howdy folks, i decided to build this simple amp as a first step into building amps,,its a dean markley chip amp but the preamp is pretty cool,its capable of driving 10 watts into a 4ohms speaker.
i included the circuit here for anyone who's interested
anyways i have some points wich are not so clear to me in this circuit,
1/wuts the R1 for ?
2/about the caps its written for example c1= 473M,,, wut the hell is M?? microfarads?and some caps like c13 is 22/16 ,,whats that stands for? 22microF/16 volts?????
3/i didnt get what the rocker switch connet n disconnect,,can anyone discribe the way it works(what terminals it connects n dont)?
4/about the pots ,, wuts A250K B250K ?? wut ranges should they take?
5/can i get rid of c19? or its better to have it?
6/wuts the point of having R10 and C11 ???
7/any comments about what kind of speaker and op-amps are suitable for this circuit is pretty welcome
.....thanks already for any answers,,hope this amp sounds good
here is the amp
1. Since I see no shunting capacitor I believe its sole purpose is to limit current and thus protect the input stage. Anyone?
2. IMO, this is again FAQ material: First two digits=value (in picofarads), third digit=multiplier (or number of zeros i.e. 1=10, 2=100 3=1000 etc.)
Letter code marks tolerance:
C: +/- 0.25pF
D: +/- 0.5pF
F: +/- 1%
G: +/- 2%
J: +/- 5%
K: +/- 10%
M: +/- 20%
P: -0% / +100%
Z: - 20% / +80%
So, for example, 330P stands for 33 picofarads -0% / +100%. 473J stands for 47 000 pF (0.047uF or 47 nF) +/-5%. Simple?
3. They obviously tried to imply that the switch is integrated to footswitch jack. Anyway, you can picture the "block" moving left to right (and vice versa) thus connecting the center terminal to either the left- or rightmost terminal. When drive is on, "input" comes trough R6 and C5 and the diodes are connected to feedback loop to clip the signal. When drive is off the diode clipper is removed from the loop and input comes through R5 & C4, which obviously reduces gain.
4. The letter marks the taper.
Old code: A=linear, C=Logarithmic (audio) and F=Antilog.
New code: A=Logarithmic (audio), B=Linear, antilog pots are no longer in production
5. Theoretically you could get rid of it. However, I think they have a good reason for putting it there. Anyone?
6. Same as having R9 and C10: It's an RC supply filter. Essentially it lowers the supply rail a bit, filters it and decouples the opamp supply from the supply of the output stage's power opamp. This reduces interference between different stages of the circuit.
7. Opamps can be any generic model (i.e. 4558, as implied by the schematic). Use dedicated instrument or pro audio (PA) speakers. Do not use consumer grade HiFi speakers.
thnx a lot man, u answered all my questions..
i was thinking to do any modification to the circuit but it looks pretty cool i dont wanna change anything
ill try to omit the c19 and see how does it affect the sound,, i dont think its more than a protection against dc .
since there are q's can be included in a FAQ ,why dont u guys make one? who is in charge here?since it saves a lot of questions its a good idea to go with.
again there are some caps in the circuit thats drawn different like c15
whats that shape means? and are they ufarad/volts ?
Yep, in that case those are uF / volts. If you look closer you will notice that they have a marking, which indicates the correct polarity. The alternative shape simply distincts them from non-polar capacitors. I guess both IEC code and a normal format of uF were used simultaneously because polarized caps are often electrolytics, which have the capacitance printed to side of them in the latter format. Non-polar ones tend to indicate the capacitance using IEC code.
If you omit C19 be sure to measure DC offset in the output before attaching a speaker. I have feeling it may vary when treble or bass control is adjusted.
By the way, if you make R8 (5.1K) in the tone control variable you got yourself a "mid-range" control. I think 10K linear potentiometer wired as a variable resistor should do the job.
Other mods: I would hookup two diodes from IC2 output (pin 4) to rails like this: Diode #1 cathode connects the positive rail, anode connects the output. Diode #2 cathode connects the output, anode connects the negative rail. This will protect the output from inductive peaking. In normal conditions the diodes do nothing but when the output signal exceeds that of the rail voltage (plus voltage drop over diode) the signal is shunted to supply and clipped. These diodes must be able to take some current: For example, 1N4002 is a proper device choice for the purpose.
Hook a similar diode arrangement to the pin 5 of the input opamp and you have yourself an overvoltage protection, which will protect the input stage. Cheap amps often lack these simple protections. 1N914 does fine here since you need less current handling capability than in the output.
Bypasss supply pins of the opamp (pins 8 and 4) to ground via 100 nF capacitor as close to the pins as possible. These caps will compensate the inductance of wiring and act faster to transient current draw than the main 220 uF caps. This will increase stability and decrease the possibility of the opamp oscillating by default. You can also hookup pins 8 and 4 together with a third 100 nF capacitor.
As shown, the headphone out feature is rather useless. You need to wire it in such a way that plugging in the phones will disconnect the speaker. There are phonejacks that come with a suitable integrated switch for the purpose. It is strange that they chose to take feed for "line out" before the capacitor.
Yep, it looks like a very basic, simple and cheap amp. Expect it to sound like one as well.
Quoteif you make R8 (5.1K) in the tone control variable you got yourself a "mid-range" control
thats a nice shot,,but im not pretty sure of the base range and the midrange in this circuit as long as the -3db of the highpass of r4/c3 is about 372 hz ,,,why on earth they cut the frequencies below? wouldnt that affect the bass control,,lookslike it needs too much boosting to have base ,,, :o
same thing for other c2/r4 a lowpass with about 300,000.0 hz as a cutoff ,,why that much? isnt better to make it somthin about 30,000.0 hz ??????????????
you got also the lowpass of the volume control pot with the c6,,the band pass would get smaller or bigger as we vary the volume. wuts the point??protection against high amplitudes at high frequencies?could be?
in general the whole circuit looks fine to me but the cutoff frequencies of the filters look a lil strange as i indicated above.
QuoteHook a similar diode arrangement to the pin 5 of the input opamp and you have yourself an overvoltage protection, which will protect the input stage.
what should cause an overvoltage at the inputs? too much gain?
QuoteAs shown, the headphone out feature is rather useless. You need to wire it in such a way that plugging in the phones will disconnect the speaker. There are phonejacks that come with a suitable integrated switch for the purpose. It is strange that they chose to take feed for "line out" before the capacitor.
yea you are right and i decided to not include this headphone plug at all,about having the feed before the cap ,,mmaybe headphones'speakers have their own protection against dc,,or low dc voltage would do no harm.
QuoteYep, it looks like a very basic, simple and cheap amp. Expect it to sound like one as well.
yea it is but isnt chip amps supposed to sound better than discrete ones?
why would it sound cheap? a chip power amp with an op-amp preamp
,,i guess that should sound fine...maybe using the TL072 would make it better
If you want a really killer small amp, build a Dean Markley K-20X. The schematic is still on dean markley's website. I bought one as my first amp 7 years ago and it is still my favorite practice amp. I hooked it up to a 10" dean markley cab (it comes with a built in 8" speaker) and it is loud enough with only 14w RMS to keep up with bass and drums. Plus it has excellent tone.
They may have introduced so much low frequency cut to make the overdrive sound less muddy, who knows... It can be a mistake as well. (Those Dean Markley schematics seem to be full of them). The low-pass filter becoming more prominent at higher volume settings may be beneficial in removing "harsh" high frequency content from the signal at higher volumes. Anyway, I don't like the arrangement that drives an opamp stage with negative gain (attenuator).
The "overvoltage" protection protects (or at least tries to protect) the input from sources that introduce a too high input signal amplitude. For example, if someone decides to "experiment" by hooking up a power amp out to the input of this amp. ...Or something. Even static electricity can cause high voltage spikes. These are poison to the input stage of the opamp.
About that output cap: Headphones do not have a DC protection (they're just speakers) - and yes, low DC voltage, i.e. few millivolts, does no harm but it does no good either: It can't move the speaker cone so it only increases the heat dissipation of the voice coil.
On the other hand, many "line in" inputs are AC coupled so they can be fed from a point before the coupling cap.
Still, I don't know why the included the output capacitor (if it's a hi-pass it would have been cheaper and easier to implement it at the preamp side). Like I stated, I assume it is a DC protection since messing up with the tone control values may upset the DC offset of the output. They still could have compensated that by other means, i.e. by AC coupling the tone control from the power amp.
Anyway, since the output has nearly zero DC volt potential that capacitor needs to be non-polar. Those are expensive and rather unreliable components. They are also lot less linear than ordinary polarized electrolytics but using such a low value capacitor will cause distortion already. In HiFi amplifiers this arrangement would not be acceptable. In guitar amps where linear reproduction is not a demand it's, well... It doesn't look like good design. However, neither does the whole amp so why bother with guessing the designer's motives.
The tone issue: The chip amp will pretty linearily reproduce the signal fed from the preamp (but with higher power). That is it's function, which it does a lot better than average discrete amps. However, the preamp is the problem: Its basically just two gain stages, one with switchable gain and diode clipper. These drive a fender-type tone control circuit. Don't expect any miracles from such a simple arrangement: Likely the clean sounds are good (they can be even excellent) but I think the O/D channel may have characteristics of a crappy fuzzbox. That's because the circuit practically just puts a couple of clipping diodes to the signal path. Such simple arrangement nearly ever sounds any good.
Alternative opamps can't make circuits to sound any better if the voicing is done badly in the first place - at best they just decrease noise or something.
Tone is highly subjective issue so it's pretty vain to comment or speculate about it. Let's just say that I'm anything but impressed by this circuit. By default, these low power practice amps are results of little money being spent on their design. They are not especially "high-quality". Complement them with a cheap speaker stuffed in as virtually small cabinet as possible and you got yourself an amp you will begin to hate after few years. Simplicity is a virtue all right - but not when it has an effect on tone, performance and features. IMO, the amp circuit suggested by Joecool85 looks way better than this one, allthough it still contains many of the flaws seen already in this circuit (no supply bypass for opamps or the power amp chip, opamp stages running at negative gain, that strange output cap).
Since you are obviously after a DIY project that will provide a satisfactory outcome why not do it right from the beginning? Throw away the idea that the amp schematic should look simple. Learn a new way of thinking. With little and simple mods those amps can be improved dramatically from the original. You can put in features, like footswitches, channel indicators, FX loops etc, that they never even decided to implement (to such a cheap amp) in the first place. Those still can't change the fact that these amps are far too loud for bedroom use and far too silent for serious practice or gigging with a band. Anyway, it's a start. If you make a good preamp now you can later sub the power amp section with something that has higher output power. If you build something cheap you will eventually find yourself hating and never using it. - And likely you just throw the thing away in a year or two. Trust me, that has happened to me many times.
Quotethe preamp is the problem: Its basically just two gain stages, one with switchable gain and diode clipper. These drive a fender-type tone control circuit. Don't expect any miracles from such a simple arrangement
whats the bad stuff? the negative gain? how about modify that for a positive gain, and modify the resistors values,,would it help?
or the bad thing is the fendertype tone control ?? how about making the tone control an active circuit ,,any better?
i dont wana talk now about protections as much as i wanna talk clean sound,,
about the negative gain again all i know is that it shifts the phase 180 degrees,,so wut distorion would it make if i put two negative gain stages like its included in the design so i have the exact copy of the input signal with positive gain(180+180 = 360 = 0) any wrong with that?
good idea about ac coupling the power amp from the tone control ,lookslike we got rid of the cap this way.nice shot :tu: .
QuoteSince you are obviously after a DIY project that will provide a satisfactory outcome why not do it right from the beginning? Throw away the idea that the amp schematic should look simple.
well i was goin for simplicity cuz its gonna be my first built amp as i told you before ,,and yes nothin wrong to apply mods to the design and i will,like changin some resistors values(filters) and have the midrange pot u suggested and bypassing the op-amps with diodes also ill think of better supply,use the TL074 ,i love this IC's s;ew rate,, <3)lol,thats what i think of now.
,if you suggest me another circuit that sounds better than this and is fine for a first build Please let me know.
maybe i should go for joecool's suggestion,.thanks joe for the notice n thanks teemuk for your mods.
I find my K-20X perfect for bedroom use but only because it stills sounds good turned down. It is quite loud with a good speaker. And the built in overdrive is one of the best overdrives I've heard in a while, best for old school rock tones and not much else however.
Quote from: benzer on September 22, 2007, 07:41:01 PM
whats the bad stuff? the negative gain?
Oops. My bad: I should have used a term "less than unity gain", which is more correct. That is prone to instability and opamp manufacturers generally suggest that such configuration should not be used. Why build an active attenuator anyway...
QuoteOops. My bad: I should have used a term "less than unity gain", which is more correct. That is prone to instability and opamp manufacturers generally suggest that such configuration should not be used. Why build an active attenuator anyway...
sorry but i dont see any less than unity gain configuration.
do u mean that the vol pot of the second stage can cause a unity gain?
even though ,,at this level the sound would be very very low so it wouldnt affect much.
well anyways i think ill go for the K-20X as long as joecool thinks that its a GREAT AMP...thanks joe and any of you guys wanna talk its circuit? cuz it looks impressing to me.
Well, isn't it obvious? Roughly the gain of the second stage is determined by VR1 (potentiometer) divided by R5 or R6 (depending on the "drive" setting). This is the basic gain math for inverting opamp configuration. Now, when VR1 is set to lower value than R5 or R6 (again depending on "drive" setting the gain becomes less than unity. For example, 25K/33K = 0.75. Since VR1 is logarithmic this would be the gain with volume at five on "clean" channel. Any lower volume setting would also be attenuating. So, when it comes to clean channel it is pretty realistic to expect that you run this opamp stage mainly as an attenuator, Drive channel is a bit better since R6 is considerably lower than R5.
I don't see what a possible risk of oscillation has to do with volume levels.
Anyway, it is obvious that the designer(s) got away with this. However, if you end up with a sloppier layout or an opamp that's not particularly happy about running it with less than unity gain setting you might end up building a medium power oscillator. I'd rather try to avoid such scenario in the first place.
The other amp...
K-20X looks pretty allright since it at least tries to "voice" the clean and "O/D" channels a bit differently. It still has the same less than unity gain "flaw" for clean channel but I like the way how the gains for different channels are arranged. Since the circuit is not too simple it will also lend itself to more extensive modding if you don't like its tone.
Some other comments...
- The lack of feature to switch channels with a footswitch sucks. If I were you I would spent some time in engineering this feature because it's extremely useful
- The input impedance is, IMO, too low (only 220K)
- It has those protection diodes in input and output I was talking about earlier (and in the particular schematic they at least managed to draw them correctly). Good.
- In case you didn't know, that dual LED can be replaced with discrete red and green LEDs. Using different colors causes asymmetric clipping, which is pretty much the intention. I would try this also with ordinary diodes and see which ones I prefer. The latter might create a bit more aggressive distortion, although you loose some volume in comparison to LEDs. If you like both setups make them switchable.
- Still has that output capacitor. For some reason its larger now...
- The amp uses a current feedback scheme that increases output impedance "to mimic tube amps". It also allows creating a "damping" control if you put some variable series resistance to R14. Very popular setup nowadays.
- Opamps and the chip amp still lack the close vicinity rail decoupling that decreases risk of oscillation
- The schematic has incorrect labeling for bass and mid-range controls
- They at least have tried to correct DC offset versus tone control setting issue by selecting R13 = R12 and using R11 to separate the chip amp input from the tone controls (besides its function as LP-filter). I would still AC couple these two circuits.
- The headphone jack feature of this amp is finally usable since it disconnects the speaker. Value of R18 is not shown
- Make that mains switch to switch off both "Line" and "Neutral". It's safer that way.
- The Zobel: R17: Use non-flammable, medium power (i.e. 1 - 5 W) resistor. C13: Use polypropylene self-healing cap. If the amp oscillates these parts can save the chipamp but will take a lot of punishment.
- The volume control for drive channel is linear so it might be a bit inaccurate in controlling the volume level (i.e. Volume rapidly changes to very loud even at low settings but after that the control seems to have less effect). It's a cheap and annoying trick that is used to fool people to believe that the amp is louder than it actually is. Does your amp suffer from this issue Joecool85?
I guess that was pretty much all I have to say about the K-20X circuit.
some points about the k-20X :
1/ that input configuration R1 R2 C1 C19 , thats a second order bandpass filter ,,well it doesnt make any phase distortion as long as the frequencies it passes are in a very big bandwidth ,,
but still dont know why use this kind of configuration.
2/the cutoff freq of the lowpass C2 R4 is about 225 khz ,,why that much???? couldnt it be just about 40khz???
3/ nice configuration for the second op-amp,a third order filter for the overdrive,but, are the switches SW1A SW1B both switch in the same time?
4/ why use a lowpass to couple the tone control with the chipamp??
nice idea about using two different color leds teemuk ,ill try to include it and make a switch for this,,also the footswitch.
QuoteIt still has the same less than unity gain
cant we modify the values of C3 R5 to solve this? and let the pot take lower values than before
Quote from: teemuk on September 24, 2007, 03:19:03 PM
- The volume control for drive channel is linear so it might be a bit inaccurate in controlling the volume level (i.e. Volume rapidly changes to very loud even at low settings but after that the control seems to have less effect). It's a cheap and annoying trick that is used to fool people to believe that the amp is louder than it actually is. Does your amp suffer from this issue Joecool85?
I guess that was pretty much all I have to say about the K-20X circuit.
Actually yes it does. It doesn't really get any louder past 1/2 volume or so, maybe 2/3. It also picks up volume very quicky on the low end of the knob. For instance, 1/4 volume is relatively loud. I've gotten use to it, but if the pots wear out I will probably replace them with the log pots it should have had to begin with.
I'd also leave the OD circuit as is, at least at first. I love the tone it puts out, it sounds sweet. In fact, it's nice enough that a couple Peavey fanatics I know are jealous of my practice amp.
Let's see if I can answer some of your questions...
1. The input stage: Some likely needs for bandpass here are reducing RF interference and AC coupling the input in case the source happens to have any DC offset. Rolling off some bass and high-freq outside the frequency range of guitar is also very usable.
2. I suppose the value was chosen mainly to suppress oscillations. Lower corner frequency undoubtedly works as well.
3. Yes, the switches are on at the same time. It's a dual switch.
4. In non-inverting opamp configuration the DC offset is lowest when DC resistance from non-inverting input to ground equals the resistance of feedback resistor (R12=R13). However, without the capacitor there is an alternative DC path to ground through the tone control circuit. This is in parallel with R12 and since the tone control circuit's impedance is rather low it imposes a considerable effect. A capacitor can break this path. If you wish to get rid of the output cap then adding this coupling cap may become a considerable tweak.
The unity gain issue and C3 & R5. No, this can't be modified but you can add a 100K series resistor to VR1, which limits the gain drop to unity when VR1 is turned to zero ohms. However, unlike the drive channel (which already has this limiter) the clean channel has no individual volume control. I suppose you could (similarly to drive channel's arangement) add one to "upper" signal path that goes to SW1b.
Unfortunately I can't take credit for the idea of different color LEDs. It' just the inherent nature of the dual LED package or "bi-color LED". The following is a direct quote from wikipedia:
"Bicolor LED units contain two diodes, one in each direction (that is, two diodes in inverse parallel) and each a different color (typically red and green), allowing two-color operation or a range of apparent colors to be created by altering the percentage of time the voltage is in each polarity."
It is a pretty nice invention because a digital control signal (either 0V or 5V) can easily control the color of a "single" LED (see attachment). Nice way to tell when things work (green) and when they don't (red) using a single indicator. The schematic displays the component a bit strangely (it should have three pins) but anyway, I suppose you got the idea.
QuoteHowever, unlike the drive channel (which already has this limiter) the clean channel has no individual volume control. I suppose you could (similarly to drive channel's arangement) add one to "upper" signal path that goes to SW1b.
isnt that VR1 is supposed to do? its a gain modifier for the clean sound so ???
QuoteThe schematic displays the component a bit strangely (it should have three pins)
well its 3pins but 2 of em are connected together and to ground,,right?
onething more is that the tone control itself ac couple its input with its output,there is just no dc path . so is it the output cap put there for the dc offset caused by the power chip?
and also wuts the cool thing about having R14 R16?some idea behind having the C12 inside the feedback loop.!!?
Quote from: benzer on September 25, 2007, 08:15:30 AM
isnt that VR1 is supposed to do? its a gain modifier for the clean sound so ???
Exactly, and turning that potentiometer to less than 100K resistance (with "clean" channel selected) turns that stage into an attenuator (VR1 < R5). Since it's logarithmic any (clean channel) volume setting less than "5" runs the stage as an attenuator. Anyway, I think this issue getting a bit too much attention. You can try that configuration and if it oscillates then just try it with a different opamp (which may cope better with this circuit). The bottom line is that you should get it to work but personally I avoid this kind of configuration since alternatives exist.
Quote
and also wuts the cool thing about having R14 R16?some idea behind having the C12 inside the feedback loop.!!?
I already mentioned it. The circuit is explained here:
http://sound.westhost.com/project56.htm
some important q :
how the heck can we calculate the powerchip voltage gain?
its a lil complicated thing to do.
ill try to get its transfer function n plot it by matlab but it doesnt look easy
and is c11 there to attenuate high freq signals?
K-15´s are great little amps. They sound specially good when coupled with a somewhat larger 10" or even 12" guitar speaker. What saves the distortion, which "should" be buzzy, is that it has no Master control, so when you distort it you´re also clipping the output stage, which clips the buzzy peaks; besides you can´t distort without putting out at least 8/10 watts.
Some answers :
1)R1 suppresses RF interference by a)working with the OpAmp´s input capacitance and b)"Killing" or at least degrading the input junction diode which would have demodulated parasite RF.
2)473=47+"three zeros"=47000 pF=.047 uF
22/16= 22uFx16V
3)Clean gain= pot value/33k or approx. 250/33=8x max.
Dirty gain: 250/1.5=166 ; besides, because of the diodes, the clipping peak voltage lowers from almost 15v to around 700 mV .
4)Log pots preferred (same as Fender uses)
5)I think C19 is a drawing error, maybe carried over from an earlier single supply version. It iss not necessary and can actually harm the sound.
6)R9/10 and C10/11 lower slightly and filter the preamp supply.
Good luck with your project.
PS: the distortion footswith is very crude but actually works; remember to isolate it from ground and use a screened cable, with the screen connected to C5.
PPS:I almost forgot: the gain of the output chip is =(5100/180)+1= around 30. Part of the magic is that the distorted signal (which is attenuated by the tone control) barely drives the power amp; the mix of preamp (buzzy) and power amp (flat and boring) distortions gets interesting: smoother than one and sharper than the other.