Salas DCG3 preamp (line & headphone)

Your UltraBiB PSU has a CCS current limiter so it will protect the load (DCG3) from a fault higher to your R1 mA setting anyway. If surpassed it will drop the PSU's output voltage and you will instantly know by the PSU's LEDs turning off. Not only useful in test, but forever, its a built in feature. I put the CCS there primarily for dynamic separation reasons but it works as a CC protection mode too.
The dim bulb tester will sense the 6800uF reservoir capacitor's inrush current and try limit it but once charged it will not react for brightness any further. Because the PSU's current draw will be maintained constant no matter what gross fault may occur after its current limiter. Naturally it won't protect what's located before it i.e. the transformer, the bridge diodes, and the reservoir capacitor.

Salas,

Thanks. I tested it without the bulb. However, the neg. reg. leds are not illuminating when I run both sides of the DCG board. I haven't received my 2R yet for R1. So, with the 12R in R10 (~100ma bias) I am exceeding CCma.

Or I have a short.

When I run one side of the DCG3 board the neg. reg. does fine. I am measuring ~1vdc across R10 when both channels are powered. But, only did that to check voltage. Don't want to run the neg. board if there's a problem.

The relay did close, and all three LEDs on the DCG3 board did illuminate when I tested both sides.

Cheers,

Greg
 
Member
Joined 2006
Paid Member
Salas,

Thanks. I tested it without the bulb. However, the neg. reg. leds are not illuminating when I run both sides of the DCG board. I haven't received my 2R yet for R1. So, with the 12R in R10 (~100ma bias) I am exceeding CCma.

Or I have a short.

When I run one side of the DCG3 board the neg. reg. does fine. I am measuring ~1vdc across R10 when both channels are powered. But, only did that to check voltage. Don't want to run the neg. board if there's a problem.

The relay did close, and all three LEDs on the DCG3 board did illuminate when I tested both sides.

Cheers,

Greg

Unloaded from DCG-03, does the UBIB Negative side light up fine and send proper voltage?
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Greg's regs have been already used in his BA-3 so they should be alright. He still has 3R CC set resistors (~200mA) so he is right on the verge for powering 100+100mA DCG3 channels. I wouldn't worry there is an anomaly before he will be setting the regs to 300mA CC for proper evaluation of the system. If in a hurry, some handy 10R power resistor across each now 3R makes 2.3R as a parallel combination for a confirmation test before receiving the proper 2R replacement ones.
 
I got really close to finishing this yesterday. Unfortunately, I'm headed to Stockholm, would rather stay home a play with audio!

0713181542.jpg

Hoepfully as Salas said when I increase CCma it will be fine. They worked perfecrly with the BA-3.
 
Salas/Tea,

Okay, really having a problem with my power setup.

I’m running single mono BiBs and I’ve switched R1 ro 3R, so plenty of CCma but any time I try to run both sides of the DCG I’m shorting the neg psu board. I’ve followed the single mono psu grounding from section 6. In the DCG3 manual. I can run one side of the board just fine.

I had no problems powering the BA-3 board with this set up.

Any ideas?

Thanks,

Greg
 
diyAudio Chief Moderator
Joined 2002
Paid Member
How did you solve your PSU set-up problem? Wasn't 2R good enough? No don't switch pots yet, let it play. After the system settles I will let you know how to bring it spot on to your preference without changing the pot with one more than double its value, that may make it soft. You can tune finer just by changing the value of the DCG3's input series resistor.

*Sorry for not replying till now, I was at an island for a short holiday.
 
Salas,

Got everything working. It has been running for a couple of days now. Absolutely amazing pre. Still a lot of hours of break-in, but these are my mods so far.

1. BC327-40 (per post # I don't remember)
2. BJT, fully discrete op-amp (huge soundstage and smoothness increase)

This is by far the best pre I have ever heard. Still a bit bright. So let's see how it does.. Very dynamic, a lot of micro detail. Not as emotional as the BA-3, but way more interesting to listen to. It's just a treat to listen.

I will do a full review later.

A huge thanks to both Salas and Tea.

Cheers,

Greg
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Salas,

Got everything working. It has been running for a couple of days now. Absolutely amazing pre. Still a lot of hours of break-in, but these are my mods so far.

1. BC327-40 (per post # I don't remember)
2. BJT, fully discrete op-amp (huge soundstage and smoothness increase)

This is by far the best pre I have ever heard. Still a bit bright. So let's see how it does.. Very dynamic, a lot of micro detail. Not as emotional as the BA-3, but way more interesting to listen to. It's just a treat to listen.

I will do a full review later.

A huge thanks to both Salas and Tea.

Cheers,

Greg

Hi Greg

Its ~1MHz fast right now with a 20K pot (post#20 scope pics) and if your system's tone has been formerly tuned around a softer rise-time incoming signal it may sound fast/bright. In most builders systems here its agreed as tonally neutral around the 20k pot mark. With one or two notable exceptions who decidedly preferred 50k. The loudspeakers tone and their room position being the more influential factors.
In any case try a 50k pot if handy to get the feeling. Maybe its all you need. But if it will lean toward a too soft/round opposite, change R1 on DCG3 from 3.3k to 6.8k so to make it behave like seeing a 34k pot when using a 20k pot. That may hopefully be your point of preferred tonal balance. Also there are build quality differences between pots, like a 20k budget switcher pot is subjectively sharper than a 25k Goldpoint etc. The pot's value and quality is intentionally a tuning area in this pre, there is no source impedance active buffering, so see it as a matter of system synergy/personal preference and its up to you. Even a carbon track super cheap Alpha pot may prove to be your leaning. No wrongs.
 
Salas,

As far as pots go, I usa a dact type with smds. It's very neutral and transparent. I only have 2 values on hand: 10 and 20k. I switched to 10k and it is almost perfect. I know this is isn't in the proper range for your design, but it really works well for me. Now, I can do what you said and change R1 and go back to 20k. In fact, its the same pot I used with my DCB1 for many years, with no harshness or glare; very smooth sounding.

My 2 channel room has never been bright. What i'm hearing is definitely a component change. With the 10k, I can now enjoy and listen. It really is a truly phenomenal preamp. But, I need to address the slight sibilance I'm hearing. So, I'm all ears for suggestions.

Unfortunately, my tube dac has a very high output impedance of 20k. So, would 50k not be way too bright if 20k was so? I'm open to any suggestions because I love this thing otherwise. There has to be a fix with either a pot and or R1. Maybe a trimmer on R1 to adjust tone? Or would that be stable?

Cheers,

Greg
 
Last edited:
diyAudio Chief Moderator
Joined 2002
Paid Member
50k should be less bright vs 20k unless its a pot quality issue. But 10k is even harder loading the DAC than 20k. Maybe the older pot had better resistors and the new one is harsher. Because with 10k it also means faster dcg3 vs 20k but you prefer the particular 10k and rise-time seems not to be your main bright factor after that experiment. Try various high quality R1 either for type (tant carb MF) or step by step value increase until perfect. Install female pins for R1 to make life easier until hitting your golden tone mark. Keep those resistors legs rather short. A trimmer test may be indicative for value. Watch not to go too low for value because the circuit may oscillate.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
The pot & R1 are surely a strong tuning area but do try the AD823 servo in the socket again. Don't forget to recheck DC offset. I won't go into caps substitutions like C6/C7 FC to Nichicon ES auditions etc. because too much rework is involved if the boards are installed with nuts and bolts already.
 
Yeah, a cap is passive item, way down thr list:
1. Design
2. Active parts
3. Passive parts

Your design and short signal path makes a very, very revealing pre. I'm very impressed. I think once I get the tone right this is an amazing setup.

Anytime you build something you just have that chance that some tweaking is necessary. In my case, I was like "uh oh". Then I switch to 10k, and then my jaw dropped. Its as revealing as the my DCB1 �� with much better dynamics and not just a unity gain device.

Wish you guys could hear it driving my Montis speaker via my class a USSA-5 amp. It really is amazing.

I will order a 50k. The 20k is the one I have been using for years with my DCB1. I dont think its the pot. I must have done something wrong. I have checked all of the Rs on the board. Dont see any wrong values. Plus, Tea put the kit together.

Anyway, its all good and just have to try the 50k.

Cheers,

Greg
 
diyAudio Chief Moderator
Joined 2002
Paid Member
BTW try coax internal wiring first if its not. This is a very fast preamp it may reveal EMI/RFI you were not aware it was there before. There are many builds done with 20k but with the dual mono DCSTB and AD823. Not bright even on very revealing headphones but you got to scrutinize your own to nail the tone bcs having a great cake but missing the cherry ain't cool :)

P.S. Post a pic so we can see the wiring and grounding, maybe there is something there to do better.
 
Lol. Yes, already wired with cardas shielded. But the shield is open on both ends. I hear nothing from the loudspeaker, but any emi at low level could easily ruin the magic. I use star ground with everything on the clean side of CL-60 and a/c power inlet on the dirty side.

50k is on its way. Couple of weeks I will know. In any case, I can enjoy it while waiting and try some other things.
 
is there any way this preamp, with high enough PSU rails, gain of 10 or so, with attention to the heatsinking needs, would NOT be ideal or suitable to drive directly a f4 (without imput buffer)?

I am looking for quite a while for a preamp to swing lots of volts. First i thought i'd build a scaled down aleph, having already built a beta22, mini-f5, and some of juma's stuff.... but i landed (again) on this thread and it has «stung» my curiosity.

i also have a «slightly modified» sigma22 regulated supply that i use regularely that has the same toplogy and sounds top notch for a regulated supply, so maybe it could make a nice companion :)
 
diyAudio Chief Moderator
Joined 2002
Paid Member
It would need +/-15V additional chip reg modification for the servo and the relay not to break them and it can take +/-25V rails or more. Back in the thread some Sigma and a European high quality reg were tried by a builder but he ended up with dual mono DCSTB if I remember well? Not sure, search back pages.

P.S. Also see in post#23 the CCG3 solution if you are willing to have an output coupling capacitor. No servo and relay there, no worries, just the rails.