I have had a go at redesigning the circuit using the TI single supply handbook.
It avoids most of the pitfalls of the original though i am not expert.
If anyone in the know can comment that would be good.
I will try it with both a TL071 and a OPA134.
The OPA134 should sound better since it is a very high spec opamp btu you never know!
I can read a schematic but have no clue on how they work... just a question with your IC does this mean that nothing is connected to pins 1 and 5?
Yeah they arent used, ethier 'no connect pins or bias ajustment pins which arent used in this application
First impressions:
- I realize that I'm nit-picking but the gain of a non-inverting circuit is R3 / (R4+Rpot) + 1. Thus gain is adjustable between 2.6 and 11. (Yes, I delibrately ignored the capacitors to keep this simple).
- If this is an input stage then it has a rather low input impedance (220 kilo-ohms) for guitar applications.
- R1 and R2 should preferably have as low value as possible to keep the "virtual ground" as a low-impedance source. In practice, too low values will draw too much current and result in great losses so try something like 10 kohm instead.
- The circuit is not expected to drive a capacitive load and opamps are short circuit protected so R5 seems rather useless
- If this is intented to be a linear design then C3 (100 pF) seems a bit too high. It will be cutting off the higher frequencies.
Anyway, as far as I see there are no "real" problems in the circuit.
Teemuk,
Thanks so much exactly the sort of comments i was looking for, I am commpletely new to this sort of thing.
1 Your bang on i forgot to change the gain marking after i had changed the layout.
2 What sort of input impediance should i use? 470k i really am guessing any input greatfully recieved
3 thanks i with try 10k
4 fair enough
5 what sort of value would you suggest?
I have etched some board already so i will post the pcb layout when i get to work tomorrow.
Thansk for your input
The input impedance... well...
Generally you should aim for "impedance bridging", which means that the input impedance is kept higher than the source impedance; this provides the best voltage transfer.
The source impedance of a typical guitar with magnetic pickups is several kilo-ohms (i.e. 5 kohm - 50 kohm) so at first glance it might seem that few hundred kilo-ohms is enough. Well, to make a long story short, let's just say that usually it isn't and the "rule of thumb" is to use at least 1 Meg for passive magnetic pickups. Low input impedance values generally attenuate the higher frequencies too much, thus "sucking" the guitars characteristic tone away. For "bright" single coil pickups even 1 Meg might not be enough.
In practice, all this depends on the circuitry of the guitar: Guitars with active electronics typically have a very low and fairly linear source impedance so they can be connected to circuits with quite low input impedance without considerable losses.
However, in reality you have to prepare for the "worst": A conventional pickup (and the following tone/volume control) circuitry is not a source with "linear" impedance (in fact it has a resonant character) and thus it is very much affected by input resistance and capacitance of the stage it connects to. A typical pickup is so sensitive source that even the capacitance of the guitar cable may affect its tone.
Regretfully, too little has been written about this subject. Manufacturers are quite skimp on giving meaningful data about their pickups as well. Anyway, here's some stuff that might be interesting (and which is far too often ignored):
http://buildyourguitar.com/resources/lemme/
http://www.terrydownsmusic.com/technotes/guitarcables/guitarcables.htm
http://www.guitar-repairs.co.uk/how_guitar_pickups_work.htm
Piezoelectric pickups are quite different from magnetic ones being practically capacitors by nature, thus connecting them to a too low input impedance will result into loss of lower frequencies. For piezoelectric pickups some manufacturers suggest using as high as 10 Megaohm input impedance. However, guitars with piezo mics often have an inboard active preamp, which can be interfaced with rather low input impedance. Anyway, this is not the case always.
Practically there are only few things I consider as drawbacks in using a very high input impedance. 1) Resistors larger than 1 Meg are difficult to source (you can connect those in series though), 2) using high value resistors inreases noise (not a very good for "feature" for an input stage) and 3) high impedances are more vulnerable to capacitive effects (tone loss due to "parasitic" capacitances, instability etc.)
About capacitor C3: This is highly subjective issue but I would rather try values closer to 10 - 47 pF. In fact, I might even omit the capacitor completely assuming the circuit stays stable without it. Anyway, if audible low-pass filtering is sought after then ignore my comments.
Quote from: teemuk on September 12, 2007, 06:01:07 PM
About capacitor C3: This is highly subjective issue but I would rather try values closer to 10 - 47 pF. In fact, I might even omit the capacitor completely assuming the circuit stays stable without it. Anyway, if audible low-pass filtering is sought after then ignore my comments.
Wouldn't the -3db frequency provided by the value of C3 be determined by the series value of the 50K gain pot and R4? :-\
I calculate a 100pf cap and 60Kohms to be at 26.5KHz. Your ears (and speakers) must be much better than mine to consider that as audible. ;D ;D
It is good to hear someone talk about pickup impedance rather than "DC resistance". :tu:
I do know the guy has very expensive guitars and his main acoustic and electric both have active electronics in them. Does this mean i should leave the 220k value alone or should i still go with something more like a 1M?
Thanks a lot by the way a am learning a lot and very quickly.
Do you think this circuit on its own would be enough to drive a passive set of tone controls and then on into the class d power amp or should a put a buffer after the tone controls?
phil
220 kilo-ohms would be ok for active guitars but you make the circuit more "universal" if you use 1M.
As is, the circuit has a gain between about 8.5 dB and 20.8 dB. At least on the maximum setting the insertion losses of most tone control circuit types should be compensated and you still would have some gain (instead of attenuation). At lowest gain the gain stage and the tone control will quite surely have an attenuating effect in total.
Typical tone control circuits like to see a fairly high impedance at the input of the following stage. Likely the input impedance of the class-D amp is not very high? So yes, I would separate the tone controls from the power amplifier with another gain stage that has a rather high input impedance. I guess that is considered buffering. With resistor R5 the source resistance of the gain stage is about 1K so it will happily feed most tone control circuits without considerable losses.
Anyway, so far I have only seen a schematic of a gain stage so its quite impossible to answer anything definite to your questions. You must note that you are the one having the details of the design and therefore you should answer most questions by yourself:
Do you know the input sensitivity of the class-D power amplifier (voltage required for maximum output)? You must know it in order to rate the preamplifier's total gain accordingly. Guitars have a varying range of output signal levels (active ones usually have a "hotter" signal) but a safe bet is to assume that the amplitude of a pickup's output signal is approximately few millivolts (i.e. 20 mV). If the output sensitivity is, say, 1V you definitely need more gain than 11x because 11 x 0.02V equals only 0.22V. And that example didn't even include the rather considerable losses introduced by the passive tone control! ...More reasons to include another gain stage.
It seems to me that at this point you still have a mere sketch of the final circuit and perhaps it might be worthwhile to think about the design in its entirety for a while. It would be unfair to tell you how you to design your amplifier but what I can do is to provide some basic suggestions:
- A fairly typical configuration is a design consisting of two of those opamp gain stages (or similar) with the tone control in between them.
- In such circuit, typical options for volume control's position are between tone control and the second gain stage or between the second gain stage and power amplifier (assuming you want a volume control feature)
- Gain control is likely needed in only one of the opamp stages
- The first gain stage should preferably have enough gain to compensate nearly all losses of the tone stack because this improves signal-to-noise ratio
- The second stage needs enough gain to amplify the input signal to the level of the power amplifier's input sensitivity (or preferably a bit higher)
- Unless you are building a distortion effect keep the gain per stage at a reasonable level (i.e. 3 - 10 x). It's better to have two series gain stages with a gain of 10 than one with gain of 100.
That should get you started...
Hi Teemuk,
Yet again you provide a reallu useful post. I guess i should explain the basics of the amp i wish to build.
The amp will be used by a friend who is a professional busker on the London Tube, the real issue for such muscians is the fact that while busking is now legal and controlled the 'stages' have no power points therefore the amps have to run off batteries.
The only available one for electric guitars is the Crate Limo /Taxi series which while fine for messing about are far to fragile for day to day work.
All the buskers i know have at least 2 since they break down so often.
The amp i wish to make will be based on a tripath class-D or Class-T as they like to call it TA2021B, they make about 13 watts per channel into 4 Ohms clean.
I will be using a ready made kit from www.41hz.com, Amp 32 which is intended for hifi use.
I built several of these for hifi use an found them bo be excellent amps period let alone for the money, the bonus being the 80+% effiencey which should make the battery in the unit last a decent time. I do not know the input sensitivity of these kits however i believe that hifi typically works at around 2v, I may be wrong. This would mean that i would need a second stage with a gain of 10 judging by what you tell me.
To explain what i want in the amp, it will have two guitar preamp channels. Each with tone control, one with a Baxneldale setup for acoustic guitar, the other with a fender\mashal style dip in the middle range for Electric. I have these tone stack example you posted in my other thread.The elecric channel will have a bypassable fx unit (The English Channel) on the front.
Both these channel them go into a summing amps which will also have an input for a ipod/cd player before finalling going into the Tripath based power amp module.
From your post the it would seem to me the best option would be to have a fixed 10 x gain stage initially by removing the pot with a 1M input impedance. This should help drive the following tone controls after which we would have the opamp setup we have previously been talkign about with a pot and variable gain between 2.6 an 11 which should take use up to near 2v depending on the attenuation in the tone controls.
Temuuk, once again thanks for your help, you have been utterly invaluable in explaining these sorts of cicuits and the thing to consider.
You really have moved this project on leaps and bounds.
Phil
Hi,
Here is where i have got to thus far. Pad labled Tone goes out to the tone controls and then come back into the second stage at the point labelled Input 2. The thing I am confused about is what i do on the input of the second stage, at the moment it is a near enough a replication of the first stage but i imagine i should change the input impedance (r15) and the input cap (c8).
As always any comments greatfully recieved.
Phil
Input resistors R7 and R15 can share a common "bias" voltage divider, you don't need two separate ones.
OK that seems fair enough. What about those input impediance resistors and caps. Should i decrease the value of r15 and increase the value of the input cap (c8) surely this would reduce noise in the second stage and since it isnt dealign with the guitar pickup this shouldnt matter?
Also i have fixed the gain in the first stage at 10, the more i think about this it seems too high. wouldnt something like 3 be better.
Teemuk i read you are producing a book on guitar amp design is this in print yet, I think it would be very useful to me.
All of your questions are so much related to the way you want your design to perform - and more importantly to sound - that I feel you will never get answers. Basically, what you have already is a working circuit so the rest is up to highly subjective tonal issues.
If I didn't mention it already, I usually like to have less low frequency coupling at the first stages, thus bass frequencies have less total gain and consequently will clip less - thus the "blocking" effect (caused by low-freq clipping) is reduced. Seems to be what you are aiming for. Anyway, this kind of stuff that deals with tone of the circuit is something you should design yourself.
You said you have the tone control in the middle of the stages. The input impedance of the second stage will define how much the circuit in question [tone control] is loaded. This will affect it's frequency response and a too small load will mainly increase insertion losses. Of course that very much depends on what kind of tone control you have. Have you tried Duncan Munro's "Tone Stack Calculator" software or something similar? It would help you in designing the tone control circuit and defining the proper impedances for it's source and load.
The book is not in print as I'm still writing it. It's in english language, which is not my native tongue so there's a lot of work in correcting the grammar issues and stuff written confusingly. Also, I always seem to find more stuff that I feel should be included. I think the time when I'm satisfied to its content will come soon, though. However, the rest is still open...
Quote from: teemuk on September 18, 2007, 02:53:44 PM
The book is not in print as I'm still writing it. It's in english language, which is not my native tongue so there's a lot of work in correcting the grammar issues and stuff written confusingly. Also, I always seem to find more stuff that I feel should be included. I think the time when I'm satisfied to its content will come soon, though. However, the rest is still open...
If you need someone to proof read, I'd be glad to help. And my fiance is an English teacher and always willing to help people with stuff like this. Between the two of us we could probably proof the whole book lol.
Intereting to hear about your book, you have a garanteed sale here. I have been looking in many book shop in London and its all either pure valve stuff or just plain pot and cap swapping, or cable wiring stuff. Very limited all of it.
Thanks a lot for your advice again, so if the circuit will work i feel happier stuffing the board and actually starting to experiment. Tone control wise i will be trying a couple of circuits, from that useful site you linked to a while ago, one setup will be for an acoutic setup the other for an electric. Once i have values that work i will still up my pcb transfers if anyone is interested.
Do you think it is worth using film caps is these sorts of setups or would ceramic do to test?
IF anyone is interested here is a pcb layout for the above 2 stage preamp.
Change R4 to cahnge initial gain its current set at 4 the second stage gain is variable between 2.6 and 11.
Put tone controls between Tone pad and Input2. It is currently untested but seems liek it might be useful to people building Lm3886 based chipamp, though gain have to be changed acording to your application
Component overlay :-
Right i tried this setup last night and had all sorts of issues. I didnt have a guitar handy so i tried ipod at near zero volume, gain stages set to minimum, volume minimum. Nada. After turing up the gain and volume i got a output even with the ipod at high volume, messed around for a bit and nothign helped.
New opamps, all sorts, funnily enough swappign out the OPA134 for NE5334 actually helped a bit, not sure why. I was wondering if the 2 cap bleed resistors R6 and R14 might actually drain away the signal if the impediance of the tone controls is too high.
Really not sure what is going on, tried changing the gain in the first section from 3 to 10 that didnt help. Weirdly touching the hot connection on the input jack actually increased the volume a bit, wondering if that would indicate a cap is in reversed. Tested my ground and power voltages and tehy were all fine. Any help would be recieved very very greatfully. Put loads of work in and very dissapointed.
On a brighter not the Class d power amp in the chassis sounded superb, , i have 2 cd input going into my mixer Amp board with a gain of 3 set, that then goes to a pot and into the poweramp, that really transformed the ipod sounds from this thing. Got great volumes and it sounds excellent.
Just to let erveryone know, the PCB is shown from the "top" side (component side) so you need to mirror it.
lefizz, I'm sure you could make the layout more compact by using dual opamp and doing the mods I suggested but... Let's get to more important issues:
The layout is basically correct except,
- C2, C13 are wired in wrong polarity. The negative polarity leg must connect the ground as the "opamp side" has positive voltage in reference to it.
Another checklist besides that:
- You did rememeber that jumper from R15 to 1/2 Vcc ref, right?
- Check out that resistors are indeed correct value since color coding is sometimes hard to decipher correctly. I always measure them before fitting since I'm slightly color blind and thus telling the difference between brown-red, blue-green-purple & orange-yellow is difficult. This is one very typical problem.
- The trimmer: Their pinout varies. Make sure it's wired correctly.
- R6 and R14 indeed look pointless as they are in parallel with the tone control and input of the output stage. Their value is rather hight but still... Are you sure they are not interfering with the operation? If you remove these, make sure the polarity of the capacitors stays correct.
- General check: Capacitor polarity, solder bridges etc.
- You are sure that the tone control circuit works
HI Teemuk,
Thanks a lot words the words of advice, as always you have given some great suggestions.
I already checked for solder bridges so that isnt the issue, i will change the 2 caps around, can't beileve i did that but still. A second more trained set off eyes really helped there. I am 90% sure i have got the jumper in place but will check and i think i will remove the r6, r14 resistors. What do you mean about the cap polarity if i remove them? Just check everythign is as it should be or somethign else. To be honest i am sure all the cap are in the 'right' direction as in they are the same as my eagle stuff, its just was wrong!! I did actually bypass the tone controls last night and that didnt help.
Do you think as it was with caps in thewrong direction it would be possible to blow the opamps, they seem to go very strange. One minute the cicuit worked a bit other times not at all. I switched to using the cheap ones pretty soon cos i dont want to blow my opa134, they cost a lot.
I will finish off the design by doing as you say and using a dual opamp and removing the reduntant second set off biasing resistors, i just didnt have any OPA2134 at the time and wanted somethign to together quickly. Looking back that might not have been the best option.
Anyway thanks a million you have given me some great ideas about how to go foward.
Thanks a lot,you have really cheered me up.
The polarity issue: Most capacitors "tolerate" incorrect polarity at low voltages (read: they do not explode). However, they can work very strangely, for example, their capacitance may be lower, they may have significant series resistance, etc.
When you remove the ground referencing resistors from the other side of the caps you must make sure that the preceding stage has lower DC level than the feeding opamp stage. It's just a question maintaining the proper polarity.
It would help if I saw the input stages of the power amp and the tone control circuit but I assume that both basically consist of a resistor to ground that ties the DC level to zero volts. If there's some additional AC coupling involved you can likely remove it since you need just one capacitor for that purpose. That is the case even if those stages have a DC offset: Just connect the capacitor so that it retains proper polarity. Really simple.
If the circuit worked correctly now and then it also may be just an issue with a poor solder joint (contact) somewhere. Instead of just disconnecting the current path these may also show up as "resistors" that attenuate the signal. You say you swapped the opamps and this helped a little: Maybe this lowered the resistance of the cold solder joint (or some other contact) a bit?
Anyway, first things first: Correct the issue with wrong capacitor polarity and see if you still have some problems.
OK well the caps seemed to sort it out to a ceratin degree. No distortion and everythign appears near enough normal. I think i need to change the gain pot from log to lin since as it stands it offers very little control with all the action in the last 1/5 of a turn. My Marshal tone tones dont seem to work very well ethier i think i might need to do a bit of rewiring.
One thing that does confuse me however is the lack of gain. Even with the tone controls removed from the circuit the difference between the gain pot at 47k and at 0 ( I measure it and it was wired correctly) thus changing the gain from 1 to 11 seem really not as much as i would have expected, i mean i can easily force the cirucit into distortion but this is from an ipod at 80% volume. i would have thought much less would have been needed.
I did measure the ac voltage from it using a multimeter and that said i was getting about 0.250 volts, well if that went into the cicuit with 10x gain first stage surely with my t-amp set at a input sensitivy of 2v and a 1.5 gain in the mixer amp i should be well over full volume but this just isnt the case. I might try swapping the opamps once more to see what happen but i find it quite confusing what is going on.
Phil
You will find that an iPod will not put out as much voltage as a guitar pickup, as well as the impedance will be significantly different.
I just read somewhere that a guitar can put out up to 2.5v peak to peak, surely would means that the preamp actually need very little gain, in my setup at the moment its got up to 100X gain which seems liek it would be total over kill. I think i need to lower it, this should help the noise floor a lot and reduce the hiss i currently get and higher gain settings.
Actually my gain is even higher because the mixer amp also has 3x gain in it, this doesnt of course take into account the losses for the tone controls which will be significant.
Hello all,
I have been workign on my preamp again this weekend with less than stella results.
I wanted to change the setup so i had a fixed gain of 2 in the first section, then a volume pot and tone tones between it and the final fixed 11 x gain section. To this end i took the schematic from 'Tone stack generator' software and built this with a pot on the front for gain control.
I put this back in and the only thign that works is the volume/gain control none of the tone controls are doing anything.
Any ideas?
The simple cause of those symtoms is not having a ground connection at the 1A lug of the mid pot. Validate that it is actually grounded.
Well it is defientely connected, however i did have to bend some of the pins on the pot back into place after bending them 90 degrees back to wire point to point. This is probably the issue, i will check it out tonight.
Thanks a lot.
Oh also i set the input impedance of the first section to 1m, the second section is set the same. I imagine this isnt needed since the issue with the guitar pick has already been dealt with. What would a suitable value be for the second input impedance settign resistor? 100k?
Phil
LJ King, i found that a couple of the pot were broken, replacing them fixed the issue, thanks a lot.
I change input settign resistor int he second stage (r15) from 1m to 50k. This seems to have clear the sound up a lot, more test to follow tonight.
My amp is still pumping at high gain levels. Could this be because i haven't got enough capacitance on the preamp board? i have eliminated the speakers, tried a hifi one and got the same effect, replaced the 25watts class-t with a 100 watt board. And removed the mixer amp temporarily. The only thing the same as the first test is the preamp so i guess this is where the issue originates. At the moment the virtual ground is only decoupled with 1uf and the 25v to gnd with 10uf? I think i might put some decent 100uf in both locactions to see what happens.
Thats completly fixed the preamp,to little capacitance between the virtual ground and 0v, sound great and totally stable at any gain setting.
I am so happy!!
The 100 watt class-D really gives this setup a hell of a kick. Utterly deafening in a flat, my neighbours must be getting a little annoyed. Thanks for all the input guys, even though this thread has turned into a bit of a monologue of late.