Episode 715 Back to basics: Power Supply Voltage reference, then odds and ends Buy: www.banggood.c... Manual: www.icstation.c... Be a Patron: / imsaiguy
It's worth mentioning that modern 5.1 volt Zener diodes have a negligible temperature coefficient. A 10.2 volt Zener has a pronounced positive temperature coefficient.
I didn't know that either, just looked at some datasheets and found the tempCo is most ideal around 5.5V! For example, I look up a cheap ON Semi series zener. The "BZX79" series shows 5.2V and 5.8V being closest to 0 and diverges away either direction. Amazing stuff, good on ya oldblokeh
The particular circuit used operates the zener at constant current, which is advantageous. - input bias current of the op amp is negligible for practical purposes, so the current through the zener is is equal to the current through the resistor in series, which is fixed, though there is an aspect of positive feedback - if the zener voltage rises the current through it will rise which will further increase the zener voltage. I would be inclined to run the zener at significantly higher current. It _may_ be that the zener used is unusual, but a typical (say) 500 mW zener would be on the knee of its I-V curve at only 1 mA. I'd use 5 to 10 mA, depending on the specific type.
1:36 "the only way this works" - equal voltages across the 10K resistors because ideal op amps do not have any current flowing into or out of their inputs. You have an impressive portfolio of videos.
Another advantage of the circuit they used over the buffered zener diode example is that the zener diode voltage is fairly stable but is still somewhat dependent on current. A buffered resister loaded diode will see current fluctuations that come from loading the supply, whereas the circuit used will maintain a stable current through the diode even if the raw supply voltage fluctuates under load. Note also that the transistor that disables the output works during startup and shutdown. The -5.1 rail is designed to fall more quickly than the main supply thus the output gets shut off when the power is turned off as well as during the initial startup sequence.
Another benefit of the Zener circuit as shown is that the current is stabilized by the Zener voltage. So this has better line rejection than a resistor to V+ driving a Zener. Of course most Zeners need a much higher operating point current to achieve rated specs than an op-amp like this could provide.
Great video and learned a lot. I will be getting one particularly for the constant current capability. These days there is great emphasis on "power integrity" which deals with how stable and clean the output is when handling fast (10-50A/usec) load transients at many different duty cycles. The output impedance is typically measured with the converter in the off AND on state at no load, 50% and 100% loading. Since you have the TinyVNA and you have converters, it will be a good series if you could delve into how to use the TinyVNA to measure the output impedance of converters vs having to use an expensive lab type VNA.
Nice serie of this PSU👍 , it would be interesting to change the values and component´s to handle CC 0 to 30A or 15A to start with and also run it in Micro Cap
For a new circuit/PCB, I wonder if the OpAmp+5.1V zener could be replace by a TL431 + resistor divider, instead. I see the choice of 5.1V Zener as Voltage reference having the lowest Temp. coeff. possible, near zero. And the OpAmp provides then a constant current to it. But I don’t know how the existing circuit would compare with the “Adjustable Zener” of TL431 adjusted for 10V.
8:43 The original design (from October 1978) had LM741 opamps in it, which are internally protected from short-circuit. I'm assuming TL081 / 084s are the same (need to check).
I recommend to add your PayPal Link into the description. I am more than happy to support your lab with your great content where I learned a lot. I also forward links to this power supply series to our Arduino electronic user group in Hannover, Germany. From time to time payments via PayPal works for many of us, I cannot not do regular Patron subscription payments. Danke, Danke, Danke. Helmut
Excellent series!! One thing I didn't see, or maybe missed, is the purpose of C9 and C6. I'm guessing they're both negative feedback on the voltage control circuit, which would slow the rise or fall times of the output?
A bit too complicated for this video, but changes in voltage need to be controlled so they are not too fast or too slow, so most likely the circuit was built and found to over or under shoot or oscillate on capacitive loads so the feed back loop needed to be frequency controlled and some capacitors were added.
Thank you for the interesting discussion of the circuit. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-1CY2dr5Qbjw.html A hysteresis can be applied to an opamp with a feedback resistor to the positive input. With a capacitor to the positive input, it would be a dynamic hysteresis. In this circuit the capacitor C8 leads to the negative input so the current limiter acts as an integrator with limited speed of change at the output, providing a stable current limitation.
The way it looks to me is that this is a re-implementation of the Harris/HP 6200 series. What are the specs for this power supply? Line-load and do they compare to the 6200 series?
Isn't the diode across the output for flyback if you have an inductive load? You said it was to protect against opposite polarity, but that's the output and not the input
Great series! I am building the Hiland PCB version but the circuit is (nearly) identical. I am also studying for my Full Amateur license here in the UK and found your explanations really educational, thank you! Is there an error on their supplied schematic regarding the -ve supply for the U2 and U3 op amps? Shouldn't their pin 4s run down to the -5v pump rail? I would love you to take this series a stage further and maybe test it for ripple on both normal operation and current limiting. Maybe even suggesting some improvements to the circuit, you already intimated a few. It would be a terrific way to learn some more for you.
I have been looking at this design a bit further and ran some tests on mine. I am only running 18v AC into the PSU and the rails are pretty close to the max vcc. datasheet value for the TL081s. With a 24v AC supply, won't they will be well over their spec?
You think you can make the schematic using a second transformer for the op amps, so we can skip the negative pump and the safe power on circuit? Maybe we could skip the third opamp also?
@@IMSAIGuy this psu was analysed in forum for centuries, how to make it more reliable or how to fix it. never saw an alternative version though. it seems they made it in china for mass production but havent checked what version it is or if its even upgraded from the original version. a nice alternative is the well known elector 0-35V 0-3A high precision linear psu with sense lines. would make a nice educational video for your channel too! thank you for the analysis on this psu, i own it and i plan to upgrade it too!
@@IMSAIGuy the elector magazine precision psu is from december 1982 page 12-22 . its 0-35V 3amps with sense lines and dual transformers for the op amps. it can go easy to 5amps or more. it doesnt use relays for the coils though, thats the downside. i bulit it 2 years ago from custom pcb and its an amazing psu. i think its worth your channel time!
Sorry for the necropost, but I was wondering about the designer's choice of the TI TL08xs. You probably know all this, since it's a jellybean op amp, but it's kind of a weird one. I was curious if you looked at the datasheet for that one, or if you're already intimately familiar with its quirks, specifically it being a "JFET op amp" from 1977 when JFETs were hot. I was curious if this informed the design at all, or if they just found a good deal on an old chip. I'm also curious about op amps not liking to source current, or "ideal op amps do not source current", and if this design would be significantly different if it used a TL082s (dual) or a TL084 (quad), instead of discrete TL081s. Hope this question isn't too remedial or requires too much time to answer; it's just a particular op amp that's I've always been curious about.
Op amps source current just fine unless they are weird. They virtually invariably have a push-pull output. There are several ways that the output stage might be configured, depending on the objectives of the design. For example, you couldn't use an emitter follower or source follower for the upper transistor of the output stage if you wanted to be able to make the output swing to within a few tens of millivolts of the positive rail. More often than not the concern is to be able to get the output within a few millivolts of the negative rail for single-supply operation. Source and sink current capability may be asymmetric. See figures 6-12 and 6-13 of the TI datasheet for the TL08x series for source & sink voltage vs. current plots. The main merit of the TL081 over the others in the series is that it supports offset voltage trimming. Strangely, in this design the voltage error amp is trimmed but the current amp, which has much smaller input voltage, isn't. I'd do exactly the opposite.
Great job in explaining the operation of this power supply. As well as good trouble shootin hints. Thanks, I bought 3 of the kits with fans etc. Will see what more I can extract from them.