Just finished listening to this (also see attached circuit) DAC for about a week and am looking for some suggestions to modify the board a little. I have some experience in changing capacitors and resistors, but don't really know about reading schematics.
Is anyone able to suggest what might be some worthwhile changes to the circuit?
Is anyone able to suggest what might be some worthwhile changes to the circuit?
Replace the I/V transformers with ones from Sowter. The 9762.
SOWTER DAC I/V TRANSFORMERS
Sowter Transformer 9762 DAC INTERFACE FOR AD1865N-J
The used ones in that kit will be the weak point.
I have a AudioNote Dac 1 kit which is also a NOS AD1865 design and I am fitting a pair of the 9762s to it in a few weeks.
Might also try them on an old Philips CD471 TDA1541 design as well.
I
Wow - they look good, a designed especially for the AD1865, but are there any other cheaper options? Those transformers are more than the kit and chassis put together!
They are expensive, but you already have cheap ones in the dac and wanted to upgrade the sound. This costs money if you want to make a worthwhile difference and I'm confident these will.
I
That's an easy one!
The Sowter one is designed for the AD1865 and there's no way the old valve era ones used in the kit were designed for use with any dac chips.
"For use with most current out Digital-to-Analogue Converters. Designed for and tested with AD1865N-J . Should also work well with AD1851, AD1852, AD1853, AD1861, AD1862 . Provides current to voltage conversion with approximately 22 dB gain. Presents a low impedance to the DAC output for best frequency response. Provides significant sound quality improvement when compared to other interface methods. Ensures full isolation of digital noise in ground and supply rails."
However I am assuming the ones in the kit are actually being used for I/V conversion and are connected directly to the dac chip.
The Sowter ones are intended for direct connection to the Iout pins of the 1865 (pins 4 and 21) so you'd need to follow the tracks from these pins to see where they go to. If not straight to the transformers, then disconnect from whatever they do go and connect instead to the transformer.
(a circuit diagram would be useful!)
Thanks B'Man - from my limited knowledge I think there is a 220R resistor in between the DAC chip and the transformers (schematic is on the first post).
Yes, I should have looked at the schematic first.
The Sowter ones will just replace the existing ones but you may need to experiment with changing R4 to the values suggested by Sowter.
The 220R on the dac output will be okay.
I
A lot of old transformers like that can be as nice or nicer than the best you can get now, HOWEVER, those transformers are marked as TELEPHONE grade, 250-3500Hz +/-2dB. They must be quite a lot better than that as used in this app, but there's little doubt that a Sowter or Cinemag, etc. would be a lot better.
Thanks again - I think the transformers are quite good (I have a solid state AD1865 dac and this DAC beats the pants off it!) but an upgrade should make a big difference again.
what does the "250-3500Hz +/-2dB" spec indicate?
re the resistors, I have altered R4 in the SRPP to 2k as there was far too much gain (resulting in distortion even!).
what does the "250-3500Hz +/-2dB" spec indicate?
re the resistors, I have altered R4 in the SRPP to 2k as there was far too much gain (resulting in distortion even!).
The transformers you have are rated, as marked, for speech range, i.e., telephone grade, which is generally specified as 300-3000Hz. Those must do well beyond that range, though, if they actually sound good.
Ahh okay - they do sound good, but I guess without something for an A/B comparison there is no real context for an impression of their sound. How important is keeping the "primary 20k" and "secondary 600 ohms" spec when looking for a replacement?
Would something like this be a good test to compare?
ST-20600-OB Audio Transformer 20?:600? for DAC (eBay item 330435862267 end time 16-Jul-11 00:22:29 AEST) : Consumer Electronics
Would something like this be a good test to compare?
ST-20600-OB Audio Transformer 20?:600? for DAC (eBay item 330435862267 end time 16-Jul-11 00:22:29 AEST) : Consumer Electronics
As an update for anyone following this thread - I've made the following changes:
* swapped out the original transformers (telephone grade) for Lundahl LL1530
* changed the 220R resistors for 220R Z-Foil resistors
* soldered 0.1uF caps across all diodes
* changed the output caps from 1uF Wimas to 1uF K40Y-9 Russian pio caps
The sound has certainly markedly improved.
However I note that when I put the DAC to directly to a power amp (w/ 100k input impedance) there is a buzzing that comes from both channels - similar to the hissing noise one gets when a preamp is turned up to its maximum. Anyone know how to troubleshoot something like this?
* swapped out the original transformers (telephone grade) for Lundahl LL1530
* changed the 220R resistors for 220R Z-Foil resistors
* soldered 0.1uF caps across all diodes
* changed the output caps from 1uF Wimas to 1uF K40Y-9 Russian pio caps
The sound has certainly markedly improved.
However I note that when I put the DAC to directly to a power amp (w/ 100k input impedance) there is a buzzing that comes from both channels - similar to the hissing noise one gets when a preamp is turned up to its maximum. Anyone know how to troubleshoot something like this?
Could you be more specific about the noise you are hearing?
'Buzzing' suggests power supply related noise, to me. Which could stem from a ground loop or poor connection with the power amp, or from a regulator defect, or from induced coupling between the power transformer and some other part of the circuit (particularly, any signal transformers).
'Hissing' suggests a general white noise, to me. Which could stem from some component having become noisy. I'd start with the DAC chip and the +/- 5V regulators, since you said the noise is in both channels and the AD1865 is a dual DAC. You may need a scope to trace the source of the noise.
Last edited:
It is more like a 'buzz' than a 'hiss' - seems to be at a higher frequency, at low volumes it can’t be heard – for example with one power amp (sensitivity 3.0V) the noise is not overly annoying, but with a more sensitive power amp (300mV tripath) one simply would not want to use this dac.
It is definitely not related to the tubes themselves (I’ve tried swapping out a few and the buzzing continues), but could be due to AC interference? The transformers are quite close to the output section of the board – would it be worthwhile installing some kind of shielding? The signal wires are already twisted, as are the AC power wires, would it be worthwhile installing some kind of metal plate perpendicular to the chassis floor – ie. between the PCB and the transformers? I have seen this done in a few amps....
It is definitely not related to the tubes themselves (I’ve tried swapping out a few and the buzzing continues), but could be due to AC interference? The transformers are quite close to the output section of the board – would it be worthwhile installing some kind of shielding? The signal wires are already twisted, as are the AC power wires, would it be worthwhile installing some kind of metal plate perpendicular to the chassis floor – ie. between the PCB and the transformers? I have seen this done in a few amps....
I assume you don't have the buzzing through the Tripath with other sources? As it turns out, I built a 200W/ch. Tripath amp (2V sensitivity) which produces a low level grundgy buzzing which I've not been able to eliminate. I've tried power line conditioners and shielding the input leads within the enclosure with copper foil tape. No luck. I continue to use it because it has fabulous clarity, and a realistic dynamics that somehow remain musically apart from the low level buzzing.
Last edited:
Nope - the tripath (2024) is clean a a whistle normally - and you are right on the musicality of the tripath gear, phenomenal!
darn buzzing hey - what a curse!
darn buzzing hey - what a curse!
Last edited:
The SRPP circuit is very high gain, which means you can very easily get your described symptom with any inadequacies in the power supply or ground circuits, as well as greater sensitivity to the power transformer's EMF and external ground loops via other equipment. Without seeing the circuit board & actual build, I can't speculate about the ground paths, but the schematic shows what I'd say is below bare minimum post-regulator HV supply filtering. That 10uf is not big enough for my comfort, but it's damn near two big for that LM317 to survive charging, sitting as it is on top of a 230V output. It's lucky to survive first power up. If the discrete mosfet circuit before the 317 were turned into an actual voltage reg, which would require quite a small number of alterations, the 317 could be eliminated, and the post reg filter, C14, could be increased to a more suitable & certainly better sounding 33uf or higher. I would do at least 47uf, and also add another 10-22uf in parallel with C5 on each channel, but that's just me with my "no such thing as too much power supply" philosophy. This may not be the immediate cause of the buzz, but I would certainly do it for the sake of sound & freedom from worrying about relying on a low voltage reg chip in a high voltage circuit. If the proximity of the power transformer is the problem, which should show up no matter what the dac is connected to, then I have found shielding of any kind rarely works as well as simple distance, but a grounded panel of "T.I. shield" alloy material is likely the best material to try.
I also thought that the 10uF looked to be a problem. I've got 470uF pre-regulation, so might try altering the post-regulation filtering. Quite messy to get LM317 out of there, though.
SRPP is very fussy - I've reduced the gain resistor right back to 2k7 but it doesn't seem to have solved the problem.
Where is "TI Shield" available (never heard of it)?
I've only ever seen TI Shield at Michael Percy Audio, where I've been buying it for the last 15-20 years. He's in USA, and I don't know what his overseas selling policies are.
It looks like the mosfet is doing current limiting on turn on, which would explain how the LM317 survives at HV. You can probably go to 22-33uf without cooking the 317, if this is indeed the case. Should be enough to at least determine if that cap size is part of the buzz problem.
It looks like the mosfet is doing current limiting on turn on, which would explain how the LM317 survives at HV. You can probably go to 22-33uf without cooking the 317, if this is indeed the case. Should be enough to at least determine if that cap size is part of the buzz problem.
- Status
- Not open for further replies.