WD Phono 3 - DC on Output?

[moley]

Hi Andy,

Thanks for the reply.

I've got 2.2UF clarity cap MR polyprops on the output currently. I get a similar reading on both outputs, it would be quite unlucky for both caps to be leaking I guess?

Perhaps I need to try with a different meter, maybe the readings are erroneous. Also I have R27 giving a weird reading of 50K instead of 1Mohm while R28 is fine at 1Mohm. I'm going to remove the board and test from underneath to see if R27 still doesn't measure right.

I will also try your suggestion of a second cap in series at the output.

[moley]

Sorry I only just saw your second post thanks for that.
I had the dummy load resistor still in place! R27 and R28 now both reading 1Mohm.

[bikerhifinut]

The idea of a second cap was only to eliminate the possibility of leaky output caps. I wouldnt do anything other than put one end of a known good cap on the positive meter probe and the other on the live pin of a phono plug inserted in the output socket, and really its a last resort test.
Measure your 1 meg resistor out of circuit if possible, then you know for sure, but for the 10p one costs I'd put another one in anyway. The value isnt massively critical its there to bleed residual charge to earth that gets stored in the 2.2uf output DC blocker, thats quite a large value so when you switch off there can be most of the 120 or so volts on the cathode stored up in there.
it has no real effect on the output resistance of the Phono3, that is set by the output impedance of the cathode follower in series with the 1k cathode stopper.
It will have a slight shunting effect as its in parallel with the input resistance of the next stage, which is quite possibly a 47k potentiometer, negligible in other words.
I think others have advised you to try a different output cap. that will eliminate any likelihood of leaky 2.2uF outp[ut caps, i'd just stick a cheerful 0.1uf or similar value plastic film cap in there and measure for DC leakage. Then If necesarry i'd look for something else to replace your clarity caps if they are dodgy.
What I would do here is insert "veropins" in the pcb holes where this cap lives and thus swapping different ones in and out is easier.
Bob and I both really like the original "Orange drops" and its worth noting a 250V rated one will be fine on the cathode but 350V or above are better practice.

A.

[Phil Y]

Hello Moley,

I really don't believe you have any DC leakage.
The 'phono' side of your o/p cap sits at about +150v or so, yet you are measuring a small + AND - reading on the op side. If the cap is leaking, how can it possibly be leaking a negative voltage from its +150v side to its output side????

I still think you are seeing low frequency AC on the output(cause not yet identified). Your meter can not tell the difference, don't think that because your meter is on the DC range that you can ONLY be seeing DC.

Think about a really low frequency AC signal for a moment, say 2 minutes per cycle. It will be on the +ve side of the 0V line for a whole minute, then on the -ve side for a whole minute. What will you see if you connect your DC meter when it is half way through the +ve half cycle? A slowly changing +ve DC voltage. It is still an AC signal.


Phil.

[bikerhifinut]

Originally Posted by Phil Y

Hello Moley,

I really don't believe you have any DC leakage.
The 'phono' side of your o/p cap sits at about +150v or so, yet you are measuring a small + AND - reading on the op side. If the cap is leaking, how can it possibly be leaking a negative voltage from its +150v side to its output side????

I still think you are seeing low frequency AC on the output(cause not yet identified). Your meter can not tell the difference, don't think that because your meter is on the DC range that you can ONLY be seeing DC.

Think about a really low frequency AC signal for a moment, say 2 minutes per cycle. It will be on the +ve side of the 0V line for a whole minute, then on the -ve side for a whole minute. What will you see if you connect your DC meter when it is half way through the +ve half cycle? A slowly changing +ve DC voltage. It is still an AC signal.


Phil.

Think you're right Phil.
I had a lot of LF instability issues with my phono3S. I'd be looking at the PSU as the cause. In the end i cured mine by the rather drastic means of building a regulated HT PSU. But if its really really low and its not affecting the sound then maybe just ignore it? Or use a smaller dc blocking cap on the output?

A.

[Phil Y]

Yes A,

That is what I suggested in post 11. Not necessarily instability as such.
A 250v unregulated supply drifting around by a few tens of mV sounds entirely reasonable to me and then just getting coupled through the cathode follower stage to the o/p.

It would be interesting to scope the HT and the o/p at the same time to see if there is any correlation.

Phil.

[moley]

Thanks for the replies.

I put a 4.3UF in series with the output using clip test leads from the RCA interconnect. The 2.2uF is still in there for the time being as I haven't had time to do much more than test everything so far. The dummy load resistor I used was a 55K.

Measuring at switch on the spike appears though it is lower at around 20volts DC. Once this has settled down I get the changing reading of -20mV to 50mV.

Once I get chance I will change that 2.2uF and put in a lower value, perhaps a 0.22uF.

It would surely mean that there is no DC leakage given the addition of a second cap? This would support the idea that it is AC that my meter is picking up.

Checking this on a scope was also mentioned earlier in the thread. I don't own one and have not used one before. Would it be a good idea for me to buy one? There appears to be a range of fairly cheap options on ebay at around £20. Would it be a good idea to pick one of these up or are they likely to be useless?

Would it potentially help changing the voltage regulator from the standard LM1084 to LT1084CT-5#PBF, (Low Noise LDO Voltage Regulator, 5A, 4.9 to 5.1 V, TO-220 3-Pin) as some on this forum appear to have done?

I have one in my parts box I think I bought it a while ago with a view to modding my PSU III but never got around to it. I can't seem to find the thread about this now but I recall a mention of putting tantalum bead caps across the input and output of the 1084T regulator also. I have some 1µF/35V tantalum bead caps in the bag with my LT1084 as I must have kept them together for the job.

[John Caswell]

Hi all,
This in not DC leakage through the capacitor but the normal drift due to the ac signal flapping up and down in the phono.
If you view the output on a scope set to ac and high gain ie 5mV/div with no signal input and say a 50K resistor across the input for MM or 100R for MC then you should see the traces flapping about a bit due to Johnson and shot noise in the circuitry. It is possible to reduce this by - careful selection of the ECC83/12AX7/5751 valves, changing some of the resistors on the input circuitry to metal film/bulk foil(gulp at the prices)
If you change the scope input to dc then the scope traces with drift up/down in sympathy with the shot/Johnson noise.
You won't need to change the output capacitor leave it at 2µ2F, and leave the pull down resistor on the output at 1MΩ
A regulated HT supply is not really necessary as the current draw is very low and virtually constant.
The LDO regulator for the heaters was changed to alleviate hum due to regulator "drop out" in low voltage areas. Changing to it will not harm and you can then set the heater voltage to 6.3V easily. The tantalum beads were normally added as some of the 3 leg regulators had a tendency to oscillate at around 55MHz. Have a look for the data sheet on the web, as it gives some good ideas re hum reduction.
I think you are worrying a little too much about DC on the output.
PhilY has said pretty well all there is to add to this, just sit back and enjoy the music.

John

John

[bob orbell]

Well there we have it from the man that know's, do take John's advice Moley, he is a professional, and a very nice man, . BOB

[Richard]

Hi John, all,

Yes we'd concluded it was an AC signal not a DC offset.

What we haven't established is,

if it is measurable with a meter as Moley finds, will it be be passed through a DC coupled power amp he plans to build

and can it be made stable or blocked more effectively?