A COMPLETELY different perspective on "soft clipping"...

[Mallette]

...or perhaps just insane ravings. You decide.

Some of the discussions on another thread caused me to think of this, but I've had enough of that one. Some folks don't know when to shake hands and return to their corners...

I'll be quite interested to see if ANYONE makes sense out of what I am about to say. It is TRVTH, and I interviewed many who testified to the fact and, at least IMHO, provides another way of looking at what many perceive as the benefits of soft clipping in VT amp designs. It certainly changed my attitude.

Back in the 1980's I worked for a UNOCAL subsidiary that made a very special isotropic graphite of strength more like metal that like graphite. It had originally been developed as a possible heat sheild for the shuttle program, but lost out to the tiles. In search of a market, the Electrical Discharge Machining (EDM) industry became a prime target. EDM uses a capacitive/resistive relaxation circuit to create sparks between a workpiece (metal blank) and an electrode (in this case, graphite machined to the reverse image of the desired mold from the metal blank). While the number of shops using this technique is relatively small, the impact is HUGE and touches everyone. The majority of all the plastic devices in your lives come from molds machined by the EDM process. Hopefully, that is enough to set the stage...

I was tasked with writing a technical manual about electrode material selection. As I traveled the country, I found a rather extraordinary number of old vacuum tube based EDM machines in use. For a while, I paid little attention as I assumed that, like most machine shops, people will use anything that does the job as long as it works. However, I then visited an EDM machine builder, and found they were PRODUCING VT-driven EDM machines again due to customer demand. This was a shock, as none had been made in many years and the machines were already going Computer Numeric Controlled (CNC). It was explained to me that the demand had grown for them. Seems that while the SS machines were more efficient and cost less, there were certain jobs where certain old timers had the screwball idea that the results with the old VT machines were superior.

To make sense of this, I need to provide a bit more explantion of the EDM process. Surface finish is determined in the EDM process by frequency. The lower the frequency, the more power delivered per arc between electrode and workpiece (sound familiar), the higher the frequency, less power is delivered but more arcs. So, for fast metal removal, low frequencies but rough surface finish, and for fine finish, high frequencies and more time to remove metal.

So far, so good. What the old timers were finding was that they got superior finishes from the same high frequency settings using the VT machines than from the SS. Nobody new why. I did. While I'd not been convinced as yet, I'd followed with interest those in the audio magazines (no forum yet...dark ages) who were speaking of "soft clipping" and "less EDGINESS (caps mine)" from the old VT amps. Now, these guys are NOT hobbyists and operating a machine shop at a profit is not an easy task, so they do nothing unrelated to producing what the customer ordered, on time. No room for "romance" in that business.

Bottom line is this: The phenomenon observed by so many audiophiles and so often derided turns up in a field where decisions are not based on taste or tradition, but on observable outcomes. Whether in audio or in metal, it seems that tube provide a better "finish."

Again, I apologize if there is not enough above to make sense to many of you, and my explanation may have some technical gaps in it due to the 20 years having passed since I wrote that book. If there is any interest in this I'll see if I can dig up a copy at home and provide more info.

Dave

[JBryan]

Sounds completely plausible to me but as most folks here know.... tubes are biased. 

[jbsl]

If a person prefers solid state to tubes or tubes to solid state in their system then that is fine it is their system and their ears!

I use both SS and tubes in my system and for me the tube amp just sounds better but if I add a tube preamp I do not like the sound as much, it still sounds good but not as good as my SS Luxman preamp with my tube amp.

To others SS just sounds better and to others tubes sound better. Everyone is right for their ears and their system.

No one is wrong until they start trying to tell people why tubes are the better equipment or solid state is the better equipment.

I bet alot of you drink coffee and like it alot, maybe even love it.

If it was up to me coffee would be eradicated from the planet!!!! I hate the SMELL of it[N], makes me sick[:S], and moste of all I hate the taste[][][N]. Did I say that I hate coffee?

I like hot tea.

But if you drink coffee I am not going to try and convince you tea has a better flavor. But it does have a better flavor for me [].

Everyone is biased[] when it come to their system just don't tell others what their bias should be.

I

[Mallette]

The post was in NO way a sales job for tubes. Just another perspective from about 90 degrees. In my case, I listen about 90% of the time through a 127.00 class D amp and find it as good as anything non-tube at any price. Now THATS an attitude for you...

However, you'll pry my ST-70i from my COLD, DEAD hands!

Dave

[DrWho]

So the SS machines were designed to be more efficient and cost less and there were absolutely no compromises made in the process? And the "same settings" actually had the exact same effective output? (ie, two drills set to 1200 RPM on their dial may not actually be spinning at exactly the same speed).

I propose that older mics sound edgy...

[Mallette]

If I understand what you are saying correctly, the answer is "yes." Same amps, same frequencies produced molds that had qualities master machinists (golden ears) found more to their satisfaction.

As to the "older mics sound edgy," my piano mike of choice is my RCA BK-3A, vintage 1936. A number of "golden ears" have found my recording of Stuart Wayne Foster's piano recital the best, most natural piano recording they ever heard.

Dave

[Mallette]

This pic may or may not be useful. I scanned these from my 1985 pub "Guide to Electrode Materials Selection." One necessary explanation to the circuit reading types is that the workpiece and electrode are submersed in a light oil that provides a dielectric. It is that dielectric value the circuit works against at a given volt/frequency setting to create the sparks and erode the workpiece to the shape of the electrode. The VTs are from a "revival" VT amp-based EDM machine of the 80's. The sparking is through a window I had built into an EDM machine tank so I could get photos of the action, and the finished mold is a VHS cassette mold. The grain pattern is a direct result of the EDM process and that which the master machinist (golden ear) would judge to determine the quality of the process. Any VHS cassette (and many other things) you have around the house will carry the reverse of what you are seeing here.

This may be taxing some of you and I'll try to deal only with direct questions or comments as they relate to audio on this. As you should have figured out by now, there IS a direct relationship, IMHO. BTW, there was (and probably remains) a large number of EDM machinists who said it was all crap and these guys should get with the latest technology.

Dave

[capo72]

Interesting perspective. I have been in the "trade" and involved with EDMs for 16 years. I'm 34 so this has been my entire working life. I am a Tool & Die maker that has worked header die, mold making, and currently for a steel stamping and die building company. I have been an EDM guy from day one. I ran some very old machines in the begining that all had power supplies that had been updated to SS. I have worked with a few "old timers" and have never heard anyone long for the days of VT technology. I'm in no way disputing what you are saying and I really respect your opinion on that other thread. I just want to let you know that there is someone here who understands exactly what you are saying. I love VT for audio, maybe because it provides an emotional connection to the equipment for me, but I digress. From all my years of experience in EDM I can say that from a cost prospective, which like you said is why we are in business, tubes have never been missed. The maintainence and unreliablility compared to SS would never be accepted from a cost standpoint regardless of how the "old timers" felt about the results. Again i'm not saying you are wrong. Our experience lies in different decades. Mine started in the early 90's and things have changed rapidly in my 16 years. Today I deal with mostly wire EDM, albeit different, the technologies are very similar. Today's new machines used digital generator technology to produce much higher frequencies than SS generators providing far superior speed and surface finish over any technology previously available. In my industry faster and better is what it is all about. My boss would laugh in my face if I suggested we went back to VT because "I just like the finish they provide", or "the tubes just look so cool when they glow."

I guess my point is In this industry it's all about the bottom line. Profits are not subjective like sound. We have evolved from VT to SS to Digital machines and there is no looking back. I guess as long as music reproduction is subjective there will always be people that put emotion ahead of the bottom line. If we could all get digital amps for $5 that produced equivalent liner results as the best SS and VT equipment out there, why wouldn't we? In this subjective world of audio it is not worth arguing about.................... But sometimes it is fun! Great post Dave.

Jeremy

[Mallette]

One of the rare! At the time I was writing that book, there were only 25,000 EDM machinists in the US.

Oh, I would not even consider disputing your position, and, perhaps even more so than in audio, it was only a few who either insisted on keeping or returning to VT EDM machines for certain jobs. They were considered "a bit queer but harmless" by many but included some of the acknowledged "artists" of the field. There was enough to create sufficient demand such that new machines were built by makers who'd abandoned VT long before. Wire EDM was just coming into the market but had significant penetration, especially those CNC machines built by Mitsubishi.

I've no idea what happened since then. Perhaps all these old guys died off. Now, it's interesting you should mention DIGITAL generators as replacing SS. That's exactly what happened to me and SS... I found Class D to have the accuracy of SS without the "edginess." Strange, huh?

Dave

[ben.]

Well I thought it was interesting.

And yes - a U67 sounds edgy. That's why it sounds great.

[capo72]

One of the rare! At the time I was writing that book, there were only 25,000 EDM machinists in the US.

At the time I got into the trade that # was alot higher but has really declined over the last 8-10 years due to Asian competition. That foreign competition is another reason that we have no room for "quirky" technologies. I can't imagine the heat generated by a bank of tubes like those pictured above. It would be quite a sight to see in action though. Thanx for bringing this topic up Dave. I think I have some research to do!

Jeremy

[Mallette]

Check www.poco.com and see if they still have a copy of my "A Practical Guide to Electrode Material Selection." I've no idea if anyone is left there from my days. I visited about 10 years ago and there were still a few. I took a quick look at the site and it may have been concatenated with the "EDM Technical Manual" along the way, or it may simply have gone out of print due to technology changes. However, from that quick visit it did not appear that much had changed in the electrode. Ultrafine was introduced while I was there.

The "Guide" had perhaps the only history of the process ever published, at least to that time, as well as some rather unique photos.

That has been awhile now.

For those looking at this wonder what it has to do with audio, start at the beginning... Anyway, it's about done.

Dave