Вставка
- Опубліковано 8 лип 2024
- In this video Shahriar upgrades an Agilent 3400B RMS meter to add an LCD-based GUI interface and USB digitization capabilities. This excellent instruments offers high RMS frequency range of 10Hz - 20MHz and can reliably measure 100uV RMS - 300V RMS! The instrument uses purely analog techniques to accomplish this.
The theory of operation of the instrument is presented in details along with the challenges of adding digital circuits among such a sensitive analog domain. The upgrades use two modules which communicate through a serial interface and use an ADC to digitize the measured analog RMS value. A new GUI interface is developed to display a wide range of useful information simultaneously as well as historical data trend line. The instrument is used to perform several measurements including measuring the output noise of a DC/DC converter module.
www.TheSignalPath.com
/ thesignalpath
www.Patreon.com/TheSignalPath - Наука та технологія
Very nice project Shahriar! Thanks for taking the time to share that with us!
*Note: I mistakenly referred to the pair of the thermocouples & heating elements as "coupled". What I actually meant to say was "matched". In fact, these elements should be completely thermally isolated from each other. Their only interaction is the series connections of the thermocouple pairs They are "matched" in a sense that their thermoelectric responses are as similar as possible. They also have fairly small thermal mass which is why the unit can react to an AM modulated signal.*
Ah, thanks for clarifying---I was wondering about that! :-)
Completely right. Also you have to consider balancing issues on the AC side as the capacitance between the thermocouple and heater is quite some picofarads because of their close proximity. Those imbalances and coupling are actually one of the largest contributors in measurement uncertainty when doing precision ac measurements. A way of solving them is to have an ac driven guard on the tvc output to eliminate the leakage.
Thank you. I was wondering about that while watching .
Anyway great work as usual Sir.
Another little mistake is that the two termocouples are in anti-series connection, their voltages are not added together but subtracted, that creates the negative feedback.
I see you have a keysight E 36212A bench power supply at the end of the video. Does yours also step values (voltage or current) in a non linear fashion? I've noticed stepping up the set value does not change the output linear.
That mod was absolutely pristine.
Excellent! Thank you for taking us through all the steps in the project: really nice! :)
Amazing project I absolutley enjoyed it !
Fantastic Shahriar!
Great to see another upload from you Shahriar!
Nice job Shahriar!
Thank you for explaining this project. A really useful addition to the meter. Fascinating to hear about the lengths required to avoid noise injection.
wonderful addition to a classic instrument!
This was a super fun project, thanks for showing it. I'd love to see some more equipment modifications and upgrades like this (altough I think I can imagine how much work this took).
Very cool integration of old and new here with a unique design twist. Really cool meter; only seen one of these but didn't get the chance to use it. Great content as always. Keep up the old school gear content! Betting lots of us audio nuts, and power lovers love stuff like that!
Very nice proyect and execution! You are the fastest talking, yet most understandable one. I enjoy very much the wide variety of the toys you got, not limited to rf magic, to which I am an outsider
Great video, enjoyed watching it.
Awesome project!
wow - Thank you Shahriar and thanks to the Patreons for pushing for this. great!
That's what I was waiting for! As a patreon I like it so much! Thank you!
Nice project! We enjoy watching this type of videos! 🥰
One of the best videos I've seen in a long time. I had abandoned "the signal path" for a while. but it's good to come back and see these gems. As always, thank you Shahriar for the time and effort you give us in these videos, they are very inspiring.
I enjoyed this video thoroughly. Thank you very much
I absolutely LOVE the concept of this video! I would certainly enjoy more 'hacking' videos like this. Thank you for your content!
I have a search notice on Ebay for a 3400B, I never saw this unit...you beat me out! Great video with some very sophisticated design work.
Thanks. Most ingenious ,I have learned much.
I am green with envy of your equipment.
Nice Mod works Great
OMG!!!! I just watched a TSP video and actually understood the ENTIRE thing without having to re-watch it multiple times!!! This is quite a novel experience for me!!!
Very cool.
Very cool! Similar to this project I added a USB adapter to a few of my Keithley 177 and 179 for datalogging. I got lucky that there was a connector populated for an optional GPIB add-on (that I did not have) so I only had to decode the digital signals.
nice work - from a retired EE
Remember when HP made top quality kit in the USA and you could get schematics for your instruments? How the world has changed.
The CEO's Merc wasn't luxurious enough, and the private jet was having some paint scheme colors, so they had to ship production to uplifted countries where pay per product costs less.
"CHANGED" is not the right word..."worsened" seems more appropriate...regards from Italy.
It would be really wonderful if you could make some tutorial for beginners who don't know this level of complexity.
Cool upgrade project! For the electronics design I would recommend using either a 2-layer PCB:s for ensuring close-by ground return path for any digital signals (i.e ADC, display and serial port). Second best world be routing ground traces between each digital signal (or at least between each pair of signals). Soldered wires or isolated copper tape could be added on top of signal traces for improved ground returns on the current PCB designs. For future PCB designs, adding series termination resistors (33-100 Ohms) close to the each driving digital output would also significantly reduce ringing and radiated noise/emission from digital lines switching state by reducing signal slew rate, dV/dt. Keep up the good work, really like the content you provide!
Very nice
The thumbnail of the "vicious predator attacking prey" is what brought me here! I have no idea how you "survived" that vicious attack! xD On a serious note, your videos are ALWAYS top tier and greatly appreciated! Thank you for sharing your expert knowledge with the world! 🙏
Glad to see you survived the cat attack on the thumbnail ;)
Wow, you have your own PCB milling machine! Great job, this is next level
- also happy pooch the kittycat made a cameo-
it's the main reason I follow this channel , RF design is an interesting second tho!;) especially the millimeter wave / antenna design stuff like those automotive radars you did.
This machine has been reviewed on the channel. :)
Like, uhh.. WOW. This was really cool at every level.
بسیار عالی
OPA188, huh? You have fine taste, my friend. I used one of those to drive the OCXO tuning voltage in an old Datum GPSDO that I repaired and modified.
Great Content as always.
But a fact on this Voltmeter: Put it on Steroids and you will get out the mighty Fluke 5790 AC Measurement Standard. Actually all Fluke 57xx instruments use a very similar AC/DC conversion techinique to achieve their outstanding accuracy for AC voltage and current.
The basic principle of this goes way back in history and correctly done it is still today the most accurate way of measuring AC Signals. The name of the technique is „AC/DC transfer Difference“ And can give you uncertainties in the single digit ppm range for either Voltage in high impedance (with something like a Fluke 792A), current into a known shunt (Like the Fluke A40B) or effective power into a defined load.
Extracting the Thermo coupled device into a self contained unit and you get a so called „Single junction Thermal converter“. Their output is usually quite low, around 5-10 mV with their heating wire at nominal design load. The HP 1104x and 1105x series are good examples of those.
A more modern Converter which is still mainly used on National laboratory level is the „Planar Multijunction thermal converter“ which is a thick or thin film based elements on ceramic or quartz substrate. They achieve outstanding long term stability and up to 100MHz of bandwidth. They also have a considerably higher output of up to 100 mV and have an almost perfect square law response compared to classic SJTCs.
My first thought was to opto-isolate all the digital signal connections to minimize noise.
This was a fabulous project! I vote for more of this type of content. You mentioned that you were going to source the data output from the ribbon cable. You also pointed out the test connector on the schematic. It appeared to me that you could have taken the data from the test connector and that doing so may have been easier than tapping the cable. Just thinking...
Linear Technology used to offer the LT1088 for implementing this RMS measurement principle. It uses matched PN junctions i.s.o. thermocouples. Unfortunately it was discontinued long ago.
Wow, that's a beautiful instrument and a great project on top of it.
It's just a little over my head, but I know I'll be returning to this video!
Definitely would enjoy more projects like this shown end to end, from the design considerations to completion. I can imagine it took a lot of work, but damn, it was well worth it.
(3400A, "later fully transistored version" is $500 on ebay. Quite an "ouch" from me.)
Yeah, it took more than 6-months before I found a 3400B on eBay...
28:00 the cat
That project was really awesome, i really enjoyed it. Just one small note - it would be nicer from a UI standpoint to have the power in to 50/75/600R all on one line. That way they would be a bit more grouped together. Right now everything is a bit scattered.
Thanks! You are right. I can easily move the display around because it is fairly modular in code.
@@Thesignalpath You're welcome. It should be easy yeah.
Thank you again for the great content you provide.
Very nice project. Using an Adafruit Metro Mini looks to be super cost effective too. Interesting that you didn't use some opto isolators for the digital interfacing though I do understand using what you have at hand! Hope you and Pooch are doing well.
You are right about the optoisolators. It would make a good addition.
Hi Shahriar , I also have HP3400B , great instrument, BUT I never do such uppgrade to it , I will keep it as it is, and if I want all that value in dBm , vpp, rms, and trend , I can use my Datalogger 8 channel from pico technology and do all converting in that by connection to HP3400B , it has USB connection to PC too for trend.
The links in your description don't seem to work. The youtube link goes to a 404 and the other links are not clickable.
As always, some awesome content right here! Thank you for sharing it with us.
I got one question though: to keep the noise from the MCU isolated from the range logic, you've done some magig with some 74x14 and the 1k - 100uF low pass filter at the supply of the IC. Since those signals are very low frequency and low current, would'nt it be sufficient enough to filter them by a huge time constant low pass filter (like you did with the power supply)? Of course you would have a lot of additional parts,.. but in general it should be possible right?
I think it should work as well. I just really wanted to isolate everything and avoid having to make another PCB.
👍👍
Thanks for making this video! its very cool and inspiring to see the details of the design requirements and the steps taken to compensate for the unexpected side effects after the first approach. What is that tool you use to point that looks like a metal pencil?
Possibly an opto isolater would be a simple noise filter.
I agree! I should buy and have some handy in my parts list...
Why does the title say Agelent? This is real Hewlett Packard. Nice instruments... proof that some of the past is better than some of the present.
Amazing upgrade! Thanks for sharing this. I would like to see the source if it's available. Thanks!
Goog project. I love it. I would totally add a blinking red Led on the front panel of the 3400 to notify the user that it is in autorange mode. What happens if you change the range switch position?
The AD8307 + Arduino works pretty well for a simple power meter ...
Really nice project, it's amazing. By the way, have you ever had a look to the R&S URE3? It's an obsolete device but I think it might also fit this need.
Thumbs up for you Shariar.
Thanks! I will take a look.
Interesting that this popped up in my feed, I have one of these meters and was wondering how I could use that voltage output (we used to use them to monitor a baseband level indication on analog multiplex to a strip chart recorder) in new way. They really are a great meter, back when HP made great things (and manuals).
Thanks for sharing
Wouldnt those multiple earth connections to the same point of the PCD result in earth loops? That's a really cool upgrade for sure!
Yes in fact that was one of the original problems. The main GND is captured right at the rectifier point. The rest of the ground connections offers isolation and other noise suppressions.
How would one cleanup the background noise getting into system while measuring power supply noise? 36:20 You have mentioned it can be cleaned up. Could you please guide me into correct direction where i can read about it?
LOVE this. Elegant solution.
Is the code for this project available?
Interesting video as always. Any chance you do a video on your RTO2000?
Also at 24:00 did you put the code on your website or is this a patreon only thing?
I will put it on the website. There is nothing behind a paywall, there has never been. :)
Very cool! Leave it to HP to have such innovative designs. Can you log to a PC? (maybe I missed it...)
And looking forward to your review of the R&S SMBV100B! We use quite a few of the 'A' version at work but I think the 'B' supports NAV testing. (ILS, VOR, etc)
Yes, I can log into a PC from the USB port.
Very impressive. How long did it take you to do all this?
About a weekend or two of scattered work.
This is a relatively simple yet hugely useful mod! Have you wrote the USB code yet? Since the internals of the instrument is sensitive, my opinion would be to put the USB interface into the front panel instead of the delicate sampling board. That could have further advantage too (after writing serious amount of code): possibility to edit and upload display configurations, like custom color ranges, turn on/off coloring altogether, switch perspectives like full screen trend with small textual results...
Is that a touchscreen? Some of these profiles could be edited on PC, uploaded, then changed on the device itself.
Embed some sort of transfer function interpreter into it, then you could upload a custom measurement function too, alongside (or instead of) the hardcoded 50/75/600ohm load power calculators.
That was great but give us more of that juicy GHz.
Soon!
That is an awesome mod! Also the instrument is very cool, and build very easy to mod. A hobbyists dream :D.
Why do you display power into 600R? Why is that interesting?
That's also a standard for some measuring instruments.
personally, knowing usb, I would have put the usb output on the outside unit instead of inside (but if it works, it works)
Yeah, I agree. I just had this particular USB extender in my parts list and it was fitting better inside.
You should do a vector voltmeter review.
Any particular one you had in mind?
@@Thesignalpath 8405A
I have TWO of the 8405A models ... one is in service on a bench yet ...
Very strange problem with the digital noise. Did you consider Opto isolators?
Can you please scan the schematics at good quality and post them online?
The FONTs need to be upgraded. Please share the schematic of the Project. Thanks.
🙏👍🙏
What is the difference between an HP3400A and HP3400B?
At the end of the video, the reading is about .09 mVrms lower every time he puts his hand onto the instrument's chassis.
Are you prepared to release the design and software for this? I have this unit myself in need of a recap. Once restored, I wouldn’t mind trying this mod myself…. Thanks for bringing this to us…
Yes. I’ll put the on the website. I’ll update the video description tonight.
@@Thesignalpath Thanks!
Thank you so much @@Thesignalpath
Is it possible to measure the RMS and Noise using an Oscilloscope? If yes, why do we need such a RMS Voltmeter?
So... you have a good-quality Sindoh 3D printer, but you outsource this to your brother??? :-)
I should have said, he designed the 3D model and I printed it with my printer here in the lab. :)
I'm curious how many of the EMC issues you observed were the result of shoehorning the design into a single layer. If you had fabbed a 2 or (ideally) 4 layer board with a solid ground plane to avoid signal and power return currents stepping on each other, I suspect all of the overkill power filtering etc may not have been needed. Looking at your board layout at 22:33, the ground looks like Swiss cheese.
Sure, doing it in house is quicker and might save money, but how many hours did you spend redesigning the board when an extra $10 and a week of lead time would have eliminated them the first time? I used to home-etch boards but after doing some cost-benefit analysis realized it was just not worth the effort. Worse quality, lots of time spent on the process, and single layer is really limiting.
Was about to comment something similar. It is a really well polished project but it seems he spent most time fixing analog noise from the ratsnest of modules without ground plane..
I did a lot of experiments on this. The problem (as far as I could see), was not the ground plane. There was some ground loops issues which was causing something like a 1mV of drop across the main PCB. But there is no doubt that having a two layer board would be an improvement.
OK, that was easy! (JK) Now, how about an "A" model design! It's all gravy until you get to the range switching! Jealous...
Can you please share the code Sir... It will be really helpful🙂
Does Mr Carlson know what you're doing with these antiques?!
I don't think there is thermal coupling, usually quite the opposite, you want thermal isolation between thermoconverters. You want only known DC signal power to relate to unknown AC signal power, not thermal effects coupling. Let's measure +50 dBm from my calibrator into 3400B :)
Yes! I don't know why I kept saying "coupled" when I should be saying "matched". Thank you for catching that. I will add a pinned comment.
You and I still have to meet. I am sorry it has taken so long... :)
سلام آقا شهریار عزیز، من به شما افتخار میکنم. یکساله ویدیو های شما رو می بینم و لذت میبرم. البته راستش یه کمی هم حسادت چون من یه میزکار کوچیک دارم با چند تا وسیله ساده و وقتی وسایل و ابزار شما رو میبینم حس حسادتم گل میکنه، البته خودتون میدونید ابزار چه لذتی داره واسه یه مهندس الکترونیک، من هنوز درگیر گیر آوردن یه مولتی متر رومیزی ۷ و نیم رقمی هستم ولی خوب پول ماها به حتی دست دوم هاشم نمیرسه حتی اچ پی های قدیمی و دوست داشتنی، خوش به حالت که از این خراب شده رها شدی و رفتی از ایران. ماهایی که حتی نمیتونیم یه وسیله کوچیک از چین بخریم بخاطر اوضاع ایران. موفق باشی، عاشق ابزارهای تستت هستم
Thank you my friend. I do not take my fortunate circumstances for granted. This why I created this channel. To give back to the scientific and and engineering community from around the world.
اگر فکر میکنید این کشور خراب شده است شما هم تشربف ببرید. ولی هنر اونه که آدم خودش کمک کنه و شرایط رو مهیا کنه والا رفتن و سر سفره ای که دیگران سالها زحمت کشیدن برای ساختنش که هنر نیست. نگذارید تبلیغات زهرآگین غرب شما رو زمین گیر کنه. بچه های بی ادعا تو ایران دارن کارهایی رو انجام میدن که اگر شما بدانید شاخ در می آورید. موفق باشید. ایران وطن ماست. خراب شده نیست. البته بهشت هم نیست. ولی لطفا با احترام نسبت به کشورتون صحبت بفرمایید. موفق باشید. ضمنا من با تمام احترامی که برای آقا شهریار قائل هستم و ایشان را در حد پرستش دوست دارم ولی ببینید تبلیغات زهرآگین غرب چه کار کرده که ایشون حتی جواب شما رو به فارسی ندادن. در حالیکه از لهجه ایشون کاملا معلومه که ایشون در بزرگی رفتن آمریکا و قطعا بلد هستن چند کلمه فارسی جواب شما رو بدن ولی نمیدن.
کوس کش هایی مثل تو نمیذارن من برم. بابت باقی جملاتتت هم: گوزززززززززز
@@Bluebirdiran There are a lot of bad faith arguments here. First, I did not respond in Persian, because my Persian is not strong enough to express myself adequately. I have not really spoken Persian in over 25 years.
Second, I do not identify myself as anything which is established through accident of birth. This includes things like nationality, skin color, etc. I identify as whatever I have had a part in shaping. I am a scientists. I also sympathize with any and all people who are less fortunate because of circumstances in their lives they cannot change. Speaking English is not a sign of "western propaganda", to say that is silly at best.
@@Thesignalpath Hi Shahriar. All I was trying to say to this chimp was that it is the effort you put into something that brings success and not necessarily the place where you are in. He with his peanut brain thinks that you have got to where you are simply because you left Iran and went to the US. He cannot acknowledge the hours and days and weeks and years of hard work that you put in to be at the cutting edge of science and technology. These are what I call lazy chimps who are not willing to climb the tree to pick their own coconut and want it free of effort and when they cannot get it, instead of blaming their own laziness, find it easier to blame the government or anybody else which then makes them feel good. As for speaking English, unfortunately due to the venomous propaganda against Iran in the western press, I have seen many many Iranians who have even changed their first names to Jack and Jill so that they are not recognized as Iranian let alone not speak farsi. Of course you are not one of them. I am aways proud of successful iranians where ever they are and wish them all the best. It's the lazy chimps that I cannot stand. Good luck.
All your stuff is great, but these & the tutorials you do/did were the prime (a lot up to ep 30, the FMCW stuff, the power sensor stuff, the breakdown of the Siglents one port net analyzer hack). The repairs are pretty darn good. The product reviews are ehh. I mean they're really better than most, but some of them really seem like shilling for a vendor and the 'experiments' feel like 'well now we have 70ghz, look what we can do wth this extended bandwidth',
Though, some of the real upcomers (The Signalhound guys, the people who make the really clean synthesizers, etc) are real interesting just because they come at it from a real different start up position, with limited resources, so you can see what compromises they had to make during design. That clean of a 10 mhz tone..insanity.
You and Joe Smith are by far the best technical channels on youtube
All my videos, including reviews, have major educational elements in them. Even if someone is not interested in purchasing the instruments I review, they can learn how they are used, how they are built and what can be done with them. It is much more than just "extended bandwidth". To say *any* of my reviews are "shilling" is a disservice to the huge effort that goes into preparing them.
Why display dBm for 3 impedances at the same time? Two will always be wrong. A single display with settable or selectable impedance would be much better.
I’m no noise expert, but I would think that those 4 ground wires are not a good idea to that board. I think it would be better just to have 1. Gauge seems appropriate as I doubt it draws that much power.