Jump to content

Sustainer Ideas


psw

Recommended Posts

Alright, news. Went out and bought 0,2mm wire. Oh joy, it works better on low strings and does its thing on highs too. But it does it weak. The strings sustain but still don't pass the infinity mark. Maybe it's because I'm using 1/2 of a humbucker to drive the ruby. Does that give too little output? I don't think so...

It could also be because I'm using rather long wires to connect pots, jack, perfboard, the driver, etc.

What do you think?

-----------

Jimbo

Link to comment
Share on other sites

That's why...the "magic's" in the driver. If you check out the patents, there's all kinds of wacky frequency dependent phase compensated circuitry. I've avoided all this by working on more efficient driver designs. The principle is simple, getting it just right is harder...that is...if I hadn't already given explicit instructions based on hours of experimentation:

Slim blade driver, 8 ohms 0.2mm wire, very well potted on a steel core of about 3mm width and about 4mm deep. Low watt amp, no neck pickup connection, mounted near the end of the neck, either with it's own magnet or a suitably strong enough magnetic field from the pickup.

All the wisdom you need is in the quote above...

Pekko

Thanks!!

I see that the whole phase shift frequency compensation things coming back to the fore over at the DIY stomp thread...hmmm...

Perhaps the first paragraph is even more important than the formula...The whole point of my design idea is to dispense with complicated and expensive circuitry by designing a driver around these issues. I'm sure that there are other possibilities and I've made a few variations, all of which have worked and are based on the basic slim driver idea.

There are other obvious advantages to the concept: easier installation and the pickup/driver combo to name just two...but the design principle was really to make the thing work. All my drivers (4 now) based on this design with simple amplification have worked well, I can only say that with each guitar there will be differences.

Jim...I do think you'll need more preamp power with half a humbucker, Jim...just as I did with my lower powered strat...personally I'd use both coils to the driver circuitry for power and noise reduction to improve efficiency.

But, the main problems seem to stem from not sticking to the design principles above. Pekko has done wonders though it seems by sticking to the formula...

I got all the strings sustaining easily, also the high e. I was able to turn the gain trimmer of LM386 to minimum and there was still more than enough gain available for sustainer. Only problem was that the system was prone to squealing, especially in harmonic mode. I added some shielding to driver wires and experimented with magnetic shunting with piece of steel. I found out that the steel piece was most effective when placed as near the bridge pickup as possible. After these procedures I could increase gain enough to sustain every string in fundamental and harmonic mode without having to fear about squealing.

See, very similar in design. Now there's a whole bunch of things that could be explored besides the steel shunt to stop that squeeling, but hey, if it works...

-----------------------------------------------------------------------------------------------

Anyway, I'm very proud of you guys for taking the project up so enthusiastically and keeping the tradition of the occassional long post to explain it all...there's been thousands of recent hit's and that's pretty amazing.

Now, you may not have noticed, but I've gone a little AWOL. Unfortunately, I'm going through a servere personal crisis and don't have access to my stuff (as I'm currently isolated on an island...in the middle of winter...)

I had some really great plans for this thing but they will be on hold now...at least for the forseeable future. I believe I may have had a breakthrough on the circuit, just before I left. This would be very small and efficient and completely "pop-less". I also was preparing moulds to make slim drivers exactly to the size of various pickups. The intention is that a kit of a premade circuit and driver could be made to order. If I survive these other matters, I may well return with these things...but only time will tell, and it could be quite a while...

So, I hope that you all continue the work and I'm just glad to have been able to present something that does work and is an approachable DIY project. I'll be looking in from time to time when I can, but computer access is limited too (like heat and radio reception), so for now...it's a kind of goodbye...at least for now...

best of luck with it all, and to you all :D

psw aka pete

Link to comment
Share on other sites

Thanks, Tim - now I guess I'm gonna have to rig up a winder of some kind!  :D I'll put it on the list!

Hope you don't think you need a winder to make the driver...even though mine are slim, there are not a lot of turns (couple of hundred) and the wire is not fragile. All my drivers have been made without them, it's quite easy to do and allows you to pot as you go...If you try that with a winder you'll likely spray the PVA, nail varnish, or whatever all over the place.... :D

BTW...I suggested PVA (woodworking white glue) as it's non toxic, drys clear and scrapes easily off a plastic bobbin or further reinforces a paper/cardboard one. However it's not ideal, it's a little flexible and becomes more so with heat (which the driver can generate in it's windings).

A completely epoxy potted driver would be better but it's less likely to seep all the way into the windings if not done while winding it. If you try that though you'll find your finger's stuck to it. PVA is good also because, if there is movement in the windings, it wont strip the enamel and short the coil like something really tenacious....

Potting is very important, and potting of other pickups would help prevent some of the squeeling too...I have good reason to believe that much of the woes of the test strat was due to some microphonics in the pickups, as they weren't so apparent in say the LP. Even the slightest movement of windings will generate some kind of signal...I imagine that this is something that the EMG's will have an advantage over other unpotted pickups.

BTW, you need some kind of a buffer to send one pickup's signal to the Ruby and output?

Is that a question...? The fetzer/ruby has a preamp,buffer,gain boost. The standard ruby also has a buffer (no boost). Boost is better (unless EMG and high powered pickups perhaps), but the main thing is to prevent loading on the pickup. All the circuitry for the sustainer is completely separate from the signal to the amp...so there is no buffer for the guitar's output, if that's what you mean.

Basically there's a line hardwired from the pickup to the circuit with the hot disconnected by the switch, a line is also taken off for the pickup to the guitar's output. The activation switch has a bypass function to switch off other pickups. On a one pickup guitar this wouldn't be necessary, just have the switch turn on the circuit's power...

Hope that helps...p

Link to comment
Share on other sites

Now, Pete mentioned something about having a different preamp for his 386 amp for more gain. I'm still not getting satisfactory results but I'm getting closer and closer. I thought of rising gain capability on the Ruby. How to do that, what resistors to swap?

-----------

Jimbo

Link to comment
Share on other sites

I'll be looking in from time to time when I can, but computer access is limited too (like heat and radio reception), so for now...it's a kind of goodbye...at least for now...

best of luck with it all, and to you all :D

psw aka pete

Hey LK...there's no way of stoping the "tonesuck" on a standard LM386 type amp (ie no preamp is there)...???

Jim...

You'll have to go back a few pages...I did post a link to the modified kit, etc that I use. There's a schematic and really good circuit description on the PDF too! Alittle large for my liking, otherwise it works well. It has quite a bit of filtering capacity to knock out stuff out of the audio spectrum and the gain can be adjusted by replacing a resistor, which I did. This would not be necessary I feel with a standard HB or other high output pickup. I've even had it working without a preamp at all but you do get "tonesuck"....!

LK your man for circuit mods...I've never used the Ruby at all, and if I get back to it, I'm intending to ditch the LM386 too..>! :D

For you Jim, I'd be taking the full power from that Humbucker instead of only one coil...there's a whole lot more noise free power right there without preamping further. You can still split the pickup for the actual output to the amp for the single coil sound...so why not...?

Anyway...back to the wildlife and the stary skies at night....P

Link to comment
Share on other sites

So yea, I built it like HB -> Sparkle boost -> poweramp of Ruby. I must say, I'm getting some good stuff going on now.

I have the sustain control between the boost and the 386 section, with the Sparkle gain hard wired to max. One bad though... I can only turn the sustain knob up to the limit of infinite feedback. If I get too much further than that, it starts oscillating. I see two solutions: either try to trim the gains of Sparkle boost and poweramp to the optimal clipping/power balance, or totally isolate the sustainer circuirity from the signal one. What would stop the oscillation? I need to start working on it because we're hitting a studio in a few weeks and I'd be really happy to complete this gadget before that.

The good news are that it looks very promising as it is.

---------

Jimbo

Edited by JimRayden
Link to comment
Share on other sites

Is the oscillation really high-pitched squealing? It may be that taming the high end with a bit of lowpass filtering before the 386 might take care of the problem. There's also that Zobel network that the datasheet recommends (a resistor and cap in series from output to ground) that might be worth a try. IIRC, it's not included in the Ruby design, but it's in the sample setup on the datasheet.

To answer your question, Pete, yes, the 386 is not going to perform as designed with just a guitar input - it requires a substantially lower input impedance, so some kind of buffer or preamp is going to be necessary.

Link to comment
Share on other sites

Well, that means pre-filtering probably isn't going to do any good. Try adding a 10 ohm 1/2 watt resistor and a .047uF 25v cap in series between pins 4 and 5 (the Zobel network I was referring to earlier). It might also benefit you to bypass the power supply pin to ground with a 100uF or larger electrolytic (make sure you get the polarity right!) and a .1uF film or ceramic cap in parallel, as close as physically possible to the pin. Also try to make sure your driver cable doesn't come anywhere near the guitar electronics or the pickups - you're pumping 300 to 500mW to the driver coil, so it's going to be fairly easy to bleed a little signal back to the input, and it only takes a little to start oscillations. Sorry, but without seeing your setup, that's all I got, and those are at best "educated" guesses. Anyway, maybe some combination of the above will be a solution. :D

Link to comment
Share on other sites

Don't overlook the possibility of plain ordinary microphonics either, Jim. I gather you've not got a neck pickup, but those symptoms are at least similar to what I was experiencing untill I arranged the switching to completely (ground and hot) remove the other pickups from the circuit...

So...does your sustainer now get the whole HB power now?...You may very well get improved performance from making sure there is no loose windings in that HB that may be generating this noise. Try tightly wrapping electricians tape about bot coils of the HB.

The driver too may be the culprut... Any loose windings in the driver will create such noise signals which, due to there highly electromagnetic nature, will surely enter the signal chain. Pekko used an iron block to shunt such signals...that's an option too.

Again...keep refining the driver, especially this potting aspect for optimum results.

On the circuit, you could try the Zobel Network we had discussed if it isn't already present in your circuit...there are other caps that can be added to the LM386 circuit that are intended to help too...there's some schematic somewhere here of what I'm using that has these extra caps to ground (essentially) intended for that purpose. For instance have you simply bridged pins 1 and 8 with wire for 200x amplification, or have you used the suggested 10uF cap to help with such oscillations.

Also...brute power will simply radiate more EMI so the answer may ultimately lie in better efficiency...again, look to the driver!

Another thing is the output capacitor...mine is a 220uF but I've seen larger and have tried combinations myself of smaller and larger caps to make a suitable value (in my case in an effort to reduce size with SMD's). Thjis will lower the high frequency response...

One bad though... I can only turn the sustain knob up to the limit of infinite feedback.

I'm not sure what that means...you turn it up full blast to get infinite sustain or the opposite. With my LP experiments I found that I had a little too much preamp gain and had to have the "sustain gain" knob all the way down. Reducing the preamp gain allowed a better response, better performance and a full range of control allowing the LM386 to do the poweramp work as it should...

So yea, I built it like HB -> Sparkle boost -> poweramp of Ruby. I must say, I'm getting some good stuff going on now.

So, your sparkle boost is the preamp...not sure you really want it hard wired to the max...how about a trim pot as in my "sustain box"...a 1k between pins 1 and 8 along with the 10uF cap will adjust the LM386's gain nicely...so you'll have ultimate control.

.... or totally isolate the sustainer circuirity from the signal one.

This concerns me...what does this mean, a separate pickup for the sustainer??? The thing really should be separate from the signal and taken befor any controls on the guitar...a straight feed from the pickup and should be invisible to the guitar's signal...that's why true bypass is not an issue...the guitar's signal does not go through the circuit as on a stompbox...

If it's not isolated from the signal try adjusting the guitar's tone control to see it's effect. I was able to cure these problems by this means and then added such a filter to the input to the circuit...this is not ideal as it changes the tone of the guitar while in sustain mode...eventually, by removing the neck and mid pickups, the problem was solved on the test strat and I've not had to do this again...mainly as I discovered the problem was with the unused pickups and was fixed by removing them from the circuit!

Anyway...must run...goodluck...p

Link to comment
Share on other sites

LK or other's...

It was discussed that the fetzer stage of Galaga Mikes Tutorial circuit has a trim pot to adjust for manufacturer inconsistancies of the transistor so that it's supply is balanced at half the supply volatge...

I wonder though...what is the effect of an imbalance of supply and how unbalanced is it likely to be. I've alway's messed around with an op-amp solution because of it's internal balance and percieved sophistication and forgiveness, but then nothing I've made seems to have worked properly...hmmm

So I've thought I might try working with the discrete components but the need for a trim pot kind of puts me off...not sure why (size possibly)...in my work (should I get back to it) on making a universal standard amplifier for sustainer purposes.

As always, the preamp is a stumbling block for me...

Jim, what is this "sparkle boost"...I take it that it's a preamp boost...got any details, schematics or links for it...??? Does this too need the trim pot or is it IC based...

p

Link to comment
Share on other sites

Well Pete, you can use a trimpot (or a regular pot) at first to measure how much resistance you need for biasing the circuit, then replace that pot with a fixed resistor. You won't need to re-bias the circuit unless you replace the transistor, so a resistor will do the job.

By the way, i'm now officially working on a sustainer :D

...

emre

edit: Forgot to mention that there are also some circuits that bias themselves, like AMZ's mini booster. But they're more complicated (more than one transistor) circuits as a result of self-biasing.

Edited by transient
Link to comment
Share on other sites

Fantastic emre...

I've had some people ask if it would work with piezo's on an acoustic...are you likely to be doing this or able to test it out with your ghost saddles?

Thanks for the tip on the resistor...I'm actually looking towards manufacturing a standard circuit and want to keep it small and simple as possible. As always the preamp is a stumbling block...I haven't given up on the IC option...but a single transistor is pretty simple and inheirently small...except for that trimmer! It's mainly the hassle, I dont want to install and adjust the thing for each "unit"... :D

On that subject...would people be interested do you think in such a device...it would seem to open the door to more involvement if the circuit was sorted...it would certainly give more room for development if the circuitry was mare standardized and smaller.

As what I'm considering using is not as commonly available or as cheap as the LM386, but the performance far superior, it's probably the way to go to have the thing supplied and tested. It's looking to be a block about 35mmx35mmx5mm so it should fit just about anywhere... :D

If there is enough interest I have a contact that makes custom effects boxes with reliable internet sales outlet that could make and supply them...

Any thoughts on this idea???

psw

Link to comment
Share on other sites

What transient fails to mention is that you'll have to trim and replace that resistor for every single JFET you use. The only feasible way to use them in production is to do what the audio manufacturers have done for years, buy many thousands and sort them for the ones that will bias in your circuit. Why not use a simple bipolar transistor pre, a la the LPB-1?

LPB_style_preamp.png

(this isn't actually a real LPB-1, but it is very similar)

It works fine with magnetic pickups (might be iffy with piezos), and it's low-parts-count, quiet and gives almost 30 dB of boost. Above all, a 2N5089 is going to bias reliably close, regardless of when or where it was manufactured - even if you lose a little high end (unlikely, but possible) and the distortionisn't quite as pretty, it really doesn't make any difference if you're just driving a sustainer, because that signal is never going to get anywhere near the output signal (if you're lucky! :D ). Just my $.02US, adjusted for inflation and devaluation.

Link to comment
Share on other sites

Yes, LK is right (isn't he always?) about that. I thought you wanted to build a single circuit for yourself. When it comes to manufacturing, FETs are hard to bias, as every unit will have to be biased differently.

I plan to use the piezo output for the sustainer. I want to install the sustainer in the middle pickup position on my fretless, and it would be hard to get it to work fine there if i feed the magnetic pickup output to the driver.

.

e

Link to comment
Share on other sites

Well, that means pre-filtering probably isn't going to do any good. Try adding a 10 ohm 1/2 watt resistor and a .047uF 25v cap in series between pins 4 and 5 (the Zobel network I was referring to earlier). It might also benefit you to bypass the power supply pin to ground with a 100uF or larger electrolytic (make sure you get the polarity right!) and a .1uF film or ceramic cap in parallel, as close as physically possible to the pin. Also try to make sure your driver cable doesn't come anywhere near the guitar electronics or the pickups - you're pumping 300 to 500mW to the driver coil, so it's going to be fairly easy to bleed a little signal back to the input, and it only takes a little to start oscillations. Sorry, but without seeing your setup, that's all I got, and those are at best "educated" guesses. Anyway, maybe some combination of the above will be a solution.  :D

I copied this post to my notepad and am starting to try the tricks right away.

I think the major problem is that my driver wire goes with all the wires and even passes the bridge pickup. I'll try taking the wire externally. The guitar will look like Civil War afterwards but at least it'll be working. :D Oh boy, here's my chance to use duct tape on my guitar. (I've always wanted to have a reason for that. B))

Thanks to you too, Pete. Too bad that your posts are too long to post. But I'm certainly taking notes from all of them.

----------------

Jimbo

Edited by JimRayden
Link to comment
Share on other sites

:DWell, once is enough for any of my posts...always cut and paste... B)

It's all a bit frustrating as so many of these things I've had to go through myself and no amount of "posting" seems to help...it seems experience and experimentation are the only answers...

I'm still keen to get feedback on the idea of a manufactured circuit for the purpose...Any comments....

Also...Thanks for the stuff with the preamp circuits...Although the part's count seems low, that type of circuit still has more complexity than I'd like and I suspect an easier, lower parts count and superior performance could be obtained with an op-amp...so perhaps I should after all continue with this option... :D

Emre...Check out Dizzy's mid pickup sustainer at diystompboxes...

DIY Sustainer thread

The harmonics are less predictable and the circuitry and everything (twin drivers, etc) is way more complicated (and secret) but the general ideas are detailed in the patents...

I'm really keen to see how a piezo source would perform with the sustainer concept...there maybe a need to calm down those quacky, high frequency sounds but I don't see why it wouldn't work fine...

Another idea that no one's really explored yet is having the driver alone on the guitar and everything else in a stompbox with the signal going to the guitar via a separate lead or some novel kind of lead arrangement. An interesting idea but deceptively tricky to pull off...might be worth further exploration though...imagine a wha type pedal which controls sensitivity and a harmonic stomp switch combined with a proprietory driver or pickup/driver arrangement....hmmm

Your problems will be with interferance between the output and the EMI inducing driver leads if you attempt to use a single cable and possible problems with the length of the driver leads... A separate box though would solve the onboard power problems though and the necessary miniturizing of the circuits....food for thought....!

psw

Link to comment
Share on other sites

...Although the part's count seems low, that type of circuit still has more complexity than I'd like and I suspect an easier, lower parts count and superior performance could be obtained with an op-amp...
:D OK, just for giggles, let's look at a minimal opamp preamp.

opamp_pre.png

The parts count looks identical to me, and it's decidedly more complex, at least as far as wiring is concerned (5 pins versus 3). The four resistors, two caps and pot are going to take up the same amount of space, so unless you an find an opamp with a smaller footprint than a single transistor, there's no space savings to be had. And as for performance, since this signal should never reach the output, it's really a non-issue. Finally, a single transistor is almost always going to draw a lot less current at any feasible output level, and it's a less expensive solution, especially in quantity. Again, just my too scents! :D

Edited by lovekraft
Link to comment
Share on other sites

OK...you got me...I'd forgotten something else about my "new" non-LM386 circuit idea...it's got more power so a simple buffer would do the trick...similar to the original ruby buffer stage I guess...any ideas on this???

As a buffer, though I still don't quite get why, the tranny stage doesn't require the trim pot either...

Anyway...good point...once established where I can return to work on this I'll get right on it...p

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...