TGM Amplifier ?

[Bigun]

A little further research finds that the Vishay pots are not easy to find in 6-gang version. The 2-gang version are out-of-stock with a 1 month lead-time, so the only option would be single gang pots, six of them. They are held together with rivets, which will have to be ground off to dismantle them and then re-assembled with bolts.

I didn't design the layout of everything inside my chasis to allow a shaft to run from the front panel to the back, which is where you'd like to place the pot. So the pot would likely end up on the front panel with cables running from inputs to the pot and back to the amp boards (might not be so bad if I use individual shielded cables).

The digital potentiometer approach avoids these issues and for the same cost. But further research finds that it likely suffers from having no zero-crossing detector, so the volume changes will most likely cause 'zip' sounds on the output.

Nothing is as simple as it looks !

 

[currentflow]

At least with the digital control version you would have more options open to explore.

For example, using three PGA2310 dual-channel volume controls, it would be possibile to add a remote control and a display showing your exact gain settings. Of course, this requires a microcontroller...

Saying that, the Atmel microcontrollers are fairly cheap and free to develop. If you have a PC with a parallel port, the programmer consists of nothing more than a length of cable and a few resistors.

See similar projects at: Suzy Jackson and Mark Hennessy.

Regards,
Steve

 

[Bigun]

Those linked projects are really impressive, some very nice work. I would agree that the digital approach isn't that complex, the ic's are not that hard to use. But all these projects take up quite a bit of chasis space and are dedicated pre-amp units. I'm essentially building an integrated box where space is a bit limited. This box is destined to be dedicated to one source.

I've ordered the Vishay pots as a starting point. It will allow me to get something up and running and to start playing with the amplifier design. I have the option to upgrade the attenuator later

What I may find, assuming this thing gets finished and works, is that I can add a simple 3-way switch to the design to select from a) quiet, b) loud, c) adjustable. The first two positions select fixed resistor attenuators who's values I'll decide on after listening to the thing. The benefit is simply that it bypasses the pot and is very simple to implement.

 

[currentflow]

What I may find, assuming this thing gets finished and works, is that I can add a simple 3-way switch to the design to select from a) quiet, b) loud, c) adjustable. The first two positions select fixed resistor attenuators who's values I'll decide on after listening to the thing. The benefit is simply that it bypasses the pot and is very simple to implement.

Very often, the simplest solutions take a lot of beating

Steve

 

[MJL21193]

I can add a simple 3-way switch to the design to select from a) quiet, b) loud, c) adjustable.

There's something like this:

An externally hosted image should be here but it was not working when we last tested it.

$53 at Digikey. It's a 6 deck, 12 position rotary switch. Not much fine control but might be the ticket if you believe the sound quality is degraded by the pots.

 

[nigelwright7557]

There's something like this:

An externally hosted image should be here but it was not working when we last tested it.

$53 at Digikey. It's a 6 deck, 12 position rotary switch. Not much fine control but might be the ticket if you believe the sound quality is degraded by the pots.

If one pot degrades the signal then what does the other dozen or so resistors inside the amp/pre amp do ?

What about contact resistance in the multi way switch ?

 

[MJL21193]

Options, my man.

I would just use the pots already.

Listen to the music, not the equipment.

 

[Bigun]

Truth be told, one of the reasons I liked the digital approach was the idea of a continuous rotation mechanical encoder for a vol. control - you can just keep spinning it around !

Anyhow, pots for now.

I really must get these bits finished to establish a signal path and power source in order to start debugging the amplifier design itself so I have a finished system. I'm pretty confident that the amplifier design is OK since it's based on a proven topology, it also simulated pretty well. The risk may be more in optimizing some specific resistor values and my pcb layout which has not been used by anyone else before.

The prototype amp board I built in a hurry a couple of weeks back didn't use matched transistors - what are the consequences of this (simply degraded distortion characteristics) ?

 

[MJL21193]

It's nice to get the input pair as closely matched as possible. Current mirror T's should be well matched but you don't have those.
Outputs I've never bothered. I use .3 ohm emitter resistors though.

 

[Bigun]

If the LTP devices are mismatched what will I hear differently ? - are we talking a bit more H3 ?

 

[Bigun]

I promised Lumba not to give up on some enhancements so whilst I build the TGM1 it's perhaps worth some time to think about TGM2, and eventually TGM3.

Now that I have spice, I made a model of TGM2:

 

[Bigun]

and it gives some preliminary results only.... I had to fiddle with the resistors a fair bit to get it to work and it took me awhile to realize that I had to tell Spice to use an LED model instead of a generic diode or the voltage drop wasn't enough for Vbe and then no current through the LTP...

but this design will have to wait for my next project, a stereo only project. As MJL says, I may be hooked and a 'next project' would confirm that

but maybe I should look at symasym, OS and Carlos have some good things to say...

 

[Bigun]

Regarding the volume control.

It doesn't work, the resistances don't track well and at least one wiper is unreliable. I suspect the process of breaking them down and rebuilding them was too stressful.

So back to square one, with a lighter wallet

 

[Lumba Ogir]

Gareth,

I promised Lumba not to give up on some enhancements

You are amazing.
R29 does tremendous good for linearity.

 

[ostripper]

Now that I have spice, I made a model of TGM2:

TGM2 is lookin' alot like frugal 1B , right down to the CCS and bootstrap divider resistors.

Have you built the 1 yet? I already have 2 of what you posted for the 2 ... I think you will be pleased . I am going to do 2 more of the 2ish clones , but add 4700uf railcaps and the bridge right on board(amp /PS combined). Just hook up the 30-0-30VAC , 2 rear channels.

Besides R29 (I use 22r) , R28 retains the aska/dx sound while also improving PSRR. (some just hate current sources)

but maybe I should look at symasym

If anything , make your center channel a symasym , it really brings out the vocals..WOW.

 

[MJL21193]

Regarding the volume control.

It doesn't work, the resistances don't track well and at least one wiper is unreliable. I suspect the process of breaking them down and rebuilding them was too stressful.

So back to square one, with a lighter wallet

Nothing to lose (except postage) by sending them to me to see if I can fix em up.
I'm pretty handy...

 

[Bigun]

Have you built the 1 yet?

I have one prototype pcb, but it's never been powered up. The trouble is, I'm starting from scratch, have no kit so to speak. So I need to complete the chasis, power supply and signal routing. I'm pretty much done with the psu so I can start to put some wiring in place to start playing. But the signal input needs to come in via the volume control - still trying to find a solution for that.


If I like the TGM1 enough, it might be what gets used for the HT setup. Then I can relax a bit and use it as a reference for TGM2. I can take a bit more time with TGM2. I am hoping to make a stereo amp based on the TGM2 (or 3) where I can put a bit more into it than the HT system.

Whilst I get TGM1 working I want to explore the design of the TGM2/3. I'm ok with the idea that this might evolve into a TGM4 before I commit to building it - but I like 'simple' and am not necessarily looking for better 'technical' performance for it's own sake but rather a satisfying listening experience - which is why I started with AKSA. Your experience of the sound of the Frugal1B compared with other designs might help steer my plans.

p.s. I might couple TGM2/3/4 with a DIY speaker build. The idea is that I would upgrade the power of the amp design to 100W and look at making a clone of the PMC DB1+.

 

[Bigun]

Nothing to lose (except postage) by sending them to me to see if I can fix em up.
I'm pretty handy...

That's really kind of you ! - you have a way of making a problem feel smaller.

Well, I have to be honest, I think I botched it. I used a small pin-drop of superglue on the location pins to hold the stack together - I am betting that I contaminated the pots by doing this even though the location pins are well away from the moving bits. Dis-assembling what I've built may be tough. The thing is sitting in Acetone, if this doesn't melt the conductive plastic I may be be able to strip it down and try cleaning it, but I wouldn't be a betting man !

 

[ostripper]

Your experience of the sound of the Frugal1B compared with other designs might help steer my plans.

I must say that despite its simplicity , the 1b/aksa/dx (all clones of
the original rca app notes 1969) is the most reliable all around best sounding amp a DIY'er can attempt/do.
It has a very long term "livable" type of sound (never a "bad"sound) , and with careful compensation , a VERY good sound.
As opposed to the symasym , that is a hard comparison. The original sounded good with a small system , but lacked bass "slam" as they call it. The new monster symasym has nuclear bass , but it is another amp. Both have all the others beat as far as midrange definition and HF detail (brushes , cymbals, etc.) That is why i said symasym = center channel , very forward vocal imaging.
The abomination has the same sound , same general topology , the VSOP also has this sound.. but the symasym is much simpler.

 

[Bigun]

That's reassuring. I guess this is where it's important to consider the speakers as well as source material.

My current HT speakers are DIY full-range. Not the usual 'big and powerful' option that people like for their HT, but my interest started first with speaker DIY rather than HT. These full range speakers lack nothing in terms of HF detail, if anything they need to be tamed. Perhaps TGM1 is the best fit for them. Bass comes via a powered sub.

For my 'next project' where the TGM2/3 comes in, the speakers are not yet decided on, but I'm attracted to 3 options, 2-way TL, 2-way based on full range with woofer assist, and a full-range single driver in closed box with the amplifier response set up to achieve bass eq. And if I really want an excuse to build more amplifiers I'll bi-amp.

So, for now, it seems that TGM2/3 is on the right track.

What's the general opinion on the sound potential for TGM3 which uses FET-BJT input pairs ?