I assume you had 2 channels of series/shunt set up. Then the most total current it can draw is around 60ma. So, even 102ma is way too high. ou probably had a different circuit. Can you draw a picture of it?
I heard that many DMM's are not very accurate to measure small current. It is much accurate to measure the voltage of a series resistor and use ohm's law to figure out the current. I found this is indeed quite true especially with my el cheapo DMM's - can't afford to use many flukes at once 🙂
Missing trimmers?
Uriah,
Are you recommending that these be left out? I am keeping up with your developments.
Just noticed they were not there and wondered if they were left out for clarity of the wiring or it is felt they are not needed.
Thanks,
Rick McInnis
Uriah,
Are you recommending that these be left out? I am keeping up with your developments.
Just noticed they were not there and wondered if they were left out for clarity of the wiring or it is felt they are not needed.
Thanks,
Rick McInnis
Rick,
We just found a weakness and are trying to track down what is going on in the circuit. So right now its all conjecture.
I had no need for trimmers in my last matched set. You probably dont need them either. The matching is really good. However, a trimmer will not get in the way or disturb anything if it is on the LDR side of the pot. You can try to use it for balance control if your room needs it.
A trimmer in front of the pot could be messed with as an experiment if used as a variable resistor and not a pot. You would want its total value to be 100R or less. It will change the impedance of the circuit.
Right now as always, the way George lays out the circuit is great. It works really well. Maybe we should have picked a different thread for this messing around but we are already posting so much on it it would be pointless to move somewhere else.
Uriah
We just found a weakness and are trying to track down what is going on in the circuit. So right now its all conjecture.
I had no need for trimmers in my last matched set. You probably dont need them either. The matching is really good. However, a trimmer will not get in the way or disturb anything if it is on the LDR side of the pot. You can try to use it for balance control if your room needs it.
A trimmer in front of the pot could be messed with as an experiment if used as a variable resistor and not a pot. You would want its total value to be 100R or less. It will change the impedance of the circuit.
Right now as always, the way George lays out the circuit is great. It works really well. Maybe we should have picked a different thread for this messing around but we are already posting so much on it it would be pointless to move somewhere else.
Uriah
Hi Uriah,
When I measured the total current through the all 4 LDR's, I measured the voltage across the 4 100R's and used ohm's law to calculate the each current and sum them up. I never got the total above 70ma at any position of the pot. I was quite surprised to see yours settle at 102.
If I am not mistaken, the DMM should be in series with the circuit. In your diagram, it is in between the pot and the 5VDC. How did you get the current reading?
Cheers,
Fred in my diagram I did not show the DMM.
I put the lead of the DMM on one side of the resistor and the other lead on the other side of the resistor. One lead on each leg.
So what I SHOULD get is about 28mA so X4 would be 112mA. I wonder if somehow I was measuring total current of the circuit.
So now I am messing again with LM334 and thinking up a PCB for testing LDRs with LM334.
I prefer pots for matching because they are easy to dial in very precisely. But it might be nice to use .01% resistors with a switch. But I have never dialed in as bad as .01. Usually I am .006% with the trimmers since I use 25 turn pots its easy, but it makes matching take a real long time because I have to dial them in then let them sit then check them and fix any readings that have changed then test, then change them all to the new reading. I was thinking I could use .01% resistors in parallel but man they are expensive. I will figure it out.
I think I am pretty well convinced that George's circuit with maybe 200R instead of 100R is the best Pot/Resistor solution and that it might be nice to put a 100R trimmer in series with 5VDC if you want to tweak the impedance.
Fred, Gianni, do you guys concur or do you think there is more work to do here?
I think any improvements lie outside the pot/resistor circuit.
Only other thing I can think of right now is maybe use one LM7805 per LDR pair and see if we get rid of that weirdness where the 4th LDR in the circuit drives resistance . I should test this.
Uriah
I put the lead of the DMM on one side of the resistor and the other lead on the other side of the resistor. One lead on each leg.
So what I SHOULD get is about 28mA so X4 would be 112mA. I wonder if somehow I was measuring total current of the circuit.
So now I am messing again with LM334 and thinking up a PCB for testing LDRs with LM334.
I prefer pots for matching because they are easy to dial in very precisely. But it might be nice to use .01% resistors with a switch. But I have never dialed in as bad as .01. Usually I am .006% with the trimmers since I use 25 turn pots its easy, but it makes matching take a real long time because I have to dial them in then let them sit then check them and fix any readings that have changed then test, then change them all to the new reading. I was thinking I could use .01% resistors in parallel but man they are expensive. I will figure it out.
I think I am pretty well convinced that George's circuit with maybe 200R instead of 100R is the best Pot/Resistor solution and that it might be nice to put a 100R trimmer in series with 5VDC if you want to tweak the impedance.
Fred, Gianni, do you guys concur or do you think there is more work to do here?
I think any improvements lie outside the pot/resistor circuit.
Only other thing I can think of right now is maybe use one LM7805 per LDR pair and see if we get rid of that weirdness where the 4th LDR in the circuit drives resistance . I should test this.
Uriah
Is this means at each lead after the 100K resistors?
An externally hosted image should be here but it was not working when we last tested it.
BTW I put extra resistor before the 100K pot as since some say 100R is not enough to protect the LDRs...what do you think Uriah? This is just like I said previously to you...
Last edited:
No I tested on each resistor on both sides of the individual resistor.
I think I will not use the 1k any longer Gabanyaya. It gives that crazy impedance curve. I would either leave it out or put in tiny resistance of 10-30 Ohms. This will change the impedance curve a LOT. For instance putting 22R can change 11k into 35k or somewhere around there.
There is no protecting the pot. It cant be done without changing the curve a lot. Just use a pot that works for you and replace it when its dead. George says his LSA has been in service for years without damage to the pot so lets not worry to much. Lets worry about the LDRs. In 5 years will I be here matching for anyone ? I hope not 🙂 We can all do it ourselves but it is a pain and time consuming. So I suggest you consider a higher value in front of the LDR if it is of concern to you. A ~250R resistor will be perfect for protection in front of each LDR. I dont think it will present the wild impedance swings we get with the series resistor. I have not tested this.
Uriah
I think I will not use the 1k any longer Gabanyaya. It gives that crazy impedance curve. I would either leave it out or put in tiny resistance of 10-30 Ohms. This will change the impedance curve a LOT. For instance putting 22R can change 11k into 35k or somewhere around there.
There is no protecting the pot. It cant be done without changing the curve a lot. Just use a pot that works for you and replace it when its dead. George says his LSA has been in service for years without damage to the pot so lets not worry to much. Lets worry about the LDRs. In 5 years will I be here matching for anyone ? I hope not 🙂 We can all do it ourselves but it is a pain and time consuming. So I suggest you consider a higher value in front of the LDR if it is of concern to you. A ~250R resistor will be perfect for protection in front of each LDR. I dont think it will present the wild impedance swings we get with the series resistor. I have not tested this.
Uriah
Uriah
Nice thing that you keep working on it, and others too
I like the conclusion of using series res on each led, like George recommend
And sure, to change a pot is easy and cheap
Changing LDR is not
As long as a worn out pot doesnt burn something
Ok, we are back to 2-gang pots
With the first 500k LDR, is 500k pots still needed ?
Though, 500k 0.25Watt is available in 2-gang
I suppose 0.25Wat 2-gang is as good 0.5watt mono pot, in terms of power
And it has a mute switch when closed fully, which in Lightspeed would cut off the 5V power to LDR 🙄
Nice thing that you keep working on it, and others too
I like the conclusion of using series res on each led, like George recommend
And sure, to change a pot is easy and cheap
Changing LDR is not
As long as a worn out pot doesnt burn something
Ok, we are back to 2-gang pots
With the first 500k LDR, is 500k pots still needed ?
Though, 500k 0.25Watt is available in 2-gang
I suppose 0.25Wat 2-gang is as good 0.5watt mono pot, in terms of power
And it has a mute switch when closed fully, which in Lightspeed would cut off the 5V power to LDR 🙄
Last edited:
Tinitus,
I think a single gang pot is not the answer. It works for sure but it does not work as well as dual.
500k is the size trimpot I tested with. Strangely when 3 LDRs are in the attenuator circuit they behave as in my testing. When the 4th is inserted they ALL jump in resistance considerable. They still match well but are now nearly 3 times higher in resistance. Thats a lot. So I think you can do well with a 100k up to maybe 150k. Try whats in your drawer if you have a few pots but 100k WILL work for almost all of the matches I sent out. A few would be to low with 100k but almost all of them would be in that 7k and above range.
Remember Tinitus, you can use 100k and then put that tiny resistor in series and get rid of any 1k resistor in series with the 5VDC. Then you for sure need no other pot than 100k.
Uriah
I think a single gang pot is not the answer. It works for sure but it does not work as well as dual.
500k is the size trimpot I tested with. Strangely when 3 LDRs are in the attenuator circuit they behave as in my testing. When the 4th is inserted they ALL jump in resistance considerable. They still match well but are now nearly 3 times higher in resistance. Thats a lot. So I think you can do well with a 100k up to maybe 150k. Try whats in your drawer if you have a few pots but 100k WILL work for almost all of the matches I sent out. A few would be to low with 100k but almost all of them would be in that 7k and above range.
Remember Tinitus, you can use 100k and then put that tiny resistor in series and get rid of any 1k resistor in series with the 5VDC. Then you for sure need no other pot than 100k.
Uriah
Can I add that a 100k pot will let 2mA pass (using appropriate resistors) without damage to the pot?
I personally prefer 2 resistors after the pot and before the leds so to get rid of resistors tolerances with 1k and 2.2k i use. They will feed 2 LDRs at a time, really no problem and tested.
Uriah, I changed the pot, using a 220k 1/2 W and lowered down that 2.2k. Now i use 1k + 1.5k (series and parallel in the same order). I ended up with a stable 15k log volume, perfect to me. This situation is wonderful for the sound and for the life of the pot.
Very very little sound when full throttle down and 1,5mA is so safer for the pot.
At max position it draws 2,5mA, but when you just move 5% down, il will show less than 1mA again. I don't like 5-10k final volume value, i think it's too low for some sources.
15-20k is the best for me, it will work with source that have up to 1k out and i think any correctly designed amp. Even my_ref, sensible to input impedance, will work.
I agree that george layout is the best, but the matched LDRs that worked with 500k have so many advantages that we can use in favour of ourselves...
Regards,
Gianni
I personally prefer 2 resistors after the pot and before the leds so to get rid of resistors tolerances with 1k and 2.2k i use. They will feed 2 LDRs at a time, really no problem and tested.
Uriah, I changed the pot, using a 220k 1/2 W and lowered down that 2.2k. Now i use 1k + 1.5k (series and parallel in the same order). I ended up with a stable 15k log volume, perfect to me. This situation is wonderful for the sound and for the life of the pot.
Very very little sound when full throttle down and 1,5mA is so safer for the pot.
At max position it draws 2,5mA, but when you just move 5% down, il will show less than 1mA again. I don't like 5-10k final volume value, i think it's too low for some sources.
15-20k is the best for me, it will work with source that have up to 1k out and i think any correctly designed amp. Even my_ref, sensible to input impedance, will work.
I agree that george layout is the best, but the matched LDRs that worked with 500k have so many advantages that we can use in favour of ourselves...
Regards,
Gianni
Each resistor on both sides.....ok then...
~250R you said.....!!!! Since I already soldered 100R into a tiny board then i guess it's quite difficult to add ~150R into it now...
Hi Uriah,
According to your diagram, there can only be 2 LDR's drawing full current. So, even 112mA is way too high. I don't think the way you measured the current was correct. You DMM probes must be in series with the 100R, not parallel. I don't think this number is meaningful.
The LM334 idea is interesting. How would the wiring look like in the series/shunt case?
Hi Tintius,
0.25 is 0.25 watts, regardless to single or dual gang. However, as long as you don't leave the pot at the full closed and wide open position to long, the pot will be OK. In the normal operating position, only uA will pass through the pot. Everything we talked about is to protect the LDR's from accidental acts such as leaving the pot at fully closed position. Also, pot is cheap, too.
That is expected because when a LDR is added, it draws additional current which means the pot will consume more voltage; hence, all the LDR's have less voltage across them which effectively increases the resistance.
When I did my final round of matching, I matched 4 LDR's at a time with 4 DMM at each one of their legs. As I turned the pot, I watched the resistance readings. If any one of them varies too much from the others, I replaced the odd one with another candidate. As soon as I yanked the odd one out, all resistance readings immediate dropped. Once I put the new candidate in, they all came up very quickly but not stabilized until the newly added LDR settled.
I have to head to bed. To tired to think straight anymore.
I agree Fred the numbers must be meaningless in regards to mA.
So, Gaba, dont sweat the circuit. Everyone who has built an LSA so far has used 100R if they followed any of the advice on this forum. Yours is fine I am sure.
Gianni its good we have a few different values people can try if they have the same trouble with their 500k. If we have another pot lying around its probably in a range we can use.
Back to Fred, I ordered 2.2k pot, dual linear. Linear because it is what they had at that low resistance. I could probably find log no problem but I want to see the difference to.
So LM334 feeds 2 LDRs. So 2 LM334. Dual gang pot, each gang acts as Rset for one LM334 with A to W parallel to Adj and -. LDR + goes to LM334 - and then LDR- to GND.
Uriah
I agree Fred the numbers must be meaningless in regards to mA.
So, Gaba, dont sweat the circuit. Everyone who has built an LSA so far has used 100R if they followed any of the advice on this forum. Yours is fine I am sure.
Gianni its good we have a few different values people can try if they have the same trouble with their 500k. If we have another pot lying around its probably in a range we can use.
Back to Fred, I ordered 2.2k pot, dual linear. Linear because it is what they had at that low resistance. I could probably find log no problem but I want to see the difference to.
So LM334 feeds 2 LDRs. So 2 LM334. Dual gang pot, each gang acts as Rset for one LM334 with A to W parallel to Adj and -. LDR + goes to LM334 - and then LDR- to GND.
Uriah
the standard way to measure current passing a resistor is to monitor the voltage drop across the resistor, then use Ohm's law to calculate the current. I = V/R
Thanks Andrew, That the only thing that makes sense to do here as the DMM is not giving an accurate read.
I always appreciate your comments.
Uriah
I always appreciate your comments.
Uriah
Well a new testing mechanism is in the works. I am considering making mini test mechanisms for guys that want to test their own. PCB based. LM334 and parallel resistors which are switched in with a rotary switch. Another switch selects the LDR under test and reads the resistance. My setup at home will be a bit more elaborate. Testing of either 24, 48, 0r 96 LDRs at once not sure yet with onboard DMM.
Do you guys think anyone would have a use for a super accurate Pico/Nano Ammeter?
I am building one, obviously I need it, and wonder if you can think of reasons anyone else would use it.
Uriah
Do you guys think anyone would have a use for a super accurate Pico/Nano Ammeter?
I am building one, obviously I need it, and wonder if you can think of reasons anyone else would use it.
Uriah
- Status
- Not open for further replies.