In RF land we call that a transmission line transformer. At low frequencies, the ferrite coupling does the heavy lifting, and at high frequencies, the transmission line coupling does the heavy lifting. Simple in principle, but getting the crossover really flat in amplitude and phase takes a good bit of experimentation.
On the transformer, twisting the primary and secondary wires together before wrapping them on the core is a common technique called bifilar winding. On the main unit, of course there are no directional couplers because this thing operates at way too low a freq range for that. What they must be doing is digitizing both the voltage and current, then computing the s-parameters from the impedances. And the board material does look weird, but I[m sure it's not Rogers material. You'd only use that for stuff in the GHz range at least. But I liked the video. Thumbs up.
$10000.25! That's really expensive just for marking the problem... My tariffs are better, just $1000 for solving the problem: Hammer - $10 Transportation - $40 Knowing where to hit with hammer to solve the problem - $950
A russian guy once repaired a cheap lab power supply and the fault was in the mains fuse holder, it was just broken. He didn't have any holders so he just soldered the fuse. "The fuse holder can't brake if there isn't one".
Russians know how to fix stuff. If you take away the part that breaks and the product still works, well you fixed it and increased reliability! Heck, you don't even need fuses, that's what the rest of the components are for! :P There's a funny saying that expensive transistors are designed to protect fuses.
For $500 dollars, I would at least expect the label to be applied straight on my project box. Beautiful transformer though, wouldn't surprise me if those windings were spaced apart by hand during calibration to optimize. Specialized kit and the knowledge to get it where it is costs special money.
looking at the graph at 1:28 the frequency response falls off quite a bit at low freq. -15db at 1hz. it doesn't go back up to 0db until 100Hz which is much more doable I think.
I agree with your point about the 10V at low frequency and core saturatrion . It does not make sense . Usually for loop response one would use less than a millivolt to avoid saturation of the error amp .
I'm surprised by all of the negative comments. This equipment is more common than many engineers must think. Every gadget which has a power converter, switcher or linear, should be using one of these to check stability of the control loop under numerous conditions prior to going into production. Cellphones, tablets, computers, other test equipment, etc, everything that uses volts and amps needs something like this to confirm the stability of their power supplies. This equipment can be used for many other measurements as well. Other competitors to this bode box are priced in a similar range for their equipment. Large companies have dozens of these. Our large company has at least 15, most of which are in use almost constantly.
The term you were looking for on the transformer was "bifilar windings" it is a common trick for making a broad band transformer. As an injection transformer, it is usually better to have a 10:1 step down so that the 50 Ohm output impedance of your generator looks like 0.5 Ohms on the system under test. I was slightly surprised by the small cross section of the core. Generally, the more massive the core the more energy it can store. On the high current side of the transformer, you can have significant DC current leading to significant stored energy in the transformer's core.
$15 for the parts, $485 for the virgins.. Its a specific piece of test equipment. If you need one you don't particularly worry about the cost; it's a drop in the bucket compared to all your other test gear
Back in the 1970s I had a part time job at a music retailer doing bench repairs. That was the first time I saw the insides of a guitar amplifier made by a British company called Orange. All the transformers (input, output, and power) were wound on circular cores. Also, all the signal resistors were oversized to minimize thermal noise.
The wire is specially manufactured. The wire drawing machine uses normal copper but the machine has to be oriented in such a way so the wire is drawn from north to south in the Northern Hemisphere. The machines in the Southern Hemisphere needs to draw the wire from south to north. This will align the copper crystals for proper conduction in one direction. Look for arrows on the wire jacket to see which direction the toroidal core needs to be wound. :)
I'd really like to see the ToE in DaveCAD for this one! Also probe various points during acquisition, starting with the connectors and working inwards, all the way to the USB data stream. Edit: What I'm looking for is ways to make better use of simpler and cheaper tools, such as the Red Pitaya, by adding some external components and software.
Top tip for anyone wanting to make a transformer with better performance than this: Use binocular ferrites. Hard to get, but if you are serious about being a manufacturer of quality test gear, it is viable when you are charging hundreds of bucks for a wideband flat transformer. If you are being *really* snazzy, use a pair of binocular ferrites.
Are you sure this is a superhet? Looks like a classic direct conversion receiver(near zero IF) for these low frequencies VNAs. The 2 DDS are usually just offset by the bandwidth of the ADC...
Adil Malik my thoughts also, low IF (each DDS a few tens of kHz apart) plus a pair of Tayloe quadrature mixers on the RX sides, explaining some of the analogue switches. Low IF so a low speed ADC. Low IF avoids mixer DC offset and phase noise of downconversion LO. So it is “superhet” as there is frequency downconversion rather than direct conversion.
T think the transformer is just a Amidon T240-2 iron powder core with a bifilar winding. The wire is most likely PTFE insulated to get around 100 Ohms of impedance. That is a very common construction that is used in hamradio all the time.
The T225-2 (the closest size available in -2 material from Micrometals/Amidon) has an Al value of 12 nH/N^2. With 40 turns (as seen in the video), this gives a magnetizing inductance of 19.2 µH. With a 50 ohm source impedance, this gives a lower 3 dB cutoff frequency of 420 kHz. This injection transformer has a lower 3 dB cutoff frequency of 7 Hz, so the core must be something with much higher permeability, most likely nanocrystalline.
It's called a bifilar winding! The toroids look like Amidon powdered iron cores ur = 10. I don't think that any windings are squeezed or stretched to optimize frequency response, it seems that they just have been evenly spaced! Pretty underwhelming !! A better core geometry would be a binocular core and i think that a balun on the floating side might give some more improvement !!
Hi Dave! About resistors in B-SMC. As I see, all resistors there is 4R7 1%, except one between Source BNC and CH1 BNC. What are the colors of this resistor? I can't determine. Regarding the disputes about the core material of B-WIT 100, I can say with sure that it's an VITROPERM 250F.
That transformer in a box after testing is worth $50 at best. How they came up with $499 unbelievable. Their course lessons in extortion have obviously been adhered to.
The price of the surface mount tester probably comes from the gold. I wouldn't put it past them to use solid gold, or very thick plating instead of thin gold plating. Make it so if ever needed, a clean surface is easily revealed.
Could post frequency response of your DIY injection transformer? There more into it than just number of windings and core. Yes, part of the magic is in the core, but even lid taken off changes frequency response.
3:08 There is just a fuse under that black heatshrink to protect the transformer from (most) "user errors" ... Oh, never mind, he gets it open later ... And no, an "inline fuse holder" would be a bad idea, because every "connector" adds more "things that can go wrong" and possible "crappy connections". That's why the fuse is soldered. 9:15 Oh, that annoying FTDI chip ... I hate those things ... 11:59 Fun fact about those through hole resistors: The length of the leads between the board and the actual resistor body makes a big difference for things like reflections! 13:49 Yes, there are is tons of switching going on! Coupling, filters, range switching, termination switching, lots of things that can be configured by the user via the UI. No, the relays don't really need to be shielded, because you aren't really dealing with "ultra small signals" ... 17:12 " ... rather simplistic ..." Well, a simple solution is an elegant solution ;-)
You are paying for software. The hardware is basically a first gen SDR from the early 2000s. And yes the cost of it is obscene. For anyone that wants to complain and BS about low volume and research: lookup HAM radio SDR transceivers, 10x the complexity, 5x more expensive hardware and at least 20x more complex software.
For those price tags people would always prefer the hand job and avoid that hole in their pockets. Of course people would choose cheap thrills. They last longer for that amount of cash.
I imagine they can also get away with this price tag because it's a rather niche product, but also because of the (I presume) various levels of certifications it went through, whether it's self or 3rd party, they probably run various standard tests on these to produce some graphcs and other stats and then each one has to match etc. Probably a lot that goes into that. And of course the software too. But yeah I'm sure someone that wanted to could probably produce this for a much lower price tag.
Hey Dave, How about you make a video about that dickey microphone cable analysis video using the vector network analyzer. That way we can learn more about the importance of vector network analyzer in component selection and system design. Thank you
3:52 - Yet :) This thing seems doable to me. With a lot measuring and twiggeling, but in the end this thing could be pretty much done cheaper. It's no nude virgin moon juice instrument anymore - but who cares, if it's $50 instead?
Not sure about the nude virgins. Looks Like a simple #2 iron core with bifilar turns on it. Good "Q" to 30 MHz. I suspect that core wound was put together by a ham. Made many myself. Material 43 might have been batter choice though. The AD9851 is the jellybean DDS chips for making VNA's for the HF bands. Working with them myself. Although the prices at DigiKey do not reflect the jelly-beany-goodness.
I said it in the video and I'll say it again. If you can make and sell a good low cost wideband injection transformer cheaper than you'll sell 10's of thousands of them.
Dave, I have a little project idea for you. Since all these ppl think that the transformer is $5 piece and can be easily made, make one. Generic core and replicate the winding. Then characterize it :D
Price seems more than fair for what it can do. Vintage Tek and HP test gear cost the equivalent of a house back in the day; the magic has shifted from complex circuitry to the software. Hi end test gear has always cost big money. Whatever, with both your paying for precision and dependability and these cost money, it's not an Iplonk or Wankows OS that will need numerous updates, fuck up every 2 weeks and leave you weeping and reaching for a hammer.
The twisted pair is a type of bifilar winding. It insures the two transformer coils occupy on average exactly the same magnetic space as each other allowing for an almost perfect one-to-one ratio. A typical accuracy of such a winding can be better than 1 PPM.
Can you please explain the purpose of these devices? I have a bachelor's degree in electrical engineering, but i still have no clue what these things do.
Actually, I have a degree in robotics, but there was still a lot of electronic involved. What I meaned by this comment was, that it might not be obvious even for people who know something about electronics. I can recognise a toroidal 1:1 transformer. But what is it good for? You won't be using 500$ transformer for galvanic isolation.
MrTad98: You are going to use a transformer like this for isolation but not for power transfer. Imagine you have some system that you want to measure the performance of with regard to RF coming in on some connection. You want to send RF in there but not let it go through some other path in the total system. You use the output side of the transformer to apply the signal.
im shocked the 9851 is doing the DDS, i have so many of those chips laying around and i never thought of them as stable for instrumentation of this class
Stable? It is a mature known product and is 99% digital. I guess 10-bit 180MSPS is all you need, but for a $5K class machine, that's really kinda cheap. I suspect the design has been around for some time and they just haven't needed to update it.
well simply the AD9851 DDS is not a precision part ... its jitter is quite high, its nonlinearity is not great for this application and its voltage accuracy is variable with temperature
Got any citations for that? Jitter is listed as 80ps, way outside the 50MHz LPF of this application, and DDS code jitter isn't rocket science to work around in software, especially at this sort of accuracy. Absolute voltage accuracy isn't particularly a problem either, as you are measuring ratios, measuring both your own incident and then transmitted / reflected wave, it effectively calibrates itself out to a large extent.
Would be nice to see the bottom of that toroid. The colour kinda reminds me of Micrometals' "-2" iron powder material (red body and one clear face). T130 (1.3in OD) size, perhaps?
And if you can design injection transformer like like that and wire it by hand while getting less than $600 (for whole project) you should really reconsider your current workplace.
Actually I don't think it is, because bifilar implies the same return wire in the pair (as a coil usually). This is a transformer with separate pairs twisted.
When I was fifteen years old, I thought it would be cool to turn my project boxes upside down so you couldn't see the screws. WHEN I WAS FIFTEEN YEARS OLD.
That transformer is nicely wound but its definitely not worth the price. That is price gouging is ridiculous! I would've expected a panel mounted fuse holder in the side of the box for that price. If someone had a torroid the same size and some solid core copper wire (ethernet cable is a cheap source of solid copper wire) in their scrap parts box (which is likely) and counted the number of turns of this transformer in the video, they could easily make an identical transformer like that and put it in an off the shelf aluminium project box for a lot less money. Less than $100 including labour.
when you factor in production hours with calibration, qualificiation etc; you'd probably get close to 100-250 dollars worth of hours, and there has to be a fair margin of profit regardless, so I don't think its too unreasonable of a price
No you're not the only one. Those plugin boxes are a total utter rip off just on their own. The second one. $600 for _that_! Its just a box with resistors (probably 0.1% resistors) that cost a few $ at most. The box probably costs more to manufacture than the components inside it and it is just an off the shelf aluminium box. I could build that in a couple of hours for less $50 including labour.
