Little Gem MKII not working on the BB

[duck_arse]

If true polarity bothers you and the pre-amp is inverting, you only have to swap the wires for the speaker jack to put it right.

now jim, you know that's not fair.

[anotherjim]

Stephen, I claim special exemption for bridge amps. Since both ends of the speaker are live outputs. Honestly, I'd never suggest that with single end amps, even though you'd get away with it. Ground on tip is never right!

[Ben Lyman]

You guys are talking over my head but I'll figure it out someday.
Anyway, for now I like the idea losing R11, tossing the buffer, changing vol pot back to B100k, so it's just ->booster->tone->amp.

About power filter for the booster, would that be something like this? Imagine this schematic without the buffer on it:

[Ben Lyman]

This might be a winner. I noticed the need for the buffer seems to somehow be related to the B10K volume pot. When I changed that back to 100K, no buffer needed. Does that make any sense?
I also changed the values of the tone stack to get as close as I could to the BMP and it sounds really good now. The mid scoop in the middle seems to be trying to get close in sound to a blackface Fender maybe.

Let's say I wanted a mid-boost toggle switch, is that even possible?

Still not sure if I understand the power filter, here's what is on the BB now:



I like how simple it has become, thanks for all the help Jim

[anotherjim]

Getting there.
That's the power filtering exactly.
This is now the classic basic guitar amp layout. Preamp - tone stack - power amp.

Off the top of my head, you might use a switch to modify the tone by adding a bypass from C10 to the wiper of the tone pot. Switch in some resistance & reduce the mid scoop. The tone pot will still work to an extent, changed depending on the bypass resistor value. Put a cap in series with that resistance, or a cap only, and the bypass can add a treble boost. I'd call it a boost switch. A centre off switch could do 2 different bypasses.

True polarity - when the signal out of the guitar swings positive, the speaker should push forward. Most amps keep this standard. Use a Rangemaster for example and the speaker is reverse - because the RM is inverting. Nobody minds.
The pre-amp here is inverting, so it isn't true polarity.

[Ben Lyman]

...when the signal out of the guitar swings positive, the speaker should push forward....
The pre-amp here is inverting, so it isn't true polarity.

Ahh, this I understand. Very cool info, thanks.

Here's a band pass that I tried on the BB, big mid/volume boost. Am I over thinking this? Seems like there should be an easier way.
EDIT: just changed R2 & C1 in the band pass to 10K and 10uF


It is getting there. Sounds good. One weird thing that might just be BB layout and maybe no big deal; but if I share the One-Spot with a pedal (bypassed or not) AND turn the BMP tone below 12noon, mass oscillations ensue. Not gonna worry about it until I have soldered it all up and discover it continues... then I'll just not share my One-Spot anymore.

Now all I need to test out with it is how it handles a power switch, LED indicator and reverse polarity protection, none of which currently reside on the BB. Any thoughts? Try a a 1n5817 in series with +9v? or just a 1n4001 from +9v to ground? Something else?

[anotherjim]

The IN5817 in series should be ok. I don't think you'd lose much with a series 1N400x type either. Me, I don't like shunt protection diodes unless I can put them after an RC filter, so the resistor in that limits reverse fault current. You can't have significant resistance in the chip-amp supply, so series protection diode makes more sense.

Are you sure you have 10k & 10u in that filter? That passes all audio. It should just slam off all your signal. Do you mean 10nF?

You're finding out that amps should have their own power supply.
RC filters can only do so much. They won't completely iron out very low frequency supply ripple, such as sag when you whack a big chord. Turning the tone to the bass side showed that up.

When the supply voltage regulator some distance from the load, any resistance in the round trip to the load and back (wire, plug contacts), can prevent the regulator from doing its job properly. It's partially blind  - it doesn't know what voltage the load is actually getting - it only knows what it's sending out from its end. Additionally, devices sharing the same supply must suffer any voltage sags caused by any other device that's current hungry.

Best thing is to have the power supply in the amp. Get the shortest route possible between it and the load.
The small chip amplifiers don't require huge power, but it is more, and more erratic in its nature; than most fx pedals impose. For these amps, a normal power brick supply should be fine, so long as it only powers the amp.

[Ben Lyman]

Thanks Jim, I'll have a closer look at that band pass today when I try a series diode for protection.

I was looking at the BMP Analysis:
http://www.electrosmash.com/images/tech/bmp/bmp-tone-control.png
And I was trying to pass everything below the (top side of the) mid scoop, around 1.6KHz,
then shave off everything below the scoop, from about 725Hz and down.

I made all that up as I went along 
How can I pass only the lost mids through the tone stack?

Can I buy an internal power supply from an electronics supplier or is that something you guys usually make yourself?

[anotherjim]

You've got that right. The BMP is HP & LP in parallel. The area where they overlap causes the scoop. It can be set without overlap so no scoop. HP & LP in series does the opposite - take out above the mid with the LP then then below the mid with the HP.

You might like some tools to find the values. A single pole filter is easy to do by guess and listen. A complex filter has too many variables - you need calculation to find the right sort of values, then confirm with listening.

First, there's Duncan tone stack calculator. That's a free program that needs installing. It has a BMP tone emulation (you can try different values and see on a graph what they do. Most other amp tone stacks are in there too.

Then bookmark this...
http://sim.okawa-denshi.jp/en/Fkeisan.htm
In the filter tools there are simple RC LP & HP. You type in values and get the corner frequency (and a lot of other techno gumbo). It understands 10u, 1M, 47k, 100p etc shorthand, but you must use decimal points for fractional values, so for 2.2n, it doesn't like 2n2 -  it will swear at you in Japanese if you try.

Keep R values >10k <100k if you can. You should rarely have to use a C over 100nF in tone networks.

[Ben Lyman]

Thanks again Jim, I like the sound of the values in it right now so I'm going to go ahead and build it like this. I can always make another one if I'm not happy with it; the components, wires and perf are only about a buck anyway, haha.

I bookmarked the link, it looks like it could be very useful, thanks.
There is a site called "learning about electronics" and I used their calculator to come up with those values; they have a bunch of calculators too.

I made a layout, components are going on the perf and should be soldered up by tonight for a trial run before I "temporarily" install it in a recycled 1590BB that's already been many different pedals before this. Ultimately, the layout should easily fit in a 1590B with power switch, LED light, Mid switch, Vol and tone knobs all on top; guitar in, DC, and speaker jacks all on the sides.

Edit:
Jim, you're right about the 10k and 10u. I don't know what I was doing that made that somehow work on the BB, maybe I had a dodgy ground connection or something. Anyway, the perf is all soldered and I had to cut those two out of the mid boost section and now it's all working.

I'm not sure what it makes my mid boost switch now, just a 1n and 220k passing the tone network but it sounds good. Next time I'll try a 2n2 and 100k in the mid boost. I need to study up on the band pass thing a little more because I really just wanted to accentuate that one mid zone, which is what I thought I was doing 

[anotherjim]

I think your bypass is treble boost -  if it starts low enough, it will reduce the mid notch too.

You may like this if you can run to another pot...
http://www.muzique.com/lab/tone3.htm

[Ben Lyman]

Thanks Jim, ya, treble boost or treble bleed I guess. It seems to bump the mids well enough to make my amp a sort of tweed/blackface... or blackface/Marshall... or something 

Here's a new schematic even though I have assembled it already. Notice the 1N4001 shunt... sorry folks  the series diode made the amp distort too much, not in a bad way, just too much.


And here's a vid of it in my 10x repurposed 1590BB.
Strat -> TDA2822M -> 2x12" Eminence Patriots.
Then w/Rat at 1/2 gain.

[anotherjim]

Very cool Ben. You should be proud of that. It sounds a lot bigger than a Champ.

[Ben Lyman]

Thanks for all your help Jim, much appreciated.
I think I might have figured out what went wrong with my attempted band pass filter, I'll be experimenting some more with it when I can.

I made a new layout that is narrow enough for a 1590A but I'm not sure about squeezing in all the pots, switches and jacks.

I almost forgot to make a vid of the Little Gem MKII, here it is, same set up with strat, 2x12" and then with the Rat. I think the LG MKII could benefit from the booster/tone combo, it's just a little quieter than our TDA2822 creation

[duck_arse]

it's always fun listening to your demo's ben, even when they are only running at "half-rat power".

and the "I" setting makes me wonder - if the idea of a bridged amp is to feed the same signal only bigger to the speaker, what does the speaker think/do when it gets clean on one terminal and half-filth on the other? sum? divide? average? something else?

[Ben Lyman]

and the "I" setting makes me wonder - if the idea of a bridged amp is to feed the same signal only bigger to the speaker, what does the speaker think/do when it gets clean on one terminal and half-filth on the other? sum? divide? average? something else?

Good question Duck. I did ask about it earlier on in this thread before committing to it. I think the answer was that there shouldn't be any problems and so far it seems fine.
I don't know what it's doing, maybe a subtle tubescreamerish thing with a little bit of clean signal, barely audible and mixed with the gritty signal?

[anotherjim]

The speaker only moves the cone in response to the difference in volts across its voice-coil terminals. As long as there's a difference, it moves.
Having 2 different signals across the coil gives an average mix. It might well sound like TS distortion when one is distorted and the other clean. It's a good trick really. The added harmonics from the distorted amp won't get the near x4 power advantage of a bridge amp, since it's only single ended.

Then I think you may be worrying about the "clean" amp's health - it's output is being dragged about by its evil twin with only the speaker impedance in the way. It has the same power handling, it has the same output impedance. It can handle it.
Really, an amp is a variable power supply. When you use an op-amp to buffer a voltage divider for a stronger Vref, it's the same kind of thing. In fact you can use a chip amp as a high power Vref buffer. One amp can be silent (so long as it's output is sitting close the working amps Vref), or doing it's own thing.

[Elijah-Baley]

It seems kind of I missed this thread. Nice works, Ben Lyman!

I'd like to ask you something. I think you changed something in the last version of your Little Gem amp. Did you post the final schematic?
And... if you had the layout to show, will be great!

Thank you.