McCurdy Phono Pre and Replacement Capacitor questions

[ArchMus]

Hi all.

After prowling around DIYAudio for sometime (and learning quite a bit...thank you) I've decide to post my own dilemma.


A bit of background.

Recently I purchased an old McCurdy phono pre (vintage 70's, possibly early 80's). The design is such that there is a single power supply feeding two independent Left and right circuit boards. The issue that I'm having is signal bleeding from one of the boards to the other. It doesn't happen the other way though. Because of the units age and the signal bleed issue I thought that I'd replace the caps as a make work project on my McCurdy guinea pig.

Each channel has:
2x Nichicon 330uf, 35v Radials
2x Nichicon H7648L 22uf, 35v Radials
1x Siemens 220uf, 63v (Brazilian made) Axial
1x Siemens 100uf, 63v (Brazilian made) Axial
1x Philips 030K0 10uf, 63v Axial
1x Siemens 47uf, 35v (Brazilian made) Axial

under a shielded housing (for the RCA jack, gain & Hi Freq Trim) are:
1x Philips 030K0 10uf, 63v Axial
1x Philips 47uf, 63v Axial


My thoughts.

Options that I found for replacing the axials are:

Vishay BC AML 138
or
Nichicon VX.

And, due to lead spacing on the circuit board, replacing the radials seems to come down to:

Elna Silmic 2
Nichicon ES
Nichicon PX


Questions:

Given two different caps are rated exactly the same, does physical size play any part in longevity, sonics etc...

The original 220/63 & 100/63 have tolerances of -10/+50%. The Vishay and Nichicon have +-20%. Will I be fine with my selection or do I need to find the original tolerances?

Can a cap 330/35 be replaced with a 470/35 or a 330/50?

The Nichicon PX is rated for a very long life, low impedance and a high ripple rating. The Elna's, not as much. Is there a real difference when the parts are in a personal phono stage? Can you get 10, 15 or 20 years out of Elna's?


Whew, that's it for now.

Any suggestions would be greatly appreciated.
Thanks,

Mitch

[SY]

Dumb question- how have you determined that the signal bleed is through the power supply?

A schematic will help for parts recommendations- different caps have different ESRs and a supply with regulator X can be unhappy with caps that are perfect for regulator Y, and both are a different pot of pisces than an unregulated supply. Caps for the raw supply filter will have different requirements than caps to bypass voltage references. And so on and on.

[h_a]

different caps have different ESRs

Hi Sy,

the phono stage is from the 70ies/early 80ies...any standard cap from today will be something like ultra low ESR in comparison to what was standard back then.

Any case, a schematic would be nice

I doubt the signal bleeding comes from the caps, while it certainly will not hurt replacing them. In my experience the caps are almost never a real issue for small signal analog circuits, even when 30 years old. It matters a bit more for PLL-circuits (turntables), though.

Change the caps, it's fun just take care to not overheat the pads. For the old paper boards that quickly means a mess.

Hannes

[ArchMus]

Hi Sy & h_a.

Well I finally got a scan of the schematic. I hope posting a companies schematic is not a problem.

I've circled all of the electrolytic caps that I'm looking at replacing. The schematic is for a single channel. The two channels sit above a single power supply. Since the channels are discreet (as far as I can tell) I speculated that somehow signal was getting to the other channel via the power supply.

Also, I've uploaded a screen shot of my 1k sine wave test tone@-44db that I ran into the right channel (Orange) only. The left channel (purple) activity is some how related to the right channel signal. If I play the test tone in the left channel, the right channel is silent and shows no activity. As you can see there are issues. In the middle graph (Oscilloscope) the purple line should be flat like the green line and the orange line should be a SINE WAVE!

I thought that I'd post a picture of the general layout of the unit. For the pic it's laying on it's side.

Just to complicate things, is it possible to make the RIAA circuit more accurate? The specs claim RIAA +-0.5% Frequency Response from 30Hz to 15Khz. Can +-0.2% be achieved?

[h_a]

Hi,

the lack of scales makes things hard to understand; I guess you refer the -44db to 2V, so your input signal is about 12mV which should be fine.

From the looks of your amp it is from the late 80ies or so as it uses a fiberglass pcb, not paper. This somehow doesn't well connect to your schematic saying copyright 1996-2011?

Any case, please upload somewhere a larger schematic, this is almost impossible to read

Hannes

[PRR]

> schematic saying copyright 1996-2011?

The copyright notice is clearly digitally-stamped over a much older blueprint.

The era is 1970s, though this is conservative broadcast stuff and may have been in production well into the 1980s.

I've never seen broadcast stuff with paper boards. My 1970 RCA had glass-base boards which could have served as armor-plate. My LPB console had a lower grade of board but not the cheezy paper found in consumer junque.

The "signal bleed" is NOT on the boards, since these are one-channel boards with excellent supply rejection. Much more likely in external wiring.

> The specs claim RIAA +-0.5% Frequency Response from 30Hz to 15Khz. Can +-0.2% be achieved?

0.5% is 0.0433db error. This can be a VERY accurate RIAA. There are multiple options to screw it up for more punchy sound or to roll-off scratchy highs.

[ArchMus]

PRR is absolutely correct. I added the copyright, which I got from their web site, just to add clarity of ownership. The manual that the schematic comes from is dated April 1978.

"RIAA +-0.5% Frequency Response from 30Hz to 15Khz. Can +-0.2% be achieved?" is actually supposed to say "RIAA +-0.5db Frequency Response from 30Hz to 15Khz. Can +-0.2db be achieved?"

The reason I thought the problem was in the one of the boards was because if the boards were switched, the problem followed the one board.

I've posted a (hopefully) larger and clearer schematic.

[h_a]

Much more likely in external wiring.

++ I like

0.5% is 0.0433db error.

Actual error is certainly smaller than that as they have to consider manufacturing tolerances etc.

Hannes

PS: where is the large schematic ?

[kevinkr]

I'm just wondering given the extreme complexity of this design whether or not it is really worth the effort? (Lots of transistors and early, poor performing op-amps in the signal path) I suspect one of the reference designs for the LM4562 or similar would be much less complex and run circles around it sonically. If you want something more interesting check out one of the Salas phono pre's.

My apologies for advocating another approach.

[ArchMus]

h_a - I posted a new version of the schematic. I hope it's large enough. It seem fine to me but I zoom in to look at details.

Kevinkr - Ah great! Now I'm looking into replacements for the Harris HA2645's A quick search brought up the Burr-Brown Opa445.

I've also posted photos of the circuit board. The left pic is the circuit under the shielded enclosure (as the schematic says), then the middle section and then the right side of the board.