Stereomour II 45 Conversion - Anyone Do It Yet?

[Deke609]

@PB - I installed the 8K2s in parallel with the 2Ks.

VDC between pins 1 and 4 is now 2.76 (so out of spec; spec is 2.5 +5% tolerance).  What do you suggest I sub for 0.15R's?

Other voltages:

1-  365-366
2-  386
6-  60.6
7-  191.3
14- 192
15- 60.4
19- 386
20- 365.5
21- 61
22- 62.8
24- 59.8
25- 60.8
27- 398
30- 402-404
31- 404-405
34- 398
36- 61.1
37- 60
39- 63.3
40- 61

Many thanks,

Derek

 

[Paul Birkeland]

VDC between pins 1 and 4 is now 2.76 (so out of spec; spec is 2.5 +5% tolerance).  What do you suggest I sub for 0.15R's?

It shouldn't change based on the 8.2K resistor installation, but either way a 0.155R or a 0.16R resistor in place of each 0.15R resistor should get you back down to the proper range.

 

[Deke609]

Thanks PB.

While you were posting, I tried to calculate the value myself, but got the wrong answer. Can you point out where I went wrong?

My reasoning was as follows:

If 2.76V is dropped across a 0.3 ohm resistor (0.15 + 0.15), that means a current of 9.2 amps (that seems high to me).

If I want 2.5V, and R = V/I, then I get a combined R value of 0.27ohms - or 2 * 0.135 ohms.

Many thanks,

Derek

 

[Paul Birkeland]

Now that I think more about it, you did those filament measurements with the 8.2 ohm cathode resistors, so everything coming out of the power transformer would be very low. 

If 2.76V is dropped across a 0.3 ohm resistor (0.15 + 0.15)

The 45 filament is in there too.

You need to drop an extra 0.26V. 45 filament current is 1.5A.  0.26V/1.5A=0.173R.  This is the total addition you need, so just increase each resistor by about half that, or 0.08R. Something like this resistor would work ok:

https://www.mouser.com/ProductDetail/Vishay-Dale/CPF1R16000GNB14?qs=sGAEpiMZZMtlubZbdhIBIN9c7v76BNFlN04yaEgXCqY%3d

Do note that when you move this far down the trail, unlike the Stereomour I, you really don't want to plug 2A3s into an Steremour 2 that has been converted for 45 use, especially if you use the 1W dropping resistors to trim the filament voltage00. 

 

[Deke609]

Thanks PB - yeah I not only forgot about the filament resistance, but failed to take into account that when measuring pin to pin, I'm not measuring the drop across the 0.15Rs b/c I'm measuring between them.

I have 0.33 and 0.3 ohm 3W resistors on hand that if combined in parallel will give me about 0.157 ohms. I am going to try that.

cheers and many thanks,

Derek

 

[Deke609]

My magic R-value for replacing the 0.15 ohm heater resistors turned out to be 0.25 ohms , for a combined value of 0.50 ohms.  With a 0.25 R swapped for each 0.15 R, I get 2.52V (avg) on one tube and 2.535V (avg) on the other.

[Edit - May 16 2019: Now that the amp has been burned in, I checked the heater voltages again and found that they had crept up a bit (about +0.02V). My new magic numbers are a combined 0.57 - 0.58 ohms per pair of resistors on each socket : 0.30 and 0.27 on one socket and 0.33 and 0.25 on the other. This gets me to a heater voltage averaging 2.5V +/- 0.005]

In case it is useful to someone, here are my DC voltages between pins 1 and 4 of the EML 45Bs for various R-values I substituted for the 0.15 ohm Rs:

0.157 ohm: (0.33 and 0.3 in parallel): 2.7V (avg)
0.18 ohm: 2.67V (avg)
0.20 ohm: 2.62V (avg)
0.22 ohm: 2.59V (avg)
0.25 ohm: 2.53 (avg)

The above data may suggest that the drop in voltage was not linear (e.g., going from 0.18 to 0.20 resulted in a 0.05V drop, while going from 0.20 to 0.22 resulted in only a 0.03V drop. This is because I tested the increasing R-values on alternating tubes - i.e., 0.18 on Tube 1, 0.20 on Tube 2, 0.22 on Tube 1, and so on.  The same 0.25 ohm value on both tubes resulted in a 0.015V difference between the filament voltages. Factoring in the 0.015V difference suggests a linear drop of approx. 0.02V per 0.01 ohm increase.

*** Note: the above values were reached after 5+ min and then became stable. I tested the VDC for the 0.25 ohms resistors by waiting 20 min just to be sure. For anyone looking to do the conversion, I would recommend waiting until the amp is well warmed up before settling on a final R-value. In my SII, 10 sec or so after power-on my reading with the 0.25 ohm resistors was only 2.4V

These measurements were taken on a fully upgraded SII after making the following mods: 3K resistors replaced with 8.2K resistors, removal of 0.13 ohm resistors that ran in parallel with the 0.15Rs, and swapping red and black wires from the plate chokes. I did not make any changes to the 130 ohm Rs.

 

[Paul Birkeland]

My magic R-value for replacing the 0.15 ohm heater resistors turned out to be 0.25 ohms.  With a 0.25 R swapped for each 0.15 R, I get 2.52V (avg) on one tube and 2.35V (avg) on the other.

Does this follow one particular tube if you swap it between channels?

 

[Deke609]

Sorry PB. Typo! I meant 2.52V on one tube and 2.535V on the other, not 2.35.

I am treating the difference as trivial.

 

[Paul Birkeland]

Ah, yeah, that's better. 

 

[Deke609]

Quick update: the SII-45B amp is up and running and sounds pretty good. I'm hoping that, at worst, I only took some lifespan off of the tubes with my resistor value fiasco. Still early days. I'll post some impressions after a month or so.

I have another SII kit in the early stages of assembly that I am building as a regular 45 to run with some used Raytheon 45's I have coming -- just to compare with the 45B amp. And then, depending on which of the 45/45B I prefer, convert one of the SII's to 2A3 to compare against my JJ 2A3-40s. Although I wonder if this is a good 2A3 comparator? I love their sound, but I've read that sound-wise they are closer to a JJ 300B than a 2A3.  If so, does that make a fully upgraded SII w/ JJ 2A3-40's a baby Kaiju?

cheers,

Derek

 

[Paul Joppa]

It's hard to get a consensus on what is a real 2A3. By far the majority of 2A3s ever made are old-stock biplates (which look a lot like they are two 45s in parallel and in a single bottle). But the original version was a monoplate with the "harp" filament. It apparently had reliability problems; only a few were produced and they are now rare and expensive. The modern ones are mostly junior 300Bs, with a few like the JJ which is actually a 300B with a 2.5v filament. Can one actually talk about a 2A3 sound as distinct from a 45 or a 300B? I have my doubts.

A baby Kaiju based on the Stereomour kit would (IMHO) want 3K or 2.5K output transformers, and I'd add a larger plate choke as well. It would require 2A3-40s. or at least modern monoplates with the higher dissipation rating (usually 22 watts) - it would be too tempting to plug in regular 2A3s, which would overheat. That makes it less than practical for a kit. It might make 6 watts, which seems to me not enough more than the 3.5W of a 2A3 to be worth the effort.

My two cents.

 

[Deke609]

The newly built SII 45 hasn't even fully warmed up yet, and has zero burn in, and the Raytheon 45 tubes only test "good", but ... it sounds wonderful.  Wow. Liking this a lot. More detailed, but without being dry/analytical; instead more expansive/ethereal. Only early impressions; it'll be months before I get a good sense of the new amp.

Listening to Link Wray's self-titled album from 1971. It was recorded on a three-track in a converted chicken coop. It's kind of raw. I've listened to the album through the SII 2A3 (JJ's) and through 45B (EML), and it sounded great with both amps.  But it sounds absolutely stellar on the new SII 45. Sheesh, this is good.

... I might be a 45 convert.

Derek

 

[Tom-s]

... I might be a 45 convert.

And Derek, are you?

Love to hear your comparison of the 2A3 vs 45 Stereomour II with speakers and headphones.
With and without your balanced speaker outputs? Strong points/weaknesses etc.
Are they good/different enough to keep them both?

Keep us updated!

 

[Deke609]

Hi Tom - I'm going to give it another month or two before posting more listening impressions.  I also want to give the EML 45Bs more time to burn in. I may not have damaged them with my resistor value fiasco. They are sounding pretty good as drop-ins for regular 45s and getting steadily better.

Am I a 45 convert? Yes ... but not completely. As of right now, I can't see converting the other SII amp to 45. My overall impressions haven't changed: the 45 really is magical/immersive and like being "inside" the music; whereas the 2A3 is more of wide-awake caffeinated experience where you as listener are very close to, but still separated from ("outside"), the music. While I mostly listen to the 45, I still sometimes prefer the 2A3.  And I no longer think I want monoblocks for listening through my LCD4s.  Putting the BeePre in front of the SIIs gave me a big boost in dynamics and perceived volume -- so no need to double up. 

cheers

 

[Deke609]

Just a quick follow-up now that the SII 45 and EML 45B tubes have had more time to burn in.

I'd say I'm now a 90% 45-convert - meaning that 9 times out of 10 I'd prefer to listen to the SII 45 over the SII 2A3.

I still stand by my previous comments to the effect that, to my ears and in my setup, the SII 45 produces a magic and more enveloping sound whereas the 2A3 amp presents music in front of the listener as an object.  This is not a criticism of the 2A3 amp - just an observed  difference. 

The SII 45 produces a sound with more body and "mass". I don't know whether this is attributable to the 45 tube or to the fact that the 45 modification allows the OTs to go deeper in the bass, or some combination of both.

One place where the 2A3 amp continues to shine is on the attack of notes.  It punches quickly and decisively. In comparison, the 45 amp seems ever so slightly sluggish: as if the notes swell, taking slightly longer to reach full force. But I don't know which is more "faithful" to what was actually recorded. And let me be clear, the 45 amp has plenty of punch - it's not soft by any means. To use a boxing analogy: the 2A3 amp punches like Muhammed Ali - like a fast sharp knife. The 45 amp has a bit more Joe Foreman: slower but with a ton of follow-through.  On the flip-side, the 45 amp is so easy to listen to for extended periods; whereas the 2A3 amp by comparison can be a bit fatiguing.

Finally a big thumbs up for the EML 45B tubes. I am using them as a straight 45 drop in with no mods to the SII amp to take advantage of the tube's greater power handling capacity (I tried that initially and didn't notice much of a difference -- but I may play around with this again now that the tubes have burned in more). The tubes have taken a good while to come into their own (they are still improving every week). When brand new, I thought them clearly inferior to my various pairs of NOS and used OS 45 tubes from the 30's to 50's.  No longer. I now listen to the EML's almost exclusively.  And they barely get hot - I can hold my fingers on them with only minor discomfort (whereas the 2A3s and regular 45s get too hot to touch for more than a split second) - so I'm hoping the EML's will last a good long time.

cheers,

Derek

 

[Deke609]

@PB or @PJ: Question re parafeed cap value: In the initial post to this thread, the instruction is to increase the 3.3 uF cap to 5 uF, with 4.7 uF being fine. I went with 4.7 uF. But I just came across an earlier 2017 thread where PB suggested 8-10 uF.  What would be the mean of the target range: 5 or 8-10?  I'm guessing that the 8-10 uF applies if the plate choke is not rewired for 40H and instead left at 20H?  But I want to be sure.

My slow descent into audio nervosa has begun. 

MTIA,

Derek

 

[Paul Birkeland]

PJ's recommended cap value would be 5uF, with a pretty wide variation being OK.  I think in that post I used 3K for the transformer impedance accidentally instead of 4K.

 

[Deke609]

Many thanks PB ... my nervosa is subsiding.

 

[Deke609]

Just a quick note on parafeed cap value in case this is helpful to someone.


I originally installed a 4.7 uf 600VDC Clarity Cap, and subsequently added a 0.1 V-Cap CuTF in parallel for a total of 4.8uF.  PJ's recommended target is 5 uF.


I wasn't totally happy with the Clarity Caps -- I found them detailed but lifeless -- and so this past wknd I replaced the 4.7 UF caps with 2 X 2.2uF V-Cap ODAM caps, keeping the 0.1uF CuTF "bypass" cap, for a total of 4.5uF.  Result: a significant loss in bass weight -- the punch was still there, but the bass had less "mass" and "body". So I added another 0.22 uF CuTF cap in parallel, for a total of 4.7 uF and the bass returned. I was suprised that such a small difference (0.22 uF) could have such a big effect (I haven't done the math, so it is entirely possible that this is exactly what one would expect).


I am using my SII-45 with headphones that have a nearly flat frequency response to about 10 Hz. For those with speakers with a bass roll-off somewhere between 30 and 40 Hz, I suspect the effect of the parafeed cap value on the bass could be very pronounced. If I were using speakers with good bass range, I'd probably try to hit PJ's target 5.0uF and maybe even go a little higher -- e.g., 4.7uF and 0.47uF in parallel, for a total of approx. 5.2 uF. ... Just my thoughts and speculations ....


(Early impressions of the ODAM caps: detail is similar to the Clarity Cap CMR but with much more life and body).


cheers,


Derek

 

[Deke609]

When you do the 45 conversion, part of that involves connecting the entire winding of the plate choke between the 45 plate and B+, which is 40H instead of the 20H in the 2A3 configuration.  This is not possible with the 2A3 because it draws more current, but the reduced current drawn by the 45 allows this connection.  This change in conjunction with changing the parallel feed capacitor will work to improve the low end frequency response.  The power bandwidth of the OT-2 doesn't seem to be the limiting factor in this situation.  If you want to go for the high power 45B, then the plate current goes back up, and you need the other connection.

The Lundahl LL2743 may work, but it's a little off in terms of the options for the air gap. The Sowter 8985 would be the most appropriate aftermarket choice, but it's 50H/40mA, so you're not gaining much over what we provide. 

It kind of sounds like what you may want to do is build the Stereomour stock, then sort out the mods to get to #45s, then do a scratch build with all aftermarket parts.

Once I get the BeePre rebuilt, my plan is to rebuild the SII-45 (w/ EML 45B tubes) to see if I can get closer to the Kaiju's performance. So, I've ordered: (a) Lundahl 70H/160 DCR plate chokes rated for 60mA DC; and (b) Sowter "mumetal sandwich" 5K parafeed OPTs (getting custom 4Ks was too pricey and would also have necessitated a lot more parafeed capacitance when paired with the 70H chokes - also pricey when using V-caps).

Question: @PB or @PJ: can you tell me the DCR of the stock PC-3 configured for 40H, if known?

I ask b/c I'd like to estimate the effect on plate voltage of the new 160 DCR chokes. Eventually I will try B+ of 450V-500V, but that will require a new PT (I think I can squeeze at most another 20V out of the stock HV power supply by replacing the 130Rs with lower resistance chokes - but that only gets me to 400V and change). But I will stick with the stock PT for the initial test builds.

many thanks in advance, Derek