Recent Posts

Pages: Prev 1 2 3 4 [5] 6 7 8 9 10 Next
41
Beginners / Re: LM317: divider values and minimum load current
« Last post by pope on Today at 03:50:36 pm »
In other words, if the actual load is 15mA or 150mA or 1.5A, does it matter whether R1 is 120R or 360R or anything in-between?

If you have a load with a guaranteed draw of >10mA then you can use higher values for the divider resistor.  However, the other factor you have to look at is IADJ or the current on the adjust pin.  The TI datasheet for a regular LM317 shows this to be as much as 100µA with a variation due to load or voltage changes of up to 5µA.  With 360R for R1, you'll have 3.3mA and that 100µA would represent 3% of the divider current and thus a proportionate error.  The typical IADJ current and variation thereof are much lower, so that 3% would be the worst-case.  For production you have to worry about worst case, for something one-off or low volume that you can 100% test then you could probably use 1k for R1 and still be fine.

Thank you for the detailed explanation but unfortunately due to my luck of English skills I'm having a bit difficult time to understand wht you mean.

In simpler terms, using 360R is OK or not really? And if so, would the suggested 240R is better?

I'm straugglin to understand this "With 360R for R1, you'll have 3.3mA and that 100µA would represent 3% of the divider current and thus a proportionate error.  The typical IADJ current and variation thereof are much lower, so that 3% would be the worst-case. "
42
(Although you'd be in the right place on Facebook, I guess.  :P)
Yup, that's another target of mine.

I just finished compiling a fitness music playlist from my library which is exclusive original 1970s through 2010s for those who want nostalgic songs they can exercise to instead of modern club dance mixes.

I figure on X and Facebook, I will release 3 songs every other day, 1 good for warmup/first burn, one for mid burns, and 1 extra high power song for that final power burn.  I have enough music to run my workout music playlist stream for 2 years without repeats.

It is a strategy to try to get followers and thumbs up on a regular flow basis.  I just need to fine-tune the presentation and I'll probably use youtube-music links as they will play the full song for anyone whether you are logged in or not, though an add will come every other song.

43
Beginners / Re: requesting schematic review (newbie warning)
« Last post by Terry Bites on Today at 03:46:10 pm »
Is this a laser down link receiver for a satellite for testing on the bench. Not orbiting, I mean. That needs a ground station with some serious optics and super low noise fibre amplifiers and a large telescope no?

Why are there two photodiodes and why do they need separate processing? How does the receiver hardware reject ambient visible light and IR? Filters will only get you so far.

So, the APD amplifier with the higher output gets routed to an output or if not, ignored? Why does the FPGA need to know about the analog level from the APD amplifier other than as signal strength monitor.  Check out logarithmic PD amplifiers. They give you huge range compression making that comparator function simpler to implement. Your ADCs will thank you too.

My guess is that you'd get a lot of jitter that will mess with the analog switching and the FPGA inputs. Small noise voltages and tiny glitches at the inputs will flip the thresholds unpredictably.  Some “latching” action inside of the comparator output state is a good idea as is some "hysteresis" see for example a TLV3603. A latch function lets you decide when to output test of the input voltage using an external clock (aka strobe) rather than instantaneously and chaotically.  External logic is likely the simpler route and probably far less design effort. A couple of comparators and some "glue logic" Why do you think that would be overly complicated.
I think all that is needed is a very basic "state machine". Synchronous logic can save your day.

Can you digitise both APD amplifier signals separately and do the comparator action numerically inside the FPGA. Once acquired you can massage the numbers any way you like. That sounds like some trivial code to me. I'm still not understanding the need for signal routing of the APDs.

The comparator in your circuit is not tied to any reference Zener or not? If you do need a precise reference, use a TL431 or similar. If your power supply is properly regulated, a simple voltage divider may be good enough.

With bit rates up in the 10-100s Mb/s, transistor switches may be far too slow. You can get wideband analog switches but the time taken to turn them on or off is not usually that fast either. So, a switch might boast 100MHz (even GHz) of signal bandwidth but only be able to switch on or off at maximum speed of a few uS. Still, have a look at an OPA1S2385IDRCR ic. Pd amplifier and analog switch combo.







44
Beginners / Re: LM317: divider values and minimum load current
« Last post by pope on Today at 03:46:08 pm »
If the external load will always be above the minimum value, then the resistors in the voltage divider only matter to determine the effect of Iadj, similar to the effect of bias current and feedback resistors in an op amp circuit.

So, R1 should be adjusted so as the Iadj is 100uA (max value per datasheet)?
45
Sorry, I was wrong in the first place: the motor should run on 98.5V, not 85V. In this case the 10Ω resistor is more protection/fuse then real dropper.
46
There is no hard coding. Just setting a more sensible default. It is difficult to tell what Siglent did but I have created multi-language products in the past and Linux has some pretty nifty ways to make it easy for software developers to add support for multiple languages. It doesn't take recompiling the software to change the texts; update the file with translations, restart the program and done.

Yes, I think we are nitpicking now. My whole point was that it needs some changes by the product vendor, even if you nowadays can take ready made pieces of open source software and just "configure" them, there can be a lot more to it and as every change or feature addition to a product, you have the whole R&D process and testing activities involved. And even if you insist that en_US.UTF-8 would be a good default, did you already test with the interface changed to e.g. Chinese? If you are a small vendor, it might be fast and easy. If you are large, it's a process that costs money (I work in a big company R&D). I think all has been said, I have no further input.
47
Beginners / Re: LM317: divider values and minimum load current
« Last post by bdunham7 on Today at 03:43:21 pm »
In other words, if the actual load is 15mA or 150mA or 1.5A, does it matter whether R1 is 120R or 360R or anything in-between?

If you have a load with a guaranteed draw of >10mA then you can use higher values for the divider resistor.  However, the other factor you have to look at is IADJ or the current on the adjust pin.  The TI datasheet for a regular LM317 shows this to be as much as 100µA with a variation due to load or voltage changes of up to 5µA.  With 360R for R1, you'll have 3.3mA and that 100µA would represent 3% of the divider current and thus a proportionate error.  The typical IADJ current and variation thereof are much lower, so that 3% would be the worst-case.  For production you have to worry about worst case, for something one-off or low volume that you can 100% test then you could probably use 1k for R1 and still be fine.
48
That was just a random Google, the point was some inserts ARE designed for press fit; they are not all for injection moulding. I have some M3 inserts at home, but they're the split type.

Hex inserts isn't a bad idea though sounds like a bit of tinkering might be needed to get them to fit properly. Or are they heated & pressed in?

A reliant car is designed to be driven on the road, will you buy it ... yes :-DD these kind of inserts very often rip on the surface when tightened and often have a small play and is not aceptable mechanically speaking you have zero dimentionnal precision.
For this reason I NEVER heat a printed part, hex spacers are not precise parts and i have to measure all of them each time to ensure a tight adjustement (no play at all) in order to have a good dimentional accuracy.
49
Repair / Re: 8 button IC replacement of Pasapair desiccator
« Last post by Tarien on Today at 03:41:16 pm »
I translate datasheet and im not sure if i correctly understand the D8-D1 function, as inverted i understand inverted time 500/1500us. For k8 detected 1500us and inverted time for d8 500us. Its that correct or im making some mistake?
50
Programming / Re: algorithm to quickly find a pattern in a block
« Last post by Tation on Today at 03:39:51 pm »
512 byte patterns are "long" patterns. In this case Backward Oracle Matching (BOM) algorithms and derivatives are thought to be the most efficient.

But I do not think that the benefit over Boyer-Moore (maybe around 50 % in selected cases, definitely not orders of magnitude) worths the effort.
Pages: Prev 1 2 3 4 [5] 6 7 8 9 10 Next