Just to add.........., it's been a while:
I've not tampered with filter values but I replaced the buffer op-amps (AD843) with LME49710, and to my ears it made a substantial improvement.
At < 0.001% THD before the change it's difficult to tell if distortion is now lower on my pc based analyzer but the mid and treble is subjectively more open, more pure and more resolved than before.
Unfortunately, apart from the op-amps everything is smd, so it's difficult to identify ceramic capacitor values. It would be nice to replace with film types in specific areas.
Art, I was hoping you would add some clarity regarding Bricasti M1 mods.
As I understand it:
There are four AD834 on each analogue board. The two nearest the front of the unit should be VI converters and those were driven into class A using one CRD on each device. Do you remember the exact device used so I can purchase those and what series resistor value to use?
The two AD834 nearest the rear are used as output buffers. These were initially driven into class A as well but later you replaced these two devices with LME49710. Were the CRD's used with the LME49710 or were the CRD's removed when the AD843 were replaced with LME49710?
Like you I have owned Berkeley DACs and the one area they exceed Bricasti M1 is sweetness of high frequencies. We also both run B&W with Diamond tweeters so may be hearing similar things in the top end with these DAC's.
Thanks for your time, Dale
As I understand it:
There are four AD834 on each analogue board. The two nearest the front of the unit should be VI converters and those were driven into class A using one CRD on each device. Do you remember the exact device used so I can purchase those and what series resistor value to use?
The two AD834 nearest the rear are used as output buffers. These were initially driven into class A as well but later you replaced these two devices with LME49710. Were the CRD's used with the LME49710 or were the CRD's removed when the AD843 were replaced with LME49710?
Like you I have owned Berkeley DACs and the one area they exceed Bricasti M1 is sweetness of high frequencies. We also both run B&W with Diamond tweeters so may be hearing similar things in the top end with these DAC's.
Thanks for your time, Dale
Hi Dale,
It's been a while so I can't recall exactly......and I've moved on from the Bricasti too.
That said, I'm sure they were 2mA CRD's. I can't recall the exact resistor value used but you should be able to work out an optimum value from the data sheet.
Btw, only the AD843 buffers were replaced. The 843's are better suited to IV due to the higher slewing and faster settling times.
And, FTR, I didn't use CRD's on the buffers, only the IV op-amps. I can't recall if I connected from output to negative or positive rail, but you need to pull bias in the same direction as per op-amp output stage config, so best to check the AD843 schematic to be sure.
Hope that helps.
AV
Edit.... CRD's connect to the -ve rails.
Last edited:
Not sure what's that supposed to mean. All opamps have a complementary output stage and the AD843 is not different. Biasing from the negative rail is used in order to get single ended operation using the npn transistor only.
While i usually hear a slight improvement this way, it is not nearly as audible to me as per example changing the nfb resistor types.
AD843 is a vintage opamp, looks to be from before the days of true complementary processes. That'll normally mean it has a slow lateral PNP in the output stage and the bias is best to V- to keep that lateral PNP off.