Ok, so now I'm in analysis paralysis.
Am I decoupling the supplies, V+ and V- or decoupling the inverting and non-inverting inputs? I've read a few datasheets and looked at lots of circuits but can't tell definitively (usually the details of V+ and V- are omitted, and the caps attached to the +/- inputs could be a part of the circuit design, not decoupling).
I'm trying to figure this out to learn (best way), but I do need and much appreciate your advice.
Am I decoupling the supplies, V+ and V- or decoupling the inverting and non-inverting inputs? I've read a few datasheets and looked at lots of circuits but can't tell definitively (usually the details of V+ and V- are omitted, and the caps attached to the +/- inputs could be a part of the circuit design, not decoupling).
I'm trying to figure this out to learn (best way), but I do need and much appreciate your advice.
Buying from a recognised supplier is the best way... digikey should be fine... but the devices should still be warm (they shouldn't be cold).
For desoldering, its just pin 8, leave all the others soldered, and then carefully solder a 1 ohm to the board and the other end to the free unsoldered pin. All done in less time than it takes to type this. Honest 🙂
Decoupling refers to the voltage supplies on pin 8 and pin 4. I'm sure you will see small caps from each of those pins to ground if you look near each chip. That should be sufficient on a unit like this. You could if you wanted add a small (say 4.7uf 63v) electrolytic soldered in series with a 1 or 2.2 ohm 0.25 watt resistor across pins 4 and 8 of each opamp. You shouldn't need it for the 5532 and TL072 though.
I did not read beyond the first page. On top of the other concerns given, TL072 is a jFET input that has high input impedance, it works with higher resistance signal path, LM4562 is BJT input that draw a lot more current, you really need to look at the schematic to verify that this does not present a problem.
LM4562 is a whole lot faster than TL072, but if you said it sounds OK, I guess you don't have oscillation problem.
You must be a very slow typer 😱
Anyway a got the leg bent out, and put together a franken-resistor (4 x 4.7ohm). I may have mangled the leg too much to resolder, but that's why you buy extras 😉
So here are the results:
R=1.25 ohm (measured)
V=12.19mV (measured)
That leads to 9.8mA (approx).
Should I be comparing that to the quiescent current in the datasheet? If so, then it's good - datasheet says 10mA is typical.
I won't resolder and close it up just yet.
10ma is bang on the money. They must be warm at that because 0.01amps times the total supply voltage of 34 volts is 0.34 watts. Maybe your fingers need a bit more gain 😉
You should try solder braid if you intend working much on this kind of gear. One second per leg tops to desolder one of these little suckers.
You should try solder braid if you intend working much on this kind of gear. One second per leg tops to desolder one of these little suckers.
That is good news.
I use braid and a sucker - that's the easy part. Using tweezers to bend and pull the leg out of the hole was the tricky part. Would have been as quick as you say if it were an smd part.
I use braid and a sucker - that's the easy part. Using tweezers to bend and pull the leg out of the hole was the tricky part. Would have been as quick as you say if it were an smd part.
For my own edification, is the quiescent current the benchmark spec I should be comparing it to? The measured numbers didn't seem to change based on volume or other factors.
Ah... I see. I meant to just solder the resistor to the free bit of leg poking through the hole (lol, that sounds a bit risqué).
Yes, the current won't change much under signal conditions. Even if you had a volt or so signal it will probably be working into impedances around the 10k mark so very little change... and its a dynamic quantity which the meter wont show accurately.
@Mooly - I just saw your other thread on swapping opamps. Wish I had found that earlier, it is very good.
I'm going with jean-paul's suggestion of a Wima cap between +/- voltage supplies. It's just more space efficient than your suggestion of a cap and resistor in series.
Digikey doesn't carry Wima so I'll get an equivalent. I'm also going to pick up a cheap oscope probe.
Anyone have any other suggestions for my cart?
I'm going with jean-paul's suggestion of a Wima cap between +/- voltage supplies. It's just more space efficient than your suggestion of a cap and resistor in series.
Digikey doesn't carry Wima so I'll get an equivalent. I'm also going to pick up a cheap oscope probe.
Anyone have any other suggestions for my cart?
The Wima's will be fine. Scope probes don't come cheap these days and I wouldn't know what to recommend without trawling the catalogues tbh. You do need a divider type (10:1) preferably switchable as that gives the option of using the attenuated setting which gives low capacitance loading to the circuit under test.
I suspect you are probably looking for something like this as a minimum,
76-109 - TENMA - PROBE, O'SCOPE 60MHZ | CPC UK
I suspect you are probably looking for something like this as a minimum,
76-109 - TENMA - PROBE, O'SCOPE 60MHZ | CPC UK
OPA1641 (single) and OPA1642 (dual) are J-FET inputs, with 1.8mA quiescent per stage - same as TL071/2. Only surface mount though, so adapter needed, and they're quite expensive.
Haven't tried them yet but will soon, as replacements for TLs.
Haven't tried them yet but will soon, as replacements for TLs.
Hi all. I've finally got it all in place, thanks for your help!
I'd like to say the mids sound more airy and the bass is...different. I can't say that for sure. But I can say definitely that I had fun making the changes, learned a lot, and that's a large part of the diy experience for me.
Unfortunately I forgot to take some nice pix of all the changes. I do have some though.
I'd like to say the mids sound more airy and the bass is...different. I can't say that for sure. But I can say definitely that I had fun making the changes, learned a lot, and that's a large part of the diy experience for me.
Unfortunately I forgot to take some nice pix of all the changes. I do have some though.
Attachments
Thanks for the update (and pics 🙂). Pleased to hear it has all worked out OK, and I think on extended listening you will notice a difference.
My thoughts, exactly. I have found many counterfeit LM4562NA and RC4558P available. Be sure you are buying genuine/authentic OpAmps.
RC4558P counterfit? They are only 25 cents each from RS. I thought these were the ones that get remarked as something more exotic
They are now running warm. My only concern is the labeling on those on one of the boards has faded. Labelling is fine in the other board. I have no worries about them being fakes.
Purchased from digikey
Purchased from digikey
Amazingly, yes. All along the board you will find knockoffs. I have built my business selling, only, authentic, genuine brand name electronic components. Thats how much these fakes have effected my projects.
How do you test for fakes? Your prices (lm4562na) are lower than digikey, which normally would send up red flags. If it's easy I'd like to test one of mine, I have extras.
I try to keep my pricing cost effective and respectable, since I supply many hobbyists and DIYers which I can directly associate with.
With respect to testing, manufacturer datasheet tolerances are my key reference. I look at Intermodulation distortion, THD, SINAD, noise floor, CMRR, frequency response, full spectral analysis, thermal properties under variances of supply voltages... and I can go on and on.....
There are many cases where you can pick out fakes by simply comparing Silkscreen differentials. For instance, many PTC PT2399 IC fakes are silkscreened with "DTC" as oppose to "PTC" cleverly.
With respect to testing, manufacturer datasheet tolerances are my key reference. I look at Intermodulation distortion, THD, SINAD, noise floor, CMRR, frequency response, full spectral analysis, thermal properties under variances of supply voltages... and I can go on and on.....
There are many cases where you can pick out fakes by simply comparing Silkscreen differentials. For instance, many PTC PT2399 IC fakes are silkscreened with "DTC" as oppose to "PTC" cleverly.
- Status
- Not open for further replies.