The datasheet states many examples per supply (for 15VDC it's 7 - 8.4mA) but the following graph (page 9) tells the whole story
So for 12.4VDC of my reg and Temp around 20 C, the 7.2mA is little bit high (I'm seeing something 6.5mA or thereabout)
So for 12.4VDC of my reg and Temp around 20 C, the 7.2mA is little bit high (I'm seeing something 6.5mA or thereabout)
That graph is for split supply (+/-). I don't know if there's a difference for single supply?
Strange behaviour: while I can setup both channels to below 1mV without opa, when I put the opa back, measuring after few minutes I get 0.1mV at one channel, the other channel slowly climbs up to and little above 5mV. I've changed the Jfet pairs between channels and the same figures come up.
Tired and went to sleep, I'll try tomorrow with a 5VDC supply to compare with datasheet but I'm already set my head in buying a new chip. SMD version are easy to find but the DIP ones are becoming scarce (and more expensive).
Tired and went to sleep, I'll try tomorrow with a 5VDC supply to compare with datasheet but I'm already set my head in buying a new chip. SMD version are easy to find but the DIP ones are becoming scarce (and more expensive).
Thousand thanks, it's great to have such a community where you are not alone in your endeavor, although I'd try locating one in European eshop, my favorite TME only stocks the smd version. Do allow me some time for research
I run a test in various frequencies (100, 1K, 10k) and volumes up to 1.2VAC and nothing came up, offset DC at the channel in question, was steady -1.4mV.
After that, I've decided to do another testing with input shorted and now the offset dropped to 1.4 (measured after 10 minutes) with open lid and with closed climbed to 2.3 (in another 10 minutes). The other channel remained with much better result, around 0.1mV.
Room temperature is about the same and I haven't touched anything from yesterday.
How do you like them apples !
Last edited:
Maybe it’s the scopes bw limits. Or maybe not. Strange current draw in opamps is generally oscillations. What decoupling to rails is used and what resistance do you use on the output?
Back to this project tonight. No smoke, no flames & no funny business. Excellent!
Supply is +/_ 15.02 VDC rails
148mA across R10
151mA across the other R10
I have 0.000 VDC offset out on both channels with OpAmp running. Prior to that it was drifting 0.000 - 0.001 - 0.003 around that neighborhood.
I underestimated the power required. I think I have solutions. More on that in a bit but everything checked out like the Salas BOM - Guide instructed.
Perhaps I can get a signal going and measure THD.
Cheers & many thanks.
Supply is +/_ 15.02 VDC rails
148mA across R10
151mA across the other R10
I have 0.000 VDC offset out on both channels with OpAmp running. Prior to that it was drifting 0.000 - 0.001 - 0.003 around that neighborhood.
I underestimated the power required. I think I have solutions. More on that in a bit but everything checked out like the Salas BOM - Guide instructed.
Perhaps I can get a signal going and measure THD.
Cheers & many thanks.
That's the ringing of the audiocard's squarewave synthesis. Some cards are even worse. Digital lab generators are way smoother. But no alteration is seen from the preamp.
I don't think it's the scope or the rest of the circuit, which is configured for line out use (50R at the output) with UBIB serving as regs with 6800uf per rail, normal 100uf decoupling at DCG3. But still think something fishing is going on and I'm gonna blame it on the opa (BTW I dislike e opas in the signal way)
I've tested today with my other multimeter (Voltdraft VC840) and provided 5VDC and the current was 5.13mA, almost very near to the 5.2mA indicated in the papers.
I think is safe enough to continue my Jfet testing but I'm gonna get my hands on a new AD823 just to sleep without worries.
My suspicion is that maybe when I took out the opa, I've bend the legs too much, even with using a special tool the one side came off first bending the other side's legs little bit. Continuity test shows everything is good and optically seemed no harm came to the legs but it's the only explanation I have.
Anyway with the new base I can do my test easily and one bloody opa won't get in my way, no sir !
Thanks guys.
My open-face-sandwich charcuterie board-style Salas listening session went exceptionally well today. The power supply has CRC LM317/337 and is a diyaudio.com circuit board design called FlexReg - rhthatcher - Rev 1.0b 1/2023. It is a good power supply but the heat sinks just barely keep up with both channels of the DCG3 but I have a remedy for that heat, or so I think. Later on that.
I played a few good listens and I was delighted with what I heard today. This design is right up my alley. I feel fortunate to have focused on this design and it paid off putting it in the top of my project list. I'm gonna say it sounds like vacuum tubes when they are on their best behavior. I can't wait to listen to it on my old Symasym Bitner amp and also my headphones. I couldn't find a headphone jack today and I didn't feel like cutting good cables. Bench tests can wait a few more days.
Old CDs on a Toshiba XR-9058, DCG3, JHL1969 custom special and Dayton Audio bass reflex two-way with Audax tweeters. Oh and a generic volume control box from China.
The JLH1969 with the DCG3 may be coloring the ultra top end of the spectrum but I'm not sure. Over 15kHz is hard for me to hear(I'm getting older). I have to play certain other musical passages to continue my excellent evaluation.
It is obvious to me, at this stage, to say this is exactly what I like and exactly the way I like it. It will probably shine better in other configurations, I'm sure of it.
I think my rail voltage, LED selection and relay need a slight tweak. It is latching lazy. I can fix that before I box it. Not an issue.
Good times!
Thanks again.
I played a few good listens and I was delighted with what I heard today. This design is right up my alley. I feel fortunate to have focused on this design and it paid off putting it in the top of my project list. I'm gonna say it sounds like vacuum tubes when they are on their best behavior. I can't wait to listen to it on my old Symasym Bitner amp and also my headphones. I couldn't find a headphone jack today and I didn't feel like cutting good cables. Bench tests can wait a few more days.
Old CDs on a Toshiba XR-9058, DCG3, JHL1969 custom special and Dayton Audio bass reflex two-way with Audax tweeters. Oh and a generic volume control box from China.
The JLH1969 with the DCG3 may be coloring the ultra top end of the spectrum but I'm not sure. Over 15kHz is hard for me to hear(I'm getting older). I have to play certain other musical passages to continue my excellent evaluation.
It is obvious to me, at this stage, to say this is exactly what I like and exactly the way I like it. It will probably shine better in other configurations, I'm sure of it.
I think my rail voltage, LED selection and relay need a slight tweak. It is latching lazy. I can fix that before I box it. Not an issue.
Good times!
Thanks again.
Attachments
Congrats !
I'm surprised, that piece of aluminium don't seem adequate for DCG3, maybe for winter use. I had something similar and couldn't touch it. Keep that in mind if you gonna box it sometime.
I assume that later you gonna change the cable arrangement of the psu, one channel powered via another is not optimal.
You may solve your heat problem and maybe achieve better sound quality by adding another of these FlexReg psus for the other channel. Again, remember again that you gonna box it sometime and there starts the problem with heat.
And not only that: I assume that you decided on +/-15VDC because the rails sag too much for the regs to output +/- 17VDC, right ?
With another transformer on the play, you minimize the load so less sag which means more headroom in order to push the output voltage to the recommended +/- 17VDC.
Me too. Ha, this is not the final setup or wiring. My boxes are going to be part of the heat sinks, so I think I have it figured out. I'm waiting for some aluminum plates to arrive to test the final arrangements for heat dissipation. I did not decide on +/_ 15 VDC, that was where it was set previously, so I left it there for testing. I don't think the rails will sag and I should be able to deliver the specified 17-volt supplies but I'll find out soon. I need more heat sink first.
I will perform extended load tests into 32 ohms with the rails up to 17 and all the gear stuffed in the enclosures. I think it will work well. There will be testing on the Quant Asylum QA401 before that stage. Lots of detail before it is finalized. Soon...
Cheers!
The 1/4" aluminum I used works perfectly. I bolted the aluminum stock to the aluminum shells with 3mm hardware and heatsink compound. The room temperature is 22.5°C and the power supply shell is 37°C on top and 38°C on the bottom. The preamp shell measures 34°C on top and 41°C on the bottom. This is all buttoned up and running for many hours. I still have the rails at +_15 VDC and I will change that soon to +_17 as specified. I am exceptionally pleased with the thermal performance of the aluminum enclosures. This is a better result than I was banking on.
With the volume pot in the 3 o'clock position, I can barely stand the SPL on a pair of AKG240's, lots of power and tons of gain for me! The current source is FLAC from a Windows PC. This is the only source for headphones I have used so far today. More sources soon. I had a wonderful time rocking out to Magdalena Bay - Imaginal Disk. Loads of fun to listen to on headphones. An interesting recording to say the least.
I am starting to wonder about this clever circuit being used as a front end to an audio power amplifier. It's just so wonderful to listen to.
Cheers & thanks!
With the volume pot in the 3 o'clock position, I can barely stand the SPL on a pair of AKG240's, lots of power and tons of gain for me! The current source is FLAC from a Windows PC. This is the only source for headphones I have used so far today. More sources soon. I had a wonderful time rocking out to Magdalena Bay - Imaginal Disk. Loads of fun to listen to on headphones. An interesting recording to say the least.
I am starting to wonder about this clever circuit being used as a front end to an audio power amplifier. It's just so wonderful to listen to.
Cheers & thanks!
Attachments
Congratulations. Nicely shoe-boxed and great that you like it. Some even held it at +/-15V for slight subjective preference reasons. When having more than enough loudness with your headphones, holding it there may anyway help with keeping dissipation lower. Do you ever get periods with 28C room ambient?
Thank you. I love it. I shall consider keeping the rail voltage as is.
Taking into account extended periods of use (heat/poor setup/in the sun), I have decided to vent the enclosures, both of them. These little clam shells do not afford much room to do anything but I have a plan to execute and keep them cute looking and not butcher the little boxes. I can get the two boxes lower in temperature with a tiny bit of airflow/convection, no micro fans are needed. Some delicate surgery is required. I had stainless steel mesh here but I took it to work last year. I need to find what is left of that or order more to proceed with cosmetic vents.
I have an interest in changing the volume POT and the RCAs. Minor tweaks. I still want to run it by the Quant Asylum QA401. So lots of little things to tinker with but overall a huge success for me.
Exactly a year ago, to this day, I finished building a Maxim/Analog Devices MAX97220 Headphone driver/line amp. I should compare it for fun.
Wünderbar. Salute!