His Master's Noise: A Thoroughly Modern Tube Phono Preamp

philips ecc88 2nd page near bottom.

http://drtube.com/datasheets/ecc88-philips1958.pdf



telefunken e88cc 6922 cca 3rd page half way

http://www.r-type.org/pdfs/cca.pdf


electro harmonix 6922eh (just for completeness if anyone else is interested) bottom page 1 and top of page 2.

http://www.thetubestore.com/lib/thetubestore/eh-6922eh.pdf



Siemens d3a page 7

http://www.pearl-hifi.com/06_Lit_Archive/08_Tube_Data/Siemens_D3a_7721.pdf


Im still a bit unsure. Lets take the RCA, Siemens and Telefunken example you found.
limits -60v and +120v

If we elevate the heater to 65v then doesnt one side of the triode or the other fall outside this spec?

Does -60v mean
1. 'the cathode can be up to -60v below the heater'
2. 'the heater can be up to -60v below the cathode'

I think some of my confusion is coming from each data sheet seems to use a different notation and way of expressing this specification.

As we need to elevate the heater. Im guessing it has to be the second option. But this doesnt explain why not elevating the heater would strain the other triode in the ecc88(and variants). Is there a cathode to cathode voltage limit or something?

If we do elevate the heaters, this still leaves the cathode on the D3a more than 60v below the heater, and the sheet says cathode negative max 60v. I find this language confusing. It seems back to front.

Reading between the lines Ive come to this conclusion.

Am i correct if i were to say:

Typically, tubes of this nature can
(a) have the heaters upto ~120v above the cathode
(b) have the heaters no lower than ~50v below the cathode

And therefore when I read the spec sheet, however the spec is labelled, assume:

the higher voltage rating refers to the max heater elevation above the cathode

the lower voltage rating refers to the max difference when the cathode is raised?

Thanks
 
and it gets better.

put the tubes in and the 260v reg sparked at the output. One of the small transistors or maybe the voltage source fell apart and the 8k2 resistors started glowing. Didnt do that with the lightbulb tester!

Oh well. Thats another reg ill have to build then.

I think ive had enough of this for a bit. It all seemed to be going so well...
 
I think I may have done something a bit daft but still not quite sure of the exact mode of failure.

When disconnecting the transformers to test which one was humming the ht supply was taking ages (well over 5 minutes) to bleed out so my bright idea was to whack the umbilical in to add some extra drain. There was a bit(!) of a crack, so i left it a bit. I assumed that it came from the connector.

Im wondering if this weakened either the mspa42 on one side or the lm329 on the other. When plugged everything back in the light bulb limiter was limiting the current and the voltage was considerably lower probably stopping the thing getting too hot. Thinking about it I think the raw b+ was quite a bit lower than when i initially tried it the other day.

As the thing seemed to be working but at a lower voltage I assumed it was still ok like before without the load. When i plugged back in direct to mains the thing popped. Im guessing that the massive drop resistor protected the 160v reg.

I will double check my current draws on the ccs and try again!

Im going to do a post mortem as the failure seems to be different on both sides but yet both 8k2 glowed.

Oh well. I guess this is how one learns!

Yes, the pass resistor is heatsinked and they seem to be ok. Ill replace them anyway as i have a load of them. Just in case.....

At least i can just snip the components of the perf board. No desoldering required (hooray!)

Thanks
 
Thanks jack. Frank had picked up on that. I changed the 68k to 77k (2x150k in parallel selected to give closer to 77k, I got about 76.something) and the 220k to 270k.it was giving 220v before and with these alterations it was spot on 260v before I destroyed them with my all conquering idiocy.
 
I took a look to the datasheets.
The E88CC is a enhanced longlife-version from the ECC88. So the values are unequal.
The datasheet for the CCa is the same which I got. You are looking for maximum ratings.
Better is to use the values for typical operation.
The negativ Ufk is not interresting for us, cause UA is not so great.
If you´r working with heatervoltage lifted on higher Dimensions, 250V, then you can think about the neg. Ufk (most problem with SRPP). So we can save the trouble.
 
I assume you're talking about the ccs and that i don't need to use heatsinks on q1 q3 and q5. I'm already using dn2504 probably the n3-g, that sounds familiar.

I heatsinked the lot as I possibly had heat issues in the impasse preamp and it seems that this preamp is going to generate a lot more heat in a smaller box.
 
It works! Woooooooooooooo. Amps arent even warm yet and it sounds better in all respects! Was feeling a bit down having spent a ton of money on a gyrodec, audiomods and benz cart and being a bit underwhelmed.

The gain for this is spot on for my benz cart at 0.3mV at 3,54cm/s.

Thank you all so much for your help and patience. its really appreciated.

Sadly i will have to get better transformers. Most of the hum is the heater tx but a little underwhelmed that my custom wound hums a little. Any recommendations from the uk? I may also try using a bridge rectifier for the ht on a single wind 250v instead.

Thanks again. Wooooo!!!!
 
Swapped out the heater tx for a torroid that I had kicking about. It doesn't hum so I'll use that for now.

I checked my plate voltage and although the drop across r3 is almost spot on 20v the plate on one D3a is about 140v and the voltage on the other is about 126v. Under load the 260v reg is 260 on one side and 256 on the other.
Any ideas?
 
Hi Dave,
thanks for your mail. You wrote to me, Va = 146V. In your post here, 126V. Whats the correct value?
I would first change the tubes from right to left and then the PS from right to left. So you see, from where the wind is blowing.
Another problem in my view is the connecting of the voltage dividier for the heater to one channel.
A solution should be, to connect the voltage dividier for the heater before the regulated PS. I remember, you got 317V. I will do it in the same way.
You have to calculate new values. I think 100K and 27K are good values. This gives you 70V (and needs 2.5mA). Better is to use greater values. So the current the dividier needs is decreased (470k/120K (64V) 150K(76V) 0.5mA).
 
Right, finally got round to swapping the tubes round.

I think the correct value for the plate on the D3a should be 140v.

Im getting about 138v on one and about 126v on the other. It seems to slowly work its way down to that.

After swapping the D3a tubes around the lower voltage followed the tube.

Is this a sign of a worn or faulty tube?

I remember commenting on one of the other projects that my plate voltage didnt match even though everything else did and there does seem to be a range of acceptable variance but i have no idea what that would be for this.

I measure about 1.18v across the IR diode on both channels. Close enough to each other.

Thanks
 
Administrator
Joined 2004
Paid Member
<snip>
Is this a sign of a worn or faulty tube?

<snip>

Thanks

All is well, this is just an indication of the parametric variation tube to tube, and as the tube warms up and settles down to its warmed up operating point. High transconductance types tend to exhibit quite a lot of variation sample to sample.

In some of my early designs I had to match them to the circuit, I learned not to make that mistake as my stash of D3As shrank, and designed circuits that are less sensitive to parametric variations in the tubes.

Mu is essentially independent of operating point variations because of the CCS loading, so this is not something I would worry over.