Active pickup build

[nakedzen]

@curtisa I'm getting pretty much the same amount of emi from the active pickup as from passive humbuckers, from pc cases and speakers. EMG's are dead quiet in comparison.

I didn't have ground planes built into the prototype pcb, so that might explain it.

[curtisa]

Are you wiring it up as per EMGs original outline? Note that this is not the typical way a standard humbucker is connected (from the Electrosmash website):

Note that the midpoint of the two humbucker coils are connected to ground, which is unusual for a normal humbucker in a typical (passive) guitar. Not shown in the diagram, but the polarity of the coils is important as well, as it relies on the two outputs of each coil to be in anti-phase with each other. You may have to experiment with the wiring of your pickup to get the right polarity for each coil.

[nakedzen]

I'll need to check how I did them, I can't remember right now.  Been too busy fiddling with the pcb layout.

[nakedzen]

@curtisa You're right, I had them wired wrong on that side. Unfortunately my soldering iron decided to crap out, so I can't truly try how it affected anything, but just by taping the connections I hear it's now about the same in terms of emi noise as the EMG neck pickup in the same guitar.

But I think I'll still do a third prototype PCB with the ground planes around the components as a faraday cage, should be much better in terms of noise. My first prototypes only had the ground as traces which isn't really too smart in hindsight.  

[Prostheta]

On 5/30/2023 at 9:54 PM, mistermikev said:

well no expert here... but I can relay what I've read, and that is that the emg version of a difference amp actually is running the two coils of a humbucker into sep inputs... then dropping voltage for anything they have in common, and amplifying anything they don't have in common.  This keeps the noise floor really low.  it requires some really underwound singles.

the schematic you linked to... that's just a garden variety booster.  anything at the input is being boosted -noise included.  Just like a pedal running after the guitar.  Nothing wrong w that... just isn't anything like what is going on in an emg pickup (not being critical here... just observation).  you could, essentially take any booster circuit and put it in a guitar.  I've done it with micro booster, linear power booster, mid boosters, sho boost... and enjoyed it quite a bit.  IMO it's a very useful and cool thing to do.

 

All of this. Interesting project! From what I am aware, the original EMG preamp based around the LM4250 is a circuit that is current-limited by function of that silicon. This has the effect of reducing high frequency response, which is fine in the audio spectrum.

A bit of an aside, but here's a preamp from a 1980s Aria Pro II SB-Integra which I reproduced a few years back:

This preamp was external to the pickups, however those pickups were also made similar to the approach EMG use in their 81s by being hugely underwound, or at least using lower gauge wire in the coils. The circuit is hard-coded for 6db of amplification per pickup (10x gain). This also works as a differential amplifier. Rather than having the circuitry within the pickups themselves (a P and a J) we had this outside of the pickups. The silicon is a plain old JFET TL072CN, but in principle could work using an 074 or even an 064 for lower current with a bit more noise. Like @curtisa states, the function of a differential amplifier is to reject common mode noise. This can be done by removing reference to earth and running a differential amplifier off a split supply or creating a floating reference as per EMG and this circuit with a single-ended supply. The Aria Pro II SB bass project I made a while back uses the latter approach, so has 2x PP3 batteries for a -9v/0v/+9v split supply, with a nice amount of headroom. The pickups are floated with no reference to ground and they're inky-black silent. It also worked wonderfully for the single coil in my 5-string 1951 P-bass.

[Prostheta]

The LM4250 doesn't really offer anything specific beyond being able to be massively current limited for immense battery life. I mean, that's a great thing but I started running all my EMG circuits with 18v as they have very little headroom. I don't know whether it's the current limiting and frequency response that makes them run less linearly, or whether the headroom within the circuit is less than ideal (I suspect the former) but 18v opens them up a lot, especially with basses but also the 81 as well.

Interesting aside. Fishman Fluence don't discernibly change in response to being run with 18v which makes the double Gotoh battery box on the back of my white Mirage a bit superfluous to needs....

[nakedzen]

Could just be the way Emgs are constructed, lower wound mini humbucker coils and steel bar instead of slugs. Although I did read somewhere that the coils are actually wound pretty hot around 10k. Who knows, too much misinformation around.

 

We'll see what happens in the future since the production of LM4250 has ended. Maybe emg just makes copies of it. But afaik the chip shouldn't have any "sound" to it as long it's kept clean and not driven into distortion so any chip works that has the same pinout?

[Prostheta]

DCR is only a rough hint at the output of a coil, however it does conveniently sidestep the complicated aspects of a pickup such as the field strength, inductance, etc. 10k is still pretty hefty.

Yes, it looks like the LM4250 is out of production so I wonder how EMG are sidestepping that one? Yes, a chip shouldn't have its own sound in the regions where it is running linearly however the tradeoff of the LM4250 is that it has a lower frequency range when run in ultra-low current ranges (). Without researching further, this may exhibit high-end rolloff, crossover/slewing distortion or perhaps reduced headroom. Not sure. None of these things are good in the way that we otherwise characterise the aspects of other pickups.

Given the schematic that Andrew @curtisa posted above, the quiescent current of the LM4250 with R8 at 1MOhm is ridiculously low. Using a different op-amp would indeed work in principle, however you'd be lucky to get that silly battery life that EMGs provide. Sticking with TI, I would consider something more modern and quality. Briefly looking at the site, I'd say that the OPA1637 might be an interesting one to test. It is very much a precision part but not super hungry either.

https://www.ti.com/product/OPA1637

Not the most expensive op-amp I've seen (I use some that cost 3x that in my "no holds barred" preamps) and certainly worth a look. I'd snag some TSSOP8 > DIP8 PCBs to do a bit of testing.

[nakedzen]

Couple sound samples. (Metal)

Steel bars, alnico 5, brass baseplate

Pole screws, alnico 5, brass baseplate

 

Another thought, I've been wondering how a preamp board for single coils would look like? This pcb seems to work when I connect the hot to pin 2 and ground to ground but probably not the best way to do this?

[Prostheta]

I've never seen preamps set in singles before, however from what I am aware Fishman's Fluence singles have a "master" pickup that the others slave into, presumably sharing one preamp. It still requires that one pickup has the preamp, however this does imply that you can consider moving the preamp offboard. This isn't anything novel, since a preamp separate to the pickup itself is no different to one being onboard....in principle an EMG81 could have the preamp externally to the casing, however this then becomes a question about how many connections the preamp needs to the coil. Fluence humbuckers are probably very deeply integrated with the "coils" at several points during their "wind" in order to manipulate the voicing.

I'm of the opinion that a preamp within the pickup form factor isn't necessary unless factors such as the Fluence humbucker necessitate it. Actives also have a really narrow variety compared to passive. This is where I'm getting into personal opinion. I'd rather have an in-cavity preamp paired with readily-available passives unless a certain active delivers out of the box. I'm yet to weigh in with the Fluence Classic neck pickup and how smoothly it pairs with a rolled-back tone control, specifically with how well compares to a simple Seymour Duncan '59 SH-1N, my favourite neck pickup.

On 9/13/2023 at 1:59 PM, nakedzen said:

@curtisa I'm getting pretty much the same amount of emi from the active pickup as from passive humbuckers, from pc cases and speakers. EMG's are dead quiet in comparison.

I didn't have ground planes built into the prototype pcb, so that might explain it.

Did you manage to get past this specific issue? An op-amp circuit running in true differential mode connected to either end of the coils with no reference to ground being made usually works very nicely. I can send you a preamp (albeit one that requires 2x 9v batteries) to experiment with that does this. Bang the pickup coils either side of two pins and instant inky-black quietness is achieved.

[Prostheta]

Amusingly, these are all in a Fishman Fluence box.

[nakedzen]

5 hours ago, Prostheta said:

Did you manage to get past this specific issue? An op-amp circuit running in true differential mode connected to either end of the coils with no reference to ground being made usually works very nicely. I can send you a preamp (albeit one that requires 2x 9v batteries) to experiment with that does this. Bang the pickup coils either side of two pins and instant inky-black quietness is achieved.

Well yes and no. The blade style pickup is dead quiet, but the pole piece style not. They're both built identically so I'm not quite sure what's up.

So it would be better to disconnect the ground wire from the baseplate in your opinion? I don't have the noisy pickup here atm but once I get to it I could try that. Sorry I'm a complete pleb when it comes to circuit design.

But next steps are pretty clear:

- PCB with ground planes built around the components to act as a faraday cage.

- Solder a proper bridge ground, now I just have a piece of string taped between the bridge and the pickup height adjustment

- Shielding on the pickup and pcb

- Wax potting

[nakedzen]

Blech, I found the cause of the noise. There was some sort of ground fault connecting to the shielding paint on the guitar. I thought something was off when I got noise even without any pickups connected.  

[Prostheta]

I used to build in a brass mesh to my preamps bent over the board as an attempt at a Faraday cage. I didn't notice any difference. A strong ground plane both sides seems good enough. Your recordings sound great. Are you noticing any particular characteristics, such as sweeter sustain, blooming notes evolving in sustain, etc? 

[nakedzen]

I usually at least try to avoid fooling myself with audio so I haven't thought about those differences really, you start hearing what you want to hear (best pickup ever!!!?!)  I gave the pickups on loan to a friend to mess with them and give me his input.

Offtopic but here's a good one: https://www.youtube.com/watch?v=BYTlN6wjcvQ

But I feel it is more rewarding to play somehow, like when I swapped an 85 in the guitar it just felt "dry" I guess is the word. A bit more harsh by comparison.

[Prostheta]

I think that's a common description, for better or for worse. Whether this is down to any tonal colouration by the preamp or the coil being very "by the numbers and uninteresting" is debatable. I certainly think that both can be tested if coils can be extracted without destroying them. I'm of the opinion that taking the old EMG preamp, running it at 18v and it using a passive pickup like say, an SD Invader would be a nice combination. I don't see how the preamp is a bad design aside from the disco'ed parts. I can't recall off the top of my head if the EMG coil output is comparable to a passive, or whether they are lower and the preamp pickups up the slack or not. There's little unique magic going on here aside from the ultra-conservative current limiting.

Regardless, I'm a strong believer in active circuits even if the electronics simply function as a "line driver". This being said, the compromise is losing some of the niche things that passive enthusiasts love, such as the interaction between coils when dialling in/out the volume/tone pots. This can be allayed by putting the active electronics downstream of the coils, switches and pots but by that point noise is baked in and just becomes powerful low impedance noise instead!

I'm in the classroom today, so can't shouldn't check out the sample until later.

[nakedzen]

Another thing I just noticed about the pics on electrosmash, it almost looks like the blades of the pickup are connected to the copper of the pcb?

Could be just from taking it apart that the pcb tore at those spots but curious anyway.

Then there are those yellow parts over the coils on the leftmost pic. Some sort of shielding?

 

[Prostheta]

Could be. I deal more with electronics than making pickups, so not really my end of the pool.

[henrim]

To me it looks like that there is epoxy (or whatever) in between the blades and the pcb. Could those yellow straps be just for holding the bobbins together when casting?

[nakedzen]

That makes sense too.

Another thing I noticed, is the ground connection seems to be from the other side of the pcb to the output center pin, and ground from l4250 pin 4 (upside down in the pic) goes through a via to the other side. So probably the entire pcb backside is a ground plane.

 

[Prostheta]

A cursory look at the schematic suggests that this might be the case. The resistor 1004 (1MOhm) and the adjacent capacitor don't have corresponding links in the schematic other than earth. The through hole plating will go through to a common plane, so yes, ground plane. The tape is probably just to hold the steel bars in place, even though the magnet under the bobbins clipped into the PCB should do this anyway.

[nakedzen]

I think I'll swap my design for the TL071H, since the availability of LM4250 seems to be sketchy. Same pinout and it doesn't need the bias resistor so I get a cleaner design on the board too.

[Prostheta]

How about a TLV9041 instead?

https://www.ti.com/product/TLV9041

Ultra low operating current and can be run from a lithium battery. 5,5v maximum supply is more than you need.

[nakedzen]

I'd rather make these as a universal swap-in for other actives. Changing from 9V to 5.5V would require rewiring (or routing) the battery holder.

The noise figures show those have 16dB less CMRR, I think I'll go with the TL071H, it's tried and true, at least for this first project.

Those could be interesting if I ever decide for multiple chips, but I do like the form factor I have now.

[nakedzen]

Couple samples of the TL071H version. Sound changed pretty significantly from just changing the assembly. New version has a thicker brass plate. The bottom plate is now mdf since my laser couldn't cut through the 4mm plastic I had in mind for the bottom, need to upgrade my laser.