/diy/ - Do It Yourself

>>214879
oh man that's one wonky setup, but if it does the job, well in.

>>214883

Not my amp. Putting the tubes right next to the magnet looks pretty dodgy to me. It looks like a cludge in general.

Out do him, folks!

It really depends on what kind of music you play. I play black/thrash so I could care less about that myself. Be sure to use a choke though. You could also wire it for tube rec, and just drop in a silicon bridge rectifier and see what you like.

I'm not an expert, but I can't think of a way to get your B+ to the necessary 350V or so. What output tubes do you want?

>>214892
That is a disgusting looking job. Here's my x88r

>>215168
Should have said
I can't think of a way to get your B+ to the necessary 350V or so, without a proper transformer.

Built & Designed (mostly ripped off other designs) when I was 15

12AX7 Preamp, was going to add another stage of 12AX7 for distortion but didn't get 'round to it, and a 12BH7 final.

Oh, the main reason to use a tube rect is for the "sag" characteristic. With one, as you play louder, say, hit the strings hard, the current demand increases and the tube rect gives a sponge-like sound/feel. I personally like it for bluesy stuff but for a Marshall-like tone I would advise against it.

>>215196
Is that an ECC99 equivalent? How many watts do you get?

>>215203

Pretty much. The ECC99 is like a 12AU7, and the 12BH7 is just a 12AU7 with more current tolerance. AFIAK the same gain and all.

>My amp

interested instructions here:

head-fi.org/t/561994/sripboard-design-starving-student-millett-hybrid-vacum-tube-amp

Oh, OP, fella... I do not know, but please, try the thing with the generic transformer and tell us if it works. I would be very happy, because around me parts, it's a wee difficult to find the right transformers

>>215189

A Buck Boost is a practiced method, but very messy.

What I'm looking for is a list of the Power voltages used in the early small tweed amps - stuff like the Champ. If it's a case of using a cheap transformer to step it up the needed value then we can simply regulate it back down relatively cheaply. I'm talking shaving a 50 dollar power supply down to maybe 20, but that goes a long way.

So the tube rectifier thing is lefit? Cool! Has anyone here tried the Diode with a resistor in series subtitute?

I forget the link, this was the "this": http://www.justinholton.com/hotrod/sag.html

Are you making many of these amps, OP? Not sure why $20 is so important on a hand built amp.

If your xfmr can get you the plate voltages you need and handle the current draw, it's good.

>>215395
I have. I made a little drop-in package with the socket from an old tube, a little heatsink and some RTV. IIRC, I didn't pay for any of it, it was all salvaged junk.

If you're looking at adapting a transformer outside of it's intended purpose, be forewarned that it will be drawing quite a bit of current and will need to be able to handle that.

Filament taps for pre/power tubes need to be taken into consideration, with a draw of around 0.3A per 12AX7 and .5A for a 6v6. The filaments are supposed to be seperate for a tube rec IIRC, and those vary anywhere from 1.5-3A

>>215464

Spare cash is hard to come by and I consider being ecenomic one of the pleasures of engineering. I probably can't improve the sound of a single ended Fender amp, but whose to say I can't make it cheaper?

>>215472

Thanks! Highly subjective, I know, but did you feel it worked in a similar fashion to the Tube Rectifiers?

I'm aware the current handling might have to be quite high and heatsinking may be required. We happen to have two 10 x 4 inch heatsinks my brother found in Maplin for 80 cent a pop, so the means are not beyond us!

Some high voltage zeners could be the ticket and, lo, a pair of 150v, 10Watt zeners gets you mighty close. But at 10 bucks a pop? Not worth it.

The last hope is, a voltage divider. But that just seems stupidly wasteful. ALL OF THE CURRENT.

It's that or a MOSFET being switched very quickly for very short times to act as a SMPS. But that just strikes me as stupid and noisey.

I think the separate rectifier filaments often are for tubes that require differing voltages, such as a 5u4.

I commend your efforts, OP. I enjoy refurbishing or recreating amps from old organ and pa scraps. For 50-75 bucks you can pick up a chassis, xfmrs, and sockets on ebay. Tubes if you're lucky.

Tons of old practice amps use 35w4 rectifiers with a 12au6 and 50c5. These old amps used no power transformer! While these aren't entirely safe, isolation transformers with enough VA rating for the job are only $10.

I think the rectifier is an important aspect, but many great sounding amps get by with just 4 diodes. For how much they effect your final sound, it's hard to place them so high as a design priority.

OP here, I have a solution worked out but not built. As I am severely hungover, I will post power supply schematic and components list+priceing. Design costs around 40 Euro, just a little cheaper than an multiple tap transformer, but should also have signifigantly lower noise. There's also enough headroom to double up the 6V6 in a Fender Champ, allowing you to double your load driving or setup a Class B stage.

Anyway, verbose bump!

>>214879

No offense, homie, but you could have done that about 150% more neatly. The dude who taught me how to work on and build amps would beat me with that thing, if I showed it to him.

>lrn2leaddress
>try again

>>216615

Not mine, generic google pic. I criticised it myself!

>>216615
1) Electrons couldn't care less
2) More than 'neat' it has to be properly soldered and secured in place
3) In reality, for analog apps curved tracks/wires are better because rectangle-angled ones tend to become antennas in some cases
4) The shorter the connection the better, even if 'not neat', though in low-frequency amps (such as audio) it doesn't matter much
5) The only real problem I have with OPs pic is that capacitor touching hi-wattage resistor, which is probably going to get hot

>>216635

Actually, a poor lead dress can lead to excess noise and other extraneous bullshit. Not to mention making it a clusterfuck for whoever has to fix it later when it cooks. Just take your time and be neat. Geeze.

>>216713
I take it you haven't seen many '40-'50 radios? The neat thing about them is how everything is done to minimize costs, and yet work reliably. Though I agree the repairs were often difficult, if only because there was no easy way to access the solder points with the iron without burning some wires and whatnot by accident.

I've yet to see any LW radio or audio amp being ruined by such design. If you know how to route wires (to avoid any interference) and how to solder properly then there's no problem.

Please don't misunderstand, I like neat as well. But above all the project has to work properly. It's easy to make things worse by going for good visual look, which in the end doesn't matter much inside the case. The neat part is the last thing to worry about, and completly optional.

OP Here:
Component cost from Farnell would be about €40, not including Resistors, Caps or shipping. I picked Farnell to give you a fairly sour figure, could only get pricier if you went to Maplin! Shop around.

However, considering the cost of transformers I looked up for Europeans was 50 euros usualy plus shipping - and that doesn't include a rectifier, so say 55.

Now that's a pretty measley saving, except that you've the extra wattage for a 2nd 6V6 stage - so you could parallel a pair of them and drive a larger load if you wanted. Anything under 10 Watts should be happy with this, really. The filament should receive a fairly clean current, too, so power supply noise should be minimal

Did I forget to mention a hit sink? It's going to want a heatsink.

Some of the voltages are loose - that depends on PS filtering. The caps around the Bridge Rec are also optional and can even be substituted for a choke if you have one spare.

Not sure if it's right or worthwhile but here it is anyway!

Rec = 1.92x4
http://ie.farnell.com/fairchild-semiconductor/gbu8j/bridge-rectifier-8a-600v/dp/1017705

12v Zener: 14.20
http://ie.farnell.com/solid-state/1n2976rb/zener-diode-10w-12v-do-4/dp/1672839

100v Zener Diode: 0.79
http://ie.farnell.com/on-semiconductor/1n5378bg/diode-zener-100v-5w/dp/9558349

Transformer: 16.00
http://ie.farnell.com/hobut/ct160-5-f7/transformer-current-10-5-ratio/dp/1758723
*note, use secondary as primary*

Filament Regulator. 0.36
http://ie.farnell.com/taiwan-semiconductor/ts7806cz/ic-v-reg-6v-7806-to-220-3/dp/7173997

Captcha = Amuse.

>Nope.jpg

>>215516
No not at all, it works exactly like a solid state rectifier. I put it in my friends 1974 Marshall that I had to totally overhaul. He was looking for a Celtic Frost type tone so he needed the headroom.
Did you manage to source out a nice choke yet?
I like this idea, OP. Be sure to keep us posted.

>>216635
In simple little stuff like OP's picture it shouldnt matter THAT much, but once you start adding gain stages layout becomes just as important as component quality. The difference between a great sounding amp and a noisy piece of shit entirely depends on grounding and lead dress.
http://www.freewebs.com/valvewizard/

>>217613

My designs in the post above you, cost and parts above that.

No need for a choke with this design, unless you want a really expensive DC filter. Between the crowbar zeners and the 6.25 Volt regulator, it should be a seriously low noise supply with some decent caps. The choke in an amp, as far as I'm aware, is generaly to allow for Tube Rectifier's distaste for high ripple current.

>>217632
My most popular mod so far has been putting the bias mod and a 4H choke in 5150s. Tightens the low end and gets rid of a bit of fizz, and thats a SS rectifier

>>217635

Mhm, but the choke is part of the power supply surely? The tonal benefit probably came from some extra DC stability and maybe some hi-passing from the large inductive load seen leading to ground.

Either way the a larger reservoir cap and a smaller crowbar should essentialy do the same thing. Inductors in series and capacitors in shunt are largely interchangable when it comes to AC signal conditions. Try a 100uF cap. If you still want more bass/less noise, 220u. But you'll be messing wit dem V rails!

So, I don't really know much about music, but what is the difference between a tube-based amp, and a transistor-based amp? Don't we have excellent transistors today?

>>217649

This is where the rest of the electronics world snorts at our psuedo-romanticised science. But in short:

"We like Tubes because they suck".

They're prone to distorting, they've got terrible Gain to Bandwidth, the impedances are generaly shit and you have issues such as power sag compressing the signal aswell as the cost, maintenance and ball-achery.

However, that happens to suit a guitar. Alot. If you ever get a chance just plug an electric guitar straight into your soundcard or DI it into a mixer. It's not always very pleasant.

The bandpassy high-thd, lo-fi quality of the Tube Amp is what makes it what it is. Are their very high quality or revered solid-state amps? Sure, there's the Roland Jazz Chorus - an incredibly bright solid state amp. And they're really liked too. But it's a different type of amp for a different player. Tube Amps are liked by some for extremely heavy but pleasing distortion, some people like them for their ability to just break up on the harder notes-chords and stay clean when played softly. It adds a dynamic. For an electric guitar player, the amplifier is inherently a part of his or her instrument.

Some clever people are even coupling Tube stages to Solid State Class B and even Class D amplifiers for gain efficency. There's debate to be had here over Tube Power Amp distortion, transformer saturation etc. effecting the spectra and dynamic of the guitar.

Personaly, I don't see the need for more than a 30 watt amplifier - that's what a PA is for.

>>217654

Aha, so, it's like, even though tube amps have worse characteristics, that's exactly what makes the sound better?

>>217672

Exactly!

>>217640
I'm not sure what you mean. In the "pi network" just after the rectifier there is only the two filter caps and a high wattage 400ohm resistor. I replaced it with a 4H choke and it sounds substantially better.

>>217640
what do you mean by crowbar?

>>217654

the thing that makes tube design shine is the treatment of harmonics. linear circuits just don't fuck up and amplify the harmonic content like a tube does. The Trainwreck and Dumble amps are a good example of a pretty well known circuit that has been massaged to enhance the oddball harmonic content.

Amp noob here, is there a very simple and cheap amplifier that is easy to build? I want to get into electronics and my brother plays guitar so it'd be fun for both of us.
>Herg blerg thats what she said.

>>217694

By a crowbar I mean a Zener diode reverse biased so it conducts after a certain voltage.

So if I "crowbar" a ripply, 12 source at 9 volts the 3 volts of crud will be sent to ground - thus removing alot of the ripple. This is wasteful, but it works.

In mine you can see I crowbar off anything over 100 volts - the 300 down to the 320 line and then crowbar anything over 7.5 to the regulator. To make a correction to my drawing, the 7.5 line and 6.25 v reg should be the centre part and the 320 line the lower half.

Now you're probably thinking the voltage rails will be ripply, and they might be, but we can further filter at this stage with smaller (cheaper) caps if need be.

>>217705

If you want to get into electronics I would recommend starting with pedals as amplifiers are troublesome - I've built some boosters, a tremolo, ring modulator and I've modified amps and pedals but the sort of power we're talking here is not yet safe for your virgin fingers.

www.diystompboxes.com is a great resource. If you want to build him something, try the Super Hard On. Tasty little booster! Then move onto something more difficult like maybe a Green Ringer, an Octavia or a Luna Trem.

2ndly, post here. There's a few of us here that have electronics experience and are mega guitar fags. We'll be glad to help!

>>217687

I see, that makes sense.

Does it also have anything to do with noise? Some friends are telling me that tube amps are more noise-resistant than solid state amps.

>>217740

No, that's horseshit, sorry. Tell them I said that, too.

Like everything else in electronics, low noise in a guitar amp is down to:

Good shielding.
Good DC conditions
Good AC conditions
Low johnson noise
Sensible gain staging
Ensuring the device is both easily driven and can drive the intended loads easily. In a guitar amps case it's gonna receive a 100K or so load, fuck all current. Solution? High input impedance. Like, at least 1Meg Ohm. Fucking this up will be both noisy and skew the LRC characteristics of a guitars frequency response. A guitar pedals gotta be able to drive an amp, definitely, but surely it should plug into the Line In of a mixer just fine? People do this stuff, you've to make sure they can do it without life being a pain.

Same theory applies to microphones, radios, DC motors. They're all reactive and they need to see what they're designed for or the overall impedance changes. This can be creative..or messy.


Look up the use of things like Balanced Feedback, Bootstrapping, Constant Current sources. Douglas Self's Small Signal Design goes into exhaustive detail on low noise, linear amplification.

>>217740
Most solid state (guitar) amps are just plain shittier, so are probably "noisier".

Don't bother, looks like a mess of wires and shit that will probably short out and burn your family's house down while you sleep, biggest waste of components... you probably could have got an already built one from china for 1/3rd the price and superior quality then this mess of shit.

Don't bother, china makes it for cheaper and better quality.

>>217755
>>217778

Ah, thanks, that makes sense.