onboard buffer for BASS guitar (asking for advice) [documentation]

[jeepe]

hi,

I'd like to ask for advice,

I'd like to put a buffer in a bass, but as it turns out, it is not so simple, cause despite the common knowledge that bass pickups will generate signal of about 100mV... or maybe up to 750mV.... people who use stroboscopes keep mentioning
voltages way over 3 volts, peek to peek (and I think I understand that 750mV means 1.5 Volts peek to peek)...

and the other thing is that the j-FET buffers, which can still be made as some shops still sell j-FETs,
have a limitation in terms of biasing voltage, and because of that, headroom...

OpAmps, as far as I understand can be easily used regarding headroom, as the source voltage can be doubled and by that the headroom would be "doubled", too...
but they consume much more, and drain the battery relatively quick...
therefore I'd not go in this direction...

there is Bajaman's workout of the Music Man preamp using that LM4250, consuming 44microAMP, which is a cool circuit, no doubt, but I'd like to go the easy and elegant way, like this guy at talkbass, having built a Tillman/Kreuzer j-FET buffer with a BF245 see here

there is a topic wit R.G.'s onboard guitar buffer, here, someone built it, and even posted oscilloscope data,
showing no clipping at 4.8V peek to peek, although it is optimized for 6V source voltage

I have a hot pickup, SPB-3, and I'd like to try other hot pickups in the future...
what would be a good onboard buffer for anyone in this situation, including me?

thanks in advance,

Peter

PS:
I was reading a thread here about buffers over at diystompboxes forum, and DavidRavenMoon (a person whom I'm getting to respect more and more for his knowledge and experience) mentioned that he had put a buffer in his bass for the first time in 1977

[FiveseveN]

All the commercial active pickups, onboard buffers and preamps use op amps and nobody complains. You have low power consumption ones too. Or go with R.G.'s if you want to be fancy.

[mauman]

Your pickups are passive, and even if they're overwound they are probably not any hotter than guitar humbuckers. I'd figure less than 500 mV (1V peak to peak) so you would have plenty of headroom with a 9V supply and 4.5 V biasing.

[jeepe]

I think I'd try R.G.'s....
especially optimized for 9v, or 12....
in this thread he offered to explain how to do that to anyone who would be interested... unfortunately nobody replied "please!"

[luix]

The more you beat your strings the more will be the signal level, I've measured humbucking guitar pickup with scope and the level goes to 2.5V peak for small time, since our hears are really sensible to any distortion my advice is to use the classic BJT bootstrapped buffer circuit (AKA Cornish) supplied with a voltage doubler like LT1054 that works with ultrasonic frequency.

Since the input impedance is bootstrapped your pickup will see something like ten times the 120k input resistance (R3), lower resistance means lower noise so you can lower that resistance and adjust the cap around meeting to your taste, maybe use LTSpice for simulating it, a good choice could be limiting the lowpass frequency to 15kHz if the charge pump works at 20kHz.

To improve filtering the power supply after the charge pump you can use:
1) a capacitance multiplier filter
2) a CLC filter, inductance will be really small, 47uH+47uF form a lowpass at about 3.4kHz and since the current consuption is really low you can use that resistor-like commercial inductance.

Using a low noise BJT like BC550C will help also, since it is the first stage of the chain the noise contribution will be multiplied by all the gain stages of the chain so it should be as quietest as possible.

Please remember that DIY is not about cost saving so don't skimp on component quality or design solution, leave the cost saving tasks to the commercial pedal builder.

[jeepe]

[...]I've measured humbucking guitar pickup with scope and the level goes to 2.5V peak for small time, since our hears are really sensible to any distortion my advice is to use the classic BJT bootstrapped buffer circuit (AKA Cornish) supplied with a voltage doubler like LT1054 that works with ultrasonic frequency.

thank you very much!! this is exactly what I needed...
a sudden push in one direction with some palpable doze of knowledge, so to speak
I really didn't want to put together a nice one j-FET buffer hoping / believing that
it won't clip...

I'll try a 7660s...

Please remember that DIY is not about cost saving so don't skimp on component quality or design solution, leave the cost saving tasks to the commercial pedal builder.

thank you for reminding me!! absolutely love this concept!
I go for 7660 cause I have them, I have NE555s, too, so, the 7660s is the winner,
however, I couldn't even buy LT1054

a further question:

will doubling the voltage just raise the headroom, like in OpAmps' case?
I mean, will I have to change some resistors?

thanks very much!

[Tassieviking]

Have you looked at Passinwinds preamp at TalkBass ?
https://www.talkbass.com/threads/the-pa ... p.1259692/

[jeepe]

Have you looked at Passinwinds preamp at TalkBass ?
https://www.talkbass.com/threads/the-pa ... p.1259692/

Not until now, but I see no schematic there anyway...
and I'd like "only" a buffer
if it comes to an onboard preamp, I'll go with the Stingray by Bajaman, or the Sadowsky
but thank you!
I'm already convinced, looking at the price and the descriptions, that the LT1352/51 must be very-very good chips!

[deltafred]

Link to wiki with all schematics in the first post - https://www.talkbass.com/wiki/pw3b-lpf- ... rd-preamp/

[Jarno]

As a bassplayer, I've tried discrete preamps, but still had the signal run into the rails in signal peaks. Really fond of the musicman stingray preamp that Bajaman posted, it sounds great, and uses very little current.

[jeepe]

Link to wiki with all schematics in the first post - https://www.talkbass.com/wiki/pw3b-lpf- ... rd-preamp/

I have Only didn't realize the were "clickable" thanks!!

EDIT: I couldn't find the schematic,
but sure, if you have the bill of parts, the PCB layout, you have the thing

EDIT 2: I am a headcase... "The current schematic is always accessible from this link: passinwind.com/PW3B_LPF_OpenSourceLicense.html" ...
the "opensourcelicence.html" mislead me, but I am a headcase, no excuse

[jeepe]

thanks, Deltafred and Tassieviking for the Passinwind project link,
I think it is beautiful project!

[luix]

a further question:

will doubling the voltage just raise the headroom, like in OpAmps' case?
I mean, will I have to change some resistors?

thanks very much!

Doubling the supply voltage will avoid clipping in case of hard slapping on the string.

Probably the bias resistors should be adjusted but in this case LTSpace will help you.

[Jarno]

Not a headcase at all Jeepe, initially I also couldn't find the schematic (or maybe we both have cognitive disfunctionalities ).

Interesting schematic of that preamp, I myself am also working on an onboard preamp with a lowpass filter. Using the schematic by cAMPUS:
https://www.musikding.rocks/index.php?a ... plete-pdf/

But have two of these on a board (one per pickup) with a mixer on the end, still needs some components so board is not verified yet. Alembic series-esque functionality

[jeepe]

a further question:

will doubling the voltage just raise the headroom, like in OpAmps' case?
I mean, will I have to change some resistors?

thanks very much!

Doubling the supply voltage will avoid clipping in case of hard slapping on the string.

Probably the bias resistors should be adjusted but in this case LTSpace will help you.

my question was stupid, but I meant, whether I'd only have to raise the voltage and the headroom would rise, and that'd be it, or
i'd have to tweak with biasing.... as this appears to be the case

as a Linux fan and believer, I can't use LTSpice (I do have a copy of windows on one of my machines though)

but which resistor is to be adjusted? R3/ R4/R5? and how? if I may ask?

PS: Looking at web search results, the cornish buffer is supposed to work with the range of 9-18v.. although some say, different voltages produce different sound quality...

[luix]

The academic bias point for symetric and maximum output swing is half the supply voltage, so if you have 9V you should have 4.5V on the output of the buffer (the emitter in bootstrap case).

Symulation says that wit 9V supply and the values above the emitter is sitting at 4.7V.

CNBuffer_9V.PNG (13.41 KiB) Viewed 884 times

Charge pump will realistically give you 16V so the optimum is 8V, using the same value the simulation gives about 9V so you can left the circuit in this way or lower R4 to 150k to have about 8V, using your hears will be the best way to solve this question

[jeepe]

The academic bias point for symetric and maximum output swing is half the supply voltage, so if you have 9V you should have 4.5V on the output of the buffer (the emitter in bootstrap case).

Symulation says that wit 9V supply and the values above the emitter is sitting at 4.7V.
CNBuffer_9V.PNG

Charge pump will realistically give you 16V so the optimum is 8V, using the same value the simulation gives about 9V so you can left the circuit in this way or lower R4 to 150k to have about 8V, using your hears will be the best way to solve this question

thank you, luix!!

for the explanation, too!!


PS: R4 now, the one that has to be lowered to 150K is in fact "R3" in the simulation schematic, right? (200K to ground)

[blackboarcult]

as a Linux fan and believer, I can't use LTSpice (I do have a copy of windows on one of my machines though)

I've ran LTspice on several distros through Wine/Winetricks/PlayOnLinux, no prob. There are a couple of minor limitations or bugs, but otherwise works like a charm. Currently using Fedora and couldn't be happier

[george giblet]

FYI:

The reason you want a buffer is to prevent loading on the input a signal. That can be caused by either a resistive or a capacitive load.

There's some hidden problems with *low current* buffers. In most DIY circuits the biasing is set with a resistor or current source. Suppose the desired voltage swing is 4Vpk-pk (2V pk). For a non inverting buffer like say RG's circuit the output device is biased at about 100uA. For positive signal swings it can output more current than the bias current. For negative signal swings it cannot pull down any more current than 100uA. That means it cannot drive a resistive load less than 2V / 100uA = 20k, otherwise it clips. That's OK to drive an amp and probably OK to drive most pedals. When you drive a capacitor the peak current is i = C_load * 2 * pi * f * Vpk, or C_load max = i / (2 * pi * f * Vpk). For a capacitive load instead of clipping you get slew-rate distortion. To drive a long guitar cable you might be driving say 1nF. If we assume an upper frequency limit of 20kHz then we need i = 250uA. A bass isn't going to be pushing out a full 20kHz. The pickups might only go out to 7kHz so that means i = 88uA. So you can see the 100uA in RG's circuit is about as low as you would want for driving most pedals and cables but maybe not both at the same time. You will get a bit more out of it because it's unlikely you will need full swing even at 7kHz.

What's the point of all this? The point is while low operating current seems desirable if you go too low you can end-up with distortion problems - the buffer fails to buffer sufficiently. If the buffer was driving a Baxandall tone control then you might have to consider even lower impedance loads.

The op amp based circuits don't suffer as much from the above issues because they have a push-pull output stage. The output stage has a strong drive for both polarities.

[jeepe]

as a Linux fan and believer, I can't use LTSpice (I do have a copy of windows on one of my machines though)

I've ran LTspice on several distros through Wine/Winetricks/PlayOnLinux, no prob. There are a couple of minor limitations or bugs, but otherwise works like a charm. Currently using Fedora and couldn't be happier

thanks a lot, blackboarcult!
of course I know of wine, only I have a switch in my head, which I set "off" years ago, meaning I'd never use it
BUT times have changed!
some years ago I never wanted to use what I used when using windows
but now this has been full success... windows really is past (if only this was true for 1/10th of the Western World!! )
so, I really have nothing against using wine anymore!!

so, thanks for reminding me of the great possibility offered by open source developers !!!
I'm gonna try it!