Dear Smarter people than I,
I built the ACA with my daughter and we are delighted with the results. I have since been slowly chewing my way through the articles available on Firstwatt, the threads here etc. In other words this experience has piqued my interest and I am now trying to learn.
There were a couple of things that came out of the Zen articles:
1. If I add a resistor and cap in series between the gate of Q3 and the positive end of C1 can I turn the constant current source to an Aleph current source..? and can anyone advise as to values?
2. one could add a cascaded depletion mode LU1014 Jfet for improved high frequency performance ??
I built the ACA with my daughter and we are delighted with the results. I have since been slowly chewing my way through the articles available on Firstwatt, the threads here etc. In other words this experience has piqued my interest and I am now trying to learn.
There were a couple of things that came out of the Zen articles:
1. If I add a resistor and cap in series between the gate of Q3 and the positive end of C1 can I turn the constant current source to an Aleph current source..? and can anyone advise as to values?
2. one could add a cascaded depletion mode LU1014 Jfet for improved high frequency performance ??
1. in place of limited rails voltage , let's say that Mu Follower (ACA output stage) is better than Aleph CCS ; in any case - observe resistors series to output in Aleph CCS , to understand full function of it ....... so - answer is - you need those besides cap and resistor , and you also need to short all those central resistors in ACA , to make conversion
2.improved - compared to what ? I'm not finding lacking high freq. in ACA , comparing to F3
2.improved - compared to what ? I'm not finding lacking high freq. in ACA , comparing to F3
At what voltage does the Aleph CS start to make sense?
With the input Jfet seeing the full voltage there is a practical limit and from what I've read this is ~30V before one has to start getting clever.
The comment about the cascoded depletion Fet related to ZV9 article which noted the improved high frequency distortion performance as the result of removing the Mosfet capacitance / Miller effect.
With the input Jfet seeing the full voltage there is a practical limit and from what I've read this is ~30V before one has to start getting clever.
The comment about the cascoded depletion Fet related to ZV9 article which noted the improved high frequency distortion performance as the result of removing the Mosfet capacitance / Miller effect.
I did ACA with linear supply. Dual mono. Trafo toroidy.pl 250 va 2x 20v. 40 000uf per channel with o.1 ohm. 28v cc. Rock stable, dead quiet, sound was really good. I`ve noted that elna silmic from vcc to gnd is rated 25v. Exchanged for BG 100u 50V. Also change generic 3300 uf on out to gold nichicon 4700 bypassed by 0.1 ero foiltype. Sound became even better. Works with 100 cm heigh wharfedale 86dbM 6ohm really good in 20 square m
Last edited:
Myself as well am very much anticipating another release of the original ACA. This amp is what will allow my foray into diy amp building which I never would have thought possible otherwise.
Just want to let the ppl who are responsible for making this happen know that we are truly appreciative of the hard work and dedication it requires to do this.
thank you.
Just want to let the ppl who are responsible for making this happen know that we are truly appreciative of the hard work and dedication it requires to do this.
thank you.
Pass DIY Addict
Joined 2000
Paid Member
As Zen indicated, the ACA doesn't lack in treble performance. When the project first came out a few people commented on the strong treble presence so I built mine with IRFP044 output mosfets (these sound a bit "darker" than the IRFP240 due to higher capacitance) that were part of a group buy back in 2002 for another amp.
With the IFRP044 in place, I measured the following output:
At 1kHz into 8R, 1.0vac on output (same input amplitude used for all readings, just changed the input frequency)
At 10kHz in to 8R, 1.079vac on output
At 20kHz into 8R, 1.241vac on output
At 50kHz into 8R, 2.15vac on output
At 100kHz into 8R, 3.13vac on output
Above 130kHz into 8R the output starts to oscillate
I am running my ACA in stock configuration with the exception of the 044 output devices and it sounds great. Makes a great amp for my office.
My question came from reading one of the Zen variations where the distortion at high frequencies was improved when using a cascaded device with lower capacitance. However, since that time I have also listened to all of the BAF videos and Nelson makes a comment at the end of one of the presentations that he had done something different with the ACA design to get around this problem (that exists with the Zen amp). It was ot followed up on at the presentation but I am assuming that this is the addition of the little jfet on the front actively driving the output stage. I am open to an explanation...
The ACA does produce more THD at high frequencies. I build one with a pair of (TO-3) IRF120s replacing Q1 and Q2. It sounds mellow compared to the standard ACA, I think I posted a THD plot in this thread. I had to turn up the tweeter level control to make up the difference. I like to listen to it with the feedback disconnected.
Hi Eric,
Am I understanding correctly that with the IRFP044 you are getting rising gain
as the frequency increases, until it oscillates at 130K?
The freq response in the article with IRFP240 seems very well behaved:
http://www.diyaudio.com/forums/diyaudio-com-articles/214808-amp-camp-amp-1-a.html
Maybe I'm misreading what you wrote.
Thanks,
Dennis
Am I understanding correctly that with the IRFP044 you are getting rising gain
as the frequency increases, until it oscillates at 130K?
The freq response in the article with IRFP240 seems very well behaved:
http://www.diyaudio.com/forums/diyaudio-com-articles/214808-amp-camp-amp-1-a.html
Maybe I'm misreading what you wrote.
Thanks,
Dennis
HI Dennis,
The ACA distortion curve shows an uptick at ~5-6kKz and if I recall this is due to the Miller effect and capacitance of the device multiplied by the gain. Zen v.8 shows the use of a cascoded Jfet and a dramatic improvement in THD vs frequency. Maybe this could be achieved by different choice of device, or just copy zv8.
In one of the BAF presentations Nelson Pass did allude to a different design choice being used on the ACA versus the Zen amps to improve this, but obviously in this case its not a cascade so maybe the input Jfet helps.
Beardy
The ACA distortion curve shows an uptick at ~5-6kKz and if I recall this is due to the Miller effect and capacitance of the device multiplied by the gain. Zen v.8 shows the use of a cascoded Jfet and a dramatic improvement in THD vs frequency. Maybe this could be achieved by different choice of device, or just copy zv8.
In one of the BAF presentations Nelson Pass did allude to a different design choice being used on the ACA versus the Zen amps to improve this, but obviously in this case its not a cascade so maybe the input Jfet helps.
Beardy
I have done my first amplifier....wow sublime sound vs Marantz PM66SE
Thinghs used:
- Amp Camp kit from ebay / China
- Marantz PM44SE cooling and 2 AMD coolers
- 2 DELL 65W psu 19,5 V chargers
- Thermalright thermal paste
Measurements - 10,20 -10,10 V ( can not stabilized it yer to 10 V) and 52 degrees C ( 20-22 room temp)
Ordered another one to bi-amplify 2 way bookshelf speakers - System Audio 1105 88 db sensitivity.
Sound: huge bass and profound, crystalline hights, sound stage almost very good
Thinghs used:
- Amp Camp kit from ebay / China
- Marantz PM44SE cooling and 2 AMD coolers
- 2 DELL 65W psu 19,5 V chargers
- Thermalright thermal paste
Measurements - 10,20 -10,10 V ( can not stabilized it yer to 10 V) and 52 degrees C ( 20-22 room temp)
Ordered another one to bi-amplify 2 way bookshelf speakers - System Audio 1105 88 db sensitivity.
Sound: huge bass and profound, crystalline hights, sound stage almost very good
