Amp safari continues

[HW_Hack]

So several weeks ago I started out planning to build a battery powered amp - low watts - and was looking at the Ruby and the Noisy Cricket. As my background is all digital circuits I wanted to learn more about the LM386/op-amps so I bread boarded up the basic Smokey amp to do some measurements - oh my how frustrating. I've got a little Tek digital scope borrowed from our high school which was donated to our school from a local university. Seems to work ok but I have at times battled noise - maybe the probes or a ground issue - unclear. And at times the noise is of the 60hz type -- argh. I'm spoiled as I used to work in a lab with $3000 power conditioners - $60,000 dollar scopes - and grounded work surfaces. Of course we were dealing with nanosecond measurements. But what I wouldn't give for an old Tek 465B ---- one great analog scope --- sigh .....

So at some point I decided to just build a circuit and trust my prototyping skills. I built a Noisy Cricket and added an input for an iPod. I used one of these for the first time and I have to give it an A+++ for durability and ease of use:

http://www.adafruit.com/products/1609

A very rugged proto board - no lifted pads here. This will be my default board for any serious project.

And on powering up I got pretty decent guitar sounds even through a crappy 3" speaker from an old CB rig. Cool !  I have not tested the iPod input but will in the coming week.

Along the way I've learned about tone stacks - that Fender / VOX / Marshall amps scoop out the middle of the tonal range - and a bit about op-amps.

So phase one is still a battery powered amp and I think I will use the Little Gem MkII as it has 2 LM386s in a bridged output (should add more poop to the output). And will slide in a full tone stack (Fender or Vox) and possible a FET stage with a slight overdrive via a pot. Will be powered by 6 AA batts or an optional 9V supply.

http://www.runoffgroove.com/littlegem.html

I've cut the pine board down to the pieces for the small cab I'm building for the 8" Jensen speaker I have (Jensen Mod-8/20). So I will start tacking that together in the next week. I'm confident this will all work out so I'm going to sell the 30watt Marshall I have as its just too big for causal playing - practice. Plus I need those bucks to finance round 2.

[Roly]

With these smaller amps in particular having a large (12-inch) and efficient (95-100dB/W) speaker in a reasonable enclosure (1.5-2cu ft) can make a huge difference.  It is just amazing what a small amp can do when given a good speaker.   :dbtu:


runoffgrove is one of my less favorite circuit sites.  The Little Gem being another example of less than impressive design.  Attenuating the output with a 25 ohm pot is seriously primitive.   :loco


ed: typo

[HW_Hack]

Yeah - I wasn't really looking at the Little Gem but the Little Gem MKII with the bridged output of two LM386s driving about a watt. But I would say 25ohms is a bit weird for a volume pot ...

So the circuit thats forming in my head is the basic Ruby amp which gives volume and gain - feeding into a tone stack - which feeds to the two bridged LM386s. Easy enough to solder up for a test ride on a perf board.

You mention 1.5 to 2cu ft for a 12 inch speaker --- is that closed or open cab ? With my 8 inch speaker my first cab shot will be a 0.6cu ft open back cab. Are these ratios listed anywhere ?

Thanks for your comments.

[Roly]

What happens with a sealed enclosure is that as the cubic capacity goes down the resonant frequency of the driver goes up, and the driver has no real output below this frequency.

In the case of an open back enclosure the bass cutoff is determined by the shortest path between the front of the cone and the rear, the roll off starting when that distance is a half wavelength.

There is a table about half way down this page;

http://www.ozvalveamps.org/cabinets.htm

...that shows how driver resonant frequency changes with diameter and sealed cabinet cubic capacity.  You can scale these values for the free air resonance of the driver you select.

I would suggest that half a cubic foot is too small for any speaker to work reasonably.  With only a small amount of power to play with you really need to give the amp a break by using the best speaker arrangement you can, ample cone area, good efficiency (dB/W), and a large cab.

[HW_Hack]

Progress continues slowly --- here are some pics of the simple cab I'm building. Found a decent piece of tight knot pine 1" x 8" by 6ft. Used all but one inch of it - the speaker mount board is 5mm plywood. Standard beer bottle for scale. The area above the speaker will house the amp module which by design is swappable --- slide in - slide out. Speaker is a Jensen MOD 8".

[Roly]

Nice job.   :dbtu:

The baffle (the bit the speaker mounts on) looks a bit thin; I'd give some thought to adding some bracing to stop it flexing like a drum skin.  Next build go for something a lot thicker, more rigid.

What is the distance from the edge of the speaker cutout around the side and back to the same point?

[HW_Hack]

Hi Roly

Thanks for your continued interest in this project. So if I measure from edge of the speaker cutout inside the cabinet around to the outside same point I get 21 inches. As to the baffle rigidity I think it will be ok - if the baffle was bigger flexing could be an issue. This is a cross fiber plywood so its fairly stiff.

[HW_Hack]

Well in the past few days I've finished the amp cabinet. Not totally crazy about the final finish on the wood but i just used some tinted polyurethane we had around - trying to save some money. I plugged in the LM386 amp - Frankenstein Amp - that I slapped together a couple of weeks ago ... and decent guitar sounds came out of it. It drives the 8" Jensen fairly well. The idea here was an amp cab with an amp bay where I could slide in different modules. So next up is a battery powered amp with an actual tone stack and about 1watt of power. Then an AC powered amp with tone stack and 5 to 10watts.

The journey continues as I finalize the amp design by slicing and dicing other designs. The amazing news is that Radio Shack still carries TL082 op-amps

[Roly]

Lookin' good.   :dbtu:

The first front-back cancellation occurs when the shortest air path between the front of the cone and the rear of the cone is a half-wavelength.  So the question is, at what frequency does a 21-inch path cancel?

Speed of sound, C;
1024 ft/sec (roughly)
12*1024 = 12288 inch/sec

wavelength = 2 * 21 = 42 inches

f = 12288/42 = 293Hz

http://www.ozvalveamps.org/notefreq.htm

This is the D above Middle-C, so most of your guitar strings are below this frequency, and the rapid rolloff below this means that it will be a pretty "toppy" cab.

One way you could extend the bottom end a bit would be a partial back, say two vertical panels each about 1/3rd of the width, leaving a vertical slot around 1/3rd the width.  This would still effectively be "open back" but would make the front-back path around the sides longer (the bottom is blocked by the floor, and the top path is already longer because it has to go over the amp - provided the amp actually blocks the top cavity).

You may also care to experiment with closing the back off altogether (e.g. back it hard up against a wall); this will push the basic resonant frequency of the driver up a bit, but you may gain more bottom end overall.

Note that no amount of EQ or brute force will get you around this basic physical limitation of air, it will only make the speaker "frap".  It's a bit like trying to drive on a slippery surface, horsepower doesn't help if the speaker can't get a "grip" on the air to move it.

However you are discovering that a very small amp like a 1 watt LM382 can produce a reasonable sound as long as it has a half-way reasonable speaker to work into.  Most commercial small amps are crippled by their tiny speakers in tiny boxes.

When it comes to speaker cabs there is an old saying; "There ain't no substitute for cubic feet" (inside the cab).

HTH