that was quite funny, and out of order
Huh? Have I been misunderstood? Not sure if this is directed at my post or not, but it seems like it was . . .
If you are in fact referring to my question about "junk" LDRs, it was an honest question, intended to elicit an answer to "When should one reject an LDR as unusable for this purpose?" and if Uriah is willing, I'd like to hear the answer. If you're talking about another post entirely, sorry for my confusion and please disregard this post.
BTW I'm totally new around here, don't want to step on any toes -- please tell me what was out of order, because I honestly can't see it.
I thought you meant that "junk" seemed typical, in reply to udailey saying that theres no "typical" LDR
Maybe I have misunderstood you
I think its a fine question. I have two different times I use the term Junk LDR. The Junk LDRs I sent to Silonex were the type that when you get them and test them they are TOTALLY screwed up. Like testing their minimum and getting 1k and at the same power that a good LDR would show 10k the Junk LDRs show 300k. So I sent those off.
Junk LDRs, like the ones I sell for cheap on my site are those that I have tested that have rather 'normal' ranges but that dont match any other LDR and they have been through two rounds of testing against 920 different LDRs. They are obviously useful as a single LDR but also will obviously never match another LDR and are useless to me. Some people buy them just to screw around experimenting. A good place to use these would be the input resistor on an amp. The Rf on a current feedback amp. An input switcher, etc. Places where only one resistor is needed and where a low value such as 100-10k is needed.
Keep asking questions as long as you need to. You may be new but you are not stepping on toes and I dont think Tinitus thought anything was wrong with your post.
Uriah
Junk LDRs, like the ones I sell for cheap on my site are those that I have tested that have rather 'normal' ranges but that dont match any other LDR and they have been through two rounds of testing against 920 different LDRs. They are obviously useful as a single LDR but also will obviously never match another LDR and are useless to me. Some people buy them just to screw around experimenting. A good place to use these would be the input resistor on an amp. The Rf on a current feedback amp. An input switcher, etc. Places where only one resistor is needed and where a low value such as 100-10k is needed.
Keep asking questions as long as you need to. You may be new but you are not stepping on toes and I dont think Tinitus thought anything was wrong with your post.
Uriah
Nah, then you would definately know, and have felt it, unmistakenly
The mod job can be difficult at times, but the worst to me is that it also tend to make it even harder to just be an ordinary member
And being a member will always be my main preference
Now thats said, lets not waste more time on this small misunderstanding
i thought i'm the only person who have problems with these LDR. i use them for input switching and not attenuation. unfortunately they just randomly change in resistance or die prematurely. i told Silonex the problem and they said i'm the only one having this problem. so much for Silonex and i still have a few hundred lying around doing nothing!
Good grief, between the two of you, one wonders if this device is worth designing around.
An unsatisfactorily high rate of out-of-tolerance rejects is not good, but if you're suffering a high rate of in-circuit failures, that is not OK. Are you driving them beyond the 20ma-40R specification? The randomly changing resistance sounds the same as Uriah's reason for avoiding sustained high current operation.
Oh its worth it! Very very worth it. I have sold hundreds of quads of LDRs and if people follow the rules about the right amount of power to them things seem to be fine. I have not had even one in circuit failure in my own devices. In testing I have had some fail and the one that exploded I was attempting to push it to its limits and thats how it showed me. My only complaint with the LDRs is that they are all manufactured by the same process but come out so different from each other. Yeah its been great being able to sell so many LDRs because people need them matched, but if they were not so individually unique I would have a LOT more time on my hands to design more circuits around them which would be a lot more fun. What IS really fun though is having people email me about how their system has NEVER sounded better. I have had two people mention that they broke down in tears over the music. Recently two speaker designers, both in business since the mid 80's, both told me nearly the same words "my speakers have never sounded so good." That is, to me, huge fun so I have a love/hate relationship with these little buggers.
I fully believe that George has had nearly zero failures if he has had any at all. His circuit works very well and protects the LDRs very well. Plus he gets to test the LDRs and pull out the weird ones. When you pull out the odd acting LDRs now you have left only the LDRs that are acting as they should which means somehow they got out of Silonex manufactured correctly.
Uriah
Very very worth it. I have sold hundreds of quads of LDRs and if people follow the rules about the right amount of power to them things seem to be fine. I have not had even one in circuit failure in my own devices. In testing I have had some fail and the one that exploded I was attempting to push it to its limits and thats how it showed me. My only complaint with the LDRs is that they are all manufactured by the same process but come out so different from each other. Yeah its been great being able to sell so many LDRs because people need them matched, but if they were not so individually unique I would have a LOT more time on my hands to design more circuits around them which would be a lot more fun. What IS really fun though is having people email me about how their system has NEVER sounded better. I have had two people mention that they broke down in tears over the music. Recently two speaker designers, both in business since the mid 80's, both told me nearly the same words "my speakers have never sounded so good." That is, to me, huge fun so I have a love/hate relationship with these little buggers.I fully believe that George has had nearly zero failures if he has had any at all. His circuit works very well and protects the LDRs very well. Plus he gets to test the LDRs and pull out the weird ones. When you pull out the odd acting LDRs now you have left only the LDRs that are acting as they should which means somehow they got out of Silonex manufactured correctly.
Uriah
If I measure the two dozen or so devices I got from you and averaged the curves, is that a large enough sample that the resulting composite curve would be representative of the middle of the range of all curves, or is the variation so big that I need a bigger sample?
I really want to start by looking at a "typical" curve, even if the variation around the curve is quite large.
Well, I have not done that. I dont see how it would help me in matching devices. But for the knowledge of it.. yes I think you could get a decent representation. Throw away the measurements of the crazy ones or they will throw off your normalized curve because you would never use them anyway.
Uriah
Uriah
Ultimately, it might not be necessary for me, either, but I need to start with a "screen" step that will allow me to select out devices that are not worth messing with, so I feel as though I need to be able to recognise the bad ones.
I (think!) I know that I need to reject devices that don't reach 40 ohms at 20ma, and I (think!) I know that I need at least 1.5 mega ohms at power off, but after that, what's left to worry about? That's why I want to start with a known 'average' curve.
Dont forget that this is 20mA at a certain voltage. Change the voltage and the current changes to. I think if you just plug them all in with a 1k resistor going to each and give some voltage you will near instantly see the ones that will NOT work at all. Weed them out and focus on the ones that are somewhat similar, using either the same voltage and different resistors or the same resistor and different voltages. Its easier to replicate if you use the same voltage and different resistors, especially since even .01V different will change the reading considerably at certain resistances.