[akiroz]

4 channel balanced Beta22:
relay stepped attenuator (64 steps @ 1dB) controlled by an ATTINY26 MCU, used a linear pot. with the internal ADC as control.
Some labeling work left to do, I'm thinking of using Dymo tape.
There IS a bit of hum when using with sensitive IEMs, I might add a bit of shielding around the transformer, see if that helps.
Used a Neutrik PowerCON instead of the standard IEC connector because I don't have a saw to make rectangular holes

 
Input options:
2x XLR (Balanced)
2x TRS (Balanced)
2x TS (SE)
 
Output options:
2x XLR (Balanced)
4-pin XLR (Balanced)
4x Banana plug (Balanced / Speaker)
TRS (SE)
 
Selectable Gain: 2x, 5x, 16x
Selectable output: Headphone / Speaker
 

 

 

 

[FraGGleR]

Quote:

4 channel balanced Beta22 w/ relay controlled stepped attenuator (64 steps @ 1dB)
Some labeling work left to do, I'm thinking of using Dymo tape.
There IS a bit of hum when using with sensitive IEMs, I might add a bit of shielding around the transformer, see if that helps.
Used a Neutrik PowerCON instead of the standard IEC connector because I don't have a saw to make rectangular holes

 
Input options:
2x XLR (Balanced)
2x TRS (Balanced)
2x TS (SE)
 
Output options:
2x XLR (Balanced)
4-pin XLR (Balanced)
4x Banana plug (Balanced / Speaker)
TRS (SE)
 
Selectable Gain: 2x, 5x, 16x
Selectable output: Headphone / Speaker
 
 
 
 
 
 

Nice!  Does the hum not show up with other, less sensitive headphones?  I want to do a single case 2channel, but worry about hum.  I won't be using with IEMs, though.  What size case is that?

 

[akiroz]

Quote:

Nice!  Does the hum not show up with other, less sensitive headphones?  I want to do a single case 2channel, but worry about hum.  I won't be using with IEMs, though.  What size case is that?

Thanks.
There's no hum with full-sized headphones on low (2x) gain and a little bit on mid (5x) gain.
It should be fine as long as there's a bit of room between the power / signal modules and shielded cables are used for the signal.
That's a standard 3U 19" rack-mount case.

 

[joeyjojo]

Good job! A couple of questions.
 
I can only see two amp boards so I assume the other two are underneath them in the same footprint? If so, are they running appreciably hotter?
 
You say you have multiple gains. How have you implemented that? Is it simply a choice of fixed pre-attenuator?
 
And a comment - you're very unlikely to be able to get rid of the hum by shielding the transformer. You can try buying a transformer with layers of shielding built in (if that one doesn't have it) and placing it inside a mu metal box, but success is rare.

 

[akiroz]

Quote:

Good job! A couple of questions.
 
I can only see two amp boards so I assume the other two are underneath them in the same footprint? If so, are they running appreciably hotter?
 
You say you have multiple gains. How have you implemented that? Is it simply a choice of fixed pre-attenuator?

Yep, there are 2 beta22s underneath.
Well none of the beta22s are actually running hot, I used 120mA at the output stage and ± 27V supplies, so that's only 6.5W dissipated in the 4 heat sinks of each beta22 board.
but it may be a problem if I'm driving speakers at high power, haven't tested that yet.
 
The gain is changed by putting resistors in parallel with R3 in the feedback loop using a rotary switch.

 

[joeyjojo]

That's the naughty way to do it, as explained in this thread (I know that's for the mini3 but the R3 resistor serves a similar purpose in the B22). I would also be a bit concerned about C2-C5 being the wrong value as they were "carefully tuned".
 
You'll find that the current jumps up with speakers (far exceeding the quiescent current) so be careful those MOSFETs don't overheat.
 
I know you said you put this amp together because you got it for cheap, but I'm still a bit sad that you've assembled one of the finest amps in the world and done a shoddy job of the soldering (from zooming in on the photos) and made several of the classic errors (one box without shielding on the trafo, offboard gain switch, inadequate ventilation). Each to his own though I suppose.

 

[akiroz]

Quote:

That's the naughty way to do it, as explained in this thread (I know that's for the mini3 but the R3 resistor serves a similar purpose in the B22). I would also be a bit concerned about C2-C5 being the wrong value as they were "carefully tuned".
 
You'll find that the current jumps up with speakers (far exceeding the quiescent current) so be careful those MOSFETs don't overheat.
 
I know you said you put this amp together because you got it for cheap, but I'm still a bit sad that you've assembled one of the finest amps in the world and done a shoddy job of the soldering (from zooming in on the photos) and made several of the classic errors (one box without shielding on the trafo, offboard gain switch, inadequate ventilation). Each to his own though I suppose.

Haha, I guess I'd take some measurements when I have time.
My scope can do FFT analysis.
 
soldering? which part?
I thought it was pretty good.....

 
I plan on adding shielding because of the hum with IEMs.
 
as for the gain switch..... well I have one resistor on-board that will never be disconnected from the loop and wires going off-board are shielded, should be fine.
 
I guess the main reason I'm not putting much effort into this is because I've already retired as an audiophile...
I can no longer hear a significant difference between amps / DACs or even headphones because tuning sound with DSPs just overpowers everything else.
and as a person into electronic engineering, I just can't appreciate this design (and any designs with "snake oil" for that matter).
Pretty sad right?

 

[joeyjojo]

Sorry about the above post, it's a bit grumpy. Crappy day at work

 
If it sounds good and isn't oscillating then you're probably safe. The website says those parts are tuned to optimise the "speed/bandwidth", which are two things the B22 has in huge excess, so a little less won't be noticeable.
 
I was just commenting on most of the parts being put in a bit "wonky" or not flat against the PCB. Just aesthetically not as nice. Most B22's are done as a "labour of love" so it's odd seeing a slightly ramshackle one

but if it sounds good that's the main thing!

 

[Beefy]

as for the gain switch..... well I have one resistor on-board that will never be disconnected from the loop and wires going off-board are shielded, should be fine.

 
Following in the footsteps of Trevor Networks, that's the spirit!
 

Quote:

but if it sounds good that's the main thing!

 
You probably can't hear high frequency oscillation, but that doesn't mean it isn't there or won't cause problems. In either case, off-board gain resistors are a bad solution to a problem that shouldn't exist.

 

[nikongod]

Quote:

Perfect! There should be more than enough stray magnetic fields from that huge transformer to bring the noise floor of the amp way up into the audible range with IEMs.
 
Just because you can do something does not prove that it isn't a terrible idea.

 
QFT

 

[Nebby]

Quote:

QFT

 
But...Beefy was told to "Learn the maths!"
 
 

Learn the maths!
Bigger transformer ≠ more magnetic field.
b-field strengths are dependent only on:

• current flowing through the wire
• number of turns
• permeability of the core material

 
And TBH, toroidal transformers provide some degree of self-shielding in the first place, the electromagnetic radiation coming from the transformer is probably very small compared to other current-carrying wires in the system.
 
With that said, I cannot empirically prove that what I just said is correct because I don't have the necessary equipment to do so. If you just happen to have a EM radiation meter laying around then by all means, take some measurements of toroidal transformers when they're operating and see if they will cause EMI in amplifiers. 

 

 
 
Honestly that extra large transformer is causing excess heat, unnecessary hum, and possibly higher than spec'd secondary voltages due to low load. Anyways, I do have to wonder why the choice to go "all out" with a 300VA toroid only to settle for a 27V rail voltage?

 

[akiroz]

Quote:

Honestly that extra large transformer is causing excess heat, unnecessary hum, and possibly higher than spec'd secondary voltages due to low load. Anyways, I do have to wonder why the choice to go "all out" with a 300VA toroid only to settle for a 27V rail voltage?

Free. simple right?
It's been sitting around my house for quite a while doing nothing.

 

[akiroz]

Quote:

You probably can't hear high frequency oscillation, but that doesn't mean it isn't there or won't cause problems. In either case, off-board gain resistors are a bad solution to a problem that shouldn't exist.

lol did you have some bad experience with it?
 
personally I find it better to have at least 2 gains because I have IEMs and speakers plugged into the same unit....

 

[joeyjojo]

Quote:

lol did you have some bad experience with it?
 
personally I find it better to have at least 2 gains because I have IEMs and speakers plugged into the same unit....

 
A far better solution is a normal fixed gain (say, 8x for the speakers) and an attenuator (just a pair of resistors as a voltage divider for each channel) to knock it down by 20 dB or whatever before the amp boards. Assuming your attenuator doesn't go low enough already.

 

[Beefy]

Quote:

lol did you have some bad experience with it?
 
personally I find it better to have at least 2 gains because I have IEMs and speakers plugged into the same unit....

 
I'm not going to engage you any further. It is pretty clear from your attitude, and your post-count-per-day, that you are more interested in talking than listening to good advice.
 
 

 
A far better solution is a normal fixed gain (say, 8x for the speakers) and an attenuator (just a pair of resistors as a voltage divider for each channel) to knock it down by 20 dB or whatever before the amp boards. Assuming your attenuator doesn't go low enough already.

 


 
No, it is better to build the correct tool for the job. A smaller, more sensible amp for IEMs, along with a more sensible amp for speakers.