The Mini-A

Member
Joined 2004
Paid Member
After resoldering the zener and rechecking the circuit I fired it up again and: Success, nothing exploded. After some time of poking around in it's belly with the DMM's probes I connected my test speakers as a load.
The amp has a visible turn-on thump and an even more visible turn-off thump (visible=the cones move). I'm not sure yet if my FE108EZ will like that...I suppose shorting the outputs to ground while turning on the amp will be a workaround (for the lazy).

After feeding my Mini-Aleph with Pink Floyd, Nas, Daft Punk, Zeppelin, Hendrix and Ray Charles for almost an hour the heatsink is quite hot but I can still touch it for like 20 seconds.

I also pulled some numbers from the circuit (which look good methinks):

D1: 9.10V L, 9.19V R
R3: 4.57V L, 4.57V R
R4: 4.78V L, 4.92V R
Q4: 4.42V L, 4.40V R from collector to emitter
4.39V across each output-FET (IRF644 with same lot-number)

Differential-pair and current-source, measured from drain to source:

L: 4.00V and 3.941V
4.29 for the current-source

R: 3.983V and 3.929V
4.29 for the current-source

This difference must be due to the higher operating temperature compared to when I matched them.

DC-offset at the output after one hour is 42mV L and 43mV R.
The supply voltage is +12.66V and -12.70V.

A big shout out to Nelson and Grey for the Aleph/Mini-Aleph and also many thanks to everyone around troubleshooting, posting and building nice and inspiring amps.

Here's a pic from the test-setup:
An externally hosted image should be here but it was not working when we last tested it.


More pics will follow.
 
Working....

Good to see / read that the issue is resolved.

Now is the time to fine tune.

Did you use a cap near the input CSS?
What value? experiment with a few values to see if
the thumb issue gets tammed? Start with 1uF up to 10uF.
(use a test speaker first)

The circuit is easily tweaked for optimum performance by
accuratly selecting the AC-gain resistor. As in every ALEPH.

Search DIYAUDIO and you will find a lot of good stuff.

Other small tip: The photo shows a little different circuit than
the schematic. For instance: it looks like you have used protection
zeners on the input. Maybe you could show the actual schematic.
 
Member
Joined 2004
Paid Member
Zen Mod said:

:cheers:

rtirion said:
Maybe you could show the actual schematic.

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


For numbers (R3, R4, D1...) please refer to Grey's schematic (the zip file).

MEGA-amp said:
Nice P2P, i'm envious!

Thank you. I had this project in mind ever since I bought the heatsink from ebay. Actually Grey sparked the idea with his "someone build it inside a pepsi can" remark...

You might also like the PSU...
An externally hosted image should be here but it was not working when we last tested it.
 
Member
Joined 2004
Paid Member
Mini-A oscillating

Hi, me again.

While trying to measure the maximum voltage-swing my Mini Aleph can do I discovered something strange.

The setup:

Sony Vaio laptop LAN-port -> ethernet RJ45 cable -> linksys WMB45G music bridge -> coax cable -> CD-player's digital input
CD-Player's analogue output -> shielded cable -> Mini-A

The music bridge is used to stream digital data from the laptop to the CD-player.
When using the digital input of the CD-player, it's output can be adjusted via a digital volume control. Output impedance unknown.


Nothing weird so far. When I connect ONLY ONE cable to to Mini's input (other one open) and play some test tones via Media Monkey nothing unusual happens.
But when BOTH cables are connected the amp starts to oscillate violently, regardless of input-signal. Bad pic of right channel, left looks the same.

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


The scope is not awesomely accurate but I'd trust it up to a few percent. The oscillation is not audible through the speakers (my 6 Ohm Sony 3-way test-speakers).

x-axis is 1µs/div, y-axis 10V/div. No offset.

The heatsink is earthed, the inputs and outputs are mounted on plexiglas. Circuit's ground meets PSU's ground and is earthed via 10 Ohm resistor. CD-Player is not earth-grounded (2-pin plug), same as music bridge. Amp, CD-Player, music bridge and scope share the same multiple socket, laptop on battery-power.

Does anyone have a clue what's going on and/or how to cure this?
 
peranders said:

Yeah, peranders.
Some of your amplifiers / power supplies surely can be called X-treme.
And this is also what you call them your self ;)
------------------

I was taking a second look at Nelson Pass A75.
Using 12 IRF MOSFET pairs ( = 24 devices ) for output is pretty extreme.
The A75 Power Amplifier - Part 1
by Norman Thagard and Nelson Pass
Schematic here: http://www.passdiy.com/pdf/A75p1.pdf
 
Member
Joined 2004
Paid Member
Nelson Pass said:
Wow! Very nice.

Thank you! It's a pleasure to build your designs. Lots of zazen, even though this one is an Aleph! :D


lineup said:


Here another image, form the Rodeo Dave album.
The guy is really good in practical build

Extreme, isn't it .. like some Nelson stuff is, too ;)

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

:D

peranders said:


Hey, you know what, that's exactly what inspired me to parallel so many caps. At first I wasn't sure about leakage current and parasitic whatever, but the I saw these boards and went for it. Mine won't (safely) deliver 180A though...

OT: Maybe you consider this cruel, but have you heard of plantochi (.com)?
 
Member
Joined 2004
Paid Member
Mmmkay, the weird oscillation is gone but remains a mystery (at least to me). What did I do, you ask? Here's the final schematic.

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


First I replaced the 750R resistor (R12 in Grey's original circuit, R17 in the one I attached in this post) with a 1k lin potentiometer wired as a variable resistor.

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


As the thing worked flawlessly with only one channel connected, I started out to adjust the current gain.
As far as I understood the idea is to have the Aleph current source deliver half of the actual current, so, in this case, the potential drop over the 0R47 sensing resistor (Grey's R18) and over the array of the two 0R47 resistors at the output (Grey's R20 and R21) has to be equal. So, with my Fostex FE108EZ horns connected, I fed the amp (one channel at a time) with a 1kHz sinewave delivered from the CD-player and adjusted the volume for 10Vpp output, measured with my Lithuanian scope. I adjusted the variable resistor for an equal potential drop of 1V.
In the picture you can see them at their initial position, roughly 500R. Both channels ended up with a value around 800R (don't know for sure as I haven't desoldered an measured them yet).

As this of course did not solve the problem with the oscillation, I installed a 1nF film/foil capacitor (C21 in my schematic). I didn't take a picture of that configuration or the result, which I regret. Cause this is where the voodoo starts.
The capacitor didn't resolve the issue. Again, with to leads connected to the inputs the amp started to oscillate. Same speed and amplitude but a smoother waveform. The cap made a difference but did not quite solve the problem.
The next step was to install bypass/smoothing capacitors to the pos and neg rails. I used a Panasonic ECA 2000uF50V electrolytic bypassed with a 2.2uF Wima MKS4 cap for each rail. Look how crowded it got:

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


And from another angle:

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


Notice the position of the pots.
Well, after firing the thing up I discovered that the oscillations were gone. lolwtf, was this a PSU issue?
Whatever it was, I'm so glad that the mini Aleph works properly now. It's an amazing amp, the speed and precision of a knife but still smooth and listenable for hours without being tiring. Compared to my Zenlite: Clearer highs, probably because of the lack of coupling capacitors. Maybe due to the same fact, slightly better imaging. The MiniA sounds faster, with a darker background. Sweet. Again, thank you Nelson for sharing the Aleph with us.

Setup: Sony Vaoi Laptop -> Media Monkey software -> Ethernet cable -> Linksys WMB45G Music Bridge -> Coax -> Pioneer PDR-W739 digital input -> analogue output -> DIY cinch cables -> BOZ -> DIY cinch cables ->modded Yamaha EQ550 -> DIY cinch cable -> Mini Aleph

One last pic, prodding the belly. The scope shows a few mV of noise at the output.

An externally hosted image should be here but it was not working when we last tested it.
 
Member
Joined 2003
Paid Member
Glad you got it worked out. I'm not sure if you have the AC current gain adjustment correct, If the way you did it is an alternate way then i'm sure someone will correct me if i'm wrong:

This from Nelson about Aleph 2 with some context:

The easiest way to do this is to build the circuit
without R21 (R17 on your schema) and operate it at 10 watts or so into a
load while measuring the AC voltage across R46-51( R20 in your schema)
which are the Source resistors on the negative half
of the amp. Put in a value for R21 (R17) which halves
the AC voltage across R46-51(R20) , and you'll know that
the current source is doing half the work.


I'm not sure if AC gain adjusted improperly will cause oscillations, maybe just the nature of P2P. If it sounds good, and doesn't blow up, then i'd say its perfectly fine.
 
albertli said:

Do I need a speaker protector for this?

Depends on if you practice Safe SS. Lot's of factors (speaker design (full range vs ones w/crossovers), preamp design, main amp design).

This topic has been covered ad nauseum on a variety of forums, yet as a newbie to class a, I got bit. Some designs/implementations will gladly amplify & pass DC from your pre. And loose fets sink ships.

So here's my take for what its worth...

Everything has a cost/benefit, risk/reward, etc. The benefits of unencumbered circuit paths (few coupling caps, no DC offset detection, no pwr on/off/delay relay) often come with a risk.

Since I have Ed's wonderful horns, you may read about this fun and some ideas here and here . Loved tube pre / mini-A / HSH combo & now I love it again.

Always learning. Things to think about that I think I read in early DIY Aleph doc that input couplers are optional. Always measure your amp & preamp offsets. If you are the only one using the equipment, follow a power on/off sequence. I put pwr on/isntant off delay in my tube pre & put 1/2A fuses in series w/my new FE126 drivers for a modicum of protection.

Fun stuff & worth the adventure!