[dave slagle]

So I was thinking about the autoformer MC stepup experience and had a lot of useless data in hand. The general thinking was if we could eliminate the haze, we might have something that approached the sound of Frank's unobtanium mic stepups. My initial thought was the grid current from the input tube of the phono pre could put some DC on the cartridge and simply going to a transformer from an autoformer might be the answer.

I wound a pair of transformers and fired them out to John, but alas still the haze. Never being one to give up I suggested he play a bit with loading to see if things improved but for the most part the MC stepups were a lost cause. Flash forward a few weeks and I get an email suggesting that the issue all along was shielding and simply placing a meter of distance from the Stepups to the electronics improved things drastically and we are back in the hunt.

At this point it seems as if we are at a coin toss between Frank's and my stepups so the project is indeed alive again

Below I'll attach what I know so far. Rumor has it I'll get franks trannies in hand to do a few more measurements which should help me confirm or deny a few hunches. Hopefully the guys will chime in with their experiences so I don't just sit here talking to myself

[galibier_numero_uno]

I'll prime this thread with John's e-mail from today, along with Dave's reply.

John's comments are the quoted parts and Dave's are in embedded, plain text.

Cheers,
Thom


Hey guys.

[dave slagle]

here are the frequency sweeps of two versions of the stepups.

#2 is currently what john has and #5 is the "improved" version"

i mismeasured the #2 version and thought it was -1dB at 40k which inspired me to move onto #5. it fixed things but #2 was nowhere near as bad as i thought. Boy those HF measurements are tough!

dave

[Johnny]

Hi Guys,

I loaded Frank's step-ups with only the 47.5K load in the Artemis phono-pre. The sounded much better, more of that live quality. I compared them to Dave's step-ups with a 7K load. They were very close. If there were differences, they were very minor. Someone else needs to take a listen. Normally, when things are this close I don't worry about it. I'll try chaning the lams in Dave's in the next couple of days.


Johnny

[Steve Kaufman]

Have you tried Daves stepups with 47.5K loading instead of 7K?

[Guest]

Hi Steve,

I tried 47.5k with Dave's step-ups and they don't sound good.


Johnny

[dave slagle]

I am now convinced that there is indeed a terminating resistor inside the cans of frank's stepups. Since the inductance suggests a 60:1 ratio and the voltage suggests a 33:1 ratio i'd guess a 15K resistor is inside.

this in parallel with the 47K load resistor nets ~12K which with a 60:1 ratio leaves you with a reflected load of under 4 ohms.

not sure what to make of this yet, but if true it adds an interesting wrinkle.

dave

[Johnny]

Hi Guys,

I got some time this morning to try different gaps with Dave's step-ups. Dave sent them to me with 12x12 gap size (40H). Next I tried the butt gap (20H). It sounded a little thin and less filled in. Next I tried 9X9 (57H). With this gap it sounded a little soft and too warm for me. Finally, I tried 18x18 (30H). This sounded much like 12x12 gap. Very open, quick, with lots of juice. I'm not sure which is better the 12x12 or the 18x18. I'm going to listen to the 18x18 for a while. If I miss the sound of the 12x12, I'll go back.

Johnny


P.S. Dave, after I drove over to Steve's to get some extra lams, I found the orignal lams behind my bass amp.

[Johnny]

Hi Guys,

I ended up liking the 12x12 gap best with the #2s. I went back to them after a day or two of listening to the 18x18 gap.
Sunday afternoon Steve and I listened to the #5 step-ups. I changed them from a butt-gap to a 12x12 to compare directly to the #2s. After listening to some vocals, Steve and I slightly prefered the #2s. Next we changed the #5s to a 9x9 and compared them to the #2s in a 18x18. This time we listened to some challenging piano music, a List sonata. It wasn't close in this comparison. The #5s were much better, the dynamics were cleaner and more detailed without being hard. Finally, we regaped the #5s with a 6x6. On this comparison, the bass had more authority than the 9x9 and the "alive" quality was as good as the 9x9.
Sunday eveving we tried the #5s with the 6x6 gap at Steve's along with the #2s with the 12x12 gap. The #5s with the 6x6 gap were much better than the #2s with the 12x12 gap. They had better focus and more of an "organic" quality.

Johnny

[dave slagle]

hey john.

thanks for the info. I must admit i find it very interesting that you liked different gaps on the two different versions since electrically they don't show diferences until past 100K.

I think this might be a situation where a cheapo LCR meter would shed some light since it is entirely possible that the restacks could net you some odd results based on the "random" nature of the gap. my first guess would be that the 6X6 gap in #5 was actually equivalent to the 12X12 in #2. That is to say if the meter spits out the same number the inductance is the same, and since the primary turns are constant between the two, we can assume the gap is the same even if the stacking is different.

an interesting test would be to place a diode in series with a variac and slowly ramp up the voltage across the secondary until you hear some noise, then give the core a few "love taps" and ramp the voltage down. next short the diode and ramp up and down. The theory is the diode will put DC in addition to AC on the core which will help shake ti together and hold it in place. then shorting the diode allows you to bring the core back to 0,0 with just an AC signal.

when people say cores need to break in, i suspect they may just be hearing the gap settling in to its "norm" rather than the actual characteristics of the material changing.

dave

[Johnny]

Hi Guys,

I think part of the difference may also be subjective. I admit that I'm not always consistent with my evaluations despite trying. I know after about 3 or 4 hours of constantly comparing differences I lose focus. I think you're right Dave. I believe the sound for #2 and #5 should be the same with the same inductance and a LCR meter would help. On the other hand, I worry that if I "expect" them to sound the same with the same inductance, they will sound the same whether I hear it or not. As as result I think we can lose some of the benefit of "ears-on" listening.


Johnny

[dave slagle]

I agree. i would simply measure the inductance of the stacking before you unstack it

gaps are very funny animals. when you get into the physical spacers of a few thousandths of an inch things become very predictable, but things like a butt gap and smaller can be all over the map.

many folks plame it on differing perm, but i suspect it is nearly impossible to repeatedly set the same gap with any precision. Two autoformers stacked off the same string of lams with a but gap can measure 15 and 20hy's unstack and switch the bobbins and the inductance will remain consistant with the core. I refuse to believe that the perm could vary so much in adjoining 1/2 inch samples from the same batch.

dave

[dave slagle]

Got franks magic trannies in to test a bit more... verrry interesting.

The turns ratio is around 66:1 and that drops to 25:1 with an 8K load.

EDIT 12/5/05 (note i was dead wrong above when i was expecting an inner termination resistor) Frank liked these things unterminated and suggested a stepup ratio in the 1:30 ballpark. My initial inductance measurements suggested the 1:60 range and internal loading seemed to be a possible explanation on the difference.

I'll let the frequency plots speak for themselves... but they to seem to show that there are two extremes at work here.

another edit 12/5/05 by the two extremes i mean that franks roll off early and cleanly and mine may have a erratic behavior but it is well beyond 100K. I agree with both approaches... just don't get caught in the middle with drastic peaks and valleys anywhere near 20Khz.

Final edit 12/5/05... I generally go as far as i feel confidant on the bandwidth (~200K) but with franks, there were no hidden things happening.... after they dropped off the were off. a good thing in my book This is something that really interests me on how/why they achieved that behavior, and if it was a goal or byproduct of the design.


dave

[berlinta]

Hi Steve,
Thanks for all the effort you've put into analyzing these transformers. Since they were supposed to be used with one particular cartridge(of a unique design) there is a strong likelyhood that the rolloff was designed "into" the transformer to compensate for a mechanical high frequency resonance of the cartridge. No need for any further extension above that as the contemporary preamp made by the same company rolls of above 35k anyway.
I just came home with some mu-metal to build shielding cages for your "clones"(the wrong term, I know...)
Btw, the only criticism of my vintage trannies I had in absolute terms is that top end air is somewhat subdued(compared to, say, the silver audioconsulting mc step ups), now confirmed by your measurements.

You all have a great weekend,

Frank

[dave slagle]

I do want to point out that the 8K load was an arbitrary one, and i suspect with the typical 47K load at the input of most pre's would tame the slight peaking and remain flat to 25-30k and then nicely roll off.

this would coincide with your experiences that they sound best "unloaded" and my brief listen to them loaded at john P's place. When we did load them we were thinking they were around a 30:1 ratio so tha actual 66:1 ratio hit us with a double whammy of reduced bandwidth and a tough load for the source.

did you get both of my MC's working?

dave

[SLK]

Just a brief note before retiring. John and I tested three pair of trannies tonight. The small chassis pair sound wonderful. However, bigger is definitely better. The two pair of larger chasis trannies soth sounded more dynamic and open. The pair that had no writing on them were tested before the ones labelled T50. As great as the unlabelled ones sounded, the t50s sounded even more open with extraordinary texture on the strings and woodwinds, not to mention vocals. However, in my system they seemed to saturate/distort on some very hot recorsings. John will try them out in his system tomorrow and see if the same thing happens.

Steve

[berlinta]

Hi Dave,
Yup, last wednesday I had to get away from the grind and devoted an evening to testing your step ups with two different cartridges. I had two additional pairs of ears here, though we didn't exchange our impressions until after all the switching and comparing was done.
We came to very similar conclusions. Unloaded, your step ups were very highly resolving, but the upper bass and lower fundamental region was lacking weight and "warmth"(not the "fuzzy" kind...just some energy)
With a 12k load on the secondary, the overall balance was prefered over the initial setting, yet the top end air seemed a little diminished(now about on par with my vintage trannies). Where the new step ups still didn't couldn't match the old ones is the threedimensionality, the "roundness" and tactileness of the sonic picture. All of this with your temporary shielding which doesn't completely eliminate hum. So, once the mu-metal housing is finished, I'll do it all over again, then also with extra short interconnects(had 0,5m Gryphon ICs at hand for this comparison...).
One last thing, your step ups seemed to be mating better with the Olympos than with the Titan(which sounds a touch brighter).

A warm and peaceful set of holidays for all of you guys!

Frank

[dave slagle]

frank,

Thanks for the report! I am reasonably sure the haze we all heard with the autoformers at the RMAF was due to the lack of shield since when i sent apair of transofmers and johhny tried them unshielded he had the same results. Physically moving them 3 feet from everything else really seemed to clear up the presentation so i wouldn't be surprised if a better shield made a difference. One thing to consider when designing the shield is the ability to restack the core. You have the same version i originally sent John and steve which they found the 5X5 stacking sounded best. This gives around 175mhy of inductance or about 5X that of your vintage units, so it may be interesting to also try a larger gap. By all means feel free to play and report back (good or bad) I just sent apair of larger cored versions to tony B and want to get a pair to maryland for you to take back with you after CES. Steve will be taking his so you can do another comparison of the small vs. larger cored versions.

dave

[Johnny]

Hi Dave,

I burned-in the "pipe bomb" you sent to Steve when he was on the East coast. I ran a CD signal through the autoformer pre into the inputs of the "pipe bomb". I loaded the secondaries with 3.1 K resistors. The black box that Steve put together with the low drc and 50 turns sounded much better. The middle frequencies seemed to be missing from the pipe bomb. The black box has a copper tube for each transformer and around that some mu metal shielding with an air gap. Both of the tubes are placed in a steel closed chassis. My guess is that the difference between the two step-ups is caused mostly by the difference in drc.
Grant got the two sets of setups you send. One is a 40 turn and the other is a 50 turn with a lower dcr than the other 50 turn step-up we tested. We'll compare these two to the other 50 turn in the next few days.

Johnny

[dave slagle]

thanks for the update... the tests with the lower DCR versions should be interesting.

I have two more pair almost ready to ship, these will have 1/2 and 2X the primary turns with gaps adjusted to net the same inductance. both will be based on the preferred geometry at the moment so they will be "mid-DCR"

the next few listening tests should zero in on DCR and push us in a direction for the number of primary turns vs. gap compromise and once we are comfortable with those results, we can look to explore different turns ratios.

I plan on having your "cast-offs" be circulated to others to repeat and confirm or deny your results.

dave

[Johnny]

Sounds good Dave, I like the idea of having others repeat the experiments. It should give us more information.

Johnny

[DuoBruce]

I'm not sure what version of the step ups I have here, came through Jeffrey, but I restacked them last night to go from 6x6 to 4x4. So far I like it better. Resolution has always been good but I was looking for a bit more warmth and this got that.

Need to try more combos but it looks Jeffrey will be dissapointed if he thinks he will get these back any time soon

[dave slagle]

[galibier_numero_uno]

Another thing as been bugging me, but I'm so bloody swamped that I haven't participated in your (John and Steve's) experimentation.

Your gain structure is "unconventional" to say the least. As far s a more universal product is concerned, a 1:26 step-up is a huge amount of gain.

You boys (John/Steve)are using a 50dB gain Artemis PH-1 phono stage, and then burning off all of the gain via a 1:8 ratio between your o/p tube and OPT.

Most individuals' gain structure will predicate an MC step-up on the order of a 1:10 ratio.

I'm not sure how the design parameters will change, but I'd love to help out during the evaluation stage this Spring.

Is John Atwood listening?

Cheers,
Thom

[DuoBruce]

My cartridge is a Wilson Benesch Analog. As I understand it is a high end Benz (ruby?) mounted in a carbon fiber body of WB design. It has the same .35mV output as a ruby. A dcr of 10 ohms with a recommended load of < 20 ohms with a step up. That sounds dull to me so right now I have the secondary of the step up loaded with 68K which I believe gives me about 110 ohms. I added dip switches and resistors from 1M on down so the loading is something I need to play with a bit more

[dave slagle]

[galibier_numero_uno]

Hi Dave,

One thing I lost sight of was that most folks with MM phono stages are looking at 36-40dB of gain, rather than the 50dB of the Artemis.

I'll bet that there are more people in this thread however with a gain structure that doesn't need the extra 8dB. Of course, I'm not trying to burn you out with too many irons in the fire (pun intended)

Yup, I need to take the trek up to hear how the 1:8 sounds on the boyz' 75TL rigs. Their energy is boundless, it's Winter, and they don't ski. I have no doubts that it's special. I'll likely combine the trip with some wood horns I have on loan from Kevin Brooks.

The concept of a 1:1 MC "step-up" might not be as strange as it sounds - for the guy who doesn't need any gain.

Perhaps this would lend some insight into some of the other attributes that iron lends to sonics - perhaps bandwidth limiting, perhaps ???

Of course, if plugging in a 1:1 into an MC stage becomes the latest and greatest, there will be those who claim you're building tone controls and that we're stark raving mad The latter may well be true of course

Cheers,
Thom

[dave slagle]

[quote="galibier_numero_uno"]

The concept of a 1:1 MC "step-up" might not be as strange as it sounds - for the guy who doesn't need any gain. [/q]

but he might still need isolation. This is ironic on so many levels for me since i swear for volume controls an AVC is preferrable to a TVC and believe using a transformer to fix a grounding issue is just a band-aid. The irony shows up when you consider the 1:1 MC isolator would still require a common ground with the cartridge.

[Steve Kaufman]

Hi Dave

We (Lynn Olson, Thom Mackris, Mr Sanders from Quicksilver amps, Johnny, and me) got together at Johnny's house and compared the different phono stepups. The results were similar to those I reported to you from Friday night. The straight wire 25 turn bobbin sounded the best. The lower dcr version sounded much leaner and 2 dimensional. The 40 turn version sounded slightly more lush but lacked the openness and nuance we heard in the 25 turn unit. Lynn wanted to hear an 18 turn version without the litz wire.

Today Johnny and I did some more comparisons between the 25 turn straight and 50 turn litz trannies both of which have essentially the same dcr. The 25 turn trannie sounded significantly better then the fifty turn unit. The lower dcr is obviously not the whole story here.

Finally, we played around with the output trannies' bobbins. Thus far the 50:1 bobbin has sounded the best in both mine and Johnny's amps. The impedance rations are 8:1 for Johnny's amp and 4:1 for mine. The new 75:1 didn't work at all in Johnny's amp. With my amp the impedance ratio with the 75:1 bobbin is 8:1. It is the best sounding bobbin we have tried in my amp. These results are consistent with those in Johnny's amp using the 50:1 bobbins.

Things are really sounding fine in both of our listening rooms!

I'll bring the 25 turn straight and litz as well as the 50 turn litz stepups to Miami and let some new ears have a listen. I'll get back to you when I return.

Thanks for all the reagent!

Steve

[dave slagle]

Hey Steve,

Thanks for the report!