PLJ-8LED (clone): fixing the other module

In the previous video, you saw that my first PLJ-8LED module was inaccurate and as these are cheap, I got another one instead. But as this video demonstrates, the adjustment problem is easy to fix. But while it can be adjusted, the power-on drift of this oscillator is a bit worrisome, especially since the other one does not show the same behaviour.
Previous video on PLJ-8LED is here: • PLJ-8LED (clone) frequ...

[@rob.-_-.]

Great info here, thank you! I personally had to remove the 39K resistor and replace it with an 82K. No matter what resistor I put in parallel with the 47K, it wouldn't let me get it in range for whatever reason? Maybe a slight board change? Who knows? I removed the 39K and tried various resistors with higher resistance. The 82K put me 1 Hz away from the 10 MHz signal coming from my GPSDO. This is with the 52K pot in the middle position so I have close to plus or minus 30 Hz. The SMD resistors on this board are size 0805 just in case anyone wants to buy replacement SMD resistors. Thanks again for the info TheHWcave!

[@genejensen8225]

Your a very good teacher, explaining how this works and its needed adjustments....A+++

[@t1d100]

A great video, as always. Thanks for making it. Yes, two SMD resistors of the same size can be stacked and soldered to make them parallel very easily. You could add some heat shrink over the new through hole resistor to protect it from shorting.

[@TheHWcave]

Thanks t1d100, I thought it might, but never tried and in any case did not have any 100K SMD resistors at hand, but good to know for the future. Good idea with the heat shrink. Will do that.

[@longeron]

I just received my low-cost unit yesterday and seeing same issue regarding pulling range. It is very close, so will leave it as it is for now, but good to know of your fix. The stability on this one is very solid. I'll look for some surface mount resistors to apply a fix at some point. Thanks again. Great video.

[@msotomaior]

Thank you very much for this video. I just made the modification by putting a 120 k resistance in parallel and it worked perfectly. I saved a lot of time with your help. cheers

[@nelsonchai7]

Thank you.😊

[@tubeDude48]

Thanks for the MOD's !

[@kelleecrisafulli4209]

Nice video :), Other ideas for improvement: I noticed the two filter caps appeared be 16V. Good design practice is a safety factor of 2x for caps. Definitely bad idea to run this frequency counter at 15V. At the low end it's hard to say what voltage a China 5V regulator will drop out (could be 5.7V to 8V) so running the unit from 9V to 12V seems optimal with 9V being a good choice to maximize life of the capacitors and minimize 7805 temperature. To run at 15V remove the input capacitor, install a 25-35V rated input capacitor and add a little adhesive heat-sink to the 7805. Because this board has a precision oscillator that is somewhat sensitive to temperature to get the most from this board consider bypassing the 7805 and run the board from an external 5V power source instead. Too much info warning: A low-cost counter from China is not likely to actually have a 20yr rated 2.5PPM VCTCXO oscillator. 1PPM = 1 part in 1 million so 2.5PPM is 2.5Hz error at 1MHZ measurement frequency. Note: VCTCXO oscillators = Voltage-Controlled(adjustable), Temperature Compensated, Crystal Oscillator. For 3.3V oscillators the VCTCXO adjust voltage is typically 0.5V to 2.5V and results in a frequency change of +-5PPM to +-50PPM depending on the unit. A VCTCXO oscillator does not include an oven as is sometimes assumed. Instead they generally have a feedback loop which adjusts the frequency as temperature changes to provide a more constant frequency. So unlike an oven-controlled oscillator which may require 10 to 20 minutes to stabilize the VCTCXO will start up and be stable almost immediately. A VCTCXO can also be expected to drift up to +-1PPM per year for aging.

[@TheHWcave]

I totally agree and thanks for the detailed write-up. Regarding the long term stability I am not too worried as I usually check and re-calibrate my gear once a year, in this case against a GPS Disciplined Oscillator. So far the counters are still very close and I have not bothered to re-calibrate (yet).

[@brucepjohnson]

I received my unit yesterday and noticed the same slight offset even with the pot adjusted to the very edge of the range. Your video is an excellent teaching tool, to explain what the problem is and how it can be addressed. I wonder if the additional warm-up time might be due to the thermal charcteristics of the added resistor. I am hesitant to risk the soldering since I have no experience with SMDs, so ordered a second one from a different supplier to see if I get lucky. Thanks again for posting this!!!

[@TheHWcave]

Thanks for the nice comments. By the way, that module still runs fine with the "normal" through-hole resistor soldered on-top as shown in the video. I now have SMD resistors as well but never felt the need to go and fix it "properly". Why fix what isn't broken...

[@brucepjohnson]

@@TheHWcave I have plenty of through hole resistors but I couldn't see exactly how I would solder that on top of the SMD resistor - can you give me an idea of how you accomplished that?

[@TheHWcave]

I assume you have seen in the video at 07:10 how I bent the 100K resistor? All you need to do is snip off the leads to the correct length and apply a generous coat of solder to each end. Then apply a bit of extra solder on either end of the 47K SMD resistor (one end at a time so it does not desolder completely). Then holding the 100K in the correct position (as in the video) with tweezers use the soldering iron to tack each end of the 100K onto the solder joints of the 47K SMD. No need for more solder for that. It is much easier than it looks (or I describe it here). Once its in place and can stand on its own, you may reflow each joint (one at a time) and add a bit of extra solder to improve the joint. You need very little or possibly none, just reheating the connection usually does the trick. While you rework one end of the 100K it is held in place by the other solder joint so you don't need to hold it during the reflow making this an easy operation. Hope that makes sense. Its kind of hard to describe this verbally

[@brucepjohnson]

@@TheHWcave Excellent, thank you for taking the time to write down some details around the operation. That is what I guessed would be the approach, but having you confirm that gives me the courage to give it a try, once my second one has arrived in a few weeks.

[@JAKOB1977]

what power does these PLJ8LED take. are they sensitive. or do you need to drop a shutload of power into them for higher frequency as they do seem to take up in the 2½ Ghz and to what extent do you risk blowing them with to much amplitude/wattage.?.

[@TheHWcave]

I have to admit that I haven't used mine for a while (I now have two other "proper" frequency counters that offer more measurements and greater accuracy) but I remember that I could comfortably drive them from my FY6600 signal generator which is no power-house.

[@manuelgti6766]

👍

[@tatavox1408]

Excelent ! Well Done !

[@davidhollfelder9940]

Thx for the wonderful/fantastic video series’s on these PLJ-8LED frequency counters. I “accidentally on purpose”, bought the same two units you got, followed along to a “T” (got all the same results). The resistor mod on the TCXO oscillator .. Perfect Fix! You made me a quick study, where otherwise “ida been” totally lost. PS, I never did find your email on the about tab. I’m an amateur radio operator, call sign KI5MEF .. I think you may be one too. If so, plz look me up on QRZ, and drop me an email.

[@thomass7877]

One question for those who might visit this video ... Can the device be adapted to measure frequencies from 0? The upper limit is not so important to me, it can be 5 or 10 MHz, it doesn't matter ... Thanks in advance if anyone has an answer ...

[@TheHWcave]

Sadly no. It is specified from 100KHz upwards. It does work reasonably down to around 10KHz but soon after starts to show increasingly wrong and fluctuating numbers. 1KHz for example is usable. If you are on the lookout for a cheap counter that works down to 0 checkout my playlist on the PIC-based frequency counter. It uses a popular and cheap ebay kit as foundation and with my modified firmware, you can do amazing things with it.

[@thomass7877]

@@TheHWcave Thank you very much for the answer ... keep up the good work ... it's really fun and relaxing to watch ...

[@tomnone8378]

Good job mate :) my question is how to adjust without accurate generator?

[@TheHWcave]

Yes its a chicken-and-egg problem. First option is to just live with the "out-of-the-box" accuracy. If you want better, then check out my very first KZbin video, about a GPS Frequency Standard. These GPS modules are not very expensive and you get a very accurate frequency standard. There is some jitter but this is usually ignored by frequency counters and there are some frequencies that it can generate jitter free (I named them safe frequencies in the video if I recall correctly)

[@tomnone8378]

@@TheHWcave what about dcf-77? Have you thinking about it?

[@TheHWcave]

@@tomnone8378 Its a possibility of course but gives you just one frequency. I wonder if making a tuned LC circuit with an amplifier to boost the output and decouple it from the frequency counter would work. But then you have the problem of making a tuned LC circuit for 77.5 KHz and a long antenna....

[@tomnone8378]

@@TheHWcave there is slow but digital signal so will be the same as pps 1hz

[@TheHWcave]

@@tomnone8378 I understand but I thought you wanted to calibrate this frequency counter and for that the 1PPS is useless unless you build an oscillator with PLL that can be locked to it. A lot of work.

[@anonhollmuller4032]

73! well done +! like this like a youtube vid for education should be 👍. Liebe Grüße aus Bayern dg2r?

[@nortonics5745]

For Christ sake people, it's a cheap ass $10 meter and a 1Hz difference doesn't amount to squat. Here's a simpler fix - put a piece of black electrical tape over the far right digit - problem solved. 🤪