I'm using Klipsch Heresy 2 speakers with a pair of monoblocks ACA's v1.6 from the store, balanced (XLR). Those speakers never impressed me until I hooked up the ACA's, previously my Krell 400 integrated, McIntosh receiver, (2015) and three different Haflers were so so and boring. The ACA's bring me to the music, the others just bring me music.
wasting a lot of energy for not much gain.
will need much bigger hardware - heatsinks, PSU.
you might double the output, maybe 10W but at what cost?
unless you double the current you still won't be able to drive low Z reliably.
but if you double the current you don't need 48VDC to achieve that goal.
and then of course the heat dissipation will be enormous.
just look at a Zen Lite @48VDC, 2A to get an idea what to expect.
not exactly but close enough.
better build a different FW amp.
did i mention Zen Lite?
all my opinion off course, you are free to experiment.
It is very worthwhile to read N. Pass’ original articles on the Zen series of amplifiers; in particular Zen I through Zen IV. The ACA is descended from this circuit. The elegance of the ACA is in the way it uses a simple JFET buffer in front of a Zen/Aleph current source and single ended output stage.
I suspect the practical limit for the ACA power supply is about 32 Volts, and this requires some modification of the circuit to avoid burning the JFET and allow the current source to operate correctly. I have approached this higher voltage with one of my builds, using IRFP140 Mosfets, and was happy with the results. Check out the ACA-220 v2 in the ACA With Premium Parts thread.
More recently I’ve been looking at the FQH44N10 for its higher transconductance with higher Vds. This part appears to be an improvement over the IRFP150, which is often used for higher power builds.
I suspect the practical limit for the ACA power supply is about 32 Volts, and this requires some modification of the circuit to avoid burning the JFET and allow the current source to operate correctly. I have approached this higher voltage with one of my builds, using IRFP140 Mosfets, and was happy with the results. Check out the ACA-220 v2 in the ACA With Premium Parts thread.
More recently I’ve been looking at the FQH44N10 for its higher transconductance with higher Vds. This part appears to be an improvement over the IRFP150, which is often used for higher power builds.
This was my option, 18 years ago.
Project 83 - MOSFET Power Follower
It has lower distortion and lower output impedance, however without the preamp stage the gain is just close to 1, as it is a power follower only.
Project 83 - MOSFET Power Follower
It has lower distortion and lower output impedance, however without the preamp stage the gain is just close to 1, as it is a power follower only.
The beauty of ZenLite is it will dissipate 1/2 the heat through the lamp, therefore the heatsink can be smaller, to accommodate the dissipation of the FET alone.
A ZenLite with a buffer ala Zen V4 is very close as an idea and performance to the ACA, it needs higher B+ to generate wattage and is very transparent.
It is the simplest class A amp and very inefficient, that's why you need high B+.
And it sounds wonderful.
I have both, ACA as a main one, and ZenLite as a headamp (but was my main amp for a long time before the ACA).
Both produce about the same output over my 95 dB series x-over 3-ways.
yes, anyone familiar with circuits would know what Nelson was up to...
p-type mosfet instead of n-type in the ACA....
i think that for the higher voltage ACA, the thing to look at will be the SOA of the mosfet device under consideration...as far as transconductance goes, mosfets are found to be not lacking...
i would like to try things differently, i clone amps but seldom do a lock stack and barrel clones...
Or try 24v smps and bias at 12.5 volts I found that's to be an enjoyable difference over the 19v setup. I bought cheap ones off amazon no issues with noise in my setup. I have one on each channel.
https://www.amazon.co.uk/gp/product/B01DDX7J9U/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1 its not available but similar are. I bought 8amp versions so I have enough juice to try bridging with another ACA.
Last edited:
One way to set the midpoint voltage (bias) for those without access to a scope is to do the following.
1/ Disconnect the speaker.
2/ Connect a suitable dummy load such as 6 or 8 ohm 10 watt wirewound resistor in place of the speaker. A 10 watt part is cost effective and even a 6 ohm 10 watt part will easily survive for the brief test that follows.
3/ Now connect your speaker via a series resistor to the amp output. This will act as an attenuator. Try something like 220 ohm for starters.
4/ You now need to play a pure sinusoidal test tone of say 400Hz or 1kHz and turn the volume up until the pure sound in the speaker suddenly takes on a hard edge. It should be very apparent when this occurs and this signifies the amp is now clipping.
5/ Carefully trim the bias to obtain the highest volume setting possible while ensuring the tone remains pure.
A test CD or a test tone created in Audacity (or other suitable software) is ideal for this kind of thing.
1/ Disconnect the speaker.
2/ Connect a suitable dummy load such as 6 or 8 ohm 10 watt wirewound resistor in place of the speaker. A 10 watt part is cost effective and even a 6 ohm 10 watt part will easily survive for the brief test that follows.
3/ Now connect your speaker via a series resistor to the amp output. This will act as an attenuator. Try something like 220 ohm for starters.
4/ You now need to play a pure sinusoidal test tone of say 400Hz or 1kHz and turn the volume up until the pure sound in the speaker suddenly takes on a hard edge. It should be very apparent when this occurs and this signifies the amp is now clipping.
5/ Carefully trim the bias to obtain the highest volume setting possible while ensuring the tone remains pure.
A test CD or a test tone created in Audacity (or other suitable software) is ideal for this kind of thing.
Many posts say to set the bias at 12.5 while the build guide says 12.
Is there any significant difference?
Edit: Mooly posted above while I was typing this post. It's a very practical way to set bias. Would the change of bias from 12 to 12.5 (or where ever it ends up landing) be audible when the amp wasn't being driven near clipping?
Last edited:
