Crack Upgrade Path?

[nsorens]

Thanks, I don't know why I didn't think of trying Digikey!  Got one coming now...

[Dr. Toobz]

I couldn't resist.

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fimg532.imageshack.us%2Fimg532%2F7758%2Fcrackleftat1v1khzoutwit.jpg&hash=10a82d9b7c33db876d4924420dbfd260872377a4)

This is again at 1V rms out - Crack with Speedball upgrade left channel blue - control red.

S/N improved to better than -70dB. THD improved to just above 0.1%. 2nd still dominant, 3rd at roughly -90dB, all higher harmonics buried in the noise, power supply completely inaudible.

That's only part of the story really, because the music just sounds incredibly "right".

I would have been curious to see a graph of the baseline Quickie before any mods. How much do you think the noisefloor and THD dropped once using CCS's as plate loads?

[Doc B.]

Look at reply #20 in this thread.

[Jim R.]

Grufty,

Can you comment on the speedball vs your PSU tweaks?  I know you got some nice results with the extra RC stage and resistor value changes, but how did that compare to the speedball?

I ask because I know that once a couple of friends hear my Crack, they're going to want one for themselves but they probably won't want to spend too much on upgrades.

Thanks,

Jim

[Jim R.]

Woo-hoo, just found a dual 100k pec pot in my "junk" box.  Definitely will put this in the Crack and save the expensive TKD for the stereomour.  Man, my junk box hs been very good to me with this project -- 4 black gate NH 150 uF 350v (two matched to within 2%), the matched pair of 2.7k caddocks, and now the PEC pot, oh, and a pair of Vampire gold over copper RCAs.

[Dr. Toobz]

Look at reply #20 in this thread.

Ha! I knew I thought there was such a graph posted already.....

[grufti]

I didn't have a lot of time with just the PSU change, but I was very happy with the improvement considering the minimal effort and the minimal price.

I paralleled the kit resistors, because it was quicker and easier, but if you build with this change in mind right from the start, then you can just go with three 150 Ohm resistors and the extra cap. The B+ voltage will be about 10 V higher than stock, but that doesn't seem to hurt.

When I first turned on the Crack after the PSU change I had to check twice if it was on at all. It was so quiet.

The combination of better signal to noise ratio and very nice final cap in the PSU also work a lot of magic on the music presentation. I felt much more engaged, drawn in, happy to be listening.

For less than $10 ... any day.

Grufty,

Can you comment on the speedball vs your PSU tweaks?  I know you got some nice results with the extra RC stage and resistor value changes, but how did that compare to the speedball?

I ask because I know that once a couple of friends hear my Crack, they're going to want one for themselves but they probably won't want to spend too much on upgrades.

Thanks,

Jim

[grufti]

I have finally entered the land of diminishing returns along the Crack upgrade path.

The latest status is mostly due to tube rolling. I found a very nice E80CC tube and a used 5998 that was both cheap and probably never used. The only other change from my previous post is that I biased the 12.6V DC heater current.

The output voltage is still 1V rms. The measurement setup is unchanged. This time I'll post both channels - the left channel has always behaved marginally better than the right channel, but the difference is really academic. The only harmonic distortion that matters at this point is just a hint of 2nd. There is no noise to speak of considering that this is a tube circuit.

Left:

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fimg143.imageshack.us%2Fimg143%2F7686%2Fcrack5998pq1vleft.jpg&hash=fa196067667406068ef28abcd1c465c31987da7d)


Right:

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fimg641.imageshack.us%2Fimg641%2F3475%2Fcrack5998pq1vright.jpg&hash=9150f780870c1e468fb042d38bb9d87322b63afa)



I am not really good at describing how these "upgrades" translate into a better listening experience. The end result to me is possibly best described as: great intensity combined with a deep sense of calm, however contradictory that may sound.

[Grainger49]

Grufti,

It only matters that you like what you have done. 

I would have thought that the graphs would show some increasing noise floor as you got below 100 Hz.  Some tilt, but it is essentially flat.  There is also no tilt in the high frequencies, no tube rush.  It certainly indicates you have a very clean build.

I see no hum sneaking in the right channel.  The "misbehaving" left channel shows a repeating 60 Hz component.  And that makes no sense.  The right and left channel are the same power supply.  For the most part they should have the same noise.  Maybe there an AC lead to the left channel that is a little too close to the audio signal wire? 

Nice graphs!  I like it!

[grufti]

At this extremely low level most of the channel differences are likely to originate in the E80CC tube itself. Some of the original data sheets have a cautionary remark telling you that one of the tube sections will have better noise and THD performance. The E80CC tube that I'm currently using actually stands out for best measured performance from the "other" side.

I certainly like what I'm hearing. You might want to go back and follow the progression from the first to the last posted measurements. There is quite a difference.

And mind you, all of my other "upgrades" so far have cost about as much as the Speedball. So I'm still below $400 total right now [$199 Crack, $99 Speedball, less than $100 everything else ... no fancy $1,000 power cords], with the help of the introductory pricing for the Basic Crack kit and the minimal outlay for a few parts under the hood and two different tubes on top.

[jrihs]

A pic of the Goldpoint Mini-V compared to stock.  The chassis plate hole will need to be enlarged and the 1.1" knob just barely covers the small hole to the right of the pot hole.  Should have gone with the next size larger knob. 

Does the stock knob fit?

[jrihs]

Now that I got one I can answer my own post. No, the goldpt post is to long. which is to bad...I like the stock knob with the finishes I went with...:-(

[STURMJ]

use a Dremel or some sort of grinder to shorten the shaft.  Make sure to cover the entry of the shaft, and any other areas where metal debris might enter the wiper mechanism with tape. You can make the shaft to fit the knob you want to use.

[jrihs]

Yes, I am considering using my drumel...a scary proposition considering the cost. Guess you just go real slooow. I was wondering why they would make it so long, then I saw that they sell knobs!

[sbelyo]

In the power supply, fitting three 220uf film caps will be tough. Instead, run the stock first and second caps with the resistor between them, then use a Triad C7X and a 10-20uf cap after. This will impart 90 percent of the benefits of big film caps without the crazy space requirements. The C7X should knock the noise down enough to get away with this.

I think I want to try this.  I've got a nice asc 20uf motor run cap that would work.  I've always wanted to try them in the signal path.  I'll just have to pick up the choke.

For the first mod $15 bucks is cheaper tham film caps for the output or the speedball.  Of course both of those things will eventually end up in the circuit anyway...