The multiple Zeners in series are meant to adjust the tempo of the whole circuit. A 5.1V zener has much less drift with temperature than a 15V zener. Great job!
@leandrolaporta2196
5 ай бұрын
Ahhaaa! So that's why you use several in series! I didn't knew that, thanks!
@mikesradiorepair
5 ай бұрын
But without parallel current balancing resistors across each one you can end up with current imbalance because the voltage across each zener diode is slightly different.
@bayareapianist
5 ай бұрын
I think @IMSAI Guy had a video about 5V zeners .
@annaplojharova1400
5 ай бұрын
@@mikesradiorepair They are in series, so the current through each has no other way than to be exactly equal, so no need for any "ballasting resistors" or so.
@Dieseleux
5 ай бұрын
Zener under 7v are one side of temp drift and over 7v is opposite. Put 3x 5v zener with one 15v make 30v zener with good temperature stability.
@bayareapianist
5 ай бұрын
One important point. The opamps usually get a separate power source because if the main lines get ripples, the ripples won't effect the operation of opamps. You placed a lm741 instead of lm709 since predates lm741 and perhaps wasn't in your CAD's library. LM709 was a nasty opamps with unbiased input and needed adjustment. Beside it diees quickly if the inputs aren't protected. I think those revered diodes were placed for input protection.
@annaplojharova1400
5 ай бұрын
This way of arranging the internal power rails is quite common for lab power supply instruments. It allows the active control circuitry (the reference,opamp,...) to operate at a convenient supply for those circuits (+10V/-5V for common opamps is sufficient), regardless of what the output range of the instrument is. And it allows the power section to be dimensionned with minimum voltage headroom (of course with all the tolerances and ripple, but on top of that just 2..3V is enough, vs at least 5V margin required towards the opamp positive supply rail). In many mainly higher power designs the power winding has switched taps, so reducing the wasted voltage even more. The whole thing needs way lower voltage margin (about half or so) and does not need any high voltage (beyond usual +/- 15V supply rating) opamp components even when the power rail needs to become close to 100V (what I would expect for a 50V output instrument when not loaded). With more modern opams (LM358 counts as "modern" here) you may design the entire control part operating even from a single 5V supply (negative of this 5V supply is then connected to the emitter of the power transistor) even when the output goes way beyond 50V. The point is, the "ground" for the whole control part is the "positive" output terminal (or better before the current sense shunt, so the shunt does not see any current not flowing into the output terminal; then it requires differential output voltage sensing), the actual voltage feedback signal then goes from the negative output terminal.
@danstiurca7963
5 ай бұрын
That thing on the left is a discrete precision voltage source with a differential error compensation.
@andymouse
5 ай бұрын
I would like to learn about this, is it popular enough to be covered in text books like the art of electronics please ? I could look up voltage sources and give that a try :)
@danstiurca7963
5 ай бұрын
@@andymouse look up ADI app notes about precision voltage sources. You'll find a bunch of versions, some very much like this one. Also, they have the relevant equations, and some explanation. ... but they're always very very confusing. Generally, they compensate for temperature changes and supply voltage drift.
@andymouse
5 ай бұрын
Thanks :)@@danstiurca7963
@miroslavstevic2036
5 ай бұрын
This is excellent. Thanks for all the effort you've put into this video. R10-D9 is used as primitive voltage reference for the Negative Rail in relation to the Positive Rail. Although these rails are taken from the aux winding, they still need some kind of common reference point and a predictable voltage diff between them. Also, windings must have a common reference. OpAmp is in a serious trouble here, as one of it's inputs is fed from the main out. I suspect the manufacturer had trouble with this, notably burned opamps during fast transients, so they've put two reverse diodes in parallel on the opamp inputs to protect it. In any case it's a mess. The other three Zeners D10, D11, D12 and Q3/Q4 are the voltage regulator for the OpAmp's positive rail and common reference. Zeners are also driven from aux winding by a somewhat compensated primitive transistor current source Q3/Q4. I suspect an error in the schematic or logic at a common point. The source's driver is Q3, with it's current is set by R8/R9/D16. As it heats up its current will increase, so I suspect they've used thermally coupled Q4 to stabilize the current and prevent thermal runaway of Q3. Anyway C6 is there to prevent potentially catastrophic oscillation between transistors. This common voltage is then used as the reference for OpAmp.
@jj74qformerlyjailbreak3
5 ай бұрын
Now I get to do this too my Lambda 422a-fm supply.
@IMSAIGuy
5 ай бұрын
that's a classic
@byronwatkins2565
5 ай бұрын
Verify that the transistor has the correct specs and was installed correctly (by a previous repair attempt). Otherwise, it is being turned on for the slightest load current.
@ivolol
5 ай бұрын
Maybe repair could involve designing in adjustment possibility to allow the meter to be accurate?
@6WTF_MAN9
5 ай бұрын
it's some interesting solution for current limiting. It looks like the minimum number of discrete components for this functionality
@xenoxaos1
5 ай бұрын
The really high voltage could be shorting out part of the primary winding. If there's 120 windings on the primary and there's a short between turn 60 and 80...then it would result in higher voltage on the output. Idk
@thanhhuynh272
5 ай бұрын
I have drawn the circuits of hundreds of pieces of gear, I don’t use pictures like you or Clive, I just do it directly off the board. I could do this in about 25 minutes. One thing you learn is not to assume BCE pin outs being the same for all transistors. Gut instinct tells me there is something dodgey with Q4…looks like the emitter and base need to be swapped. Everything else looks “Kosher” though. Q4 being wrong is what is flumoxing you with that reference supply in the lower left hand corner. I certainly would not bother with a CAD version, just pencil on paper, then I can easily rub-out and slter “mistakes” like Q4. Another thing to watch out for is to be sure if certain transistors are NPN or PNP…particularly if numbers have rubbed off or they are do old there is no datasheet anymore…sometimes you will come across small SCRs in TO-92 packages, or UJTs or PUTs..these little buggers canhave you scratching your head foe a little while! I was caught out recently when replacing tired old low HFE TIP 3055s for contemporary MJE 3055s, (All I Could Get) one is BCE the other is ECB, so got them in with BE transposition…it was a Royer/Baxabdall type circuit, so no harm done, but had to take the heatsink that was UNDER the two TIP 3055s and put in ON TOP of yhe two MJE 3055s! (Was in a BWD 845 Australian Oscilloscope.)
@cocusar
5 ай бұрын
I think the 3 5.1v zener in series generates a "sorta" 0v potential. The emitter of a pnp is connected there, and one leg of the pot as well. Weird indeed.
@leandrolaporta2196
5 ай бұрын
Nice little psu, i love it, did you have -ve on the current pot?, If there is not -ve TR will be on all the time pulling the darlington down
@Smit7ua
5 ай бұрын
Good job ❤ thanks you 👍 Hi from Ukraine 🇺🇦
@KeritechElectronics
5 ай бұрын
Photographic reverse-engineering the Big Clive way! One moment please... and resume. Nice PCB design there. I love these curvy traces! (and curve tracers) Multiple Zener diodes in series look like either used for power dissipation or temperature compensation reasons. I can't really see why either would be used here. I also can't really see where +Va is taken from.
@gorak9000
5 ай бұрын
Yeah, the print the photo of the board larger, and the back side mirrored is Clive's favorite technique! He must just chew through the yellow and cyan ink in his printer (because most of the huge board images he prints every time are green)!
@KeritechElectronics
5 ай бұрын
@@gorak9000 that's precisely why I wouldn't do it that way - not only do I have a grayscale-only laser printer, but also I'd consider it a waste of ink to print in color whenever I need to do reverse-engineering. So, it'd have to stay on screen, and then there's this pesky marking the spots in a graphics program. Backlighting a PCB has helped me when I was reverse-engineering a vintage digital delay/reverb here: kzitem.info/news/bejne/1H2Ev6aOkplydm0
@TheEmbeddedHobbyist
5 ай бұрын
Looking at the circuit, i can't see where +/-Va is connected to the +50V supply as they are transformer isolated? Have i missed something. i would expect to see the mid point of Va to be connected to 0V of the 50V part?
@68CHUCKLES
5 ай бұрын
Are there any slight improvements you could make to have better performance ? Better opamp ? Can that xformer do an amp ? Maybe in one of your next videos you could go over the topology of one of those giant 19" rack HP power supplies ?
@thanhhuynh272
5 ай бұрын
Oh, my “guess” as to why Q2, the current limit transistor is always on….broken wire or dodgey insulation displacement connection on pin 8 of the pin header causing R15 to pull the base of Q2 up, or dodgey/dirty RV3…will I be correct in part-3?!
@joeteejoetee
5 ай бұрын
Flip-it just like BigClive does on every video.
@AnnaVannieuwenhuyse
5 ай бұрын
I'm happy to see you're putting time into this. A unique power supply always makes a lab feel more organic and lived in. 🫡
Пікірлер: 31