didn't even check the numbering on the range switch because it doesn't matter, as long as the switch breaks the contact between the two in one position and makes it in the other, it will work as intended.

      Any more luck with the build? What LDR have you used? I'm just wondering because the NSL32's are 500 ohm on resistance but only 500K on resistance.

      Can you hear any tonal differences between the modes?

      Those MPQ2904's have come in and I've got some NSL32's. Do you want me to send a couple so you can try it with those?

      Delete
    7. I used NSL32/SR2's in one and it worked great. the NSL32's didn't work very well.

      Delete
    8. So it probably does need the higher on resistance then because the SR2's are something like 5M I think

      Delete
    9. Thanks Mark, no need for the NSL's, it has to work with the LDR's :P Have to check anyway my wiring. I did it with the original one: buffered and with rotary. This afternoon / evening will check again it all and see what happends. Thank you!

      J.

      Delete
    10. I used GL5539, as did finally with Harald's one, so they should work too.

      Delete
  5. I built one from Harald's layout with NSL-32SR2's and it worked great. I also recently built madbean's naught fish. it's an awesome pedal. VTL5C3's work perfect for the vactrols in that one so an LDR with that spec should work well in IvIark's. I also added a moog switch (for the filte)r, a sensisivity control, and a master volume.

    Harald's vero:

    http://johnkvintageguitars.homestead.com/Effects/others/Mutron/MutronIII-01.jpg
    http://johnkvintageguitars.homestead.com/Effects/others/Mutron/MutronIII-02.jpg

    madbean's NaughytFish:

    http://johnkvintageguitars.homestead.com/Effects/others/Mutron/myMutronIII-01.jpg

    ReplyDelete
    Replies
    1. Your picture really helped me out! That trimpot replacing the 180k resistor was the key to get the downsweep working properly. I ended up using a 300k resistor instead of 180k. Thanks a lot for the picture!

      Delete
    2. Could I build this layout with two NSL-32SR2's if so how would I wire them in? I don't think I can just plug them in where the two LDR's and one LED is. Any help yall can give I'd be grateful. Trying to not spend $5 on parts!

      Delete
  6. Hey mate, guess what? :P Yes, you can tag it, works great. I had a problem with the wiring, a stupid one (Mode 1 IS NOT Range 1 :P).

    As I built it as per original settings (buffered and with rotary) and with the inverter in the same board, I had to do a lwork of rewiring / re -link. Finally I was wrong with the Mode 1 (on board) and soldered it to Range 1. This was the problem :P

    Besides of building it buffered and with rotary, I've used a tantalum on the first 4u7 (sigmal path) and a bipolar 15uf made with two 33uf ones ans a 100k trim

    Have to admit that this baord is great, and this time has been a lot easier to build, just a day, the first one took me a month :P.

    I'd ask you to publish too the "oroginal spec" version with the buffer and the rotary. Noit really difficult to figure it if you have the schematics, but I wasted a couple of hours doing it :P

    If you include the inverter in the layout, it's still easily doable in a 1590B, so, id say is easier and you save some wiring (at the expense of 4 links more :P).

    J.

    ReplyDelete
    Replies
    1. Awesome nice one mate. Why did you do the original wiring? If it was just a buffer I wouldn't have been too bothered about it, but because it's a gain stage and the gain pot will be in the signal path of all the other pedals in your chain, I thought that was a really unattractive option. I'll certainly add the layout, but I wondered what made you prefer it that way?

      Delete
    2. It's just a "romantic vintage" idea. When I build these old effects, I like to build them the way they were.

      That's why I built the Mutron Micro as per the original, or why my replica old Big Muffs don't have filtering caps, and so on.

      You always have the time to mod or add features, but when building a classical one, I like to stick to the specs.

      J.

      Delete
    3. Fair enough mate, I'll add it later

      Delete
    4. Incidentally mate, will you do me a favour and measure the voltage you're getting between both sides of the 15uF cap and ground. In bypass and in use. I just want to see if the the voltage changes to the point where the polarity required may change which may be why the bipolar cap was recommended.

      Delete
    5. i've built them with non polar and polarized and i've found no difference.
      the madbean NaughtyFish just uses a non-polarized 1u film cap there and it sounds fine.

      Delete
    6. Same results with pedal on "bypass" or engaged_ 0,30v upper row, 0v lower row. Maybe id swapping the Dir switch and playinf at the same time, there would be different measures, but I have just two hands :P

      J.

      Delete
  7. Thanks Mark :) Just 4 links more, after 18 is nos that hard :P Youl'll see my finished one in a few minutes :)

    ReplyDelete
  8. Hi Mark.. I think im stuck at wiring the switches. Cant understand fuck all. Can you make it easier? Where do those down up middle things go?! Help!

    ReplyDelete
    Replies
    1. The black line is a link you have to make. The yellow lines are showing you which contacts are making in each position of the switch. So when the switch is down you should get continuity between lugs 1 & 2 and 4 & 5. Middle should be 2 & 3 and 4 & 5. Up should be 2 & 3 and 5 & 6. Sometimes in the middle position the opposite contact may make so it's just so that you can check you have the correct type of DPDT on/on/on switch.

      Delete
  9. Thanks Mark!!! I'll mumble over your words... Thanks

    ReplyDelete
  10. Drive Sw6 connects to nothing on the board.

    ReplyDelete
    Replies
    1. Looks like it should be up a column on the + of the 4.7uf cap.

      Delete
    2. Yes you're right. I actually had a link there in an earlier version but moved a few things around and that's obviously the result. Just make the bridge with solder so both rows are connected and I'll jig things around a bit

      Delete
    3. Layouts now updated, well spotted!

      Delete
  11. Got it working ok. Though the switches are odd. The gain is acting as range and vice versa. I keep checking I'm not mixing them up. Though the switch with the upper right connections is definately changing the the up/down effect. Which should be the range?

    ReplyDelete
  12. I might misunderstand what the gain and range are suppose to do.

    Thought the mode is only usable at one setting with the up wah. Seems to work with the down. Just not using it right maybe.

    It does work good I think. Better than my BYOC filter which might be troubled though.

    Love your layouts. Amazing site.

    ReplyDelete
  13. Lots more playing and find it to be very good. Really have to consider your guitar volume part of the effect. Lots of cool sounds. With the down sweep I can get some cool phaser like sounds controlled with picking velocity if that makes sense.

    ReplyDelete
  14. Just realized. What minimum cap voltage are used in this? Along with the Mutron phaser.

    ReplyDelete
    Replies
    1. 16V should be fine. The filter caps are the only ones with the full supply voltage on them but because they are going to ground, there is still only a 9V maximum swing

      Delete
  15. No way.. Im not that clever. Im stuck. focalized would you please post a diagram of the switches connections?
    Cheers

    ReplyDelete
    Replies
    1. with which filter are you getting problems? Range, drive, mode or all? :P

      Delete
  16. Wait! After days trying to figure it out, i might just have it. Will post if the build is successful or otherwise.. Cheers

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

    ReplyDelete
  18. My build exactly as per your layout, including polarised caps, red led, 10k trimmer, works. I used a 250k pot instead of the 150k.
    I am a bit uncertain about that high low switch.. creates a distorted signal. Is it a boost? Is it supposed to distort? Also, i will maybe try some other ldrs (i used some that were univibe leftovers from Tayda) and leds, at certain settings the attack is well harsh and unmusical (plonk, plonk)..
    Cheers Mark. Another classy post!

    ReplyDelete
    Replies
    1. Hi,

      No, the Low - High switch does not create distortion, just focuses the envelope detector on bass or treble tones. BUT, the Mutron III distorts at high gain settings (it's normal, you have to be careful :P)

      And it's true that the signal gets boosted, these are normal behaviours with the Mutron III. As opposite to the Phasor, the III is really picky with the LDR's. Works great with the GL5539, but sounds like real shit with different ones (i learnt it the hard way :P). There are many settings not musical at all (and if you try the Meatball, you'l get amazed of the noisess that you can create :)

      So which ldr's are you using? You know that a Mutron III is well built when it's very stable on the sound it produces, and the envelope is clearly triggered. To be really sure, get the positions given in the Mutron III user manual (google it) and you'll get fully sure.

      J.

      Delete
    2. Thanks Javi.. i have more or less the same build approach as you (see your comment above on the selector) re old pedals, this time it obviously was a matter of convenience. I cannot remember the type ldr i used, they were lying there. The result is not that unmusical, they are just a bit too touchy. I just ordered a bunch of the ones you mentioned here from some lemon squeezy ebayer, so in a months time i will report. I am also building a biphase, so it's ldr time!

      Delete
  19. Been having trouble with this and the Mutron phaser. They both work fine after a while but eventually sizzle out. Something to do with the charge pump maybe. First I ever used one.

    The signal just seems to fizz out like a fuzz with no power. Also for some reason the Mutron III works lightly with no power. I guess a cap is storing quite a bit of charge?

    I used a MAX1044 chip. I read these are not reliable? I can't find the chip you mentioned except at mouser type stores.

    ReplyDelete
    Replies
    1. Measure your IC voltages when it's in a working condition and then measure them again when it starts crapping out. That will give the clearest indication of where the problems are. It could be the MAX1044, but you could also have a dodgy batch of caps or something similar. Voltages always tell us much more.

      Delete
  20. Thanks I'll try as it's a random thing. And it comes back to life after a bit.

    Should I pay most attention to the 1044?

    Also the sizzled out sounds actually carries onto my amp. Even after I take out the pedal. I measured a small bit of voltage on the output of the pedal. Just for a few seconds though and can't replicate it. That must be charged into the front of the amp until it discharges. It's a solid state amp, don't know if this can hurt it. It goes normal if I turn it off and on again.

    It could be fault of mine but it does it in both the phaser and the filter.

    ReplyDelete
    Replies
    1. I wouldn't say pay more attention to the 1044, the voltages will hopefully give the biggest clue to where the problem is so you want to look at it as a whole. If you're getting DC voltage at the output of the pedal though it suggests the 15u cap (and 10u in the phaser) is faulty, but if the problem with your amp is still happening when the pedal is removed then that could be the source too. Have you tried it with another amp to see if the problem is replicated?

      Delete
  21. No, I have a couple tube amps. Was afraid this might be harmful.

    The voltage was very small, around 0.08v. It might have been more when the pedal acted up. Haven't replicated the measurement again.

    I'll keep going through typical debugging. It's doesn't always happen so that makes it hard.

    I posted to see if maybe somebody else had the problem. With these or something else with a similar charge pump which I've never used in a pedal.

    ReplyDelete
    Replies
    1. And with the filter which acts up more, it sounds like it's loosing headroom and effect before it craps out. So that gets me thinking a cap is doing the wrong thing somehow.

      Also my Pedal Power puts out about 9.3v maybe that's too close to 10v for the chips I have.

      But that doesn't explain the amp still cutting out I guess.

      Delete
  22. My build sounds now really cool thanks for all the helpful comments. I ended up using GL5539 LDRS with a red square LED (very convenient to stick the LDRs on it!).

    I have one concern, though: the LP sounds amazing, but the BP and HP sound "warbly" and somehow distorted in the high end (which leads me to investigate on the HP side of the filter (since BP is LP+HP, right?). Actually, I don't understand the filter section that much, I'm afraid, so if anyone has a clue from where to look at? or maybe it sounds just like that?

    thaks in advance!

    ReplyDelete
  23. I need help to turn the Rotary 4P3T key,,,

    ReplyDelete
  24. Fellows! I would like to hve two leds, each one feeding the two ldrs. How can i do this? Can i just wire them both to the led pins on the board?

    ReplyDelete
  25. Hi
    The wire "9V" between "SW2 Drive" and "Drive Sw5" connects where? on "9v" of voltage convecter board?
    And it has 3 wires connect to the Range switch on Range SW1, if I understand.

    ReplyDelete
    Replies
    1. It connects wherever your 9V supply is coming from

      Delete
    2. Pressed send before I finished :o)

      ... so yes the 9V connection of the daughterboard, or directly from the DC socket. Up to you

      Delete
    3. Thank you for the super fast response !!
      But what I do not understand is the two wire "9V" on the voltage convecter board. I'm not very accustomed to ICL7660 and max1044.

      Delete
    4. It's just representative of daisy chaining the supply to the daughterboard. Input from the DC socket on the left, output to the effect board on the right.

      Delete
    5. what is recommended between the LDR is led away? glued or some millimeter?

      Delete
    6. Just rest them against each other, you don't need a gap and the rigidity of the legs should keep them in place without glue (which could end up blocking some light).

      It will only work correctly when in darkness and so will be fine when the circuit is all boxed up. If you want to test it before boxing you could wrap black electricians tape around the two to cut out all light, or just do what Javi does and throw a thick leather jacked over the board :o)

      Delete
  26. I've built this pedal and the envelope works fine. But there is a high pitch squeal in the background that I assume is something to do with the 7660s. I would like to try and wire it like the original with 2 batteries to see if that stops the squeal how would I do this? Thanks

    ReplyDelete
  27. Hi, If I would to add a send/return effect loop (like EHX Q-Tron plus for example) How I can to do? Thanks

    ReplyDelete
  28. Hi, I would like to buid the original version. But I have some wondering, regarding the rotary switch first. Is that correct if I say that you only have 4 cables plugged to it (lugA, lug1, lug2 and lug3)?
    Also I would like to understand the effect of the power switch. Its only utility is to save the battery life when you leave your guitarcable pluged right? No effect on the sound, it is only convinient switch?

    Thanx a lot for this website, and answers from everyone, love it!!! You're participating making this world a better place!!! ;)
    Quentin.

    ReplyDelete
  29. Thanks a lot for this one. Not easy to wire but it's very fun to play with.

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

    ReplyDelete
  31. Amazing effect thanks guys! Please update the layout to show range 3 connected to ground (otherwise the high range mode has no envelope). Also I can't feel any discernible difference between LP, BP & HP modes. Anything I'm missing there? Cheers you lovely people!

    ReplyDelete
    Replies
    1. check you dpdt on on on switch they can be oppisite or caddy corner . i had this problem on the fulldrive . look up dpdt on on on diagrams you will see what i mean there are two types 1 and two


      http://guitarwiring.blogspot.com/2011/04/how-dpdt-switch-works-dpdt-in-guitars.html

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

      Delete
    3. There are three types of dpdt switches are you sure your using the one Mark used? http://hermeticoguitar.blogspot.com/2012/07/wiring-diy-part-03.html

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

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

    ReplyDelete
  34. Finally got this one working. Started with GL5528 LDRs but got masses of feedback. Seems to work very well with GL5539s and a standard 5mm red LED. Also definitely needs a non-polarized 15uF cappy.
    I ended up off-boarding the trim pot, adds another dimension to the effect. Almost becomes a phasor on full.

    ReplyDelete
  35. Mark I have looked at the circuit with the rotary switch and it seems that a single pole three way switch would do the job, as I cannot see any of the other poles used.
    1S3T or 1S4T rotaries are much cheaper than a 4S3T or an ON/ON/ON toggle switch. And less confusing potentially.
    I will try it out and post results.

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

    ReplyDelete
  37. Thanks to everyone who's commented on this. am about two steps from finishing and would be grateful for two pieces of advice - I've built this using the daughter board for power - does the ground wire from the bottom right connect to the ground wire marked on the left hand side of the main board? And secondly I've not been able to source any non polar caps for the 15uf - how critical is the non polar design and would a normal polarised cap work? Thanks in advance.

    ReplyDelete
  38. Is there a way to tell (other that trial and error) which type of dpdt I have - I'm using the on on in switch from tayda?

    ReplyDelete
  39. Just built this bad boy and amazingly worked first try... Boom.

    I'm having trouble trimming it though. Low pass seems fine however mid and high are on the quiet side... The closest caps I had to 1n8 were 2n2 which I suspect may be affecting things, any suggestions?

    ReplyDelete
    Replies
    1. I have exact same problem, haven't find a solution neither.

      Delete
  40. How do you go about replacing the LED/LDR configuration with an optocoupler,suck as an nsl32? I have 4 pins with an nsl32, but 6 slots to fill on the layout, 2 for each LDR, snd 2 for the LED. Do i just not use two of the LDR slots? Thanks!

    ReplyDelete
  41. So, on the daughter board, i see that it says the left side ground and 9v goes to the supply, but everything on the right side goes to the effect circuit board? Im confused where -9v goes... does -9v on the circuit board, go to -9v on the daughter board? Thanks!

    ReplyDelete
  42. Hey, for some reason my circuit has become an oscillator, and the LEDs aren't turning on. I have no idea where this can be coming from.

    ReplyDelete
  43. Can anyone tell me what is the correct DPDT ON/ON pinout?

    Looking from back, do I get:

    1 6
    2 5
    3 4

    OR

    1 4
    2 5
    3 6

    ?

    Thanks a lot!

    ReplyDelete
    Replies
    1. For switches I always assume:

      1 - 4
      2 - 5
      3 - 6

      It's only really ICs that uses the other numbering convention

      Delete
  44. Mmmm, mine oscillates, and I'm only getting -4.5V on the -9v output; must check that 7660S...

    ReplyDelete
  45. Everything's working fine now (tiny solder bridge was there issue), but I can't hear any difference between the LP and BP settings. I'm using a DPDT on/on/on and everything seem to be wired correctly.

    ReplyDelete
  46. OK, had this one on the table for a while and am looking to split an 18v supply to +ve and -ve 9v each.

    What I'd like to know is this: Does the project need the TL072, or can it be substituted for another opamp e.g.NE5534 etc...?

    ReplyDelete
    Replies
    1. You can substitute it for pretty much any other dual opamp, an 18V swing should be fine for any of them, or certainly any that I have or have used. Not the NE5534 though, that's a single like the TL071. Look at the common ones like NE5532, JRC4558, RC4558, LF353, LF1458, NJM072, TLC272, OPA2134, AD712 etc.

      Delete
  47. Hello. Is it possible to replace a 15u capacitor with a 22u capacitor?

    ReplyDelete
  48. I just built this and got it working! I used the rotary switch and, as Alex asked, I indeed subbed a 22uf for the 15uf. I’ve built many pedals from this and the dirt box page. I feel like this one is a Mt. Everest build!

    My usual process is to do the cuts and links and then jump to resistors, then diodes, caps and electrolytics. I did the cuts and links and then let it sit two weeks, too afraid to begin. Well, today was the day, and, after quite a bit of utility knife work on solder bridges, this thing is amazing! There are so many tonal variations! Patience is key, this took about 12 hours total, though I am a bit slow and methodical. I highly recommend giving this a shot, though, since it’s well worth the time invested.

    ReplyDelete