Oh, forgot to add a 'tip' - fit a a heatshrink sleeve over the 'face' of the steel transistor clips - it makes it easier to softly hold the transistors in place whilst fitting the heat transfer pads under the transistors &/or when 'fixing' the bigger transistors to the pcb for soldering ...
I eventually found some copper content 60/40 solder - highly recommended for Litz wire conmection (see Chimera Labs pages)
I eventually found some copper content 60/40 solder - highly recommended for Litz wire conmection (see Chimera Labs pages)
Hi all,
Tombo thank you so much for such interesting project I was looking for it !!
I would ask whether the R25 is suitable for DACs in general and specifically for my DDDAC1794 MK2, requiring 12V and 1A totally (mainboard + 4 decks).
Thanks
Tombo thank you so much for such interesting project I was looking for it !!
I would ask whether the R25 is suitable for DACs in general and specifically for my DDDAC1794 MK2, requiring 12V and 1A totally (mainboard + 4 decks).
Thanks
That's would be a very expensive way of working. A TPS7A4700 design is smaller and a fraction of the cost when it is about a clean ultra low noise 12V 1A. Even if a few would be needed (not familiar with that device).
I should not speak for Tombo56 but R25 is meant for heavier/heavy loads like amplifiers. Should not prevent you from doing so though but possibly the original PSU is well thought out for its specific purpose.
I should not speak for Tombo56 but R25 is meant for heavier/heavy loads like amplifiers. Should not prevent you from doing so though but possibly the original PSU is well thought out for its specific purpose.
Last edited:
Hmm, taking a look at the official DDAC power supplies, I see that there is a “Magic power supply” for 509€ and most basic imaginable supply is 90€. I think that settles any questions on price of R25 build.
As for the suitability, I’m confident that R25 supply will perform better than any of those there offered. It will work well from several mA to 10-20 A loads.
Isn’t that interesting that with any of expensive power supplies, there isn’t a single measurement or even single number attached. Only flowery words.
Tomb thanks for you quick answer, I like it and I would try building an R25 to try.
Ok, for the "Magic Power supply" I don't know,
but for the DDDAC Doede Dumas always published measures since the beginning both for DAC and power supply, you can find several posts on his blog and also on diyAudio threads (DDDAC user).
Ok, for the "Magic Power supply" I don't know,
but for the DDDAC Doede Dumas always published measures since the beginning both for DAC and power supply, you can find several posts on his blog and also on diyAudio threads (DDDAC user).
Probably not the innocent DIY stuff anymore but commercialized. I did not look at prices, it is just that I am not into hypes/group behavior but I think the guy designs with a passion and it then maybe is better to keep it a set. Also to avoid new threads and questions of very specific stuff no one can possibly answer to. "My DDDAC1794-MK3 does not sound good with Ophelia's Stradivarius 7 linear PSU (Limited Edition with r-core transformer) set at 12.46V" etc.
Last edited:
Ah, yes, found the blog. There is enough measurements and MK3 power supply is good. My respect to the developer.
Comparing on the pure technical level, R25 has much lower output impedance and noise (order of magnitude).
Sorry Jean-Paul,
I just realized I missed reading your reply...
Yes I agree with you on R25 cost, fortunately for my DAC build 1 PSU is enough
and I have a self-built Pass F6, where finally I would mount 2 R25 rails,
so I will just try R25 effects on the DAC (listening impressions only, sorry... I'm a total newbye at measurements).
Thanks for you feedback.
Listening impressions are what matters the most, as we use audio equipment to listen and enjoy music reproduction. There is no guarantee that you will like more one supply over another, based only on measured performance.
I’m glad that DDDAC author and I speak the same language, a technical one. He measures his designs and explains in his blog that less noise, harmonics and lower output impedance are better for the DAC performance.
That is why I can easily say that R25 will perform better, as it is better in all measured properties and then some more, that DDDAC author doesn’t measure.
Still, that is not a guarantee that you will like the resulting sound more.
Just to be sure - no heat sinks are needed for the rectifier mos-fets?
I am using the LT4320 for the same amplifiers I will be installing R25 - I did use heat sinks on those since it was the first time I had tried them and did not know what to expect. They barely warm up at all but I did use larger than vestigial ones. They would not fit into R25 without mounting half of the mos-fets to the bottom of the board.
I have no desire to install heat sinks but would rather do it before soldering the mos-fets in. I know that the size I used is not needed at all but wonder if "vestigial" ones would be insurance.
Please re-assure me heat sinks would be superfluous.
I hope to get the first amplifier converted today.
THANKS, Tombo
I am using the LT4320 for the same amplifiers I will be installing R25 - I did use heat sinks on those since it was the first time I had tried them and did not know what to expect. They barely warm up at all but I did use larger than vestigial ones. They would not fit into R25 without mounting half of the mos-fets to the bottom of the board.
I have no desire to install heat sinks but would rather do it before soldering the mos-fets in. I know that the size I used is not needed at all but wonder if "vestigial" ones would be insurance.
Please re-assure me heat sinks would be superfluous.
I hope to get the first amplifier converted today.
THANKS, Tombo
Absolutely no need for heatsinks at the rectifying MOSFETs. Even for the max. loads.
Very low dissipation is one of nice properties of the LT4320 active rectifiers.
Very low dissipation is one of nice properties of the LT4320 active rectifiers.
Has anyone noticed any difference between the LT4320 and the LT4320-1? The latter is less expensive even though it's "more capable" as far as frequency (DC to 600Hz instead of 60Hz). Just curious if anyone has tried both to compare.
Interesting that ‘faster’ version is 1.50€ cheaper. 
Until recently, it was get what version is available at all.
I don’t see any reason why not to order LT4320-1. Performance is likely exactly the same at 50/60 Hz. For a meaningful answer, we would need to compare weather there is any difference in level of harmonics after rectification. Not very likely but I l would like to know as well.
For the next supply, I will order LT4320-1 version and check (not soon).
Until recently, it was get what version is available at all.I don’t see any reason why not to order LT4320-1. Performance is likely exactly the same at 50/60 Hz. For a meaningful answer, we would need to compare weather there is any difference in level of harmonics after rectification. Not very likely but I l would like to know as well.
For the next supply, I will order LT4320-1 version and check (not soon).
Ended up doing more re-arranging than I expected to get this installed. Everything takes ten times longer than one wants it to.
Looking on the TTA004 and TTC004 specification sheets and they say the transistors are not thoroughly isolated so are you using pads for Qs 7&8 with the shared heatsink?
Looking on the TTA004 and TTC004 specification sheets and they say the transistors are not thoroughly isolated so are you using pads for Qs 7&8 with the shared heatsink?
No need for any insulation or thermal paste. Current TTC/TTA have plastic case with no metal areas for the heat transfer. It could be that there was a different version before.
As for mounting, designation letters on their case should be on the same side as marking on the PCB’s silkscreen.
As for mounting, designation letters on their case should be on the same side as marking on the PCB’s silkscreen.
The side with the three dots! Easier to see than the writing.
Luckily I guessed right.
I based my selection on the shape of the box silk screened - there is an offset that corresponds with the offset of the pins exiting the body. I could not make sense comparing the pcb to the schematic - too much going on for me to decipher.
This is what concerned me:
Note: Although this device is encapsulated in epoxy resin, it does not provide any guarantee to the maximum isolation voltage. Therefore, as with the case with non-isolated devices, care should be taken with regard to electrical isolation from surrounding parts.
The sheet is dated 2018 so there has been plenty of time for changes to have been made.
Thanks, Tombo.
Luckily I guessed right.
I based my selection on the shape of the box silk screened - there is an offset that corresponds with the offset of the pins exiting the body. I could not make sense comparing the pcb to the schematic - too much going on for me to decipher.
This is what concerned me:
Note: Although this device is encapsulated in epoxy resin, it does not provide any guarantee to the maximum isolation voltage. Therefore, as with the case with non-isolated devices, care should be taken with regard to electrical isolation from surrounding parts.
The sheet is dated 2018 so there has been plenty of time for changes to have been made.
Thanks, Tombo.
That warning is only for usage at very high voltage. Max. is 160V and we are at max. 60V.
No need for any extra precautions.
No need for any extra precautions.