DIYHiFi.org

Carlos & Jam Man,

What I was saying was to use like the United Chemi-Con Alu Polymer. On Jocko site he published some ESR curves on these and some of the Black Gates:

http://www.jockohomo.net/data/

The UCC 220 smd is what I was suggesting but they do have it in a leaded and smd style. They are available in Super low and Ultra low ESR.

Jam I don't agree with splitting the ground plane. I was shown years ago that these analog guys where nuts for seperating the analog and digital planes. It means the current between them is going to go through that the smallest part. Let current breath in the ground plane for the best results.

I do partition the top layer fill usually to account for the best loop characteristics because of SMD parts and such. But the bottom is usually one piece.

Thanks
Gordon

I thought about that too.
The return current from the reg goes up on the board and then down the hill.
I only connect one ground to the main circuit, so the ground pins are the three central ones, marked with a 'G'.
The other ground pins are just for shielding. In case of using coax cables the shield would only be connected on one side.

Guys, quick question:
Do you think that this PSU (100ma) will have enough current to drive the 4 74HC74s?

Ok, I think it's done.
There are some IDs on the bottom layer now, and two options for regs.
I'll post the pdf files tomorrow.

Here are the files.
Thanks to all those who participated with valuable input.

Happy reclocking.

Hi Gordon Man!

You're totally right about the split! It was a oversight! I was considing the supply entry to the Reg as the beginning of the supply gnd path / end of the return supply currents ..but the circuit isn't stand alone, and use multiple gnd connections to the main board ..so, no split!

I read all the bypass thread in the other section ..very good graphs!
I will try those NCC caps.
Did you try Pana SP caps for digital bypass?

JamMan,

I have not tried the SP for bypass. I use the FC for power supplies allot. I primarily used the Black Gates for everything until Jocko started testing them. They are good in some places, others are better in other places.

Carlos, 100ma maybe on the edge for 4 74's especially if you are clocking them hard. You know really a better way to do this might be with microgate latches. Since you have seperate power for each one and some are just D, Q and CLOCK they can work to your advantage. The only problem is they are SMD only which of course can dreg up other problems. But the nice thing is most of them are newer and can hangle higher rates on the clock.

http://focus.ti.com/paramsearch/docs/pa ... mTable=yes

Thanks
Gordon

Yes, that's what I thought.
Anyway, the board will not always be used with the 4 chips (although it can can also do S/PDIF reclock so the 4 chips are always handy).
In that case, a 7809 reg must be used (included on the final board) and the BC547/550 can be changed for a BC639 (different pinout), adapting it's pins to the board.
If using this board inside a CDP I really recommend a small dedicated 12V trafo and a simple RCRC or RCLC PSU.
Tapping a voltage from the player's PSUs is not adviced - you have all been warned.



Yes, thanks.
Although for my purpose the 74HC74s are fine.
They can be bought as cheap as chips and everywhere. And they do the job fine.

Carlos,

In defense of Black Gate there are times when these really do it where nothing else does. I mean for me the UCC are real nice at the USB receiver side. But at the dac, pll and vcxo (they run on batteries) I find the Black Gate work better. Espeically on the pll for some reason.

The big hype though with Black Gate was the L cancelling pairs which was a waste of time. They did not test any different putting them in direct parallel or reversed.

Thanks
Gordon

Is this similar?

http://cgi.ebay.com.cn/ws/eBayISAPI.dll ... gory=74813

Not UCC, but NCC...

Carlos,

Yea they look the same... They have them at DigiKey and Mouser.

http://www.chemi-con.com/files/PSA.pdf

Jocko did some testing on these that looked real nice...

http://www.jockohomo.net/data/UCC_220uF_SMD.jpg

Thanks
Gordon

Yes, I've seen that.
Up to today, it's clearly the 'Jocko capo'. :mrgreen:

Gordon, Carlos,

Would like to add: yes, those PSA caps look really good, and from the factory specs the higher capacity values can go even deeper than the 220uF tested.
Just a notice: a panaFM 680uF /10V cap tested by me here, post 160, would look ~ like that UCC measured by Jocko. This can be also a ~ translation between the different measurement graphs.
(Jocko's plot shows ~17mOhm ESR, mine shows [at the impedance minimum] ~ 23mOhm ESR, and the SRF points being ~ equal, the img. parts would be ~ equal.)
The physical size is the same.
Edit - well, not the same, 8*15 instead of 6.3*9.8. But the voltage is also higher for the Pana.

Well, this was my 37cent contribution.. check at DK..

Ciao, George

Hi George,

Thanks for clarifying your graph and Jocko's.
I'm comparing both datasheets and, at least in the specs, the NCC caps have much lower ESR, for the same capacitance/voltage rating. Also the NCC has higher ripple current spec.
Although in most cases the Panasonics have smaller case diameter (NCC doesn't spec lead space). Probably the Panasonics take this as an advantage resulting in lower ESL, so in the end the results are similar.

If you want to compare, pick for instance a 470uF/10V cap from the two datasheets:

http://www.panasonic.com/industrial/com ... 0CE108.pdf

http://www.chemi-con.com/files/PSA.pdf

Well, I order from Digi-key UK and the caps come from the USA.
They eat me alive for orders lower than £75, and then customs eat my skin when it arrives here.
Sooo... if some of you fellow americans are going to order something from Digi-key and need to make a minimum order I would be interested to jump aboard. Send me a PM.
Your USPS mail system works very well, I don't mind waiting for a couple of weeks (most of the times one week), who need$ UPS?

PS: Farnell also has Panasonic FM caps, but they cost double the price.
It's the old one dollar is one pound tradition.