opa 627bp in MHZS cd33

Hi Guys,

I'm new to this site and have been reading the various CD33 threads with interest.
Please be careful what type of op-amp you try in this cdp,  some of the fast high bandwidth types are not stable in the circuit without additional modification and may suffer oscillation causing the op-amp to overheat.
The decoupling layout is quite poor in these cdp's, theres no local decoupling capacitors close to the op-amps pins etc.

I'm going to try bypassing the op-amp same as forum member h3d.

Uk member Pinkfloyd kindly gave me his CD33, he started a few mods including replacing the 1N4007 diodes with UF4007's,  replaced the pair of 0.22uf caps after the op-amp to 2.2uf Epcos films and a few other capacitor changes.
Mods I have done so far is replace the heater supply diodes to 1N5820 and  the 4700uf 16v smoothing cap before the regulator to a Panasonic FA.
I have replaced the dacs decoupling caps and dac bias cap to Rubycon ZA,  the XO clock has been upgraded to a low noise super regged type run from a separate mini psu
kk_ho's mod of replacing the 75v zener in the psu to 100v

Hi Guys,

I'll try and post some more closer pictures later,  nothing much to see though really. Also I have only had this CDP for a few days so not had much time to try everything yet :wink:
yesterday I bypassed the op-amp,  the sound is much better in my opinion,  I did the same as posted here by h3d (thanks h3d)  http://www.hififever.com/forum/viewtopic.php?t=952&postdays=0&postorder=asc&start=250
it shows a diagram of the original circuit at the top and modified at the bottom,  notice the resistor changes too for the tubes

The details for the Zener mod are here, http://www.hififever.com/forum/viewtopic.php?t=1332   kk_ho  is correct regarding the roll off,  this mod improves High freguency  (thanks kk_ho)
I changed the pair of 100v capacitors to higher voltage types

Next I'm going to remove all the op-amps supply because its no longer used,  I'm also going to remove the dacs regulation circuit and replace with a better low noise ALW super regulation

I have modded a Cayin CDT-17a,   its not my cdp but I modded it for a friend,  the sound is quite good
I fitted Burson discrete op-amps and low noise Clock circuit,  tubes was changed to Mullard ECC88

Are these any better Alfcat :wink:

Clock circuit etc

I don't have a picture of the bypassed op-amp yet but heres a picture with notes, only do these mods if you are confident

Heres the PSU showing Zener and a few other bits,  I have a lot more planned to do on this cdp,  I'll add more when I have the time

Hi Cat,

Yes I live in the UK, this is the only forum I've found that has a modding thread for the CD33.

The clocks supply is DC, I fitted a rectifier and smoothing cap on the same bit of vero board for the small transformer.
The PFM XO low noise regulator is very good!  

I have two diy dacs I use as a reference
http://pic20.picturetrail.com/VOL1293/4634025/12749957/271121637.jpg
http://pic20.picturetrail.com/VOL1293/4634025/12749957/225303858.jpg
http://pic20.picturetrail.com/VOL1293/4634025/12749957/245458821.jpg

[quote:4e901741be="alfcat"]Dear t.,

I followed your link to the Regulator page ... but I have some difficulties understanding it.  The thing comes as a DIY project? So how can we order the parts? I was in kind of hurry today when I visited the link and as I can't get much out of it, I lost my patience.

You can take a look at the last photo I posted. I changed all the large caps in the PS part to Black Gate already. That is how I boosted the S/N ratio from 65db to 75db or so.

I haven't got any Zener at home. Viewer had promised to send me one tomorrow to try out.

All the original caps I used in the wrecked PCB are now in their original places. However, I realised that my CD33 is not performing as good as it used to be. Really strange and disappointing.

BTW, is that your cat you are using as the avatar?[/quote]

Hi Cat,

Yes, the Flea regulator is a diy type, I ordered all the parts for mine from Farnell.
Now the CD33 op-amp stage has been bypassed its psu regulator circuit is not needed so I may try removing the 15v regs and run the clock circuit from there, I'll post pictures once I've tried it.

You will notice I do not use Blackgate caps, the reason being I do not like them, they sound too Euphonic/fake  to my ears,   I've tried all the grades from standard to NX Hi-Q,  even after weeks of burn in they are not to my taste.
If you improve the regulation etc that will give far better results than BG caps  :wink:   

Give your CD33 a few days to run in, it should improve,  did you apply lots of heat using your soldering iron when you replaced the capacitors?
Capacitors internal chemicals can change when the legs are heated, it can take a while for them to settle down.
Also be careful with using lots of Sanyo Oscon capacitors, their ultra low esr can make the output of regulators go unstable causing oscillation

BTW yes, that is my cat in the Avatar

[quote:aed370592b="viewer"]HI, t.,

i buy a zener in4764a 100v 1w today.
the data sheet is enclosed for your reference.
can i use this replace the 75v zener of cd33's power supply part.

many thanks.[/quote]

Those should be ok,  measure the voltage going to the pair of 100v rubycon ZL capacitors, make sure the voltage across these caps does not exceed their rating,   mine measured 97v,  a little under the max rating,  notice the tolerance  in the Zeners datasheet states they can range from 95v to 105v .

Please be careful,  remember we are dealing with highish voltage :wink:

OPA627 is NOT suitable to be used in the CD33 without modification, the op-amp decoupling is not good,  have you looked at the supply rails on a scope when using OPA627?  

I've now fitted a VBE modified ALW super regulator to supply the dac.
Op-amp and op-amp regulation removed

[quote:524276e859="WINDWSS"]Hi T,

How the epcos cap sound characteristic compare to Vishay mkp & BC mkp??

the yellow colour film cap you use is what brand?

ss[/quote]

Hi WINDWSS,

I also like Vishay mkp and BC mkp caps too,  the guy who owned the Cd33 before me fitted the 2.2uf Epcos,  I feel these are good caps so will keep them.
I did compare them against Audionote copper PIO, Auricap.   I liked the Epcos better,  just my opinion though :wink:   

Those yellow capacitors are 10uf Arcotronic MKP C.4G ,  these was the only type I had in my spares box in 10uf

I now have a favour to ask you guys   I still have the original 6N3 tubes fitted, has anybody tried those Russian 2C51 or 6N3P-E?  I will change the tubes after most of the mods for the electronics and was wondering what types make a worthwhile sound upgrade over the Chinese 6N3's

You guys should try bypassing the op-amp,  it saves having to worry about  what op-amps to use

[quote:6f822bafd0="WINDWSS"]Hi T,

Here's some brief info on the 6N3/2C51/5670 tube which i had try in my system. (Bada 6N3 CDP, audiolab 8000A (mod with BG cap & jensen big tank, AE AEGIS1 speaker (mod with vishay mkp &mkt + damping)

Tube tested:
1) Eclipse pioneer 2c51 (later known as bendix) - neutral, good detail & good bass - suit instrument lover. suit bookshelf speaker who need more bass. sound not as fast/crispy as tungsol.

2) Bendix 2c51 - characteristic similar as clipse pioneer but more dull/dark, slightly less musical & detail

3) Tungsol 2c51- very good detail & cripspy at HF & MID but bass is slightly less
4) WE 396A - more emphasis on mid (vocal), very musical, suit vocal lover but may not suit instrument lover like me

5) GE5670 - Soft, less detail, nice sweet vocal - suit vocal lover too but less detail than 396A

6) united electronic - dark, not musical...can't impress me so far...
5) GE 5670[/quote]

Thank you WINDWSS, thats a big help 8)   
Nice write up too

[quote:6bd60c2e58="alfcat"]Dear WINDWSS,

You mentioned about a "probe". Is it something that looks like a "hook"? I just have a very cheap multi-meter and it comes with two pins only. Can I use those? Will it be a bit dangerous if I just use those?

I have to make something clear first before I can go ahead and check the voltage for you. I need to turn the CD 33 on but no need to play any CD, right? But then do I have to remove the vacuum tubes from their sockets?

Please advise this amatuer. Thanks.

CAT[/quote]

Yes,  all you need to do is measure the voltage across those pins using your multimeter,  your cheap multimeter is fine.
Your probes/hook  should have a plastic handle which you hold, one black and one red?
use the black one going to a ground connection  and the red one to each of those pins,  simply write down what each of those pins measure.
Set your multimeter to read DC

Hi WINDWSS,

My voltages are below with 100v zener,  remember there is dropper resistors on the Anode supply before the Anode pins, I also have adjusted the resistor on cathodes because the op-amps have been bypassed

Grid 8.5v
Cathode 10.3v
Anodes 54v
filament 5.8v

[quote:eb98bdbed6="WINDWSS"]
Dear T,
then your CD33 circuit is a lot diferent from my BADA HD21. wehen my voltatge at zener =100v, my anode is the same =100v, grid & cothode will increase to ~45+.

ur cd33 is using 2 tube, is they use 1 tube for 1 channel?
my bada using only 1 tube (use for 2 channel)

when it zener =75v, gtid & cathode~28+v,

so the differential voltage is ~ 54-10.3 = 43.7v

my different voltage is ~47V[/quote]

Theres 2 x tubes in the CD33,  I have had a closer look and it seems that one tube is for the first stage supplying both Left and Right and the second tube is for the second stage Left and Right.
My measurements was for the tube in the first stage,  it has dropper resistors before the Anode.
Here is the measurements for the second tube

Anode 99/99V
Cathode 52/54.6V
Grid  51.8/54.3V
Filament 5.8v

Basically  each half of the triode is for Left channel  and other half of triode for the right channel,  this second tube has different voltages compared to the other tube in the first stage

[quote:0e81a2a876="alfcat"]Dear Friends,

To avoid confusion of which pin is which one, I drew an ugly diagram to show my final findings.

I TRIPLE check this time before I take down the readings. The readings in PURPLE are mine. The readings in RED are my friend's CD 33. The main difference is his Zener is the Original One while I changed mine to 100V.

What do you guys think?

Can anyone explain why the two TUBES got different readings? The most amazing part is one has FIVE readings but the one closer to the DAC only got THREE! Aren't they doing the SAME job for EACH channel?

CAT[/quote]

See my above post, if it makes no sense I'll try and explain better,  the bottom tube supplies left and right signal which goes to the top tube, the top tube also supplies left and right signal to the coupling capacitors

Your voltages do not look right,  we need to find the problem.
You seem to have some voltages missing,  
your filament voltages are ok
Your cathode voltages are missing
Your Grid voltages on top tube look too high and the Grid voltages on bottom tube are missing
Your Anode voltages look fine

[quote:57f0b07e77="WINDWSS"]T, VIEWER,

U got any idea on the nipon chimi con oil cap in the picture (black)?
it is a film cap or electrolytic cap?

anyone kno0w the sound quality of this cap?[/quote]

I'm not 100% sure but the black capacitor looks like a PIO  ( I maybe wrong)
Also notice it has a silver stripe on the right?   this is normally the side where the internal foil is connected.

Please be careful with the old PIO,  sometimes they can leak DC voltage if used as a signal coupling capacitor

Heres voltages for both of my tubes

Hi Cat,

I'll have a think and try to post some things for you to try later, its 1:55AM  here and I really must go to bed

A schematic would be a very big help regarding fault finding so I hope somebody can find one.
I remember you saying in another post that your CD33 does not sound as good as it did using this new pcb,  I think this is because of those voltages :wink:

On a positive note my modified CD33 sounds beautiful!  lets hope we can sort yours out!

Just a quick question,  did you try fitting supplied Chinese tubes and measuring the voltages?

Excellent mr Cat     those voltages are much better!  I now understand why you had 0v for a  lot of the pins,  without the tubes you only get the Anode and Fillament voltages.

As far as I can see your voltages now seem fine,  reason being my voltages are higher for pins 3&7  is because I adjusted the cathode resistors to compensate for the use  bypassing the op-amp.

Your 5.6v reading for the fillament is fine

1:10am  and I'm still awake again!

Your Bendix tubes maybe drawing more current on the fillaments, make sure the fillament regulator heatsink is not getting too hot  (tallest heatsink on the PSU board)

Hi Cat,

No problem;) pleased to hear I could help.

I'm thinking what mod to try next,  theres a few more things I have planned.
I'm tempted to try hardwiring different tubes,  maybe something more popular than the 6n3

The fillament supply in the Cd33 is done by a simple fixed 6v regulator.
You could maybe try a new 6v reg for the easiest or built a circuit using adajustable LM317  (LM317 is NOT a direct replacement and the circuit would need to be altered)
May also be worth fitting a taller heatsink anyway as a precaution

The super regs and just pcb's I use for the dac chip are available from here  http://www.at-view.co.uk/alwsr.htm

You will notice mine looks different and uses added components,  I removed the LM317 pre reg and implemented a VBE gyrator using a darlington arrangement,  it improves noise rejection

[quote:ca8466be6f="alfcat"]Hi friends,

Can you tell me WHERE is that 6V Regulator? Ha..ha.. I don't know where it is located on the PCB ....  :cry:[/quote]

http://www.hififever.com/forum/download.php?id=7690
Top left of board, its the one using the tallest heatsink,  it should be marked as 7806
In between big caps,  another hint is to change BOTTOM 4700uf  capacitor to a 8200uf 16v panasonic FC.
Don't replace the top 4700uf capacitor with a low ESR :wink:  4700uf Elna Starget or 4700uf blackgate is ok for the top one

[quote:5a566d74f1="WINDWSS"]Hi, T,

In case Cat, replace the 6V regulator, it is still the giving 5.6V, he could just use a 7v regulator with a 6V/6.2V zener diode right?[/quote]

Yes,  just as long as the 6.2v zener diode can handle the fillament current

[quote:0f799a10ed="alfcat"]Dear t.,

Oh, the 7806 is a regulator .... I see.

What do you suggest I change to? And you said I should change to a larger Heatsink for it, right?

CAT[/quote]

Maybe something like these which can handle 1.5A current http://uk.farnell.com/jsp/Semiconductors/power+control/STMICROELECTRONICS/L7806CV/displayProduct.jsp?sku=9756094

You may still get more of a voltage drop when using the Bendix  though.
If you feel the 7806 is getting too hot using the Bendix its worth replacing it with a taller heatsink to help lower temperature, just make sure it will physically fit

[quote:2dc6315ea3="alfcat"]Dear t,

Do you have more photos of your cat? It looks like we have some cat fans here as well.

KK Ho likes cats too.[/quote]

Heres a few  

http://pic20.picturetrail.com/VOL1293/4634025/12749957/286233568.jpg
http://pic20.picturetrail.com/VOL1293/4634025/12749957/286234015.jpg
http://pic20.picturetrail.com/VOL1293/4634025/12749957/286233565.jpg