|
Post by JohnH on Feb 15, 2018 17:20:39 GMT -5
Obviously. Or if R3 had been removed as noted; Q1 would be forced on.
But just trying to build up some understanding first. I didnt want to tackle hacking into Q1 before learning from parts that are working better.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 16, 2018 6:25:30 GMT -5
Obviously. Or if R3 had been removed as noted; Q1 would be forced on. But just trying to build up some understanding first. I didnt want to tackle hacking into Q1 before learning from parts that are working better. JohnH, First, I would like To thank you for all the info. You rule, man. Second, I translated all you said To My friend. I got back home from work really tired yesterday and couldnt get back to you. As soon as you have answers To the problems I pointed out in My last post, let us all know. Along with them, it would be nice to know more examples of transistor models, since the one in the project are not available here in My city.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 16, 2018 6:30:21 GMT -5
Up to now, My friend has not told me about these Numbers, Im sorry I cant help it. If your friend is within arms reach, you can relay a message. If he's farther away than you can reach, I dunno what to tell you. Exactly, "thats from Gibbs ". Kkkkk My friend deserves that.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 16, 2018 6:36:42 GMT -5
If your friend is within arms reach, you can relay a message. If he's farther away than you can reach, I dunno what to tell you. Exactly, "thats from Gibbs ". Kkkkk My friend deserves that. Retread, I dont know why I cannot reply your last post, but here it Goes. Thank you very much for your time. Really appreciate it. Ill be posting some audio files next week.
|
|
|
Post by JohnH on Feb 16, 2018 15:04:18 GMT -5
I think we have serious difficulties on this thread because it seems that you have to translate our posts for your friend (where are you?), and neither of you know enough about circuit theory yet to understand how this circuit works. The only way this will work is if you personally can understand what is happening.
Your friends circuit is not yet working at all because of problems at Q1 and Q2. There is no point in trying to adjust tone and gain when it is not yet functional. Q2 is the easiest to fix and that's why I made a suggestion that I think should be the next step.
As reTrEad points out, Q1 may be shorted out, or it may be biased fully on and so it may be more difficult to work out. if its voltages are both exactly the same, to more digits than posted (eg like 2.20V or 2.22V), then a complete short circuit is indicated, either a blown transistor or a short between leads on the circuit board.
But on your questions, by sub-bass, I mean frequencies so low that you cant hear them. The values shown are designed to let all audible frequencies through, such as at C10 and C11. If, once it is working, you are getting a lack of bass, you might try increasing R2, which was a value chosen for my specific pickup.
Gain is set at Q1 and Q3 (Q2 and Q4 are x1 gain), and it is related (for q1) to the ratio of R14 / (R3 in parallel with R23). R3 and R23 are both 10k and are in parallel at audio frequencies due to C10 being quite large. so we have a gain of 33/5, or about x6. Reduce R23 to increase gain. But R3 needs to be set for dc conditions, so that needs to be what it is for my build and yours may need a different value.
On components, most people buy online. Where can you order from? post a link and lets see what you can get.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 17, 2018 12:32:19 GMT -5
I think we have serious difficulties on this thread because it seems that you have to translate our posts for your friend (where are you?), and neither of you know enough about circuit theory yet to understand how this circuit works. The only way this will work is if you personally can understand what is happening. Your friends circuit is not yet working at all because of problems at Q1 and Q2. There is no point in trying to adjust tone and gain when it is not yet functional. Q2 is the easiest to fix and that's why I made a suggestion that I think should be the next step. As reTrEad points out, Q1 may be shorted out, or it may be biased fully on and so it may be more difficult to work out. if its voltages are both exactly the same, to more digits than posted (eg like 2.20V or 2.22V), then a complete short circuit is indicated, either a blown transistor or a short between leads on the circuit board. But on your questions, by sub-bass, I mean frequencies so low that you cant hear them. The values shown are designed to let all audible frequencies through, such as at C10 and C11. If, once it is working, you are getting a lack of bass, you might try increasing R2, which was a value chosen for my specific pickup. Gain is set at Q1 and Q3 (Q2 and Q4 are x1 gain), and it is related (for q1) to the ratio of R14 / (R3 in parallel with R23). R3 and R23 are both 10k and are in parallel at audio frequencies due to C10 being quite large. so we have a gain of 33/5, or about x6. Reduce R23 to increase gain. But R3 needs to be set for dc conditions, so that needs to be what it is for my build and yours may need a different value. On components, most people buy online. Where can you order from? post a link and lets see what you can get. JohnH, thanks for quoting. Dont worry about my translating. I have Linguistics and English formal background, and that is not a problem for me, not also a difficulty. My friend has not had formal Electronics classes, but worked as technician for some years and is a very smart guy. We use WhatsApp all the time. The only thing is, I think, I cant keep up with your expectations as fast as you'd wish. I have lots of other responsibilities, and getting stuck on smartphones is not what I have dreamed of for me. Not even Jobs for his kids. Second, he told me one of the battery poles were not doing fine, and he fixed it. He made better results as To the voltages, but not the ideal. Here they are: Q1 source 3.5v drain 7v
Q2 source 5.8v drain 9.1v
Q3 source 5.9v drain 9.1v
Q4 source 1.1v drain 5.6v Im order To speed up things here, ive bought the components To build myself one preamp. Ill be building up one tomorrow, I think. This way, My replies will be much more immediate. I have also found the BF245 transistor. On the gain, I do agree with you. I would like To address it, but not before fixing the issues with the voltages, for sure. I have some audio files, later on Ill post' em. I think youll enjoy.
|
|
|
Post by JohnH on Feb 17, 2018 15:54:40 GMT -5
Those new voltage readings are better, showing that all stages are probably working, but with a change of resistor values, a better range of voltage swing can be achieved before distortion with those transistors. You may find something similar happening when you build it yourself. Will you be able to measure voltages too?
With the current build, if you can reduce R3 from 10k down to say 6.8k or 5.6k it will help bring the drain voltage down from 7V. As an easy test, it might actually work to just replace or bypass c10 with a wire, so R23 and R3 are in parallel to make 5k
Q4 looks OK. Q2 and Q3 look like they will work fine but if you reduce R7 and R9 from 2.7M to 2.2M it would be more optimised.
But note that these values are specific to individual transistors, so your new build may be a bit different. This type of inconsistency is one reason why jfets are not often used in commercial circuits, even though they work very well when set right. One approach is to include an adjustable preset resistor in each stage, so you can tune it without swapping parts.
I don't mind how fast or slow you want to go with your project, slow is good, don't rush. But sometimes getting some measurements is what is needed as a next step before moving forward, so that is why I was asking for voltages.
I look forward to the sound clip.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 18, 2018 4:07:49 GMT -5
Those new voltage readings are better, showing that all stages are probably working, but with a change of resistor values, a better range of voltage swing can be achieved before distortion with those transistors. You may find something similar happening when you build it yourself. Will you be able to measure voltages too? With the current build, if you can reduce R3 from 10k down to say 6.8k or 5.6k it will help bring the drain voltage down from 7V. As an easy test, it might actually work to just replace or bypass c10 with a wire, so R23 and R3 are in parallel to make 5k Q4 looks OK. Q2 and Q3 look like they will work fine but if you reduce R7 and R9 from 2.7M to 2.2M it would be more optimised. But note that these values are specific to individual transistors, so your new build may be a bit different. This type of inconsistency is one reason why jfets are not often used in commercial circuits, even though they work very well when set right. One approach is to include an adjustable preset resistor in each stage, so you can tune it without swapping parts. I don't mind how fast or slow you want to go with your project, slow is good, don't rush. But sometimes getting some measurements is what is needed as a next step before moving forward, so that is why I was asking for voltages. I look forward to the sound clip. Thanks again JohnH, my friend had already made those corrections. Now the voltages are exactly the ideal. Ill check with him on which resistors he tweaked. I understood all the theory on jfets. The only thing I cant get is the Number 5v. Why not other numbers?. I have a meter, so yes, I will be able to measure things. Your circuit is not soh difficult or complex To understand, but missing some info about jfets is what makes it almost impossible to make it work properly. But, for an amateur, starting To get the handle on the circuit is exciting. The audio files wiil be posted as soon as I get To my desk. Probably.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 18, 2018 4:26:04 GMT -5
Those new voltage readings are better, showing that all stages are probably working, but with a change of resistor values, a better range of voltage swing can be achieved before distortion with those transistors. You may find something similar happening when you build it yourself. Will you be able to measure voltages too? With the current build, if you can reduce R3 from 10k down to say 6.8k or 5.6k it will help bring the drain voltage down from 7V. As an easy test, it might actually work to just replace or bypass c10 with a wire, so R23 and R3 are in parallel to make 5k Q4 looks OK. Q2 and Q3 look like they will work fine but if you reduce R7 and R9 from 2.7M to 2.2M it would be more optimised. But note that these values are specific to individual transistors, so your new build may be a bit different. This type of inconsistency is one reason why jfets are not often used in commercial circuits, even though they work very well when set right. One approach is to include an adjustable preset resistor in each stage, so you can tune it without swapping parts. I don't mind how fast or slow you want to go with your project, slow is good, don't rush. But sometimes getting some measurements is what is needed as a next step before moving forward, so that is why I was asking for voltages. I look forward to the sound clip. Just one more thing about the resistors: could I tweak r14, instead of r3, To bring 7v down on q1?
|
|
|
Post by JohnH on Feb 18, 2018 5:05:07 GMT -5
Those new voltage readings are better, showing that all stages are probably working, but with a change of resistor values, a better range of voltage swing can be achieved before distortion with those transistors. You may find something similar happening when you build it yourself. Will you be able to measure voltages too? With the current build, if you can reduce R3 from 10k down to say 6.8k or 5.6k it will help bring the drain voltage down from 7V. As an easy test, it might actually work to just replace or bypass c10 with a wire, so R23 and R3 are in parallel to make 5k Q4 looks OK. Q2 and Q3 look like they will work fine but if you reduce R7 and R9 from 2.7M to 2.2M it would be more optimised. But note that these values are specific to individual transistors, so your new build may be a bit different. This type of inconsistency is one reason why jfets are not often used in commercial circuits, even though they work very well when set right. One approach is to include an adjustable preset resistor in each stage, so you can tune it without swapping parts. I don't mind how fast or slow you want to go with your project, slow is good, don't rush. But sometimes getting some measurements is what is needed as a next step before moving forward, so that is why I was asking for voltages. I look forward to the sound clip. Thanks again JohnH, my friend had already made those corrections. Now the voltages are exactly the ideal. Ill check with him on which resistors he tweaked. I understood all the theory on jfets. The only thing I cant get is the Number 5v. Why not other numbers?. I have a meter, so yes, I will be able to measure things. Your circuit is not soh difficult or complex To understand, but missing some info about jfets is what makes it almost impossible to make it work properly. But, for an amateur, starting To get the handle on the circuit is exciting. The audio files wiil be posted as soon as I get To my desk. Probably. Yes you could. A higher value for r33 should do it, maybe 47k? Or, a 100k preset set at abopuut that value will allow it to be tweaked. 5V is good because it is just above half the battery voltage. These circuits can swing nearer to the + battery + voltage than to - ground, so about 5V gives the maximum +/- swing available without clipping. But is it not too critical to get exactly 5V. Its just a guide, and what I found to work best.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 18, 2018 9:14:58 GMT -5
cb 1.mp3 (1.38 MB) cb 2.mp3 (652.26 KB) cb comparar 2.mp3 (250.76 KB) cb comparar.mp3 (386.14 KB) comparaao grap 2.mp3 (400.43 KB) These are the files he recorded. There are minor mistakes, im aware, but these are only for reference. Even though the audio files were a bit distorted, we can have a general idea. Now the distortion is corrected, we can record new stuff. I have asked him for some new files, and as soon as i have built my preamp, ill send you new ones. The idea here is to set out notions of progression. Even though the sounds are not well recorded, or even have technical issues, they show the piezo under the strings is highly compatible with the preamp. However, minor errors needed corretions along the way. When my preamp is ready for the show, ill post here more audio files, ill try to make a good job. Attachments:comparaao grap.mp3 (393.08 KB)
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 18, 2018 9:29:14 GMT -5
gt 1.mp3 (1.05 MB) gt 2.mp3 (401.24 KB) We have more audio files from Graphtech's preamp. Graphtech graphic shows the reason why there is not much dynamics in their preamp.
|
|
|
Post by JohnH on Feb 18, 2018 14:18:32 GMT -5
The recordings sound very nice - good work!
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 18, 2018 14:59:42 GMT -5
The recordings sound very nice - good work! Really? I would like To hear more ... it was recorded on Zoom's G5n. No eqing after. the distorted sounds are audible, especially when strummed hard. My friend used Elixir's strings gauge 9. I have that branch, lol, and even more that gauge.
|
|
|
Post by JohnH on Feb 18, 2018 22:06:06 GMT -5
So far I only had time for a quick listen on good (but not great) pc speakers, before I had to leave for work early this morning. It all sounded much better than the first tests that I did, and was very credibls as an acoustic. I'll listwn more later.
|
|
|
Post by JohnH on Feb 19, 2018 16:40:57 GMT -5
I listened with 'phones. I can hear a bit of clipping but not too bad and within range of fixing. Encouraging!
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 19, 2018 17:31:47 GMT -5
I listened with 'phones. I can hear a bit of clipping but not too bad and within range of fixing. Encouraging! Thanks, JohnH. I have started building mine. Ill finish the board first thing tomorrow. Ill be more careful with the sound clips: use better strings, better gauge, a decent audio interface and some good monitors, not the best. Budget, but decent. Ill Adjust the truss rod on an Ibanez's Rg520 and rock and roll. Of course, practice The sounds a lot. We would like To make it two modes, electric and acoustic. As you can notice in the clips, your preamp has some good electric tones, just like Graphtech's does - listen To the sound clips in which I compare both preamps. However, you have To make use of dynamics as you pluck, solo or even play chords, in order To get the strings a little more pronounced. If you take a good look at the screenshots, youll see Grahtech's frequency range is flat in the 18db area. That is because they designed The preamp To be compressed, so if one wants to play soft, theyll have The electric sound at any volume level mostly. We also want To build a compressor for The preamp. My friend believes we can keep The preamp The way it is now and add an op-amp compressor To The electric mode. The second mode, an acoustic, will not let The string buzz out, as we are gonna reduce gain. We are also thinking of throwing in an active eq To shape it and have as many steel string acoustic guitar-loaded frequencies as we can get. If you have suggestions of schematics on compressor op-amps and active eqs, fire away.
|
|
|
Post by JohnH on Feb 19, 2018 18:32:12 GMT -5
Sounds good.
A couple of things to help adjust your build.
The 150k at R2 is an unusually low value for a piezo preamp, chosen to suit the large piezo element I was using. Most designs would have a larger value, like 1M. Higher value causes roll off of bass at lower frequencies when combined with tbe self-capacitance of the piezo. The 150k was right for mine, but would be too low for a piezo of lower capacitance, where if would cut too much into audible bass. Best to experiment.
For your friends build, setting that Q1 voltage down to 5v (either increasing r14 or rwdycing r3) will likley add to the amplification of this stage. If its then too much, you can increase r23, or even remove it and also c10. See how yours works out.
Ive never built a compressor, but if you do, putting it between Q2 and C4 should work, with its own approx 100nF input cap to seperate it from Q2
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Feb 20, 2018 10:13:53 GMT -5
Sounds good. A couple of things to help adjust your build. The 150k at R2 is an unusually low value for a piezo preamp, chosen to suit the large piezo element I was using. Most designs would have a larger value, like 1M. Higher value causes roll off of bass at lower frequencies when combined with tbe self-capacitance of the piezo. The 150k was right for mine, but would be too low for a piezo of lower capacitance, where if would cut too much into audible bass. Best to experiment. For your friends build, setting that Q1 voltage down to 5v (either increasing r14 or rwdycing r3) will likley add to the amplification of this stage. If its then too much, you can increase r23, or even remove it and also c10. See how yours works out. Ive never built a compressor, but if you do, putting it between Q2 and C4 should work, with its own approx 100nF input cap to seperate it from Q2 I think well set The compressor aside for now. Ill make the circuit, check The voltages, The components. If it is working just fine, ill move on. If not, ill make The necessary corrections. we now want To focus on The eq, and We are convinced it will fill up our needs on both modes without compressor. Well spot The frequencies on Proteus and then adjust them with rc filters on The circuit - is this The right term? What do you suggest in terms of eq circuit? Any books? Articles? Schematics? Rc filters?
|
|
|
Post by JohnH on Feb 20, 2018 14:17:07 GMT -5
In the circuit, R29 controls bass, R30 controls treble and R13 pans from bass to treble. R29 and R30 can be set so that there is a frequency band that neither let through, creating a mid notch (which I needed), or an overlap for a mid boost. They only act only the piezo signal, so I have a normal guitar tone pot after normal guitar switching and volume all before the magnetic input, ie not part of this circuit. If you want a tone knob on the front panel for the piezo, you could use a normal pot for R13 instead of a preset.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Mar 25, 2018 14:31:53 GMT -5
John, all I can say is that you rule man!!!! Your information always come in handy! Yes, we tested it on the guitar. It sounds good, but lacks some tweaking in the treble and middle frequencies areas in order to sound more electric-like. we are also trying to build piezo loaded saddles to get better and definitive results. I truly believe that we're gonna make it. we're not trying to exactly match GraphTech's results, but make a stand in the middle of acoustic and electric sounds. As soon as the sound files are decent, I'll let you know. There's one more tweak to keep in mind on a piezo should you ever need it. If you put a small value capacitor (in the 1000 pF-.01 uF range for most piezo pickups) directly across (shunting) the piezo element, this pads down the output (reduces output level to better match the mag output or not overload the preamp,) can "tame" too-shrill highs, and also make the piezo less sensitive to resistive loading (giving more bass response for a given load resistance.) I'm glad as well you are making good progress on your piezo Is that possible To put a small capacitor right across The piezo? Im asking this because we are planning To place The piezo element inside the nylon material that will support each string. If you think it is still necessary, even with The nylon material, we can manage that. What Im worried about The space inside The nylon stuff, so tiny:( In order To reduce The piezo output level, we included a 100k pot right at The entrance To The circuit. As a result, we can now reduce The string buzz sound - To make it more like an acoustic. It is working wonders, but it can always be better. Im open To suggestions here.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Mar 25, 2018 14:51:58 GMT -5
Hi, JohnH, Hi Charlie.
The preamp is on its way. Were almost there. We added an input level pot right at the entrance to the circuit - in order To control The piezo element output Gain - and its working fine. The volume and distortion are also fine. And we have found a good active eq for it. We decided on a different eq because we want To design two modes, as said before, The electric and The acoustic. Soh in order To reaçh better results as To The acoustic mode, an active eq would suit The circuit Better.
However, the eq is 2n5484 Jfet loaded. Ours is BF245 based. What do you suggest? Do we keep The BF guy or switch it To The 2n Jfet?
As as soon as The eq issue is solved, well move on To The modes design.
|
|
|
Post by JohnH on Mar 25, 2018 14:58:51 GMT -5
Hi, JohnH, Hi Charlie. The preamp is on its way. Were almost there. We added an input level pot right at the entrance to the circuit - in order To control The piezo element output Gain - and its working fine. The volume and distortion are also fine. And we have found a good active eq for it. We decided on a different eq because we want To design two modes, as said before, The electric and The acoustic. Soh in order To reaçh better results as To The acoustic mode, an active eq would suit The circuit Better. However, the eq is 2n5484 Jfet loaded. Ours is BF245 based. What do you suggest? Do we keep The BF guy or switch it To The 2n Jfet? As as soon as The eq issue is solved, well move on To The modes design. Can you post tbe schematic of your EQ circuit?
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Mar 25, 2018 15:04:36 GMT -5
"Distortion could come from either having too much gain, or not having optimum bias, either generally or at particular stages. Either way, the signal tries to swing beyond the available range and so distorts. Maybe your pickups are quite powerful compared to the gain. Ultimately if the signal gets too hot, then it is limited by the power supply and must distort. So in that case you would need to reduce gain. You could test tbe signal after Q1 and Q4 to see which is distorting, then reduce its gain (eg, remove the cap/resistors that bypass the source resistors." Checking for output level is ok. I get it. But how can I check for distortion - by ears? Im asking this because when we add The active eq I suppose we may experiment output changes in The Jfets. This might come in handy.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Mar 25, 2018 15:24:19 GMT -5
Hi, JohnH, Hi Charlie. The preamp is on its way. Were almost there. We added an input level pot right at the entrance to the circuit - in order To control The piezo element output Gain - and its working fine. The volume and distortion are also fine. And we have found a good active eq for it. We decided on a different eq because we want To design two modes, as said before, The electric and The acoustic. Soh in order To reaçh better results as To The acoustic mode, an active eq would suit The circuit Better. However, the eq is 2n5484 Jfet loaded. Ours is BF245 based. What do you suggest? Do we keep The BF guy or switch it To The 2n Jfet? As as soon as The eq issue is solved, well move on To The modes design. Can you post tbe schematic of your EQ circuit?
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Mar 25, 2018 15:34:25 GMT -5
Can you post tbe schematic of your EQ circuit? Sorry, my mistake. I meant 2n5089, not 2n5484
|
|
|
Post by JohnH on Mar 25, 2018 15:46:41 GMT -5
ok. Those are not jfets. They are normal silicon transistors (bipolar junction transistors, bjt's, more common than jfets) working with opamps. Youll need the right parts to build it. There may be alternative equivalent parts that will work if needed but not the BF jfets.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Mar 25, 2018 15:54:46 GMT -5
ok. Those are not jfets. They are normal silicon transistors (bipolar junction transistors, bjt's, more common than jfets) working with opamps. Youll need the right parts to build it. There may be alternative equivalent parts that will work if needed but not the BF jfets. Ok, but Would that eq work fine in your preamp? I mean, replacing your eq...
|
|
|
Post by JohnH on Mar 25, 2018 16:05:38 GMT -5
Probably it will work but I dont know how ii will sohnd or what other issues may be caused.
|
|
bwps
Meter Reader 1st Class
Posts: 52
Likes: 0
|
Post by bwps on Mar 25, 2018 16:20:17 GMT -5
Probably it will work but I dont know how ii will sohnd or what other issues may be caused. Your certainly right about that. I have found this bf245 loaded eq circuit. What do you think?
|
|