Function generator repair (Hungarian TR-0458/B)
Вставка
- Опубліковано 10 лип 2023
- Today, let's try to fix a TR-0458/B function generator made in Hungary probably in the early 1980s. This is a 0.0001Hz to 1MHz function generator (sine wave, square wave, triangle wave, sawtooth) with 58 transistors and 4 OpAmps. Finding the faulty component, aligning it, the schematic, how does it work.
The schematic:
www.blunk-electronic.de/datash...
Please support my channel on Patreon:
/ diodegonewild
Instagram:
/ savage_danyk - Наука та технологія
Always listen to your cat before poking around a complex circuit 😊
What an absolutely purrrrrrrrrrrfect comment.👍😏🇬🇧
Fried cat
Fun fact: those paint-marked transistors were made by the Mikroelektonikai Vállalat (lit. Microelectronics Company). They started producing simpler semiconductors (transistors, diodes, 7400 series, etc) in Budapest in 1983. During the night of 26 May 1986 the whole factory burnt down to ashes under rather suspicious circumstances. There was an automatic fire extinguisher system in place but it was down due to maintenance. Many believe that the fire was in fact an insurance fraud as Lloyd's insured the company for some 50 million USD.
Jó hogy magyarok is nézik ezt a hasznos csatornát!
Alap.
És EMG forever. 😎
Thanks for the interesting side story!
Sabotage by greedy capitalist spies. Eliminating competition and preparing for the complete takeover of eastern industries by western corporations for a dollar.
This guy is really master of electronics. The true expert on electronics. Danyku smekám, klobouk dolu.
It's so good to see actual understanding and explanation of the schematic. So many new 'techs' now seem to assume the only problem anything can have is 'bad caps'. That generator is excellent!
I completely agree.
We had these (and similar other instruments) in our university lab among a lot of old EMG instruments. (BMF-Kandó, around 2003). They were already outdated but really good instruments to practice measurements.
You did very well zeroing in on two transistors, one of which proved to be the fault. Makes a huge difference having access to the circuit of course.
The voltage base to emitter isn't 1.4V, its 1.4 _bloody_ V 😁💖
Great that you managed to repair it 😊
For some reason it seems bizarrely like exactly twice the stereotyped barrier potential of silicon (700mV) qwp, I wonder if something shorted in series.
Well, there's no material in practical use (other than LEDs) that has a "bloody" bandgap pc 1.4V, so he's right here, it's a bloody weird failure to happen.
I've also never seen a semiconductor component fail this way.
@@BritishEngineerwhat I can think of is that the casing perhaps wasn't closed properly and the PNP layer got doped through oxygen - if that's even possible.
A darlington pair might read 1.4v
@@liam3284 Indeed, but the low hFE would make this transistor a very poor darlington.
I guess the transistor case cracked from age and moisture ingress damage the silicone die. Really strange failure.
Amazing video as always 👍👍👍
The gap is probably manufacturing defect, but I agree, the fault in the die was probably caused by moisture seeping through the crack.
@@mrnmrn1 That's possible
Old lab equipment like this is actually easier to use than modern equipment
He owns Soviet Осциллоскоп С1-118 "Сделано в СССР" (made in USSR!). Really popular beast in it's times. A Tektronix copy AFAIK.
Yes definitely would prefer repairing one of these over modern rohde and schwarz RF generator that cost me about $500 in parts alone to get working properly again as those HP MMIC amplifiers love to fail and had to source original and display that cost over $300 alone to get a NOS one with new RF glass. Then to rebuilt the complex switching power supply. At least now it is 100%.
Interesting device. Decades ago whilst I was stuying at Uni, (Chemistry), we had a couple of electronics techs down in the basement of the chemistry building. They had a signal generator there, its brand was, “Wavetek” and its principle of operation eas much the same as this, voltage-controlled current source and current sink being switched alternately onto a bank of capacitors to generate a triangle wave…then using a comparator with wide hysteresis to get the square wave and swith from the current source to current sink. One of the techs photocopied the circuit for me so I could go away and study it in detail…that was in 1986. Since then I have acquired ownership of two Hewllett=Hewlett-Packard generators that use the same principle and have much the same ciircuit as the Wavetek. One HP is a model 3311 and it only goes to 100KHz, the other is a 3312A that goes up to 13MHz, but I knowj there is a very similar looking model that goes to 20MHz…but all the circuits are much the same, frequency pot making DC, inverting op-amp generates reverse slope DC, both DCs go to voltage controlled current source and sink respectively, a four diode matrix in between the source and sink switch the capacitor bank, (coarse frequency) alternately, comparator looks at ramps generated and sends square save cack to seitch diode array. Fancy diode-resistor clipper makes “synthetic” sine with acceptable harmonic content…you get the drift!
Anyway, as I watched the video I soon realized this generator used the same principle as my HP’s and the Wavetek and the voltage controlled current source, or more succinctly, the voltage controlling it was somehow at fault and non-linear at the bottom end of the pot. Even before you took the lid off, I had swiped the model number and downloaded a circuit that looked pretty mich like my HP’s….HOWEVER, once you got the lid off and showed the actual innards and circuit…it was utterly different despite the circuit I downloaded having the same madel number!
I was beyond fascinating to see all the same principles used, but UTTERLY differently. The sine shaper on this is far more complrx that my HP,s and the Wavetek which just use diodes and resistors…I reckon this clipper/shaper uses the non-linearity if complementary transistor pairs to get a sine withfar lower harmonic “hash” than my HP’s are capable…but everything else was different too…but used the same principle. I don,t know why, but whenever a transistor in a complimentary pair is buggered, it almost always seems to be the PNP…and I knew it wax the source and the source uses a PNP, (the NPN is in the current sink)…so, yes I called it about as fast as you went thhough it, but the real interesting part was the utterly different architecture of it whan compared to my HP’s, the Wavetak at Uni and a multitude of synthetic generators I have seen over the years.
Definitely keep the old Soviet, Czech Tesla and other central European stuff coming. It eas funny to watch you go through that Tektronix Oscilloscope recently and have difficulty identifying “Western” components, particularlythe mil-spec tantalums…but I have the same issues with any Soviet stuff I might find..which is rare here in Australia. When yiu did the Soviet TV, the germanium PNP line output transistor with live emitter and transformer couplrd base looked like something found in a crashed UFO to me. Shango66 had a slightly different model of Soviet TV but it had almost exactly the same chassis in it…it was exactly the opposite of you inside that TektronoxOscilloscope, he was flumoxd by the Soviet components.
The Czech Oscilloscope you did recently, again I pulled the circuit off the internet before you took the sides off…I saw nearly a dozen pentode valves and a fer tetrodes and diodes in the power supply…so I was REALLY Supprisedwhen you took the sides off and…NO VALVES, not even a HV diode valve in the tube supply!
I found a HP 3312 , repaired and sold it, because I was short on cash.
I still regret it to this day…
It's entirely possible that this is a close to 100% clone of a HP, Tektronix or Wavetek function generator. EMG done that a lot. Check out the EMG 1555 (TR-4653) oscilloscope. It's a perfect clone of the Tektronix 647 / 647A oscilloscope. It even has its own page on TekWiki!
They combined the 647 and the later 647A together, the 647 was only capable of 50MHz IIRC, the 647A increased this to 100MHz by small modificatons and the use of PCB in the vertical power amplifier instead of the point-to-point wiring of the 647. EMG was somehow able to produce the TR-4653 with 100% point-to-point wiring, but it was capable of 100MHz bandwidth. Very good scopes, minus the lot of bad contacts of the transistor sockets. And shorts caused by silver migration on the surface of the Tektronix-style ceramic terminal strips (yes, they copied even that!). Even 1GHz sampling plugins were available for the EMG TR-4653, which was never available for the original Tektronix model. Although the sampling plugin set was probably a clone of a later Tektronix unit, modified to interface with the 4653's circuitry.
An ICL8038 functional generator IC works similarly, and also has frequency-independent sine approximation converter. But it is limited to around 300kHz.
Heh, I just finished training as an electronics tech, and in September I will start a job as the electronics tech for the chemistry department of the university! :p
Nice to see a correct fix with systematic fault finding! I had quite the opposite experience with my last fix, a '70s Takeda Riken frequency counter with about 500-600 transistors in it. I had the schematic and a master's degree in electrical engineeting, but I just started poking randomly around in the device with my oscilloscope probe first. And just out of sheer luck, I found the fault in less than ten minutes. A bad 2N709 transistor and a bad old school epoxy blob silicon diode. Greetings from Germany!
Nice to see more test equipment repairs… the more of us that make this kind of video the better! You did well to compress the complex repair down to the video length you got.
Great video. Tracking down the problem by figuring out the schematic and explaining it was most intriguing.
Wow, I have the same type of function generator, good to see in your video. Greetings from Hungary..! 🙂
I'm totally lost during the explanation on the circuit board yet I keep watching. The hypnotic voice makes it impossible to click away! Great job on finding such a subtle fault, very impressive!
It does take some practice, best under a master electronic , to understand circuits.
Once you crack a few, it becomes much easier. There's a "language" to it, I mean it's not all chaos.
Greetings from Hungary! :) Thanks for the cool - as always, video!
To je umění - umění elektroniky. This kind of failures are the most rewarding ones, beautiful job indeed! Thank you for these 25 minutes 😍
The bloody long and bloody complex videos are the best :)
Thats a sweet multimeter man, The update on its display is real nice.
The more bloody long and bloody complex the videos, the better they become!
*Even more complexity!*
Nice problem analysis, and also an extinguisher for burning brains of course😃
This kind of repairs are the videos that I appreciate the most. Thank you for sharing your knowledge.
Great repair video again! Its a complicated and funny mistake but you managed to find the cause and repair it! That shows you are a real electronic expert! Of course we all knew already before 🙂
this 'weird' VCO-input is for wobbling, which is used for testing filter-settings in dependence of the input frequency. You also can generate weird sound-effects, used in synthesizers and so on..
EMG instruments are always described as "they have souls", very well done repair on this one.
(The second thing is, they are often copies of western instruments. In fact, quite good copies, but usually lagging 5-10 years behind...)
The block level design is how Hewlett Packard function generators are built. The design is not intuitive, but the VCO, integrator, triangle wave generator and then wave shaping topology is how they chose to do it.
Danke!
Very huge Thank you.
I love all this analog gadgets..
..and how much too learn from it ..
Even its AGE... AmazinG.. still working..
Sure, not easy short, but over all simple fix ..
Very impressive:
the amount of functions and frequency range.
Well explained... understandable.
Good work.
As always!
Huge thanks for your support! I'm glad that my videos about old electronics are appreciated. I also tend to prefer them over the modern ones.
Amazing how we can now replace the entire engine of this instrument with a 40MHz AD9850 DDS module, with an accuracy of a 125MHz crystal, the size of a big stamp, for less than $40!!!
You mean $8?
Engine - machine with moving parts that converts power into motion. ?????, circuitry is a better descriptive word.😏🇬🇧
@@brucepickess8097 Tell that to game engines.
@@LMB222 By the time I added the rotary encoder, Pro mini, LCD, battery, switches, amplifier, BNC jack ... It came out to about $40.
@@piconano my comment stands. The term engine has no relevance to this piece of equipment as it is not software based.😏🇬🇧
That's a nice piece of equipment
Very complex device, you analysed it nicely and found the bad transistor, strange fault!
BC327 and BC337 belong to my favorite transistors.
I still remember BC107…
@@LMB222 And BC 108's and 9's. I remember BC212's I think I still have some.😏🇬🇧
Ahh its so good to see you with a good old hungarian instrument. We made so reliable ones but today they have been overtaken by their cheap chinese counterparts.
Thank you for didnt let them die and repaired it. Greatings from hungary.
Pěkná práce a skvělý starý přístroj, ten si tu záchranu zasloužil 😊
This guy is really master of electronics. Another great repair! 👋☺
There were some old uA741 made in DIL14, later ones were made only in DIL8.
You are definitely a very smart and patient person! Im impressed with the logic and fault finding. I did nearly pass out during the schematic part but vape and coffee pulled me through!😊
Excellent analysis to diagnosis. Kudos.
you should not afraid to make longer videos like before. true fans of yours will be watching carefully. i will
I agree. Don't seem to me as long video. Interesting schematic, most of LF generators a based on Wien bridge oscillator.
Very good analysis and repair. Always appreciate these.
Great troubleshooting. Fun watching you use your great knowledge and extreme logic. Thanks for the videos.
Definitely interesting failure mode. I like these :).
Nice disfunction degenerator too!
Great work! Thanks for sharing your thought process. Good idea with the brush!
Very skilled procedure. Well done.
Beautifuly designed circuit.
Thanks for the video.
03:02 CA3140S on right hand is a high precision mosfet input metal canister operational amplifier.
The sine shaper circuit basically carves a sine wave out of the triangle piece by piece with 3 stages of 2 diodes and 2 transistors each.
The 4 transistors on the top are used to buffer the result, and the 2 transistors on the bottom left (presumably) provide a reference of some kind.
It's possible to reduce the transistor count and improve smoothness by using an overdriven differential pair with a logarithmic voltage response (though you'd need a matched pair).
When I tried to recreate it, it only worked properly when I omitted the direct connection to V+ in the leftmost shaper stage (which makes sense, because then the NPN transistor would never turn on in that case).
Thank you, watched every second. I, like many others, would not mind more of these longer repair videos.
I survived...lol. Great video, I really like how you show your thought process. Makes me feel slightly more normal when I do the same thing on repairs.
Another really great video! Thank you for so much for your efforts to make this video & explain where the fault lay for such a complex circuit design. As a repair tech I seem to also attract such failed pieces of test gear & i stupidly spend hours chasing down such similar sorts of silly faults as if it's some kind of puzzle just to exercise my aging brain for no reward other than i get to understand what some very smart designers were thinking once upon a time...
*Great Repair 👍🏻👍🏻👏🏼👏🏼👏🏼 Enjoyed Full Video ❤❤❤*
Superb repair of a very complex instrument.
What a beautiful user interface!
We love this longer vids though, makes learning much better.
This is bloody insane !! Nice work man
Wow, what an interesting design, very cool, worthy of a full restoration I think...
This shows the difference between knowledge and understanding. There might be many people, who have the knowledge of each of the components, they know how and what they do. But you need understanding of whole electrical circuits to be able to use proper reasoning when making a diagnosis of what most likely causes the problem. This video shows us exactly that. Describes the issue at hand, and shows it. Then breaks down the schematic to specific areas, explaining what each of them is responsible for. Using logic and reasoning to rule out what cannot be the source of the problem, because of how the apparatus behaves and when the issue becomes visible. Then, what remains, is most likely the source of the operational anomaly. We were allowed to watch over the shoulder of a true master, who explained his reasoning and found out what was causing the problem. Knowledge, understanding and practice = wisdom.
Bloody electronic expert !
A most wonderful function generator. 😎 Thank you. The good old days.
Damn, that one has some similarities with the HP older series........great video.
They died alright, that's why they reached to the end 🤣🤣. But I on the other hand, appreciated the work done here. I don't know about everyone else, but seeing electronics come back to life, it is art in itself.
I hate components that sort of fail and limp along. Great video.
I absolutely love that device!
nice fix !
Great fix.
Finally something Hungarian!😃 de nem tudtam, hogy mi magyarok csináltunk jelgenerátort is.
Hungary made a LOT of function/signal generators, EMG alone made dozens of models, and HTSZ, HIKI, FMV, TEXO as well.
I think that failing transistor is slowly dying from corrosion that has crept up the leads and is damaging the die. Old age is getting to it! Instruments like this should have a conformal coating on the board to prevent environmental corrosion. Take note of HP and Tektronix equipment. Thanks for the great job explaining the circuit. I very much looks like a Wavetek from the 70's or early 80's.
I managed to watch to the very end, and I wish I was as intelligent as you estimated. But I'm not; yet, I'm gradually getting there, partially with listening to your explanations.
Excellent repair!!! Thank for vid!!!
Very interesting video, and very clever engineer. Awesome. 😉👍
As of today I have my own function generator to fix, a HP 3310A function generator.
I accidently touched the lead to 550V and i heard a pop and it stopped working.
I found one blown transistor and a resistor got fried.
I havent been able to test much but the frequency generator doesnt seem to work at all in most ranges.
I am not as smart as you are, i hope that i will miraculously find the components that cause this problem and be able to replace them....
You are a true inspiration to me, I hope I can actually do it.
Edit: Fixed it, the entire output stage was blown, replaced the components and it worked again.
Also a filter cap on the power supply wasnt filtering correctly and replaced that.
Download the service manual and start measuring. HP equipment is well designed and pretty easy to service. And as you know what went wrong you know were to start looking for the problem.
@@RTBNL Thank you, but its already fixed now.
the output stage only was fried thankfully, replaced the transistors and resistor and its working again.
also some capacitor was bad and caused output wripple, replaced that too with one of the same size but 4 times the capacitance, i didnt have any other that could fit in the same space.
Stupid HP put the capacitors right over the voltage Regulator were they get really hot
@@janno288 was it a cap on a power supply rail or something else? 4 times the capacity can seriously change the behavior of a circuit. For example the caps on the output of a LM78/79xx regulator should never exceed the values mentioned by the manufacturer, increasing it makes it less stable.
@@RTBNL essentially the power supply is like this
Transformer Rectifier Filter Regulator.
there was serious wripple after and before the regulator, i changed the filter capacior out, thr wripple was gone and the regulator could function again. The regulator is a descrete one made out of seperate parts, no Integrated Circuits
the cap was one of those ones that has the leads on noth ends, i didnt have one of the same value on hand i had a 4700μF 50V (vs 500μF 50V) and put that in, that fixed it, same size, that capacitor also is 40 years old now.
the old one is 50 years old.
@@janno288 in that case it doesn’t really matter. A friend complained the + and - 18 volt rails in his diy pre-amp kept fluctuating slowly. He couldn’t find out why. He mentioned he even increased the 5600μF caps behind the LM7818/7918 regulators to 10000μF and it even got worse.
Great explained, thank you for sharing. 😊
Watching from Hungary !
Siuuper-interesting fault-find video 👍👍 ,,Thanks for sharing Dan. All videos are very inspiring and educational //br, from norway
Great find. Smart guy. Excelnt!
Great job, well done ☺
Great work!
Wow it's interesting how this function generator works
this is very very interesting schematic, i have similar (a gift from my teacher) it's hand build also, the triangle to sin is made with resistors and diodes and few transistors and the integrator is also with transistors, no ap amps, no full schematic, maybe even more transistors, never counted them (i can't even see all of them). oh and a digital counter with TTL 74XX, whoever build it put a lot of work.
Huge respect, I dont think that I whould be able to fix this 🤯 at least not so fast as You
Elementary my dear Watson
Well done matey!
interesting to see failing silicon transistor faults now due to age ect , its a long time ago now for that early stuff to continue to work as in tendered ect , well done on repair ect
Fascinating!
I love my little cheap transistor tester I do have a famcy Sencore brand but that cheap one works well !
In hungary we still use these kinds of equipment in schools to this day
Why not?
When I worked at a university in 2008-2011, I had some equipment from 1983.
Made in Liechtenstein (Balzers), it was the only one whose RS232 voltage levels were the proper ±12V, not like the Chinese devices that provide only ±5V.
@@LMB222 my only problem is that they are not idiotproof and it happens relatively often, that someone messes up something.
Do we still have schools in Hungary?
nice work
Veľká vďaka za toto video, je mi sympatická táto stará technika (a to platí aj pre osciloskop v predchádzajúcom videu, aj pre digitrónový multimeter v tomto videu, aj ja mám presne taký). Aj ja som poruchu odhadoval na nejakú aktívnu súčiastku, ale tá tvoja rýchla a pritom precízna diagnostika mi skoro vyrazila oči z hlavy, neviem či by som sa ja v takejto zložitej schéme tak ľahko zorientoval. Je tu vidieť ešte jedna vec, a to úroveň elektroniky v Maďarsku a Československu v dobe socializmu, takú rozmanitosť najmä aktívnych súčiastok v československých prístrojoch, aj keď sú tiež veľmi kvalitné, proste nevidno, je vidieť že v Maďarsku bola elektronika na oveľa vyššej úrovni ako u nás. Takže toto video ma veľmi potešilo a vôbec som sa celú dobu ani na chvíľočku nenudil, priznám sa že by som ocenil viac takýchto rozsiahlejších videí na tému starej elektroniky.
I seem to recall that Wavetek 110/111 manual had a good theory section that described the effect of unbalance ( offset) in the positive and negative integrations. Just like what the low frequency behavior in this case demonstrated. Their circuit used two opposing currents, one being twice the other. But the higher one was switched, while the lower one was not. Thereby the effective current alternated from, say +1 & -0 and +1 & -2, (or +1 and -1). Based on this ancient understanding, I kept waiting when the unbalance source would be revealed. The surprise was, when one of the transistors started looking like a Darlington one. How come? I have never seen any defect developing that behavior. Something to keep pondering!
suspect some uncommon manufacturing defect
Holy shit, that thing is crazy complex!
it reminds the devices I saw - when child - on my father's shop
Awesome faultfinding and repair!
Maybe that transistor cracked due to some corrosion or something inside?
Thanks!
Nice piece of kit. Imagine designing something like this from scratch, amazing really. A modern equivalent would have a dozen chips inside with a few support components and not much else. I'm struggling to think of a use for a 0.0001Hz signal though. One cycle every ~3 hours?
One chip, and it costs 8-12 USD.
(I'm making a signal generator right now)
Very good
I'm here again to say we gonna die and rebirth but never going to get experience of electronic like you sir seriously NOOOOOOOOT easy to become like you 😢
Every video with new things
Bravo !
I've fixed a few old oscilloscopes before and I too had super dodgey germanium components...both transistors and diodes. Can be deceiving because a quick test makes you think all is OK!
bloody clever
Another great video that did not kill me. 😂
The “sine” wave reminds me of old technical books that would approximate one using semicircles for the upper and lower half-cycles. At a glance, I don’t think the circuit is actually capable of producing a pure sine wave; still better than what passes as a “modified sine wave” with invertor manufacturers, though.
When I saw the problem I almost immediately suspected that it was a problem with an integrator circuit becoming asymmetrical in the lower frequency ranges, indicating a problem with one half of the current-generating circuit (having experimented with similar oscillator circuits for audio synthesis applications).
The specific reason it went wrong was extremely unexpected though.
I see that by the 1980s Hungary was capable of producing (or procuring) high quality extruded aluminum... for the case... or were the channels where the top slides in machined? Anyway the case design, and the voltage selector, fuses, and front panel remind me of late 70s Phillips test gear... not quite as good... but ergonomic and stylish. It's a shame you don't see more of that today. Of course, no one would manufacture such a set today, with so many (any?) discrete transistors, and all the associated support components.... especially the trimpots.... and that nice case, knobs, and dial...just too expensive... maybe cannot even get the parts? In any case, fixed function, function generators aren't that useful compared to even the cheap Chinese produced arbitrary function generators. Maybe audiophiles, and the superstitious vintage audio restoration doers will push back. But, usually I can't hear them.
But still circular heatsinks look like they are from early 60-s.
Hungary has large bauxite reserves and aluminium production is significant. Like other eastern bloc countries, electrical wiring in buildings from the 1940s to 1989 was mostly aluminium.
Nice