Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 03:46:10 PM

Login with username, password and session length

Recent Posts

 

Amp using TDA2030

Started by trialabc, July 23, 2010, 02:07:51 AM

Previous topic - Next topic

J M Fahey

Download and read the AmpCan and Solo amplifier schematics; Nobel's Streetman if you find it.
I'm a firm believer in commercial amplifiers, even if simplified or stripped to the bare bones, rather than well-meaning but lacking in skills or experience amateur "designs" .
Not wanting to hurt anybody, far from that, but I'm sick of seeing the "Minibooster" or the "Electra distortion" or a poorly biased (that's an understatement) Fet with "magical/mojo" values copied verbatim from a famous "X" brand amp (absolutely ridiculous values) repeated over and over, ad-nauseam.
Learn about Op Amps, the general purpose building block, and the easiest to implement.
Besides, that's what 99.9% of commercial products use.

rowdy_riemer

He probably should follow your advice and check out the AmpCan and Solo schematics, or even the  Rod Elliott project 27 schematic you recommended in the http://www.ssguitar.com/index.php?topic=1426.0;wap2 thread. However, don't be too quick to crap all over jfet designs.

Sure, with op-amps, you don't have the biasing concerns that you have with jfets. But there are some nice sounding jfet designs out there. I think the Dr. Boogey circuit absolutely kicks ass. And while I haven't built any of the ROG circuits (well, other than a few ruby amps), they seem pretty good to go. I think I might try building a BSIAB II one of these days. Most of us aren't going to be mass producing amplifiers, and some of us do not mind spending a little time biasing a FET stage or two every now and then (or maybe we're just gluttons for punishment).  No one should be discouraged from trying their own designs either, whether based on op-amps, jfets, tubes, or whatever. One can learn only so much by copying other designs.

J M Fahey

Hi Rowdy, fully agree with you. :tu:
In fact, I do use Fets a lot, and have for years.
But I spend all of 30 seconds extra and bias them.
My heart hurts when I see somebody "copies" some classic tube design (hey, the idea is excellent, by itself) but places a 1k5 resistor source to ground "because Fender did that" or a 2k7 "because Marshall did that" or even worse, 820 ohms, which is poor biasing (too hot) even on a tube amp.
That's just magic thinking.
Then everybody lies to himself thinking that they have a 100K "plate" resistor, "just as the original did".
Fact is, that is a 100K trimmer, and the actual resistance value they set to "bias" (really they do not set bias but load) is way below that.
I encourage you to build any of those stages, and post the actual "plate" resistor value set. You'll see it's nothing like 100K.
Fender , in fact, did get those biasing values from 12AX7 datasheets, which precisely state, for around 250V supplies, 1k5 cathode and 100K anode resistors.
You change +B, you change those values, there's nothing magic or immutable about that.
You change tube type, they also change (just read the datasheets).
Imagine what happens when you do not even use tubes but Fets, which to boot vary a lot from unit to unit, and to change things even further, use 9V !!!
Then those 820, 1K5 or 2K7 resistors lose all meaning.
Of course, the allmighty 100K trimmer comes to your rescue !!! ... sort of.
I fully agree with "spending a little time biasing a FET stage or two ", precisely that's my point !!
I think that the right place for those trimmers is source to ground, because there they become real "bias" trimmers; of course I suggest using 10K trimmers.
Another way to do that (which I use) is buying a few extra Fets and Protoboard a simple test jig and get the actual bias voltage needed , then label them, and later use the right source resistor to get that voltage.
It takes 30 seconds per Fet.
As you see, I encourage actual testing, not superstition or Mojo .
There's even a simple sound test I suggest:
build a 12AX7 stage, with 1K5 resistor on its cathode (bypassed by 10uF), 100K on its plate, +250V (the real Mc Coy); and a 9V powered Fet stage, with the same value resistors (well, using a 100K trimpot) , with grid and gate connected to the same 1M resistor to ground and an input jack, outputs taken from identical .047uF x400V Polyester caps , DC discharged to ground through 1M resistors to avoid popping; then hook your guitar, play a chord and switch back and forth between both outputs. An MP3 or even a You Tube video would be very useful.
Good luck. :tu:

rowdy_riemer

Ok, gotcha. :tu:

I've wondered myself why use a pot for the drain resistor rather than the source. I also prefer using a test rig and using the proper value resistors rather than use a trimpot. It makes building a circuit a bit cheaper too when you can get resistors for a couple cents a piece verses more than a $1.00 (U.S) per trimmer. I've noticed that with a fixed Rd, if Rd is too high, you can't midpoint bias the stage without setting Rs so high that you have less than unity gain, which, of course, proves your point that one really needs to choose resistor values for each individual jfet. When I built my first headphone amp way back when (with mpf102's), I wondered why adding jfet stages made it quieter  :lmao:. I suppose since Rd has an effect on output impedance, it makes a bit of difference tone wise. If someone wants a Rd close to a specific value for a good reason (not just because some plate resistor in a tube amp design used that value), then he or she should probably hand pick a jfet that fits that Rd value.

rowdy_riemer

QuoteI suppose since Rd has an effect on output impedance, it makes a bit of difference tone wise. If someone wants a Rd close to a specific value for a....

Opps, I forgot that you were discussing magic Rs values. Still, 100k may be to much to get unity gain for many fets, especially ones with relatively high Idss.

trialabc

Guys, thanks for all your suggestions. I have scanned through the schematics of Solo and AmpCan.

Sorry that I am not that keen on circuit design for the moment. Although I would like to learn the behind working principal of the JFET stages, I could hardly understand what you are talking about...

By the way, when I study the schematic of AmpCan, I find diode D1 and D2 are in opposite directions and connecting in parallel. 0.0  Can't understand how it works.... Mind a little more help?

For this moment, I think Op amp will do. Don't wanna mess up the biasing problem.... Seems so difficult....

rowdy_riemer

When the voltage across those diode is high enough for the diodes to start conducting, they increase the feedback current dynamically reducing the gain. The result is soft clipping. Read up on the theory behind the Tube Screamer for a better description. Here's a pretty good description: http://www.bteaudio.com/articles/TSS/TSS.html.

trialabc

Thanks a lot rowdy_riemer !!

Now I really learn something instead of just copying the circuit.....

Hopefully I should have time to start building the amp by this weekend. See if I can finish it, try it and post a picture here. Oh... how to post a photo or schematic?

rowdy_riemer

QuoteThanks a lot rowdy_riemer !!

I'm glad to help. To upload an image, click the arrow next to Additional Options and you'll get the UI for attaching a file. Also, you can put a link to a url to a file using the "insert image" toolbutton above.

BTW, my 7 year old daughter is right next to me bugging me about letting her put a smiley face on this reply, so here it is.  :tu:

J M Fahey

An "off the shelf" Fet gain block that *works* is the Stratoblaster.(Google it)
It uses 9V, 12K source, 22K drain, and doubtlessly (at least on the original ones) a "good" Fet, that's to say one with low Vp (cutoff) and high channel resistance.
I know this is counterintuitive, or "against logic" but it's so.
Unfortunately most cheap and widely available Fets today have the opposite parameters, which makes them good for switching and poor for audio.
Tomorrow I'l Proto one and post what I measure, but from what I remember, gain was not much more than 5x or thereabouts.
With a properly bised Fet and +20 to +30V supply I was able to get 20x gain with available (poor) Fets.
A side note to  Rowdy about his Daughter: "start getting used to it, it's just the beginning".
Mine is 16 and it has been impossible to say her "no" for about, should we say, ¿15 years? :o  ;D

rowdy_riemer

QuoteAn "off the shelf" Fet gain block that *works* is the Stratoblaster.(Google it)

It's funny but the schematic for it on General Guitar Gadgets says to use either a "j201, 2n5457, or other FET". Wow, FET variation is bad enough using the same type from the same manufacturer.  :duh. But GGG's bad recommendation aside, the circuit looks good to go. If I built one, though, I'd still pick the Rs that most closely suits whatever fet I chose. Then again, being a first gain stage in the chain, one's probably not going to overdrive this, and with the way their doing their gain control I guess a predictable Rd to Rs ratio might be a little more important than a approximate midpoint bias. It looks like they want a minimum of unity gain (I think I got this right), and I guess that might not be so simple with varying Rd to Rs ratios.

QuoteIt uses 9V, 12K source, 22K drain, and doubtlessly (at least on the original ones) a "good" Fet, that's to say one with low Vp (cutoff) and high channel resistance.
I know this is counterintuitive, or "against logic" but it's so.
Unfortunately most cheap and widely available Fets today have the opposite parameters, which makes them good for switching and poor for audio.

I know it's harder to get much gain out of  a FET like a 2n5457 (relatively high Idss, relatively high Vp), but aren't they less noisey? It seems like when playing around with a super simple booster design of mine, 2n5457's were less noisey than j201's. But then again, I really didn't give them the comparison they deserve, so I'm not that sure. I know the first version I built has one j201 and is certainly more noisey than my two stage 2n5457 version. With the first version, I used a couple of mpf102's simply because my 2n5457's had not arrived yet. Of course, even with an 18V supply, one doesn't get much gain from a mpf102, so I added a j201 stage. Of course the noise I'm hearing might not be the j201's fault. Maybe it's the mpf102's causing the problem. Or maybe something else entirely. Since this was just something I wanted to work on while waiting for my 2n5457's to come in, I really didn't investigate. Maybe my impression that low Idss FETs are noisier (noisiness inversely perportional to Idss) is totally bogus? What does your experience with FETs tell you?

Quote
A side note to  Rowdy about his Daughter: "start getting used to it, it's just the beginning".
Mine is 16 and it has been impossible to say her "no" for about, should we say, ¿15 years?

:lmao: Yeah, she started getting her way (well alot of the time) when she was about 1, so I've had 6 years to try to get used to it. She's got a very good "puppydog face" she uses against me if I try to tell her no.

J M Fahey

Hi Rowdy.
Those Fets are "reccommended by GGG" which is only a suggestion of available Fets.
I'm sure the original ones "unspecified in the schematic" were specially selected ones.
I use a lot of Fets, but usually as switchers or limiters/compressors, so for me , finding 10 "excellent" ones and another 15 "good" ones in a bag of 100 is fine.
Maybe that becomes a little steep for an average experimenter who only needs 2 or 3 of them.
My usual crop (out of 100) and only measuring  Vp is: around 10 with Vp close to 1.5V (excellent, good to copy Tube stages, they can be biased with the "magic 1k5") they can provide gain up to 20x; around 15 with Vp close to 2.5V which *do* have some gain, say 6x or 7x; 10 real bad with Vp around 4V , useful only as switchers, and all others, Vp 3.5V, usable as compressors or switchers.
Yes, they have that much variation !!!!
That apparently high  12K Source resistor used by Alembic/Stratoblaster sort of "irons out" differences.
Besides, their intention was not making a metal monster, but boosting a relatively weak classic Strat from the early 70's.
There, 3x gain is enough to put it on the same level of a hot Les Paul or SG, its main contenders way back then, and keeping Fender twang to boot.
On this weekend I'll Proto some Fet stages and post results.
No, no simulation for me.
In Music Electronics Forum somebody dialed the datasheet specified transconductance for a high voltage Fet and a 100K load resistor, and came out with an 80dB gain.
Fact is, that transconductance is specified at , say, 500mA, but that load resistor dictates less than 1 mA. Absolutely different.

JayFett

The Vox Pathfinder 15 and 15R use the TDA2030 chip, and the amp always gets mentioned as one of the favorite good-sounding solid state guitar amplifiers on various forums. Here is a link to a schematic of the non-reverb '15' version...

http://www.diyaudio.com/forums/attachments/musical-instruments/153152d1262970400-vox-pathfinder-15-r-schematics-wanted-vox-pathfinder-15.pdf

I guess the op-amp based preamp must work quite well for this amplifier to get as much respect as it does. Folks say the clean setting sounds very tube-like. Check out various reviews.

darwindeathcat

#28
I just bought a used Jay Turser Classic 10 amp which uses the tda2030 as the power amp stage. (http://www.jayturser.com/classic-10.php). It sounds REMARKABLY good for an amp that only costs ~$80 new, and which I bought used for only $35. I have not traced the circuit, but merely looked at the component side of the board (see gut shot below). Preamp stage is a single FET front end (J103) followed with a 4558P opamp. There seems to be two red LED's in this stage, most probably for clipping. The preamp stage has one gain control, and a "higain" switch. IMO, the higain switch is pretty useless, as it doesn't change the voice of the distortion, but rather just bumps the gain in the first stage. This means that it's the same tone, but just that you can use a smaller sweep of the gain pot before feedback. There is a no-name transistor (it's marked with a number, and then "Ge", so possibly a germanium?) after the opamp, probably used as a gain stage to follows the very effective but also very interactive triple band tone control section. After this, there is the master volume control that goes straight into TDA 2030 power amp stage. The amp uses a 6" 4ohm no-name speaker.

This amp is dual powered. Either straight in from AC (discreet diode full wave rectified single rail ps to +12v, with 2200uf cap to filter), or direct from onboard rechargeable 12v battery pack (probably NiCad, but maybe NiMh. I didn't open the battery pack to confirm). The battery charges when amp is plugged into AC.

Component quality is pretty good too. 1/8watt carbon film resistors in the signal chain, 10watt metal film ones in the PS. Mostly Poly film caps (the green ones) in the signal chain, but a few ceramic in low values (eg 0.001 uf and lower). Decent run-of-the-mill electrolytic can caps in the PS and elsewhere. Could have a little bit more filtering in the PS, but the amp is very low noise as it is, so I'm not too worried about that. A very low hum is only audible with the gain and volume cranked. Plastic jacks for input and headphone out, but that does not bother me as they are all mounted very securely. There's little worry of breaking them. Plenty of heatsink for the chip (a mediumish aluminum heatsink that is also thermally coupled to the larger steel chassis). The electronics are housed in a really nice steel chassis, so no worries over interference. The transformer could be bigger, and it is a standard iron lump (ie. not a toroidal), but seems to be quite adequate. I have no idea what the VA rating is, however. The cab is pretty well constructed, and is a combination of solid wood (for the front and internal supports) and particle board (sides and back). It's VERY cool looking, with the inlaid wood frontplate and the tweed covering on all other sides. Battery life seems good. I played it for a couple of hours on battery life, and it showed no signs of draining. Probably 3 hours will be the max bat life. One could potentially increase this by building a custom battery pack to increase the AmpHour rating. Probably have to stick to NiCad (or NiMh if it's that), and not lead acid, as the charging circuits are sure to be different.

This amp sounds VERY good. Plenty of tonal variety, good volume output for only 12v power (it claims 10w, but probably more like 6-7 as JMFahey says), and the no-name 6inch speaker sounds surprisingly good! It actually puts out respectable levels of bass! One could definitely fit in an 8"speaker in the space available, and that could potentially increase volume output and improve the tone a bit, but the stock speaker sounds pretty damn good! I bought it for busking, and I think it sounds WAY better than the Legendary Pignose I used to own, or the Roland Microcube I was also considering to buy. I'll be mainly playing harp through it, but it also sounded really good with my Cigar Box Guitar, which has a homemade single coil pickup in it. You can get quite clean (although only at a notably lower volume), and you can get ridiculously dirty (metalish). This amp is best at a nice overdrive tone (bluesy). At this setting, you can get deep and dark by cranking bass and dialing down the treble and mids, or you can get bright and cutting by dialing down the bass and uping the trebel and mids. Very different tones, but all at the same gain/volume level.

Overall, I think that if you are thinking of making an amp on the 2030, you'll be better off buying this amp and modifying it. The chassis alone is worth the money, plus you'll be getting the tranny, the power supply, the battery, the pots, the jacks, etc... Having built several chipamps completely from scratch now, I think that finding a good pre-fabbed amp, and modifying the circuit a bit (ie. changing or adding components to the preamp and tone control stages) is the best way to go.

Possible mods are 1) upgrading the tranny. 2) upgrading battery pack 3) changing what the "higain" switch does 4)changing the voicing of the preamp section (need to trace the circuit first) 4) changing out the speaker 5)adding a speaker out jack to run  bigger cab. But the amp sounds quite good stock anyway, so even these things would be mostly just for "fun".


Here's a gut shot:
__ ------------------- __
__ | | | | | | | | | | | __
    --------------------