Help designing schematic for 10W+ stereo portable mini amp

[SurreyNick]

As the topic title says, is there anyone out there who can help me design a circuit and schematic for a 10W+ stereo portable mini amp?

I am new to solid state circuitry, but I have a good working knowledge of vintage valve radios, so I can follow a schematic, assemble a unit etc. etc. I have just finished making my son a 'toy' 0.5W 'baccy-tin' amp and now want to make him a decent portable amp, still battery powered though (12v 4.5Ah sealed lead-acid), with optional DC adapter also used to charge the battery. This will be for practice play and for taking to the park/camping etc. He could easily buy himself a commercial unit, but I like making things and want to have a go as a surprise gift. Oh, I should mention he's a pretty good player and into heavy metal in a big way.

I'm not aiming to make anything fancy, no special effects required, just one 6.5mm input, volume, tone and gain controls, with the option to listen via the speakers or switch to headphone socket. For the speakers I was planning on two Jensen MOD5-30's. I have designed the cabinet which is 7" cuboid and the sides open out to create two wings, each of which will hold a speaker. I have attached a rough CAD drawing to give an idea and an image to show what it could look like when closed (I would use a different leather so it doesn't look so old-man!).

So, is there anyone who can help me with a circuit?

Thanks

Nick

[J M Fahey]

Excellent design and idea.  :dbtu:

Just one suggestion, of course you decide: in your design each speaker wing is smaller than possible, because you have a relatively large back section holding the battery, and a bottom one holding amp and controls.

I would split your 7x7x7 cube in just 2  7x7x3.5 sections, one containing battery and amp (and which in extreme cases could work on its own) and the other just the speaker.

In fact, and being democratic, the battery/amp one might be slightly deeper than the other, so internal effective volume is the same.

In any case, I suggest using 2 car type bridged chipamps , each providing some 16W RMS into 4 ohms, or a single stereo one with one channel driving each 4 ohms speaker, in which case you would get some 5+5 W RMS.

I'm old style and still use TDA2005, of which I still have a few, but there are more modern ones, usually TDAxxxx which are more available (I fear mine are obsolete) , typically are simpler to use (already have the gain and feedback internal and fixed, say, 40X gain) .
I'm certain younger members can suggest a popular one.

As a preamp I suggest you straight build a distortion pedal, for 4 basic reasons:

1) they can sound clean if you wish   , just use low gain "distortion" and high output volume.
And you can add a small switch which puts clipping diodes out of circuit, in which case pedal becomes a Clean Boost.

Compare MXR Distortion+ with same brand clean booster (don't remember the name) which is the exact same thing without diodes and set up for somewhat lower gain ... what can be accomplished with the other half of the Clean/Dist switch.

2) they are already designed to work with 9V single supply, easy to get from a 12V battery ... in fact most are happy to get straight 12V , go figure.

3) they match your 2 knob requirements

4) suggested PCBs or even stripboard or perfboard layouts are available.

And you can splurge and decide on a 3 knob one  , with gain, volume and tone controls.

But 2 knob is fine, I have built many amps similar to what you need.

[SurreyNick]

Excellent advice JM, thank you

I will give serious consideration to your suggestion of dividing the cabinet into two and dedicating just one side to the speakers. My normal practice when designing, be it a homebrew valve radio or other project, is to first mock-up the unit from an old thick cardboard box. It's usually rigid enough to support the weight of the internals and all the design impediments can be overcome before the real build begins.

I have seen the quality of your work JM, and I know you make and sell professionally. At the risk of being too cheeky I note you say you have built many amps similar to what I need. I don't suppose you have a schematic you could share with me do you? Although I understood everything you proposed I have insufficient knowledge of solid state circuitry and guitar amps to design one myself from scratch.

Thanks

Nick 

[J M Fahey]

Thanks

As said before, I use the TDA2005
http://www.st.com/web/en/resource/technical/document/datasheet/CD00000124.pdf
and the circuit is that suggested in the datasheet, it works well and found no need to modify it.
As you see, it can be configured as a 2 x 10W amp or bridged 20W
It's still available but "not recommended for new designs" because there are more modern and simpler versions, I'm certain somebody living in USA can suggest what's today's amp-of-the-day .

Or search EBay, I'm certain there must be a ton of "12V 15/20W amp" kits (or ready built) for peanuts, which save you the hassle of designing and making "just one" PCB

As of the distortion/preamp search GGG (general guitar gadgets) or a couple other pedal builders sites, which offer PCB designs or the PCB themselves.

Just stay clear from Runoffgroove (I guess it's them; if not, sorry) which have a fixed idea about "cloning tube amps using Fets", not bad in theory, horribly implemented.

So in a nutshell you need an MXR Distortion Plus as preamp and a 12V 15/20W àmp (or two) 2 speakers  and a battery.

No need to design individual modules from scratch.

[SurreyNick]

Ah, now that makes a lot of sense. I have found the relevant circuits and it seems pretty straightforward too.
Thanks JM, much appreciated.
Nick

[J M Fahey]

FWIW here's Pignose Hog 20  schematic.

http://www.ssguitar.com/index.php?action=dlattach;topic=1166.0;attach=3301

Just as a guide, because it has some issues.

1) good: it uses a TDA7240 which seems to be the modern incarnation of TDA2005: much simpler to use , needs just 2 resistors and 5 capacitors to work, is already bridged internally, etc.

2) bad: Pignose implementation is wrong.
I have the schematic because somebody posted it while trying to solve an instability problem.
They *should*  have used the datasheet example, period.



3) the Hog preamp is very poor and the "squeal" (distortion)  is a joke ; the MXR is MUCH better , so go for it.

4) you may read this as a reference:
http://www.ssguitar.com/index.php?topic=1166.0

although your implementation will be much better, plus it will have 2 real speakers instead of the skimpy one used by Pignose ... for obvious $$$$ reasons of course.

[JHow]

I like the 'suit case' design here.  It has given me ideas for an old coronet case I have.   :dbtu:

[Roly]

Welcome SurreyNick.

I have a good working knowledge of vintage valve radios

In which case you are going to find this one a doddle.

You are in excellent hands with JMF here, so I just have an incidental question and a couple of general observations.


JMF's point 2).  The data sheet shows two Zoble stability networks, R1 & C6 and R2 & C7, one on each of the two outputs.  Pignose economised by using only one between the outputs.  This caused generic stability problems that require a mod back to the data sheet.


You said "stereo", and I'm a bit confused.  There are actual stereo guitar amps, two channels from inputs to speakers (and I guess these would be most useful with Chorus pedals that have a stereo output), but to me the word "stereo" means an amp with two identical channels with coupled controls and source selectors for domestic Hi-Fi use.

{my homebrew Twin-50 keyboard amp is actually fully stereo capable, but with independent controls}

 
Just as a radio is a collection of stages, RF, mixer, IF, detector, audio, so is a guitar amp, only simpler.  And just as you can Lego^(tm)brick different kinds of each stage together to build a radio, so you can Lego^(tm)brick guitar preamp stages, Fx stages and output stages together.

As always the main things to take care of are the signal level and relative impedances where these stages join.

Chip amps are high gain and wide bandwidth, so like radio stages they like to be bypassed "VHF-style", as close as possible to the chip pins with the shortest loop possible, C5 in particular.

And I really like the steampunk opening case idea.   :dbtu:

[J M Fahey]

I guess he used the word stereo because from the beginning he decided on 2 speakers into 2 identical cabinets and would probably need to use a "stereo" chipamp or buy one so labelled module on EBay.

Of course, both fed the same (mono)  signal.

But I bet google turns far more results using stereo than dual mono.   

Just a guess of course.

[SurreyNick]

Thanks for the compliments regarding the enclosure design, but I can't take all the credit for it. I came across the case being used for something else and I thought it could be nicely adapted for use as a portable amp. As far as stereo is concerned, at the time I did the first post of this thread I did have a vague idea of designing an amp that could also double as a stereo amp for an alternative audio source (iPod etc.), but I abandoned that idea pretty quickly. No, it's going to be a mono amp with dual speakers for guitar only.

With this in mind I have adopted JMF's suggestion of the MXR+ and TDA2005 and have drawn the attached schematic. However, I do have this nagging feeling I have got something wrong, but for looking I can't put my finger on it. Perhaps someone could enlighten me? I have also added a circuit for the purpose of charging the battery from a 12v wal-wart with automatic cut-off. Once I get the circuit sorted out I can then get building .

Thanks

Nick

[g1]

  The only thing immediately jumping out at me right now is that the volume pot wiper is drawn going straight to ground.

[Roly]

Thorts:

Your DC charging cutoff appears to be sensing the charging supply rather than the voltage across the battery.  Even then the idea is a bit of a concern because the voltage rise at the end of charge that you are looking for is about the same time a lead-acid cell starts gassing, and gassing is something to be seriously avoided with SLA's; they vent the gas and you lose electrolyte.

The charging socket is shown "upside down".  I can think of a couple of reasons for you doing this, but you might like to check and confirm.

Me, I'd put the fuse in series with the battery (and on the +ve end 'coz I'm conventional).  The real risk with any battery like this is that if you get a short somewhere it can deliver a very large amount of current and cause a lot of damage, melted tracks and wires, and maybe damage the battery as well.  Doesn't matter which end, but the fuse should go between the SLA and everything else.

You really should have some sort of limiting resistor in the charging current path otherwise the charging current will be whatever your supply can deliver, again, not healthy for a SLA.

I used to work on solar systems with hunking 1000A/Hr led-acid sets but I'm not really up with the care and feeding of SLA's, just know that they are a bit picky, so I suggest you have a sniff around and see what the common charge control practice is.

The input jack could be a shorting type to silence the amp when the guitar is unplugged.

The 1nF right across the input will have a serious dampening effect on a passive guitar and clobber your high end response.  100pF maybe, against RF pickup.

There are op-amps with lower noise than the 4558, but that's a matter of taste.  Put it in a socket and you can try others.

Again the 1nF across the volume control seems a bit large.  In conjunction with the 10k output series resistor this will give a top cut hinge at;

1nF and 10k

Xc = 1/2 Pi f C

f = 1/2 Pi Xc C

1/(2*Pi* 10e3 * 10e-9)=1591.54943092

...only 1.5kHz.  This might be okay for bass but I think it may be a bit excessive for tenor guitar.  This partly depends on how it sounds when the preamp is overdriven, but you may need to tweek this value down a bit (lot).

I haven't checked the power stage, but a couple of points about your headphone output; again you can use a socket with a contact to do the speaker/headphone switching, and it is normal to put some sort of resistive limitation in the headphone path so you can't accidentally blow out your headphones and ear drums if they are low impedance type - twenty watts right in your lughole ain't good.

4 ohm // 4 ohms = 2 ohms load - will the TDA2005 be happy with that low load? (check the datasheet.  It implies that each amp can drive a 2 ohm load, but in bridge I think the minimum will be 4 ohms)

HTH

[J M Fahey]

Agree.
Basically the same but seen with other eyes and worded differently ... can't hurt

1) the power in jack is upside down.
Its ground contact should be ... ground and the tip should be +V .

For power supply you will use hollow plugs and jacks there, not regular 1/4"or 1/8" ones which short themselves by design    :duh

2) you will need to limit battery loading current; simplest is adding in series a plain vanilla 12V 15W`car lamp, kludge a car type lamp holder there or simply scratch bulb base and solder it in place, a few mm away from PCB to help ventilation.

3) your overvoltage limiter is fine, but once triggered (or if you turn it on with battery disconnected) it will latch off and stay there until you unplug it.

The good thing is that on barely reaching a little over 14VDC it will latch off for good, protecting battery.

To make operation point more defined, add, say, 4k7 to ground from node 1N4001/4k7 at 2N4400 base.

Now that we are at it, label all parts, such as R1/... R10/... C1/ ... etc. for reference.

4) 

put the fuse in series with the battery (and on the +ve end 'coz I'm conventional).  The real risk with any battery like this is that if you get a short somewhere it can deliver a very large amount of current

5)

The input jack could be a shorting type to silence the amp when the guitar is unplugged.

6)

The 1nF right across the input will have a serious dampening effect on a passive guitar and clobber your high end response.  100pF maybe, against RF pickup.

True, but that's MXR original design.

A`plain 100pF is not enough, but I've had good results splitting the difference:  220pF from Op Amp input to ground.
Yes, on the other end of input 10K resistor.
I tried that (and many other values) and is inaudible, switched in/out .

7) original is single 741 op amp.
RC4558 is fine, but is double.
You have to use a single there (TL071 is fine) or neutralize unused Op Amp which otherwise will introduce unexplained instability.
Best is to short Out to -In and connect +In to C ..... ok .... I'll wait until you label parts

the gain pot is actually Reverse Audio (C3 curve) 500K .

I've successfully modded the circuit to use a much easier to find Audio pot by shorting gain pot shown (so that leg of NFB is just 4k7 in series with 4n7) and replace 1M feedback resistor with a 1M audio pot.

Works like a charm and silky smooth control range.

9) 1nF in series with 10K means 16KHz , quite acceptable
Doublecheck at: http://www.sengpielaudio.com/calculator-RCpad.htm

10)

but a couple of points about your headphone output; again you can use a socket with a contact to do the speaker/headphone switching, and it is normal to put some sort of resistive limitation in the headphone path so you can't accidentally blow out your headphones and ear drums if they are low impedance type - twenty watts right in your lughole ain't good.

Agree.
a) connect the headphone jack with both hot contacts tied together (L+R) to one TDA2005 out, adding in series a 100uF cap to stop DC (there's +6VDC present) plus 100 to 470 ohms (your choice)  to pad earphone out somewhat and not blow a hole between your ears (where the brain usually resides)
The other headphone out should be grounded and nothing else.
As shown you are shorting, both DC and AC, the lower TDA2005 vpower amp.
Remember it's a floating bridge output, none is Ground.

11) the load should be 4 ohms total for each TDA2005 involved, so you either use 2 8 ohm speakers, or just a single 4 ohms one, or my favorite weapon: I use TWO TDA2005 , each driving its own (ejem, FAHEY brand ) 4 ohms speaker .... the loudest kid on the block .

Of course, it runs the battery down twice as fast, so it's not always the best choice.
But it'ss FUN  :lmao:

12) now to the unspeakable : you would need to design and make a PCB for this amp, but for a single one (or a couple) you can very well build it on perfboard.

13) almost forgot: you may add a spdt switch which both lifts diodes out of circuit and adds a fixed, say, 10k resistor in parallel with my suggested gain pot so Op Amp becomes a very clean 3X gain preamp.

[Roly]

9) OOPS, yep, it would be helpful if my help was actually right. 

That line should read;

f = 1/(2 * Pi * 10e3 * 1e-9) = 15915.4943092

My bad - sorry.  No problem there.


11) Might as well go full stereo then.  8|




Some general guidelines for lead-acid batteries.

The capacity of a battery, C, is given in amp/hours, AHr.  A battery with a rating of 8AHr means that it can (theoretically) deliver 8 amps for one hour from full charge to "flat" (which is about 11.5V for a nominal 12V battery, not zero), or 0.8A for 10 hours, or 0.08A for 100 hours, etc.

For best life a L-A battery should never be charged or discharged at more than the ten hour rate, i.e. one-tenth of its AHr rating.  So an 8AHr battery should not be charged or discharged at a current greater than 0.8A, 800mA.  {While this is also true for most battery types such as Ni-Cd batteries, Lithium batteries are a whole other thing.}

Most L-A batteries will tolerate a continuous charge of 1/100th C.

Batteries are not 100% efficient, so they require more charge put in than you can get out, and generally this means 14 hours of charge for 10 hours of discharge.  SLA's are more efficient than typical "flooded" batteries (e.g. car batteries) so they require less overcharge, perhaps only 12 hours. 

The full charge end point is defined as a terminal voltage of between 14.1V to 14.7V for a 12V battery, but this must then be reduced to around 13V or gassing and loss of electrolyte will occur. 

A L-A battery should never be discharged below half its operating voltage.

Unfortunately all voltages around L-A batteries are somewhat temperature dependent, so the ideal charger will take account of the battery temperature, but in typical domestic applications at 25ºC simple voltage limits will suffice.  Operation above about 40ºC or near freezing are to be avoided.  The freezing point of battery electrolyte depends on the state of charge, and a fully charged battery will freeze at a much lower temperature than a discharged one.  {Temperature with Lithium batteries is much more critical, which is why they almost always have an embedded temperature sensor, and hence more than just two terminals to connect to the charger.}

L-A batteries should be fully charged before storage, and topped up every month or two (due to self-discharge).

All of the above is general and approximate and the actual manufacturers recommendations should be followed.  Batteries are a whole area of study in their own right, but if you observe the voltage and current limits that will normally be enough to give good battery life.

{You may observe that some batteries in some applications, such as power tools, violate some of these rules, in particular fast charging at a one hour rate or even greater.  What they don't tell you is that this is flogging the daylights out of the battery and accounts for why so many fail after a fairly short working life.}

[SurreyNick]

Thank you each of you for the guidance, it is really appreciated.

OK. I have numbered all the components, which I should have done from the outset, so my apologies there and I have attached an updated schematic. In the process I tried to address each of the points made by 'g1', 'Roly' and 'JM Fahey' but I confess I haven't succeeded entirely so if I could ask your indulgence again to correct any omissions and mistakes I would be most grateful. I have followed the numbering convention in JMF's post for continuity.

(0)    "the volume pot wiper is drawn going straight to ground"
I have checked a number of schematics employing the TDA2005 and each show a path to ground on the input line. I have amended the schematic to show a direct path from the volume pot (VR2) to pin 1 of the TDA2005 via capacitor C9, but should I retain a path to ground too, as highlighted on the schematic?

(1)    "the power in jack is upside down"
I have checked the schematic I was using for this part of the circuit and I am pretty sure I have connected the power in jack as was shown. It came from this YouTube video (https://www.youtube.com/watch?v=c_XkuKeSLGs) and I have attached a picture of the circuit too. Am I being dumb?

(2)    "you will need to limit battery loading current; simplest is adding in series a plain vanilla 12V 15W`car lamp"
OK, I think I have done that correctly. See L1 as highlighted on schematic.

(3)    "To make operation point more defined, add, say, 4k7 to ground from node 1N4001/4k7 at 2N4400 base"
OK, I hope I have done that correctly. See R19 as highlighted on schematic.

(4)    "put the fuse in series with the battery (and on the +ve end 'coz I'm conventional)"
I have added a fuse as I believe has been suggested. See F2 as highlighted on the schematic. The original idea was to have a fuse in the charging circuit to protect the battery in case of adaptor and/or circuit failure. Not sure if that was or was not required or done correctly. In one or both cases I was thinking of a 1amp fuse, but can I get away with less?

(5)    "The input jack could be a shorting type"
Thanks for the suggestion, I may well switch to one when I order the components for the build.

(6)    "The 1nF right across the input will have a serious dampening effect on a passive guitar"
I am assuming you are referring to the capacitor now labelled C3 on the schematic attached. Is that right? As JMF says, it is the MXR original design. Are you recommending changing it (C3) to 220pF instead?

(7)    "original is single 741 op amp. RC4558 is fine, but is double. You have to use a single there (TL071 is fine) or neutralize unused Op Amp"
I have used the attached schematic for the MXR+ which recommends the RC4558 but it didn't mention neutralising the unused Op-Amp. It would be most helpful if you could describe how to do so now I have labelled the components. Thanks.

(    "I've successfully modded the circuit to use a much easier to find Audio pot by shorting gain pot shown (so that leg of NFB is just 4k7 in series with 4n7) and replace 1M feedback resistor with a 1M audio pot"
Sorry, I don't understand

(9)    "1nF in series with 10K means 16KHz , quite acceptable"
So no change to circuit made.

(10)    "connect the headphone jack with both hot contacts tied together (L+R) to one TDA2005 out, adding in series a 100uF cap to stop DC (there's +6VDC present) plus 100 to 470 ohms (your choice)...and... The other headphone out should be grounded and nothing else. As shown you are shorting, both DC and AC, the lower TDA2005 vpower amp."
I confess the connection of the speaker(s) to the dual outputs of the TDA2005 completely confused me. The bridged application schematic (attached) shows a single speaker of which one terminal is connected to pin 10 (output 1) and the other terminal to pin 8 (output 2), whereas I would have expected either pin 10 or pin 8 to go to the positive terminal of the speaker and negative straight to earth (-). I'm afraid I still don't understand how to wire this correctly for two speakers plus optional (switched) headphones in spite of your explanations! Sorry.   

(11)    "either use 2 8 ohm speakers, or just a single 4 ohms one, or my favourite weapon: I use TWO TDA2005 , each driving its own (ejem, FAHEY brand ) 4 ohms speaker"
In light of item (10) above I'm undecided how to proceed (2x8ohm, 1 larger 4ohm, or 2xTDA2005 and 2x4ohm).

(12)    "you would need to design and make a PCB for this amp, but for a single one (or a couple) you can very well build it on perfboard"
Yep, as it's a one-off I plan to do it on perfboard or prototyping board.

(13)    "you may add a spdt switch which both lifts diodes out of circuit and adds a fixed, say, 10k resistor in parallel with my suggested gain pot so Op Amp becomes a very clean 3X gain preamp"
Sorry, call me thick but I don't understand.

Attachments:
1. Revised schematic
2. Circuit I adopted for the MXR+
3. Circuit for TDA2005 for bridged application

Thanks

Nick