Recommendation for 5-10W amp.

Hi,

Been looking around to build a nice sounding 5W amp, I have the ACA which I truly enjoy so this leads to for wanting more of Mr Pass design.

I bought some components to built the F5 but this will be later, a fall project. I have a whole bunch of NOS TO3 lateral mosfet that I'd like to use soon.

Looking for some recommendation, I heard that sound wise the Aleph 3 sounds like the ACA, makes sense ? I'd like to avoid building an amp that sounds extremely similar to the ACA which I already own, I'd like to taste a 'different wine' while building a Pass design.

I'm now looking into the Zen but a low power version (+24V-Gnd) to get about 5W. Any other recommendation are welcome.

My speakers are MLTL with FE206E and my music taste is mainly Blues, Jazz and Classic.

I own a few low power tube amps, all SE except for one.

LATEST INFO : Juma came up with a design, please see post no. 217.

Thanks for your time.
Eric
 
Last edited:
The one and only
Joined 2001
Paid Member
Laterals make good versions of the F5, I have several laying around.

Here is a simple one. You will note that it looks very much like a stripped
down version of the Selectronics product.

:cool:
 

Attachments

  • lat f5.gif
    lat f5.gif
    22 KB · Views: 3,196

fab

Member
Joined 2004
Paid Member
Hi,

Thank you all for the suggestion.

I'm looking for a compact 5W Class A amp., is it possible to use an existing Pass design and reduce the rail voltage while maintaining the overall great sound ?

Thanks,
Eric

For the one I have pointed out, You can use a regular F5 type pcb , cut the pcb traces for the on-board MOSFET drains to the output and run wires to your to-3 lateral mosfets and that about it. You can use +/- 12 vdc for about 6 wrms/8 ohms.

Fab
 
So, it's a single rail 24V power supply, clean 5W cl. A amp that you want - Laterals at the output, different sounding than ACA, right ?

This is what I like (sch. in attachment) - one more F5 modification. Input stage is rearranged to avoid unobtainum small Toshibas (all who tried it reported that it doesn't sound worse than k170/j74, some even like it better). Of course, k170/j74 can be used but in this case I'd recommend GR variant (Idss 3-5mA) since LatFETs need less current drive and have lower Vgs. In that case all additional biasing shebang is not needed (R4/R6, C1/C2).

When I made it, I had reasonably matched transistors and it worked lovely as it's drawn - 12v at the node where drains meet, Id was about 0.7A which is quite all right since with no source resistors we can enjoy the extended class A regime (thanks to FETs' square law transfer curve). If you want to experiment with Id value, use 1k or 2k trimpots instead of R3/R5.

Current through BJTs is set at about 2.5mA (it depends on R4/R6 value) which I chose as a compromise between the need to have enough drive for decent bandwith (>=100kHz) and Ib low enough to provide for decent Zin (50k is fine, good 2u2 coupling caps are plentiful and inexpensive).

Edit: I used k1058/j162 at the output but they are just k135/k50 in flat package.
And yes, C2 polarity should be reversed (drawing error).
 

Attachments

  • F5-BJT+LAT-SR24.gif
    F5-BJT+LAT-SR24.gif
    9.4 KB · Views: 3,118
Last edited:
This description fits more the one Nelson shown in post 3.

The last one from Juma should also be a good alternative although I would use double polarity supply to get rid of the big output cap...

Fab

Good idea but I have many single rail +24V supply that I want to put to use so I'll stick with single supply, plus I like the added protection of the output cap with my sensitive full range.

Thanks,
Eric
 
So, it's a single rail 24V power supply, clean 5W cl. A amp that you want - Laterals at the output, different sounding than ACA, right ?

This is what I like (sch. in attachment) - one more F5 modification.


And yes, C2 polarity should be reversed (drawing error).


Seem to me that C2 polarity is right since your circuit is a single supply type.

Indeed, to me it looks like a native dual power supply circuit and arranged pretty bad to work with a single power supply.

It seems that your circuit is not balanced at all in the actual operation, being that Q1 working as a CE amplifier while Q2 works as a CC amplifier. For which the two drive signals of the power mosfet are quite different.

Then, what is the purpose of using the network formed by R12, R13, C4, D1?
 
Seem to me that C2 polarity is right since your circuit is a single supply type.
Yes, you are right on that one, my mistake - I posted in a hurry...

Indeed, to me it looks like a native dual power supply circuit and arranged pretty bad to work with a single power supply.
If you put your mind into it, it can look like an ice cream, or whatever you want it to look like...

It seems that your circuit is not balanced at all in the actual operation, being that Q1 working as a CE amplifier while Q2 works as a CC amplifier. For which the two drive signals of the power mosfet are quite different.
Look a little harder - both Q1 & Q2 are correctly polarized and are driving the appropriate LatFET from the collector and receive the feedback to emitter, so both work in Common Emitter mode.


Then, what is the purpose of using the network formed by R12, R13, C4, D1?
I'm afraid that not understanding that one disqualifies you as a critic, but here it is:
R12/R13 create a voltage divider which central point is held stiff at 12V (half PS voltage) with C4 which filters the residual ripple from the rail and that is the voltage set through R14 to the bases of Q1/Q2 and through the feedback network to the drains of LatFETs. R14 predominantly determines the Zin of the amp (since the base currents of Q1/Q2 are very low) and D1 takes care of C4's discharge.

Build a quick prototype and be amazed how nice it works...
 
Last edited:
So, it's a single rail 24V power supply, clean 5W cl. A amp that you want - Laterals at the output, different sounding than ACA, right ?

Hi Juma,

Since this is your design, how did you built it..via perforated board, PCB or p2p. Just wondering if a PCB already exist for your design.

Is there an existing PCB for another project that has similar amplifier topology.

Thanks,
Eric
 
Here is the PCB I made for this amp - 46 x 57mm, copper side view.
It has provisions for bandwidth limiting caps in the feedback network and input LP filter cap (both in case of oscillations which didn't happen).
Output cap (10mF) is made out of 2 x 4700uF caps connected in parallel.
Input cap discharge resistor (220k) is placed on the input RCA connectors and the output cap discharge resistor (1k) is soldered on speaker terminals.
 

Attachments

  • F5SR-BJT+LAT-PCB.gif
    F5SR-BJT+LAT-PCB.gif
    22.7 KB · Views: 1,262
Hi Juma,

On your schematic (post no.8), R14 (47k) is connected on the left side of R18 (junction of C5).

On your PCB layout, I believe R14 is connected on the right side of R18.

Should R14 be connected to the left or right side of R18, does it make a difference ?

Thank you for your help.

Eric