Hi John,
I don't know about you, but I love learning new things.
I would like to make a couple comments though. All signal relay contacts are designed to wipe every time they close. All rotary switches are designed to wipe every time they are operated. Slide switches share this desired attribute. So it seems that the designers foresaw this requirement a long time ago. I'm guessing well over 25 or 30 years ago.
I don't know about mercury whetted relay contacts, or reed relays. I do have some. Can you tell me what the problems with these are please? I'd like to point out that some higher end companies did use reed relays in the signal path, normally for mute functions. Just an observation with no assertion that they are good due to this.
On cars. My brother owns several car lots and dealerships, so I have been lucky enough to be able to drive many different cars over the years. The 928 was nice to drive. The Mercedes 500 SEL was nice (heavy feeling), Jags are no fun, BMW's are a real pleasant car to drive, the 745 is a complete blast. The Chev. Corvette is very nice, and fuel efficient too! Comfy on a long trip. Speaking of long trips, those Suburbans or similar would be my choice for that. I really enjoyed the Buick Roadmaster wagon, another good distance car, and good on gas with a 5.7 liter. I had an Olds Delta 88, 1984, another great car and good on gas also. Even some of the Kia products are pretty nice, they don't do a whole lot wrong and they are very reliable. GM type steering feel, a touch heavy, but they go exactly where you turn the wheel every time. For a daily driver, I guess a BMW would be the best overall. No surprises there.
Anyway, John, even though you've been there and done that (and never miss a chance to let us know), why not teach the rest of us? If you told us everything you know, you would still be 25 years ahead of us. It's one thing to say that a switch type didn't cut it, but why not actually let us know what it was about whatever component it was that you didn't like? That would give us far more to think about.
Have you ever looked at a switch contact up close? Same conditions. Exclude any advertising material or "white papers". They really don't count! White papers should be illegal, most are filled with nonsense, or some good facts twisted to show why someone's widget is better than all others. That last one is no exception. Complete refuse.
All contacts are going to look pretty rough under magnification. It think that electrical performance should be the indicator of quality, since that is what the contact is being used for. Reliability data is even more telling, and those contacts are going to look pretty bad at end-of-life.
-Chris
Haven't you done anything since then?
I don't know about you, but I love learning new things.
I would like to make a couple comments though. All signal relay contacts are designed to wipe every time they close. All rotary switches are designed to wipe every time they are operated. Slide switches share this desired attribute. So it seems that the designers foresaw this requirement a long time ago. I'm guessing well over 25 or 30 years ago.
I don't know about mercury whetted relay contacts, or reed relays. I do have some. Can you tell me what the problems with these are please? I'd like to point out that some higher end companies did use reed relays in the signal path, normally for mute functions. Just an observation with no assertion that they are good due to this.
On cars. My brother owns several car lots and dealerships, so I have been lucky enough to be able to drive many different cars over the years. The 928 was nice to drive. The Mercedes 500 SEL was nice (heavy feeling), Jags are no fun, BMW's are a real pleasant car to drive, the 745 is a complete blast. The Chev. Corvette is very nice, and fuel efficient too! Comfy on a long trip. Speaking of long trips, those Suburbans or similar would be my choice for that. I really enjoyed the Buick Roadmaster wagon, another good distance car, and good on gas with a 5.7 liter. I had an Olds Delta 88, 1984, another great car and good on gas also. Even some of the Kia products are pretty nice, they don't do a whole lot wrong and they are very reliable. GM type steering feel, a touch heavy, but they go exactly where you turn the wheel every time. For a daily driver, I guess a BMW would be the best overall. No surprises there.
Anyway, John, even though you've been there and done that (and never miss a chance to let us know), why not teach the rest of us? If you told us everything you know, you would still be 25 years ahead of us. It's one thing to say that a switch type didn't cut it, but why not actually let us know what it was about whatever component it was that you didn't like? That would give us far more to think about.
I don't know..... yes, I have ..... depends on your definition, but I'll say no for now ...... no, no pictures. The microscope didn't have an attachment for a camera.
Have you ever looked at a switch contact up close? Same conditions. Exclude any advertising material or "white papers". They really don't count! White papers should be illegal, most are filled with nonsense, or some good facts twisted to show why someone's widget is better than all others. That last one is no exception. Complete refuse.
All contacts are going to look pretty rough under magnification. It think that electrical performance should be the indicator of quality, since that is what the contact is being used for. Reliability data is even more telling, and those contacts are going to look pretty bad at end-of-life.
-Chris
PH104 said:
You also don't get it. The pistons are made to a tolerance and no, they don't need selectively picking because they are made to a tight enough tolerance in the first place, in a design that only requires that tolerance.
As for using different tyres that is plain daft. Two different makes and models of tyre are not at all like tolerances, It is like using DIFFERENT FETs, NOT selected instances of a particular part.
What you guys are on about is NOT engineering in my book, it is about hand made craftsmanship, which in itself is not necessarily bad, but good engineering practices have made uncessary in the modern car industry. A good modern design should be able to achieve the same in electronics.
When we match fets, it is because the fet manufacturer gives us a selection of Idss current , and not necessarily the same current. It is the nature of the manufacturing process. IF we were Porsche, I guess we could demand all of their 10ma Idss fets, and not have to match later. What is the difference in reality? NOTHING!
John -- I don't think anyone is belitting the 944. I think they are a beautiful car with classic styling. There's a black one in town that is absolutely gorgeous. In fact, before I learned more about Porsches, I thought the 944 was the ONLY Porsche.
Bobken said:
Hi Bobken,
I agree that the Shallco is superb sounding and found it to be essentially transparent in my listening tests. (This is in contrast to all the relays that I have tried, which have had obvious colorations.) But the actual construction is not *quite* as simple as you say.
The inner contact is a solid silver ring about 0.030" (0.75 mm) thick. But this is welded to a pure (not beryllium alloy) copper post that is where you make the connection. And similarly, the outer individual contacts have a solid silver contact about the same thickness welded to copper posts. The "bridge" is not only solid silver, but also doubled up, so that there are two contact paths in parallel.
The self-cleaning action you mention is important. If I have left my preamp on one input for a few weeks, I can hear a small but definite improvement in the sound by "scrubbing" the contacts by turning the switch back and forth a few times. The volume control doesn't have this problem as I am constantly changing it.
I just wish I could find a relay that was even close to as good sounding as the Shallco rotary switch.
alansawyer said:
OK, I'll be waiting for you to start making transistors that all have exactly the same beta and FETs that all have exactly the same Idss. That will be great...
posted by alansawyer:
You also don't get it. The pistons are made to a tolerance and no, they don't need selectively picking because they are made to a tight enough tolerance in the first place, in a design that only requires that tolerance.
I get it just fine. I specify electrodes for an RF quadrupole spectrometer that have hyperbolic profiles cut to 0.0001 inch tolerance (yes 1/10,000"), diamond-turned to a mirror finish. We still find that some work better than others.
And about my tire example -- I was just using absurdity to illustrate the absurdity of your statement.
You also don't get it. The pistons are made to a tolerance and no, they don't need selectively picking because they are made to a tight enough tolerance in the first place, in a design that only requires that tolerance.
I get it just fine. I specify electrodes for an RF quadrupole spectrometer that have hyperbolic profiles cut to 0.0001 inch tolerance (yes 1/10,000"), diamond-turned to a mirror finish. We still find that some work better than others.
And about my tire example -- I was just using absurdity to illustrate the absurdity of your statement.
Hi Charles,
Well, you are aware that tolerances on semiconductors are being improved - a lot! The later power output transistors from On Semi are a case in point. Getting the beta spread (or IDSS spread) down on signal devices may not be that far off in the future.
-Chris
Well, you are aware that tolerances on semiconductors are being improved - a lot! The later power output transistors from On Semi are a case in point. Getting the beta spread (or IDSS spread) down on signal devices may not be that far off in the future.
-Chris
All porsche fan know that 944 have good handling ,nice car ,turbo cabrio better😉, the 911 engine back the wheels are not so good but at that time the only way to have some grip !!
I see one 944turbo black last month ,fantastic....
I see one 944turbo black last month ,fantastic....
Quote without comment: 'The Audio Amateur' LTE 1/1981 p. 45
..."We believe the dissimilar metal junctions (approximately ten) and the high cross-capacitance (30pf) in each of the mercury relays used in the Spiegel Box tend to create a false impression of similarity between units. Instead, silver contact rotary switches were used that had even lower contact resistance than the mercury relays and virtually unmeasurable cross-capacitance. Subjectively, the silver contact switches appeared to have significantly less effect on the system resolution than the mercury relays." ...
John Curl
Karen Richardson
And what I have done lately is still a trade secret.
..."We believe the dissimilar metal junctions (approximately ten) and the high cross-capacitance (30pf) in each of the mercury relays used in the Spiegel Box tend to create a false impression of similarity between units. Instead, silver contact rotary switches were used that had even lower contact resistance than the mercury relays and virtually unmeasurable cross-capacitance. Subjectively, the silver contact switches appeared to have significantly less effect on the system resolution than the mercury relays." ...
John Curl
Karen Richardson
And what I have done lately is still a trade secret.
Charles Hansen said:
No idea if this is common but the relays i experimented with sounded much better through the normally closed contacts. It takes some logic or a micro and a single relay per volume step but the result is incomparably better than any R-2R attenuator i've tried.
Just checked the analogue switches of AD and the variety is bewildering. What parameters are important? Rd?
Analog switches are a REAL BEAR! You can get them to work OK, and they sound OK, probably because they don't have a REAL junction for the signal to pass through. However, they have significant MEASURED distortion, some noise, and limit early.
Charles Hansen said:
Hi Charles,
I am (virtually) certain that the Shallco switches I have used (32 contacts per wafer) do have solid silver contacts throughout, for what it matters.
When I first investigated these switches I filed down the end of the post you mention (fortunately being unnecessarily long for the purpose) and whilst it is possible that that this was friction-welded on to the ring you correctly describe (for ease of manufacture) it is certainly not made of copper as it is white in colour throughout. I cut the rear of this down far further than any cladding or electro-plating would have 'survived', so I am quite sure of this unless perhaps solid nickel is used here which seems unlikely.
When I have some equipment stripped down, I could probably check this with a magnet on a switch in use.
I did only manage to scrape the rear surface of the other (multiple) contacts as there was insufficient parent metal to go any further without risking damage, but again this was similarly white in colour and if this is cladded base-metal or whatever, it must be unusually thick for this purpose, or again entirely white-coloured metal.
Silver is not that costly, and I would have thought that the additional operations needed to fabricate such a bi-metallic part-silver contact would be more-costly than the tiny extra amount of silver needed for a solid silver contact.
I began my career as a manufacturing silversmith/jeweller nearly 50 yrs ago, and I could still positively test this construction throughout if I get hold of another new switch sometime. They are a bit costly to deliberately destroy to satisfy one's curiosity, so I accepted that their construction is silver throughout.
Another point about the dual wiper arrangement which I like is that the wiper's tips are arranged 'edge on' so that a relatively small cross-sectional area almost like a knife-edge literally cuts through any dross (mostly silver sulphate) which will accumulate during normal use on the surfaces of the ring of static contacts.
I don't for one moment think that you will find any relay which is sonically as transparent and non-destructive to the delicate signals, unfortunately.
Regards,
Congratulations to Nelson Pass for his great reviews in April/May issue of 'The Absolute Sound' for his amp and preamp.
If you don't want to listen to Charles Hansen or me about audio design, at least give Nelson some serious consideration.
If you don't want to listen to Charles Hansen or me about audio design, at least give Nelson some serious consideration.
Hi John,
When I learn, I like to have the supporting evidence that goes with the lesson. That's why we did experiments in school and university, to have first hand knowledge and understanding. Otherwise, it becomes an argument of faith alone. That doesn't sit well, especially in view of all the audio myths that abound.
The process of teaching must include the references and proof along with the statement. I asked about what you found as a problem with mercury whetted relays.
I guess my question should have been more specific then. John, have you run any tests at all on a relay with mercury whetted contacts that you chose as a reasonable candidate for audio switching? Ignoring capacitive coupling between contacts, was there any problem with the signal passing through a relay such as this? You could always use one relay per channel per input if everything else checked out okay. They even have bistable relays that only require a pulse to change state. This would act as a "memory" if the unit were to be turned off (perish the thought!). The other benefit would be that continuous coil current does not flow (eliminating electronic noise from the coil) and the signal could be muted during the changeover.
Hey! I thought that was a good idea!
Anyway, I do appreciate the quote you dug up John, but I wanted the answer to be clear, without the other variables the situation referred to in the quote.
If you haven't tested mercury relays, that's okay. I just wondered if you knew off hand. You generally have a good reason for making a statement.
-Chris
That isn't at issue at all.
When I learn, I like to have the supporting evidence that goes with the lesson. That's why we did experiments in school and university, to have first hand knowledge and understanding. Otherwise, it becomes an argument of faith alone. That doesn't sit well, especially in view of all the audio myths that abound.
The process of teaching must include the references and proof along with the statement. I asked about what you found as a problem with mercury whetted relays.
I don't know what this "Spiegel box" is, but I'll assume that it is a source (input) switcher or selector. The higher capacitance should be expected since the mercury is a conductive fluid and remains in close proximity to the contacts. This may or may not be a problem depending on the particular relay used and the impedance levels the signals operate at. Not only that, but it is unclear whether the 30 pF figure includes all the wiring and jacks as well, or only the relay. It appears as though the problem is the "Spiegel box" in it's entirety. The fact that it used mercury whetted relays may or may not have been a factor the way this quote is worded.
I guess my question should have been more specific then. John, have you run any tests at all on a relay with mercury whetted contacts that you chose as a reasonable candidate for audio switching? Ignoring capacitive coupling between contacts, was there any problem with the signal passing through a relay such as this? You could always use one relay per channel per input if everything else checked out okay. They even have bistable relays that only require a pulse to change state. This would act as a "memory" if the unit were to be turned off (perish the thought!). The other benefit would be that continuous coil current does not flow (eliminating electronic noise from the coil) and the signal could be muted during the changeover.
Hey! I thought that was a good idea!
Anyway, I do appreciate the quote you dug up John, but I wanted the answer to be clear, without the other variables the situation referred to in the quote.
If you haven't tested mercury relays, that's okay. I just wondered if you knew off hand. You generally have a good reason for making a statement.
-Chris
john curl said:
So again, I ask - how come the experience allows only a selected few to hear the "relay distortion" while others are apparently oblivious to that "fact" ?
I mean one can hardly call people like Pass, Manley, Johnsons, Candy et al. "inexperienced"? (yet they all include relays in their cost-no-object creations)
Not to mention numerous reviewers (some extremely golden-eared!) who praise and award relay based gear.
And of course clients (but they are all deaf, of course 😉 )
- Status
- Not open for further replies.