advertisement


Starfish revival

You really need to remove the Traco and NERT to make a Starfish sing.

Pair of super teddy regs is my recommendation.

Pete

I never get around to replacing the Traco or the NERT. I know there were plenty of people that did. I do have a couple of +ive and -ive ALWSRs knocking about, I should try those.
 
I never get around to replacing the Traco or the NERT. I know there were plenty of people that did. I do have a couple of +ive and -ive ALWSRs knocking about, I should try those.

Its well worth it, inky blackness, veils lifted, micro dynamics, the wife popped in to say how good it sounds, the rug really tied the room together...


Pete
 
You really need to remove the Traco and NERT to make a Starfish sing.

Pair of super teddy regs is my recommendation.

Pete
Hmmm, I’ve got some STR PCBs squirrelled away somewhere, I may have to drag my Starfish out of retirement for a little winter project…
 
Thanks Hacker for the schematics! There are some differences between my version and Jims original, marked in red:

gZdkFsV.jpg

  1. My output caps have the polarity different from Jims original schematic: I measured -1.2V at the input of C10 of both the Starfish and my clone; there is +0.12V at C17 of my clone (if I short the feedback cap I get -6.35V, and the Starfish has -3.25V with C14 linked out).
  2. My 47uF caps go to PWR-GND instead of -12V (this was recommended here by DaveMc64)
  3. The feedback cap: the BOM1.2 says "no fit" for the feedback cap (C14), so Jim has the 1k resistor directly connected to signal-GND (indeed that's the way my Starfish board I got from sq225917 is configured)
  4. My regulators do not have these input caps shown in Jims diagram:
Ab3ENJZ.jpg


What do you think? Should I adapt...?
 
Addition: I measured voltage across the feedback cap, it also seems to be reverse biased, I think I have to reverse C14 and C17...
 
no issues with the difference you have already marked, all looks sensible (esp the red line connection for the CCS bootstrap cap c4 7 C9)

One thing I would recommend with the LM3x7 regs is -yes, do fit input side caps, and if you have them, use 100uF not 10uF.
Note these don't need to be anything special at all - I'd take 100uf of generic reputable electrolytic over worrying about 10uf of (undefined 'something better'.)
 
Thanks, I will update my circuits. Here the corrected schematic (changes marked in white):

q3ppyTl.jpg

I would recommend with the LM3x7 regs is -yes, do fit input side caps, and if you have them, use 100uF not 10uF.

OK! But I don't have much space on my completed boards (shown in this post): would it be enough to fit one input capacitor for multiple (parallel) regulators, or should each regulator have it's individual cap?
 
Last edited:
If you only have space for the one - fit the one, in as large uF capacity as reasonably fits. It's low precision stuff - do what fits/you have to hand : )

It helps, because transients on the output of any reg - is reflected to its input only 20-30B down (think- where does the reg source the energy to maintain regulation? -from the closest input-side cap for fast transients). A lump of dumb capacitance at the input side (even if shared)- does help each of the LM3x7 regs.
 
  • Like
Reactions: jpk
all looks sensible (esp the red line connection for the CCS bootstrap cap c4 7 C9)

OK, today I modified the preamp boards to have the caps referenced at -VE (only 3 track cuts and 3 wires were needed):

9eJW1yF.jpg


If you only have space for the one - fit the one, in as large uF capacity as reasonably fits. It's low precision stuff - do what fits/you have to hand : )

I only had some 68uF OSCONs in my parts bin and fitted one to each rail: there are 4 positive and 3 negative regulators per board, so for each board one OSCON for the negative and one for the positive supply. Then I found small 2.2uF WIMAs and managed to squeeze them in, they are as close as I could get them to the regulator pins:

QHkxu4s.jpg


The LEDs are for 5V output (2 LEDs in series at the ADJ pin), for other voltages I used zeners...
 
Confession time: I still haven't finished my NAP250. I can't seem to find anyone to make a transformer with the matt black tape coverings. (I'll probably end up un-covering and re-covering a Toriody, but I haven't gotten there yet.)

I have compared the HPA-1 with a NAD3020, a Bottlehead S.E.X, a Torpedo III and a Rogue Audio Cronus Magnum. The other 4 all have strengths and weaknesses; the HPA-1 is better than any of the others on every measure.
Good Day
Maybe this one ?
https://flic.kr/p/2nJo34i

Black ribbon transformer:))
 
And what is the difference in the schematic?
Your design based on old schematics plus dual rail power and 552 it’s new (old) schematic correcting for dual rail power +- but have different issue.
Please studying pcb gerber boards in my Flickr after few time I remove this design
 
Your design based on old schematics plus dual rail power and 552 it’s new (old) schematic correcting for dual rail power +- but have different issue.
Please stuid pcb boards in my Flickr after few time I remove this design

Can you send a picture of the schematic please?
 
I checked the values of the AliExpress 552 clone (the one mentioned in Avondales comment on Facebook) and think the circuit looks good. But I wonder about C2: the PCB shows 100pF, but Avondale has 470pF and the Starfish has 220pF? Also I am not sure about the polarity of C1, C9 and C12: what do you think?
 
OK, measured the polarity of the in and out caps, here is the corrected schematic:

ssLdA6b.jpeg


Some values are updated, so it's a Starfish with 552 values now...
 


advertisement


Back
Top