Markus2006 said:
The calculations, and examples given, ALL lacked a coupling capacitor on the output - so didn't apply to the 'real world' example we were discussing
It's no good quoting theoretical calculations on just a small part of a circuit, when the rest of the circuit completely changes how it behaves - as it very often does.
At the risk of been excessively 'crude' you could leave out the 100K, short out where the coupling capacitor is, and use the load resistor in the amplifier as the emitter resistor. No problems over it's value then - but it is a bit 'naughty' - which is why you don't see it done (unless anyone knows different?).
Cheers guys
mjf, thanks for your latest circuit, this version is similar to the circuit here using an MPF102...
http://www.hawestv.com/amp_projects/fet_preamp/fetpreamp2.htm
This circuit works quite well but the two stage circuit (FET input to emitter follower) looks like it could meet my requirements better regarding impedance matching and with a lower current consumption to boot.
One thing with this circuit, I am not sure if the ESD protection can be reduced to one zener however as surely this would clip the input when it swings negative and the zener becomes forward biased, limiting the input to 0.6 to 0.7v? A back to back combination will only clip when the input reaches the zener voltage of one zener + the forward voltage of the other zener, around 6.2 to 6.3v with a couple of 5.6v diodes.
I will post the circuit that I intend to knock up and play with as soon as I can and then I will get the bits and report back on my findings.
Cheers all
Ray
mjf, thanks for your latest circuit, this version is similar to the circuit here using an MPF102...
http://www.hawestv.com/amp_projects/fet_preamp/fetpreamp2.htm
This circuit works quite well but the two stage circuit (FET input to emitter follower) looks like it could meet my requirements better regarding impedance matching and with a lower current consumption to boot.
One thing with this circuit, I am not sure if the ESD protection can be reduced to one zener however as surely this would clip the input when it swings negative and the zener becomes forward biased, limiting the input to 0.6 to 0.7v? A back to back combination will only clip when the input reaches the zener voltage of one zener + the forward voltage of the other zener, around 6.2 to 6.3v with a couple of 5.6v diodes.
I will post the circuit that I intend to knock up and play with as soon as I can and then I will get the bits and report back on my findings.
Cheers all
Ray
hello ray.
i have only redrawn the schematic of post 1,but with a bypasscap c1.
yes,esd protection diodes are normally used "back to back" in audiocircuits (e. g. look at nelson pass preamps.........).clipping voltage is zenervoltage + 0,6v for a "halfswing",thats all wright.i prefer it too.
i think there is no great difference in powerconsumption............perhaps try to use a 9v accumulator(gives you 8,4v in reality....).
greetings...........
i have only redrawn the schematic of post 1,but with a bypasscap c1.
yes,esd protection diodes are normally used "back to back" in audiocircuits (e. g. look at nelson pass preamps.........).clipping voltage is zenervoltage + 0,6v for a "halfswing",thats all wright.i prefer it too.
i think there is no great difference in powerconsumption............perhaps try to use a 9v accumulator(gives you 8,4v in reality....).
greetings...........
Hi All
Seeing as it could be a while before I get the bits for the J201 version, I knocked up the circuit with what I had to hand to see if I had enough gain and how it sounded. I biased up a 2N3819 FET using Hugsley’s method to around 3.5mA source current and 4.7V across the drain resistor, and then added the emitter follower based on a BC108C and a 10K emitter resistor. As for the component values I used, the circuit is as per here…
http://www.diyaudio.com/forums/attachment.php?s=&postid=1711468&stamp=1231852917
Except…
R1 omitted
FET source resistor R3 = 750R bypassed with 10uF electro (gave just the right amount of gain)
FET drain resistor R2 = 1K5
Coupling cap C2 = 330nF
Q2 = BC108C
Emitter resistor R7 = 10K
I plugged my acoustic in to the circuit and out into the line input, set all EQ on the mixer to flat and outputted straight into the PC sound card (also set for flat response) and recorded the following sound sample...
FET sound WAV - 11mb
FET sound MP3 - 1.5mb
I think the sound is not bad considering it is dry and the circuit is definitely a goer. I will also try to find a day when I can plug it into the PA and see what it sounds like.
I will hopefully get something together with the J201 FET version soon (need to drop the power consumption) and I will try the same experiment and compare.
Cheers
Ray
Seeing as it could be a while before I get the bits for the J201 version, I knocked up the circuit with what I had to hand to see if I had enough gain and how it sounded. I biased up a 2N3819 FET using Hugsley’s method to around 3.5mA source current and 4.7V across the drain resistor, and then added the emitter follower based on a BC108C and a 10K emitter resistor. As for the component values I used, the circuit is as per here…
http://www.diyaudio.com/forums/attachment.php?s=&postid=1711468&stamp=1231852917
Except…
R1 omitted
FET source resistor R3 = 750R bypassed with 10uF electro (gave just the right amount of gain)
FET drain resistor R2 = 1K5
Coupling cap C2 = 330nF
Q2 = BC108C
Emitter resistor R7 = 10K
I plugged my acoustic in to the circuit and out into the line input, set all EQ on the mixer to flat and outputted straight into the PC sound card (also set for flat response) and recorded the following sound sample...
FET sound WAV - 11mb
FET sound MP3 - 1.5mb
I think the sound is not bad considering it is dry and the circuit is definitely a goer. I will also try to find a day when I can plug it into the PA and see what it sounds like.
I will hopefully get something together with the J201 FET version soon (need to drop the power consumption) and I will try the same experiment and compare.
Cheers
Ray
Nice. This is exactly how I would do it . The only suggestion is that if you increase the 10uF capacitor to let's say 22u, you may get some more bass. But the difference is only 0.5dB at 40Hz so I'm not sure whether this will be audible.
With J201 you will get the same results, only the current comsumption will be decreased (even to 0.2mA).
Nice music. Is it only the piezo pickup (recorded)?
EDIT: I'm not sure about the input resistor. I thought that this is a kind of protection for the FET.
Mark
. The only suggestion is that if you increase the 10uF capacitor to let's say 22u, you may get some more bass. But the difference is only 0.5dB at 40Hz so I'm not sure whether this will be audible.With J201 you will get the same results, only the current comsumption will be decreased (even to 0.2mA).
Nice music. Is it only the piezo pickup (recorded)?
EDIT: I'm not sure about the input resistor. I thought that this is a kind of protection for the FET.
Mark
Hello
Many thanks, I will include the input resistor in the final build. I did not think that it would make much of a difference considering the input impedance of the preamp, and I was concerned about blocking any RF as I remember gigs years ago where a local taxi firm's cars would often join in on a chorus or two!!
Mark, I tried 100uF as a bypass (the next largest I have up from 10uF) but this sounded a little woofy to my ears and although probably OK, I feel that low feedback may be an issue as I am using a bridge plate transducer. I was thinking on the lines of 47uF or so here, but no components to try. I will get 22uF, 47uF and maybe 68uF and try them out.
The recording is only the pickup via the preamp & mixer straight into the soundcard, there is no mic or anything and no EQ or effects. I know you were toying with the idea for a piezo Bass element, so what do you think of the acoustic sound of the preamp?
Cheers
Ray
Many thanks, I will include the input resistor in the final build. I did not think that it would make much of a difference considering the input impedance of the preamp, and I was concerned about blocking any RF as I remember gigs years ago where a local taxi firm's cars would often join in on a chorus or two!!
Mark, I tried 100uF as a bypass (the next largest I have up from 10uF) but this sounded a little woofy to my ears and although probably OK, I feel that low feedback may be an issue as I am using a bridge plate transducer. I was thinking on the lines of 47uF or so here, but no components to try. I will get 22uF, 47uF and maybe 68uF and try them out.
The recording is only the pickup via the preamp & mixer straight into the soundcard, there is no mic or anything and no EQ or effects. I know you were toying with the idea for a piezo Bass element, so what do you think of the acoustic sound of the preamp?
Cheers
Ray
Hi Mark
I have only 1x 10uF and the other is the 4.7uF on the output. Both of these are probably around 15 to 20 years old!
Message to self: "must get some stock components for fiddling around with!"
The K&K systems are basically just piezo elements that you stick down using super glue (gel type) so it may be worth trying to get a suitable sized set and trying them taped to different places on the bass and then sticking them when you have found a sweet spot.
The K&K range of transducers can be found here and are worth a try as they are not as expensive as some units...
http://www.kksound.com/findpickup.html
They do systems for both upright bass (Bull Fiddle!) and acoustic bass, but if I remember right you are trying to embellish an electric bass, right?
What part of the bass are you thinking of blending the piezo in, fingerboard, bridge?
Cheers
Ray
I have only 1x 10uF and the other is the 4.7uF on the output. Both of these are probably around 15 to 20 years old!
Message to self: "must get some stock components for fiddling around with!"
The K&K systems are basically just piezo elements that you stick down using super glue (gel type) so it may be worth trying to get a suitable sized set and trying them taped to different places on the bass and then sticking them when you have found a sweet spot.
The K&K range of transducers can be found here and are worth a try as they are not as expensive as some units...
http://www.kksound.com/findpickup.html
They do systems for both upright bass (Bull Fiddle!) and acoustic bass, but if I remember right you are trying to embellish an electric bass, right?
What part of the bass are you thinking of blending the piezo in, fingerboard, bridge?
Cheers
Ray
Hi Mark
The K&K system was about the cheapest I found at the time I was looking a couple of years ago, things may have changed since then! Maybe you could try some piezo sounder discs first to see if they can be used
I will add a 1uF to my component list! What difference will a 1uF output cap make to the circuit?
Cheers
Ray
The K&K system was about the cheapest I found at the time I was looking a couple of years ago, things may have changed since then! Maybe you could try some piezo sounder discs first to see if they can be used
I will add a 1uF to my component list! What difference will a 1uF output cap make to the circuit?
Cheers
Ray
Hi All
I wonder if you could me!?
Mark kindly introduced me to LTSpice so I could have a play around with component values and I thought I would try it out on the proposed circuit just for fun (I almost wish I hadn’t now!). I started with just the FET stage. It seemed to work but the simulation showed that I might have a problem with my FET biasing. All is OK without the by-pass capacitor on the source resistor but as soon as I add even 1uF here, my negative output swings get clipped! I am inputting 0.71v rms for 1.0v pk at 1000Hz. I tried to alter the biasing resistors to see if things could be improved and this proved only a little successful and there seemed to be less negative swing gain than positive swing gain or vice versa with the bypass cap
So I thought I would blow the dust of my scope and try this for real on the actual circuit. Sure enough the problem was there, albeit not quite so exaggerated as in the simulation. Have I got my biasing correct? I used the following method…
Measured the current from a 9V battery into the FET drain, with the gate and source shorted together. This gave…
Idss = 12mA
Then using a -9v source across a 100K pot into a 100R gate resistor, the FET required -4.13v to turn the FET off. So…
Vgs-off = -4.13v
I chose a FET current of 3mA and found that I needed -2.197v on the gate to set the drain current to 3.0mA so…
Vgs= -2.197v
This would then require a source resistor of 2.197/3mA = 732.2R so I used…
Source resistor = 750R
To get a drain voltage of 4.0v I then require a Drain resistor of 4.0v/3mA = 1333.3 so I used
Drain resistor = 1300R
Is this all correct or have missed the point entirely (happens a lot with me!)?? I need to handle an input of around 1Volt (rms or pk was not given, so I will assume pk) and as much gain as I can get out of the FET with a 9v battery source, without any clipping or distortion.
Can any of you sages point out what I might be doing wrong here?
Cheers
Ray
I wonder if you could me!?
Mark kindly introduced me to LTSpice so I could have a play around with component values and I thought I would try it out on the proposed circuit just for fun (I almost wish I hadn’t now!). I started with just the FET stage. It seemed to work but the simulation showed that I might have a problem with my FET biasing. All is OK without the by-pass capacitor on the source resistor but as soon as I add even 1uF here, my negative output swings get clipped! I am inputting 0.71v rms for 1.0v pk at 1000Hz. I tried to alter the biasing resistors to see if things could be improved and this proved only a little successful and there seemed to be less negative swing gain than positive swing gain or vice versa with the bypass cap
So I thought I would blow the dust of my scope and try this for real on the actual circuit. Sure enough the problem was there, albeit not quite so exaggerated as in the simulation. Have I got my biasing correct? I used the following method…
Measured the current from a 9V battery into the FET drain, with the gate and source shorted together. This gave…
Idss = 12mA
Then using a -9v source across a 100K pot into a 100R gate resistor, the FET required -4.13v to turn the FET off. So…
Vgs-off = -4.13v
I chose a FET current of 3mA and found that I needed -2.197v on the gate to set the drain current to 3.0mA so…
Vgs= -2.197v
This would then require a source resistor of 2.197/3mA = 732.2R so I used…
Source resistor = 750R
To get a drain voltage of 4.0v I then require a Drain resistor of 4.0v/3mA = 1333.3 so I used
Drain resistor = 1300R
Is this all correct or have missed the point entirely (happens a lot with me!)?? I need to handle an input of around 1Volt (rms or pk was not given, so I will assume pk) and as much gain as I can get out of the FET with a 9v battery source, without any clipping or distortion.
Can any of you sages point out what I might be doing wrong here?
Cheers
Ray
Hi Ray,
I'm glad that you used LTSpice. You see that you can find out a lot of information about the preamp. And this can be done much easier than with the acutal circuit.
I haven't checked your calculations but I would look at the problem in a different way. With the original version of the preamp (with J201 transistor and 3k9/43k resistors) the gain of the first stage (calculated in LTSpice) is x6,7. So if you provide 1V amplitude on the input, you will get 6.7V on the output. But this would mean (over) 13V signal swing. And you have only 9V power supply. You cannot avoid distortion. To overcome the problem you would need 18V power supply. But...
I think that the information about 1V amplitude from the transducer is incorrect. You can verify this by configuring the FET as a source follower (similar to emitter follower), recording the signal, and checking the recorded signal with some software.
When I wanted to increase the gain of the first stage, I added 4.7k resistor in series with 10uF capacitor. Decreasing the resistor causes distortion.
On your place, I would start with measuring actual amplitude of the signal, and then adjusting components in LTSpice. Of course, you need a Spice model of the J-FET you use. Let me know if you need the model for J201.
Mark
I'm glad that you used LTSpice. You see that you can find out a lot of information about the preamp. And this can be done much easier than with the acutal circuit.
I haven't checked your calculations but I would look at the problem in a different way. With the original version of the preamp (with J201 transistor and 3k9/43k resistors) the gain of the first stage (calculated in LTSpice) is x6,7. So if you provide 1V amplitude on the input, you will get 6.7V on the output. But this would mean (over) 13V signal swing. And you have only 9V power supply. You cannot avoid distortion. To overcome the problem you would need 18V power supply. But...
I think that the information about 1V amplitude from the transducer is incorrect. You can verify this by configuring the FET as a source follower (similar to emitter follower), recording the signal, and checking the recorded signal with some software.
When I wanted to increase the gain of the first stage, I added 4.7k resistor in series with 10uF capacitor. Decreasing the resistor causes distortion.
On your place, I would start with measuring actual amplitude of the signal, and then adjusting components in LTSpice. Of course, you need a Spice model of the J-FET you use. Let me know if you need the model for J201.
Mark
Hi Mark
This sounds like a good idea!
OK when you say configure the FET as a source follower does this mean I connect the drain directly to +9v and then take the signal from the top of the source resistor? If so, then do I need to change the value of the source resistor or can this stay at 750R?
I never thought of recording the signal and seeing what the output was, good thinking there! How do I ensure that when I use the sound card, the signal recorded has not gone through any gain stages?
I have WaveLab on my PC, can I use this to establish the output voltage level of the recorded signal and if so, do you have any idea how I would go about this?
Questions, Questions!!
If you have a J201 model you could give me that would be great as I intend to use this device in the circuit, appreciated!
Many thanks for your help
Ray
This sounds like a good idea!
OK when you say configure the FET as a source follower does this mean I connect the drain directly to +9v and then take the signal from the top of the source resistor? If so, then do I need to change the value of the source resistor or can this stay at 750R?
I never thought of recording the signal and seeing what the output was, good thinking there! How do I ensure that when I use the sound card, the signal recorded has not gone through any gain stages?
I have WaveLab on my PC, can I use this to establish the output voltage level of the recorded signal and if so, do you have any idea how I would go about this?
Questions, Questions!!
If you have a J201 model you could give me that would be great as I intend to use this device in the circuit, appreciated!
Many thanks for your help
Ray
Yes. Actually, you can check with LTSpice whether 750 Ohms resistor is suitableMondyT said:
. I would increase the resistor slightly (2-4 kOhms).It depends on your sound card's audio drivers. I use KxProject and I simply set the gain to 0 dB. You may try the same setting.MondyT said:
I will send you the model.MondyT said:
Mark
I use Sound Forge but I think that with WaveLab you can set the vertical axis units to Volts. If not, you can just display decibels and convert a specific value of amplitude to Volts (there are many online converters that you can use).MondyT said:
You can also do it with LTSpice (because you can provide a WAV file as an input for a simulation). For details, you need to search the LTSpice Yahoo group: http://tech.groups.yahoo.com/group/LTspice/
Even more; you can provide "clean" recording as input to the "final" version of the preamp and see what you get on the output.
Mark
I have one correction to what I have said previously regarding signal measurements of signal from piezo transducer (using source follower). With source follower the voltage gain is 0.9 and not 1.0 unlike with emitter follower. This is described very well here: http://books.google.pl/books?id=EEc...X&oi=book_result&resnum=2&ct=result#PPA105,M1
on page 108. Unfortunatelly the schematic which is on page 107 is not available in the preview. But the circuit is very simple: the source resistor is 4.7k, the gate resistor is 2.2M, the input capacitor is 100n and the output capacitor is 100uF. It means that bipolar emitter follower is better suited for such measurements. And the output impedance is (in typical configuration) about 500 Ohms, which means that with emitter follower you may get lower output impedance. Of course you need to take into account the input impedance, which is higher in case of a source follower.
Mark
on page 108. Unfortunatelly the schematic which is on page 107 is not available in the preview. But the circuit is very simple: the source resistor is 4.7k, the gate resistor is 2.2M, the input capacitor is 100n and the output capacitor is 100uF. It means that bipolar emitter follower is better suited for such measurements. And the output impedance is (in typical configuration) about 500 Ohms, which means that with emitter follower you may get lower output impedance. Of course you need to take into account the input impedance, which is higher in case of a source follower.
Mark
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.