I will not put hand into fire for this measurement in R-2R DACs without digital filter 🤔
Well done Nixie!
Playing around with all the possible combinations building a DIY dac takes a looong time. Putting the lid on is like ending an era 🤣.
Thanks!
Re 1. This is what I am thinking too, however, I have no experience to tell that this is the case.
Re 2. I plan to power CPLD from a single LP5907, so my thinking is the same as yours, I/O block should not matter. Using second I/O block for RJ output would make the PCB routing a bit easier as it would give me a bit more space, but I think I can do this with a single I/O block too. NB, if I considered using many regs with CPLD, then, my intuition tells me to use one for VCCcore an another one for VCCio.
Cheers
F
Well, I adjusted the DC offsets inside 1mV and I don't touch it anymore. A lot of patience is required, the trimmer pot is tiny and imprecise. Plus it changes with V7C heating. And it heats up a lot, around 45-50 degrees C.
Now I can only deal with the I/V, with the transformer and tubes, it's over with the opamps. I left the PCM56 board to have for those tests. I won't play with AD1862 or PCM63 until I try it with PCM56 first.
Last edited:
From the surplus parts left over from experimentation, I hastily put together a USB DAC with AD1865 + Amanero clone + Sparkos SS3602. Power supplies are Chinese modules with LT1963/3015. Filter capacitors are Samwha WL 4700uF 25V. There is room left on the power supply board for another LT1963, if one ever decides to add an SPDIF receiver. The SS3602 matches the AD1865 extremely well. An acquaintance from the local DIY forum will pack it all nicely in a decent box, instead of this cheese box. 😁
Attachments
"Into a plastic box" style is better and advanced than "onto a wooden board", good inspiration 👍😊
In my opinion, it's easier and much safer, both for the device itself and for users.
I want to get rid of those parts bought during the previous 2-3 years, and I don't have the strength or time to pack them in metal boxes. I also have a PCM1702, in a plastic case too. I kept the PCM63 and AD1862 for myself, I'm giving away the rest as it is. Who knows how to package it to look like a device, it's his. 😁
I want to get rid of those parts bought during the previous 2-3 years, and I don't have the strength or time to pack them in metal boxes. I also have a PCM1702, in a plastic case too. I kept the PCM63 and AD1862 for myself, I'm giving away the rest as it is. Who knows how to package it to look like a device, it's his. 😁
@andyjevans no way you're going to accomplish this with a dht alone, you need a transformer in-between... Something like 1:5 or 1:6... Once sowter (a UK company, so near you) had such a model, I used it for another project, think it was 9055 or 9066, but the new model had other specs...
DAC side, I would use extreme caution on easily stepping up Riv... Distortion can't get extreme...
My 2 cents
DAC side, I would use extreme caution on easily stepping up Riv... Distortion can't get extreme...
My 2 cents
Hi Michelag - thanks for the reply, but can you calculate what gain is required in the tube stage for 2V out? Like I say a DHT is around x8. But tubes have all kinds of gain up to x100 mu for a 12AX7. A 6922 is commonly used in kits, and that has a gain of around x30. I don't know if that gives 2V out or more than that. It would be helpful to know this, for instance. I have a board coming and I need to plan for what tube to use.
Say, ad1862 outputs +/-1mA, through Riv of 100ohm, we've got 0.2Vpp, to get your 2Vrms which is 2.8Vpp, you need 2.8/0.2= gain of 14. Take a value of 17 min.
Perfect for a 6dj8 and friends, less for 300b or 2p29l... But perhaps 2p29l can do it with a small helper as a source follower...
Perfect for a 6dj8 and friends, less for 300b or 2p29l... But perhaps 2p29l can do it with a small helper as a source follower...
2.8/0.4 = 7, but go for some more.
But I give the word to DAC specialists for the effect such a resistor value has on sound.
Usually the challenge here is to get a lower value for Riv 😁
But I give the word to DAC specialists for the effect such a resistor value has on sound.
Usually the challenge here is to get a lower value for Riv 😁
So depending on the I/V resistor, a DHT like 2P29L is in the ballpark.
I have some 3a5 which is a 7 pin dual triode with common filaments. Can be used just one half or both tied together. I haven't used one for quite a while - I remember the sound as being quite good but other DHTs were better. However, it does have a gain of 15. Cheap enough and not rare. If you threw money at it there's the 3A/110 for instance. One or two others for silly money like EML20B, which I heard once and wasn't at all impressed by.
A DHT is do-able by the look of it.
I have some 3a5 which is a 7 pin dual triode with common filaments. Can be used just one half or both tied together. I haven't used one for quite a while - I remember the sound as being quite good but other DHTs were better. However, it does have a gain of 15. Cheap enough and not rare. If you threw money at it there's the 3A/110 for instance. One or two others for silly money like EML20B, which I heard once and wasn't at all impressed by.
A DHT is do-able by the look of it.
I think that is too much for AD1862, 100ohm or less for passive I/V. All previous experiences that I know of say, the less the better. For PCM63, 50ohm or less. Exception is AD1865, there you can go to 200-300ohm, I don't know exactly why AD1865 is different.
Last edited:
Thanks! I'm posting on this AD1862 thread, but I'll be using the AD1865 so thats good news that 200R works OK. I read up on all the posts by woodturner fran where he loves the 3a5 sound. I've generally preferred the 2P29L in the past so I'll try both. When Ale Moglia tested the 3a5 he found that in parallel the gain was around x12.5. Not much different from the 2P29L at x9. For a non-DHT there is also the 27 mesh at gain of x9 which I already use to drive my SE output stage. That's unusually nice for an indirectly heated tube.
https://www.bartola.co.uk/valves/20...s-a-practical-example-with-3a5-dht/#more-5565
https://www.bartola.co.uk/valves/20...s-a-practical-example-with-3a5-dht/#more-5565
AD1865 & 200ohm = 142mV RMS. For the final 2V RMS you need a voltage gain of 14x.
It will also work with 300 ohms, 213mV x 9 = 1.92V.
It's just that the sound is better with a small RIV resistor in the passive I/V.
It will also work with 300 ohms, 213mV x 9 = 1.92V.
It's just that the sound is better with a small RIV resistor in the passive I/V.
Last edited: