chascode said:
Hi,
I couldn't agree more with these 2 comments. Regrettably obtaining relays which do have a significant wiping action are not so easy to come by, in my experience.
Also, as I have posted elsewhere recently and a while ago, there are other considerations when comparing relays with simple rotary switches.
When using the sonically-excellent Shallco switches, the other end of the same piece of silver material (about 4mm distant from where the input/output connections are soldered directly to) is actually the silver wiping contact, itself. Therefore, you cannot get a shorter more straightforward signal path than this, and the wiping action is right across the entire contact's surfaces from one side to the other.
Compare this with (almost) any usual relay, and you inevitably have several junctions between dissimilar metals, and a far more convoluted path for the signal to traverse. This is due to the virtually enforced requirements for different properties of the metals involved, some of which need to be springy, some hard and resistant to wear/corrosion pitting, and so on.
As I already mentioned elsewhere, if you dissemble some of the 'usual suspects' after a period of use, unless all these junctions are either spot or friction welded (and most I have inspected internally are not) some discolouration will usually be seen where the riveted metals are mated together. This cannot normally be seen without destructive dissembly and seems to be something like oxidation/sulphation, and which is likely to give rise to non-linear distortion, which is clearly undesirable.
I guessed that this unwanted discolouration was possibly due to some kind of galvanic action with dissimilar metals and the passing of currents, but with Pavel's relevant experience here, maybe he can clarify this.
For certain, the usually-espoused comments that when metals are securely rivetted/crimped together in this manner it will eliminate all contamination as air etc is excluded, are not always realistic as some of my (destructive!) inspections have shown me.
Most of the relays I have inspected in this way have been hermetically sealed, but I have not looked at any which are inert-gas filled, nor mercury filled or reed-relays. However, as this is not due to oxidation as far as I can tell, I doubt that they would fare any better over a period of time in the field, if they also have these dissimilar metal internal junctions unless they are welded (i.e. the parent metals actually melted together).
Regards,
jacco vermeulen said:
Hi jacco,
If you are suggesting by this that some sealed relays are nitrogen-filled, then I am sure you are probably right.
I am uncertain if nitrogen is considered to be truly "inert", and I am certainly not going to the bother of checking up on this, just to satisfy myself. 😉
Regards,
In the naval business it's considered inert, not according to the definition i remember from highschool.
The reason i posted the question is because the NAIS/Panasonic/Aromat/SDS (D)S series is nitrogen filled, curious of your findings of those relay series.
The reason i posted the question is because the NAIS/Panasonic/Aromat/SDS (D)S series is nitrogen filled, curious of your findings of those relay series.
Hi jacco,
For now, and unless something new to me turns up from somewhere, I have completed any relay investigations. However, next time I saw the top off an enclosed relay, I'll remember not to take a sniff! 😱 Maybe it was lucky I never joined the Navy.
I wonder if they were ever filled with Helium, would it affect the frequencies through them like with human vocal chords? 😉
Regards,
For now, and unless something new to me turns up from somewhere, I have completed any relay investigations. However, next time I saw the top off an enclosed relay, I'll remember not to take a sniff! 😱 Maybe it was lucky I never joined the Navy.
I wonder if they were ever filled with Helium, would it affect the frequencies through them like with human vocal chords? 😉
Regards,
Re: Low-level relay contacts (also known as dry circuits):
"Dry circuits are low-level loads whose voltage and current are so low they do not cause any electrical erosion of the contacts. If the contacts oxidize, however, there may be enough resistance to prevent proper operation of the switched circuit. Precious metals are used in relay contacts intended to switch dry circuits. This ensures that low contact resistance is maintained over the mechanical life of the relay, which can be twice as long as the rated electrical life. Dry loads are defined as those with a maximum current and voltage of 1 mA and 50 mV, respectively. However, once relay contacts have been used to switch high currents the precious metal coating can be damaged so they are no longer suitable for dry service."
Best Regards,
Chuck Hansen
"Dry circuits are low-level loads whose voltage and current are so low they do not cause any electrical erosion of the contacts. If the contacts oxidize, however, there may be enough resistance to prevent proper operation of the switched circuit. Precious metals are used in relay contacts intended to switch dry circuits. This ensures that low contact resistance is maintained over the mechanical life of the relay, which can be twice as long as the rated electrical life. Dry loads are defined as those with a maximum current and voltage of 1 mA and 50 mV, respectively. However, once relay contacts have been used to switch high currents the precious metal coating can be damaged so they are no longer suitable for dry service."
Best Regards,
Chuck Hansen
Thanks, Bob, Chuck and Jacco. You have saved me a lot of effort to convey this info. I hope that it clarifies some of the differences between a quality Shallco switch and a good minature relay. For a real surprise, check out (in principle) what a mecury relay is fabricated with. Talk about dissimilar materals!
Protection Relays (Krell KST-100)
Since we are talking about relays and contacts, I would like to share my experience with an amp that I own; at the same time, may be you folks could confirm if this could really cause the problem that I describe below.
I own a Krell KST-100 power amp and have been experiencing some problem when listening at low level, especially if listening to soft piano music - the sympton is; occasionally, I'll hear "ssshh" sound cutting in and out; it'll repeat the "ssshh" sound if I replay the same musical passage.
After searching the web to try to find a solution; it seems that a lot of people are suggesting that it could be bad capacitors at the pre-driver board; even contacting Krell for a service manual (Krell charge $400 for the repair); but didn't get anyway. One day, I thought, why don't I open up the plastic covers of the two heavy duty relays (10 seconds after turn on, it'll click to turn on the output) and use some fine sandpaper to clean out the relay contacts, which is what I did. After I clean the relay contacts, all problem goes away!!
I still can't believe that a relay that looks pretty normal, no oxidation or rust that I can see of, could be causing the problem. Is this what you folks are talking about when you talk about using relays in audio products?
Since we are talking about relays and contacts, I would like to share my experience with an amp that I own; at the same time, may be you folks could confirm if this could really cause the problem that I describe below.
I own a Krell KST-100 power amp and have been experiencing some problem when listening at low level, especially if listening to soft piano music - the sympton is; occasionally, I'll hear "ssshh" sound cutting in and out; it'll repeat the "ssshh" sound if I replay the same musical passage.
After searching the web to try to find a solution; it seems that a lot of people are suggesting that it could be bad capacitors at the pre-driver board; even contacting Krell for a service manual (Krell charge $400 for the repair); but didn't get anyway. One day, I thought, why don't I open up the plastic covers of the two heavy duty relays (10 seconds after turn on, it'll click to turn on the output) and use some fine sandpaper to clean out the relay contacts, which is what I did. After I clean the relay contacts, all problem goes away!!
I still can't believe that a relay that looks pretty normal, no oxidation or rust that I can see of, could be causing the problem. Is this what you folks are talking about when you talk about using relays in audio products?
Re: Protection Relays (Krell KST-100)
Nobody's mentioned the films that build up on the contacts, not oxidation. This is the reason the wiping action of the rotary switch is beneficial; breaks through the films allowing a fresh connection.
Mike.
fcel said:
Nobody's mentioned the films that build up on the contacts, not oxidation. This is the reason the wiping action of the rotary switch is beneficial; breaks through the films allowing a fresh connection.
Mike.
Such a post is not really fitting in a thread about a CTC specialty product.
As an rough example it's ok, an open style relay with silver(alloy) contacts has no place in high quality audio gear.(imo)
The LY4 relay that Krill used in the 80/90s used to be experimented with overhere by sealing and injecting it with gas, still remains a 10 dollar industrial relay.
As an rough example it's ok, an open style relay with silver(alloy) contacts has no place in high quality audio gear.(imo)
The LY4 relay that Krill used in the 80/90s used to be experimented with overhere by sealing and injecting it with gas, still remains a 10 dollar industrial relay.
Series, ladder or shunt
John,
Shallco's are probably the best rotary switches. I am using Seiden and ELMA rotary switches but will try the Shallco's in a near future.
What no one has mentioned as far as I can remember, is how to configure the rotary switch with resitors, ie ladder style, fixed series/variable shunt, or series attenuator (DACT style). I have been using all three versions, but my last series attenuator was only used in the mid 80s.
How do you prefer to connect a stepped attenuator using a rotary switch?
Attached is my latest stepped attenuator (K & K Audio's fixed series/variable shunt using a Seiden switch, I have used PRP reistors for it.)
Sigurd
John,
Shallco's are probably the best rotary switches. I am using Seiden and ELMA rotary switches but will try the Shallco's in a near future.
What no one has mentioned as far as I can remember, is how to configure the rotary switch with resitors, ie ladder style, fixed series/variable shunt, or series attenuator (DACT style). I have been using all three versions, but my last series attenuator was only used in the mid 80s.
How do you prefer to connect a stepped attenuator using a rotary switch?
Attached is my latest stepped attenuator (K & K Audio's fixed series/variable shunt using a Seiden switch, I have used PRP reistors for it.)
Sigurd
Attachments
Hi Sigurd,
The UK-sourced Blore Edwards switches are very good sonically, too. One of their range is very similar to the 'usual' Shallco, both in overall design and the silver contacts etc., and 'sounds' quite similar to me in a direct comparison.
Also, for anyone wishing to use these switches in 'banks' for multiple channels etc., the Blore Edwards is easier to reduce the tension of the detents than is the case with the Shallcos.
I have modified both with success, but with the B E design it is a simple matter of just removing one of the original two springs, so nothing could be easier.
My own preference with any such switched attenuators is always to avoid any multiple contacts and/or resistors in series or in a 'chain' anywhere, as these additional soldered joints/contacts/resistors etc., are not helpful in achieving the ultimate in transparency.
If the added complication is acceptable (and this can be a pain if one is using balanced connections in a stereo set-up, or worse in multi-channel) then a 'true' ladder with switched series and shunt positions is probably the best alternative.
However, if the local circuit requirements will tolerate 'seeing' varying impedances, then the fixed series and varying shunt types also give good results. There is the benefit with the latter that the series R can be fixed (soldered) which, if anything, will likely assist with HQ sonics as this avoids a contact in the 'direct' signal path.
Regards,
The UK-sourced Blore Edwards switches are very good sonically, too. One of their range is very similar to the 'usual' Shallco, both in overall design and the silver contacts etc., and 'sounds' quite similar to me in a direct comparison.
Also, for anyone wishing to use these switches in 'banks' for multiple channels etc., the Blore Edwards is easier to reduce the tension of the detents than is the case with the Shallcos.
I have modified both with success, but with the B E design it is a simple matter of just removing one of the original two springs, so nothing could be easier.
My own preference with any such switched attenuators is always to avoid any multiple contacts and/or resistors in series or in a 'chain' anywhere, as these additional soldered joints/contacts/resistors etc., are not helpful in achieving the ultimate in transparency.
If the added complication is acceptable (and this can be a pain if one is using balanced connections in a stereo set-up, or worse in multi-channel) then a 'true' ladder with switched series and shunt positions is probably the best alternative.
However, if the local circuit requirements will tolerate 'seeing' varying impedances, then the fixed series and varying shunt types also give good results. There is the benefit with the latter that the series R can be fixed (soldered) which, if anything, will likely assist with HQ sonics as this avoids a contact in the 'direct' signal path.
Regards,
Hi Bob,
Mostly agree with what you said and have a couple of things to add:
a) The easiest way to change the spring tension on the Shallco is to take it off and stretch it. However, if you stretch it too far you will have to replace it, as you cannot "de-stretch" it.
b) I disagree with the preference for a fixed series and variable shunt. This gives a variable input impedance, which is noticeably audible to me. On the other hand, even though the idea of a string of resistors seems bad (for a traditional pot-like divider) there is almost no sonic degradation if you are using good resistors.
Mostly agree with what you said and have a couple of things to add:
a) The easiest way to change the spring tension on the Shallco is to take it off and stretch it. However, if you stretch it too far you will have to replace it, as you cannot "de-stretch" it.
b) I disagree with the preference for a fixed series and variable shunt. This gives a variable input impedance, which is noticeably audible to me. On the other hand, even though the idea of a string of resistors seems bad (for a traditional pot-like divider) there is almost no sonic degradation if you are using good resistors.
Charles Hansen said:
Hi Charles,
Yes, it is an old dodge to stretch these detent springs, but not one which I generally would use, especially if several need to be done when as you say, there is no going back! I replace them with some which are weaker, and always consistent.
As to my "preference", this is not actually as you suggest here, but maybe my choice of wording wasn't as clear as it should have been. However, I did say that the true ladder switched *series and shunt* type was "the best alternative" and of course this can be made with a constant input impedance, which we both seem to prefer.
I finalised a design (unfortunately using relays, as the application demanded this) for a high-end UK speaker manufacturer over a year ago, and by careful choice of R values for series and shunt positions, the variation in input impedance was virtually zero. The resistors are made by Texas Components, and being all separately trimmed during manufacture (at no additional cost) I opted for values accurate to 6 decimal places, with a tol. of 0.1%! An example I just picked out of the bin being 38k6212, so you can imagine that this stepped attenuator is extremely accurate both in true 1dB 'steps', and in balance across 6 channels.
Regards,
Hello Bob, thanks for the info. A quick search did not bring up any supplier, do you know whether and where these are still available, and at what price? Thanks in advance.
Hello Charles, again thanks for your info. The TKD stepped attenuator appears to be that configuration (string of resistors). Did you try them and if so what is your opinion, (also on their cheaper carbon comp pots if you did try them)? Just as an aside, TKD are also producing leaded resistors, but the only source I found is asking EUR 1,50 per piece, so too expensive for me.
Hi Martin,
I don't believe that Blore Edwards have any stockists as such, although I could be wrong here, as their switches are (or were) made to order when I last purchased some about 4/5 yrs ago.
If you cannot contact them via the Internet, let me know and I will dig out my invoices from that time and give you some contact details. They are (or were) in Wales, in the UK.
Regards,
P.S. I just did a Google search, and you will find their info easily.
I don't believe that Blore Edwards have any stockists as such, although I could be wrong here, as their switches are (or were) made to order when I last purchased some about 4/5 yrs ago.
If you cannot contact them via the Internet, let me know and I will dig out my invoices from that time and give you some contact details. They are (or were) in Wales, in the UK.
Regards,
P.S. I just did a Google search, and you will find their info easily.
One of the reasons for my moving away from ELMAs is that I think that the contact-bounce is the source for my having irritating click sounds when changing volume.
Whatever electrical solution I tried, I still have those irritating click sounds using my ladder attenuator.
Are the Shallcos better in this parameter?
To avoid misunderstandings regarding what type of stepped attenuator is which,
I attach a schematics showing what I call a fixed series / variable stepped shunt attenuator for balanced use.
Bobken,
I guess you are referring to the naked TX2352 resistors. If so, that is a very fine choice for attenuator resistors. Albeit rather costly.
I have used them extensively, and they are my favorite resistor. Not as fragile as some think. I find them sturdy enough.
Sound is so clean and detailed that other resistors seem to create a sonical fog.
Charles,
I read your comments about resistors in another thread which has been closed, and I would like to say that the Caddock MK132s are not in the same league as the TX2352s. The MK132 have shown to have an unflattering sound in the lower bass region which manisfests itself as a "fat" bass sound. Details are more easy to find with the TX2352s, and I find that the MK132 colour also the midrange and treble.
If you have the time, give them a try.
Costly but not expensive 😉
SIgurd
Whatever electrical solution I tried, I still have those irritating click sounds using my ladder attenuator.
Are the Shallcos better in this parameter?
To avoid misunderstandings regarding what type of stepped attenuator is which,
I attach a schematics showing what I call a fixed series / variable stepped shunt attenuator for balanced use.
Bobken,
I guess you are referring to the naked TX2352 resistors. If so, that is a very fine choice for attenuator resistors. Albeit rather costly.
I have used them extensively, and they are my favorite resistor. Not as fragile as some think. I find them sturdy enough.
Sound is so clean and detailed that other resistors seem to create a sonical fog.
Charles,
I read your comments about resistors in another thread which has been closed, and I would like to say that the Caddock MK132s are not in the same league as the TX2352s. The MK132 have shown to have an unflattering sound in the lower bass region which manisfests itself as a "fat" bass sound. Details are more easy to find with the TX2352s, and I find that the MK132 colour also the midrange and treble.
If you have the time, give them a try.
Costly but not expensive 😉
SIgurd
Attachments
Bobken said:
Yes, that is a good approach. Since Shallco is in the US, it is easy for us to obtain extra springs, but your solution is better for most DIY'ers. Then they can try a variety of springs until they find the one they like best. One bad thing about the Shallco design is that the geometry of the detent mechanism is such that the force in one direction is different from the force in the other direction.
In our newest design, we have custom Shallco switches with no detent mechanism. The switches are only controlled via precision stepper motors and toothed timing belts and pulleys. So the worst thing about the Shallcos is completely avoided.
Bobken said:
Sorry I didn't make myself clear. I understood your preference for a true ladder. When using switches as transparent as the Shallcos, this is an excellent approach, as the ladder design adds an additional switch contact to the signal path.
I was instead comparing a shunt attenuator to a divider attenuator. One would think that all of those resistors in series of a divider attenuator would degrade the sound. But I found that not to be the case, even with ordinary "good" resistor, such as the Roedersteins. I was a bit surprised by the result. On the other hand, the difference between 10 kohm and 100 kohm and 1 Mohm input impedance is quite significant sonically. Obviously the change of impedance of a shunt attenuator will vary depending on the design, but I am leery of such an approach.
Bobken said:
If I remember correctly, these are the "naked Vishays" that cost around $5 each. I don't know how many you had to use, but that adds up very quickly -- perhaps too quickly for all but the most costly of commercial products. However, it's not out of reach for a DIY design.
- Status
- Not open for further replies.