The Phonoclone and VSPS PCB Help Desk

[rjm]

Since I already use Gmail, I had a quick look at Sites. I don't think this will work for the simple reason that I don't appear to be able to upload content.

I have the whole website already. Basically I'd just want to ftp the files over to a web host server and register a domain name. Maintenance via ftp and the good'ol unix account will be fine.

If I'm missing something with what Google has to offer please let me know. I never found an "upload HTML file" button... or any equivalent to the Geocities file manager.

/R

[Thzar]

Subject: VSPS

Howdy.
I think I need alot of help.

A layout pic (mspaint)
Its a LM4562



Those input caps 220pf I added before and got rid off all the noise and hum, they did something with the VSPS.
I hooked it up with a Fisher turntable and nothing. Deadsilent when it suppose to play some Enya.
Checked and rechecked. nothing.
I remove the input caps and still nothing. =(

Some new measurement. Crazy
input shorted: 2000mV
input open 8000mV

I declare it R.I.P for now. =(

[Thzar]

Sorry.

Too much troubleshooting is not good for your health. eheh.

I desolder those input caps, but nothing changed. =(

After some time with frustration, I found the problem that cause the high dc offset.(8000mVdc on the output.)
A microscopic space between the power rail and the power pin on the LM4562 did that. But now I dont know if the input caps where any good.

All the changes makes the pcb card looks more like a warzone every time I make a change

The dc offset is back to "normal" 50mV and 210mV: shorted input.
And I hooked up the Fisher turntable again and I can hear Enya =)

There is one thing I want to know, why did the input caps cause the dc offset to rise to 1822 mV on both channels.

soo long...

[northernsky]

I'd like to move it to somewhere where I could have my own web address, but doesn't cost very much.

Hi rjm,

my provider Tiscali.it, gives for free email address and 100Mb webspace.
It is an Italian site (the language should be a problem for you), but there is also the British site www.tiscali.co.uk and it seems it does the same thing: for free email and 100Mb web space


Renato

[rjm]

Originally posted by Thzar
After some time with frustration, I found the problem that cause the high dc offset.(8000mVdc on the output.)
A microscopic space between the power rail and the power pin on the LM4562 did that.

There is one thing I want to know, why did the input caps cause the dc offset to rise to 1822 mV on both channels.

The op amp was probably oscillating, and the DC offsets got jacked up as a result. That can happen.

Next time save yourself some frustration and just buy the boards / parts from SkyCoral. Much cleaner.

[Thzar]

Hey Rjm.
1 pcs of VSPS pcb for 7 USD, wow thats cheap. =)
Its the same price I payed for a naked pcb with "coppar islands"

"The op amp was probably oscillating "
But the output were dead silent. no hum, noise or crackling sound whats so ever.

I think I will try the input caps again and I gonna lower the gain a little.

Thanks.

[rjm]

Originally posted by Thzar
"The op amp was probably oscillating "
But the output were dead silent. no hum, noise or crackling sound whats so ever.

In that case the op-amp is almost certainly dead.

[busoni]

Almost finished with my VSPS, will post photos soon.

I used the board from Skycoral, however I made some changes, most of which were clearly audible:


1. changed the NE5532 to LM4562. This was the first one to change and made the VSPS sound interesting. NE5532 was not bad, but a bit dull.

2. added 47uF electrolytics as coupling capacitors from opamp V+/V- to power supply.

3. used 2 12V, 1.2A batteries instead of building a power supply. I cannot commend on impacts of this change since I have not tried a power supply. However, I suppose that this is the easiest way to provide clean power to the opamp.

4. changed the output capacitor(a Nichicon 4.7u, non-polar electrolytic) to 4x1.0uF russian, millitary PIOs. This provided the biggest improvement, after changing the opamp! Dynamics are much better, voices sound crisp and naturall, highs are smooth.

5. biased the opamp to class-A by adding a current source, i.e. a 2SK170 with gate and source shorted, between the opamp output and V- in each channel. Not much improvement, the sound became a little "rich" and the opamp much warmer!

Still, there's room for more and serious improvements. I plan to change the RIAA MKT caps to silver micas, add input caps to match the cartridge's output capacitance, improve stability of the current source.

And, last but not least, fight humming which, started puzzling me again. I plan to separate power ground from signal ground. Has anyone tested this?

[rjm]

Originally posted by busoni
Almost finished with my VSPS, will post photos soon.

I used the board from Skycoral, however I made some changes, most of which were clearly audible:


1. changed the NE5532 to LM4562. This was the first one to change and made the VSPS sound interesting. NE5532 was not bad, but a bit dull.

2. added 47uF electrolytics as coupling capacitors from opamp V+/V- to power supply.

3. used 2 12V, 1.2A batteries instead of building a power supply. I cannot commend on impacts of this change since I have not tried a power supply. However, I suppose that this is the easiest way to provide clean power to the opamp.

4. changed the output capacitor(a Nichicon 4.7u, non-polar electrolytic) to 4x1.0uF russian, millitary PIOs. This provided the biggest improvement, after changing the opamp! Dynamics are much better, voices sound crisp and naturall, highs are smooth.

5. biased the opamp to class-A by adding a current source, i.e. a 2SK170 with gate and source shorted, between the opamp output and V- in each channel. Not much improvement, the sound became a little "rich" and the opamp much warmer!

Still, there's room for more and serious improvements. I plan to change the RIAA MKT caps to silver micas, add input caps to match the cartridge's output capacitance, improve stability of the current source.

And, last but not least, fight humming which, started puzzling me again. I plan to separate power ground from signal ground. Has anyone tested this?

1. LM4562 wasn't around when the VSPS was first drawn up. It would seem to be an excellent choice.

2. Ideally 100uF would be soldered from the power pins to the out- wirepads, and C6,7 would be left off the board. This is the main problem with th current layout; C6,7 are too far away from the op amp.

3. Batteries don't sound as good as a good linear supply.

4. 2x1.0uF should be sufficient. I suspect the Nichicon nonpolar sounds rather poor, though its OK in a pinch.

Hum? With a battery supply, the issue would have to be related to the turntable motor... signal, power grounds are already separated on the SkyCoral layout, i.e. already correct for a noninverting layout.

Thanks for your detailed comments.

Richard

[busoni]

Originally posted by rjm


1. LM4562 wasn't around when the VSPS was first drawn up. It would seem to be an excellent choice.

2. Ideally 100uF would be soldered from the power pins to the out- wirepads, and C6,7 would be left off the board. This is the main problem with th current layout; C6,7 are too far away from the op amp.

3. Batteries don't sound as good as a good linear supply.

4. 2x1.0uF should be sufficient. I suspect the Nichicon nonpolar sounds rather poor, though its OK in a pinch.

Hum? With a battery supply, the issue would have to be related to the turntable motor... signal, power grounds are already separated on the SkyCoral layout, i.e. already correct for a noninverting layout.

Thanks for your detailed comments.

Richard

Thanks for the reply Richard

So you think that I should invest in a good power supply? I'll certainly check it. Would an R-core transformer be better than a torroidal?

I'll also test 2.0uF, maybe less capacitance in combination with the output resistance. My intention is to try teflon caps which are available in smaller capacitances, in combination with the output resistance in order to keep the frequency of this output filter as low as possible. Input resistance of my opamp is around 200k, so I suppose there's still room to increase resistance and lower capacitance in the VSPS output. Nichicon does not sound bad at first. I didn't give it time to evaluate, so I can not comment more. But compared to the PIOs, what was "interesting" in sound now becomes "involving".

I haven't noticed separate grounding on the skycoral board. I'll check it again. Thanks for pointing out.