@abraxalito Here are the pictures of the I/V stage. I looked it over with a magnifier and also gave it a sniff test to see if anything smelled burnt but it all seems ok. Are there any particular pins you want me to measure?
Hans - at the top of your third pic I see some larger-than-stock caps have been fitted at the input end of the board. Please tell me the specs/part numbers of those as there's a possibility that changing something there could induce your problem.
@abraxalito I attached a picture showing the location of the caps and their values so you know what is where. The Nichicon ES are bipolar, in case that is relevant.
Datasheet for:
Nichicon Muse ES (Green Circle) https://www.mouser.com/datasheet/2/293/e_ues-3082323.pdf
Nichicon Muse KZ (Black/Gold Circle) https://www.mouser.com/datasheet/2/293/e_ukz-3082249.pdf
Elna Silmic II (Brown Circle) https://www.mouser.com/datasheet/2/129/rfs_e-606.pdf
Datasheet for:
Nichicon Muse ES (Green Circle) https://www.mouser.com/datasheet/2/293/e_ues-3082323.pdf
Nichicon Muse KZ (Black/Gold Circle) https://www.mouser.com/datasheet/2/293/e_ukz-3082249.pdf
Elna Silmic II (Brown Circle) https://www.mouser.com/datasheet/2/129/rfs_e-606.pdf
Thanks, next I'd like to learn how the replacement 16/18mm diameter caps have been mounted to fit into footprints for 8mm diameter caps. How are the caps wired in and supported mechanically? The caps most relevant to the noise issue are going to be the bottom row of 3 on your graphic.
@abraxalito Well... the process for mechanically supporting and wiring the oversized caps is shoe-horning and jimmy-rigging. They are about as supported as kilt wearing Scottish men. 😏
There is about a 1cm lead between board and cap. Not elegant and definitely frowned upon, but it works if you don't touch it much. I included a pic so the engineers can wince and facepalm. 🤣
Here is the "however"... I replaced the three lower ones with the original stock caps and the issue persists.
There is about a 1cm lead between board and cap. Not elegant and definitely frowned upon, but it works if you don't touch it much. I included a pic so the engineers can wince and facepalm. 🤣
Here is the "however"... I replaced the three lower ones with the original stock caps and the issue persists.
Here is the "however"... I replaced the three lower ones with the original stock caps and the issue persists.
That saves me asking, thank you. Now to rule out the Muse caps on the output contributing something unexpected, would you please do the same with those? (That's replace them with the stock 6.8uFs).
@abraxalito So I changed out all caps marked with a red X on the picture and replaced them with the original. I ran out of solder wick and am waiting for more to arrive. Then I will replace the last 3 caps.
So far there is no change. The noise persists. In case it is not one of these remaining 3 caps, what should I consider next?
So far there is no change. The noise persists. In case it is not one of these remaining 3 caps, what should I consider next?
I've been scratching my head for quite a few days wondering what the next step would be. Seeing you have noise on both channels I need to come up with a plausible mechanism which explains why both channels are noisy and to date my best guess has been it would be related to the power supply as that's common to both. I'm ruling out two noisy opamps as that's super unlikely. The output caps could possibly be adding extra stray capacitance but that's a long-shot really.
Today I plan to fire up a Celi and see if I can get anywhere with trying to provoke it to create noise. One thing you could try is adding a cap (470pF, 0603 or 0805 NP0) between pin5 and pin6 of both of the LT1028s. The cap will slow the opamps down a little just in case they're oscillating for some unknown reason. If you by any chance have access to an oscilloscope putting it on the outputs might indicate if there is oscillation, I suspect that because you said touching the output caps changed something - that is consistent with oscillation.
Today I plan to fire up a Celi and see if I can get anywhere with trying to provoke it to create noise. One thing you could try is adding a cap (470pF, 0603 or 0805 NP0) between pin5 and pin6 of both of the LT1028s. The cap will slow the opamps down a little just in case they're oscillating for some unknown reason. If you by any chance have access to an oscilloscope putting it on the outputs might indicate if there is oscillation, I suspect that because you said touching the output caps changed something - that is consistent with oscillation.
Hans - I have a question based on your pics in post #321. C19 (next to LT1028) according to our graphic is a cap that was presoldered to the board but your pics show it like it has been soldered again. Did you change it or touch up the original soldering ? This cap is one essential for stability of the opamp so it seems relevant here.
One other thing that you could check would be the supply current. If there is oscillation its possible the overall supply current would increase - normally its around 120mA from +24V. Specifically, if you would check the voltage drop across R35 & R36, that normally is ~14mV. Those resistors are in series with the power supply to each LT1028 which each draw 7.5mA on average.
One other thing that you could check would be the supply current. If there is oscillation its possible the overall supply current would increase - normally its around 120mA from +24V. Specifically, if you would check the voltage drop across R35 & R36, that normally is ~14mV. Those resistors are in series with the power supply to each LT1028 which each draw 7.5mA on average.
Last edited:
Hi @abraxalito So I took some measurements R36 17.1mv and R35 is 17.8mv. Another interesting thing is that when I touch R41 and/or R47 (I have fat fingers and can't narrow it down more than that) I get a kind of popping sound. This does not happen if I touch R40 or R46.
Can you provide me with this asc file? I think it would be fun to use it to design analog filters.
Hi @pelopidas I am away from my bench for a couple of weeks but it does look like your LT1028 idle currents are showing as too high for some reason. Suggest have a go with the 470p cap I mentioned a couple of posts prior to see if the idle currents decrease.
Here you go - its a bit messy! What kind of analog filters were you thinking of using it for?Can you provide me with this asc file?
Attachments
@abraxalito thank you I made it with a discrete opamp
It's not finished yet, but soon
The interesting thing is that the problem of needing two 74157 was solved with cpld
It's not finished yet, but soon
The interesting thing is that the problem of needing two 74157 was solved with cpld
Last edited:
Hi, any problem with paralleling film caps with all the electrolytic caps on the DAC and power supply boards? Thanks!
I can't see a problem in principle, what sizes of film caps are you going to use (physically and electrically)?
On the DAC, if you're looking for improvement in SQ that might come from upgrading the I/V and current source resistors to 0.1% thin films. One customer has done this and reported a lift in SQ. If you're interested to try I'll come up with a list of the Rs and Mouser part numbers to update.
On the DAC, if you're looking for improvement in SQ that might come from upgrading the I/V and current source resistors to 0.1% thin films. One customer has done this and reported a lift in SQ. If you're interested to try I'll come up with a list of the Rs and Mouser part numbers to update.
@abraxalito
Nothing fancy, just 63V 1 uF Wima film caps under the PCB. I don't believe there is enough room for fancier caps like UKZ/Elna Silmic, so this will do. Please do send the the location and value of the I/V and source resistors I have a lot of 1% KO metal film resistors just taking up space, thanks!
I can't see a problem in principle, what sizes of film caps are you going to use (physically and electrically)?
On the DAC, if you're looking for improvement in SQ that might come from upgrading the I/V and current source resistors to 0.1% thin films. One customer has done this and reported a lift in SQ. If you're interested to try I'll come up with a list of the Rs and Mouser part numbers to update.
Nothing fancy, just 63V 1 uF Wima film caps under the PCB. I don't believe there is enough room for fancier caps like UKZ/Elna Silmic, so this will do. Please do send the the location and value of the I/V and source resistors I have a lot of 1% KO metal film resistors just taking up space, thanks!
@RenEH101 - here are the resistors to swap out for thin film, 0.1% types and their values.
'Down' and 'up' current sources, 6 off : R17,R20,R23,R26,R58,R59 47R 0.1% 0805 Mouser : https://www.mouser.com/ProductDetail/Panasonic/ERA-6AEB470V?qs=1VWA5LkbEar0M1qs0D6xDw==
I/V resistors 4 off : R38,R39,R62,R67 300R 0.1% 0805 Mouser : https://www.mouser.com/ProductDetail/Panasonic/ERA-6AEB301V?qs=m8myXnDJXpWgbmI3heQnEA==
Output stage bias 2 off : 39R 0.1% 0805 Mouser : https://www.mouser.com/ProductDetail/YAGEO/RT0805BRD0739RL?qs=aZCcseWLxnsf7F1Qtw1uUA==
@pelopidas - I realize it's been a while but did you eventually get to the bottom of your LT1028s giving the schwoooshing noises?
'Down' and 'up' current sources, 6 off : R17,R20,R23,R26,R58,R59 47R 0.1% 0805 Mouser : https://www.mouser.com/ProductDetail/Panasonic/ERA-6AEB470V?qs=1VWA5LkbEar0M1qs0D6xDw==
I/V resistors 4 off : R38,R39,R62,R67 300R 0.1% 0805 Mouser : https://www.mouser.com/ProductDetail/Panasonic/ERA-6AEB301V?qs=m8myXnDJXpWgbmI3heQnEA==
Output stage bias 2 off : 39R 0.1% 0805 Mouser : https://www.mouser.com/ProductDetail/YAGEO/RT0805BRD0739RL?qs=aZCcseWLxnsf7F1Qtw1uUA==
@pelopidas - I realize it's been a while but did you eventually get to the bottom of your LT1028s giving the schwoooshing noises?
Last edited:
- Home
- Vendor's Bazaar
- Celibidache NOS DAC