Kel84 and a source with a small dc offset. Problem?

[Dwjames]

Hi all. I've built a really good sounding DAC to use with my kel84 amp. I'm looking at dc blocking output coupling caps for it and the concensus is that really expensive ones sound best.... but I'm wondering if I need them at all? After all, the best sounding solution should be no cap at all.

The DC offset from the DAC signal is stable at less than 200mv.
Now I can see that the kel84 doesn't have dc blocking coupling caps on the input, but it does have them between the ecf80 and el84 sections, so I'm happy that this dc offset won't make it to the el84 output stage and on to my speakers, but will a dc offset of this size effect the first stage of the circuit in any way that will cause me a problem or degrade the sound in any way?

Thanks,
James

[bikerhifinut]

Originally Posted by Dwjames

Hi all. I've built a really good sounding DAC to use with my kel84 amp. I'm looking at dc blocking output coupling caps for it and the concensus is that really expensive ones sound best.... but I'm wondering if I need them at all? After all, the best sounding solution should be no cap at all.

The DC offset from the DAC signal is stable at less than 200mv.
Now I can see that the kel84 doesn't have dc blocking coupling caps on the input, but it does have them between the ecf80 and el84 sections, so I'm happy that this dc offset won't make it to the el84 output stage and on to my speakers, but will a dc offset of this size effect the first stage of the circuit in any way that will cause me a problem or degrade the sound in any way?

Thanks,
James

200mV? Thats a LOT. thats close to 2/3 of the sensitivity of a KEL84.
It's not a good idea to have any DC on the grid of a valve. I'd have some DC blockers on the output of your DAC sharpish. I don't know what the circuit of the analogue output is but that does seem an awful big dc component.
Value of the blockers will depend on the output resistance of your DAC and the input resistance of the amplifier you are feeding. I'd say around 2 to 10uF would be ok. Just start with something like Orange drops to start with. I never thought much of them until I heard how good my Phono3 is and thats got orange drops on its output.

[Richard]

Hi James, Andy,

If it's pure DC (no AC component) it won't make much difference. Kel84 first stage has 2V bias at the cathode. Putting +0.2V DC bias on the grid is the same as taking 0.2V off the cathode. Or if it is opposite polarisation, putting -0.2V on is like adding 0.2V to the cathode bias. The valve will pass a little more or less current that's all.

Nor will DC get through to your speakers as apart from the following cap coupling the OP stage is TX coupled. (If there is any AC present though this will be amplified as signal and will go through to your speakers.)

It's usual to ensure the source has no DC output so block it there and your DAC will be safe with any input. Typically a 2.2uF polyprop is used giving a coupling down to 1.6Hz into the common 47K impedance but in practice with a valve amp you may find even 0.22uF (16Hz) will block DC without making the bass too light so try anywhere between the two.

[Dwjames]

Thanks Richard, that's great. Yes, the dc part of the signal is pretty pure as it's all ultra low noise regulators feeding the DAC. I currently have 0.44uf paper in oil coupling caps on the DAC output, but I'm going to try without now I know it's safe and see if there's any difference to the sound.
Cheers,
James

[Richard]

Yes give it a try but it may make the vol pot noisy and it's not really good practice. If there is a sound difference I'd try a polypropylene cap as that will sound more neutral than many PIO caps.

I should also have said last post that any DC will be attenuated by the vol control around -20dB = 1/10 level at the typical listening position a little before midway so the offset would then be down to 20mV.

[Ali Tait]

I liked AN coppers in that role, or teflon SCR's.

[John Caswell]

Hi all,
Putting a signal with DC into any amp/preamp is not a good idea, especially when it is directly coupled into the first stage, as when you switch sources there will be a bloody great thump though the speakers, because you are altering the state of the front end in the KEL84.
What I suggest is that in the DAC output feed you place a coupling capacitor and then a bleed resistor to ground to drain off the charge that builds up through the capacitor.

John

[david]

I am using the same DDDac as DW James and can vouch for its great sound quality. I run it into an Aikido cf pre which has a blocking cap before the first valve and I can state I do get a very slight bit of pot (Tocos with shunt mod) noise this way, but no thumps when changing source.

Without caps there is a reasonable improvement in SQ-slam, depth-compared to the supplied Mundorf MKPs.
David

[John Caswell]

Hi all,
At present sound quality is not the problem, putting DC on the input grid of the amplifier is, and as stated by David his is already being DC blocked by the Aikido pre therefore the effect will be ameliorated
I repeat this is not a good idea, on a direct coupled SS amp this would be disastrous, on a valve amp with no input DC blocking a loud thump would pretty well certainly result. This is poor engineering.

John

[Richard]

Yes, poor practice as you say John and I'd definitely use a blocking cap on the output of the dac with 1M to ground following it which should cure noise as you say.

Let's be clear though if James wants to try it to see if the coupling cap does make any difference sound-wise;

Selecting the dac then firing up the amp will not cause a problem with selector thumps. The extra bias (either way) will not be a problem. There may be a bit of noise turning the pot. Then have a listen. If he does find the "straight wire" better than the PIO cap then I think a polyprop will be a better alternative.