Instead of regular zeners, you could use varicap regulators: they have a better stability, and depending on the voltage you will only need 4 to 6 of them (their voltage is generally in the vicinity of 30V).
European examples are TDA1550 (diode style case) or TAA550 (TO18), but Japanese and US types also exist
European examples are TDA1550 (diode style case) or TAA550 (TO18), but Japanese and US types also exist
5.6V zeners in series will make a very soggy regulator. A single avalanche diode, properly decoupled is much better at regulating voltage.
Zeners work best at around 40% of their maximum current (Pz/Vz).
Calculate the normal current in the regulator tube, and take this to be 40%
of the Zener maximum rated current. Then assume 1W devices, and determine
the proper Zener voltage (1W/Izmax). Now divide that voltage into the 150V total,
to find the number of Zeners to connect in series.
Running a Zener with current less than 10% of rated will usually give poor results,
(more noise and worse regulation), as well as higher sensitivity to tolerances of
the series resistor and the voltage source.
Calculate the normal current in the regulator tube, and take this to be 40%
of the Zener maximum rated current. Then assume 1W devices, and determine
the proper Zener voltage (1W/Izmax). Now divide that voltage into the 150V total,
to find the number of Zeners to connect in series.
Running a Zener with current less than 10% of rated will usually give poor results,
(more noise and worse regulation), as well as higher sensitivity to tolerances of
the series resistor and the voltage source.
Last edited:
Avalanche diodes are stiffer than 5~6V zeners, but they have a significant positive tempco.
AC impedance can be improved with a parallel cap (unlike VR tubes), but nothing can be done about the tempco (except adding forward biased diodes which will degrade the dynamic resistance).
What is more important? Stiffness or DC accuracy?
Varicap regulators offer the best of both worlds: low dynamic resistance and low tempco.
Nowaday, they should be cheap and easily available even on Ebay, since traditional TV's are mostly gone, and significant stocks should remain.
Their nominal current is also similar to VR tubes
AC impedance can be improved with a parallel cap (unlike VR tubes), but nothing can be done about the tempco (except adding forward biased diodes which will degrade the dynamic resistance).
What is more important? Stiffness or DC accuracy?
Varicap regulators offer the best of both worlds: low dynamic resistance and low tempco.
Nowaday, they should be cheap and easily available even on Ebay, since traditional TV's are mostly gone, and significant stocks should remain.
Their nominal current is also similar to VR tubes
When noise matters, which it does according to the opening post, Zeners are probably better than avalanche diodes. I don't know about varicap regulators.
By the way, the noise of a glow discharge voltage reference tube expressed as a voltage per root Hz is around sqrt(2 q V^2/I), where V is the DC voltage across it, I the DC current and q the elementary charge (about 1.6022E-19 C). At 150 V and 10 mA, that is about 849 nV/sqrt(Hz).
By the way, the noise of a glow discharge voltage reference tube expressed as a voltage per root Hz is around sqrt(2 q V^2/I), where V is the DC voltage across it, I the DC current and q the elementary charge (about 1.6022E-19 C). At 150 V and 10 mA, that is about 849 nV/sqrt(Hz).
If I read the schematic correctly the 150 volt regulator is fed from a 225 volt power rail through a 10,000 ohm resistor. That would be a 75 volt drop or a current of 75/10,000 or 7.5 mA. I would use a 1N5383 Zener diode, under a dollar.
Adjust the final voltage after the unit has warmed up. I wouldn't worry about the noise as it is filtered on the 200 volt regulated power rail.
Adjust the final voltage after the unit has warmed up. I wouldn't worry about the noise as it is filtered on the 200 volt regulated power rail.
Last edited:
datasheetcatalog.com digikey & mouser never heard of those part numbers. mouser thinks a varicap regulator is a kind of LED. digikey refers to varicap regulator to variable capacitance diode, where I have a choice of capacitance at voltage.
Yet, they must be one of most ubiquitous type of IC (kind of: a string of temperature compensated zeners + transistor) ever made.
From ~1970, each TV set had one: that's litterally billions of pieces...
This page shows a number of equivalents: some Japanese, other Koreans, etc:
ZTK 33, Tube ZTK33; Rohre ZTK 33 ID34487, IC - Integrated Ci
From ~1970, each TV set had one: that's litterally billions of pieces...
This page shows a number of equivalents: some Japanese, other Koreans, etc:
ZTK 33, Tube ZTK33; Rohre ZTK 33 ID34487, IC - Integrated Ci
Elvee, the only easily available datasheet I can locate is for the ZTK33 - which indicates there is a quite a long start-up time for voltage regulation?
I don't recall using or seeing such a device, but then I wasn't in to TV's.
They certainly aren't cheap on eBay nowadays.
I don't recall using or seeing such a device, but then I wasn't in to TV's.
They certainly aren't cheap on eBay nowadays.
https://www.astelectronica.nl/AstCatalog/Info/141190.pdf
Seems to be good info. 33 volts ("B" model) on the TAA 550. Seems to be an amplified string of critical 'nearest-to-zero' temperature coefficient Zener-wired PNP and NPN transistors. Quite a device, if you can get a big bag of them inexpensively enough! Beware tho' they're packaged in little metal cans … which now cost more than chicken's teeth.
⋅-=≡ GoatGuy ✓ ≡=-⋅
Seems to be good info. 33 volts ("B" model) on the TAA 550. Seems to be an amplified string of critical 'nearest-to-zero' temperature coefficient Zener-wired PNP and NPN transistors. Quite a device, if you can get a big bag of them inexpensively enough! Beware tho' they're packaged in little metal cans … which now cost more than chicken's teeth.
⋅-=≡ GoatGuy ✓ ≡=-⋅
I think the confusion about varicaps is that there are diodes intended to be used as voltage controlled capacitors also called varactors. What is proposed here are the voltage regulators that were made to provide the stable voltage required to allow such diodes to work properly. As the capacitance of the varicap was controlled by the voltage biasing it any small variation or drift in the control voltage would cause the capacitance to change. This would case the selected frequency to drift. 1% of capacitance drift on a UHF analog TV station would completely change the channel!
The 150 volt gas regulator tubes would have an equivalent internal impedance as low as 150 ohms. The voltage could also vary by 2% typically. Unlike Zener diodes they really aren't affected by temperature.
The same technique used to reduce the temperature coefficient was also used in some low voltage reference ICs. Mostly replaced by band gap references, which pretty much also replaced standard cells. (A gizmo I suspect most here have never seen.)
Before fancy IC voltage references the Zener temperature stabilization technique was to use forward biased diodes in series to get a complementary voltage/temperature drift (usually quoted as 26 mV/C which really varies a bit in practice) to compensate the Zener diodes drift. When done well it did get the drift close to zero. Another technique was to use a special resistor that had a calibrated thermal change in value. Usually 3300 ppm/C. A more interesting method was to use a biased Zener diode connected to a biased forward voltage drop diode string of the same voltage connected by a potentiometer. As the voltage was the same on both sides of the potentiometer the reference voltage could be taken from the wiper. As the temperature changed the voltage on one side of the potentiometer would rise and the other side fall. All you had to do was set the potentiometer to the point where the temperature effects cancelled. A practical circuit would use a reference Zener and a lower voltage unit with a string of forward biased diodes. The potentiometer would then feed an emitter follower. The control would also compensated for the transistor's thermal shifts! Adding a resistor to such a circuit would allow it to replace a gas shunt tube.
The 150 volt gas regulator tubes would have an equivalent internal impedance as low as 150 ohms. The voltage could also vary by 2% typically. Unlike Zener diodes they really aren't affected by temperature.
The same technique used to reduce the temperature coefficient was also used in some low voltage reference ICs. Mostly replaced by band gap references, which pretty much also replaced standard cells. (A gizmo I suspect most here have never seen.)
Before fancy IC voltage references the Zener temperature stabilization technique was to use forward biased diodes in series to get a complementary voltage/temperature drift (usually quoted as 26 mV/C which really varies a bit in practice) to compensate the Zener diodes drift. When done well it did get the drift close to zero. Another technique was to use a special resistor that had a calibrated thermal change in value. Usually 3300 ppm/C. A more interesting method was to use a biased Zener diode connected to a biased forward voltage drop diode string of the same voltage connected by a potentiometer. As the voltage was the same on both sides of the potentiometer the reference voltage could be taken from the wiper. As the temperature changed the voltage on one side of the potentiometer would rise and the other side fall. All you had to do was set the potentiometer to the point where the temperature effects cancelled. A practical circuit would use a reference Zener and a lower voltage unit with a string of forward biased diodes. The potentiometer would then feed an emitter follower. The control would also compensated for the transistor's thermal shifts! Adding a resistor to such a circuit would allow it to replace a gas shunt tube.
Last edited:
If I understand the graph correctly, the voltage slightly drifts due to self-heating and the nonzero temperature coefficient. The thermal time constant is of the order of 1 minute, so it takes a few minutes for this process to settle. That is, it starts almost instantly, but needs a few minutes to settle precisely to the final voltage.
By the way, glow discharge voltage regulator tubes can take many seconds to start up if the tube has no radioactive primer.
I infer it takes 1 minute for the regulated voltage to rise to 60% of spec level when the zener current is 5mA, and at t=0 the regulated voltage starts at 0V.
That is opposite of the VR, where VR voltage rises up to strike voltage at the start, so could sit at a high level for a while until it strikes.
For some applications, that difference in start-up may be an issue - it would depend on an assessment. So it may not be as simple as using it instead of an ss zener (which is effectively instantaneous).
That is opposite of the VR, where VR voltage rises up to strike voltage at the start, so could sit at a high level for a while until it strikes.
For some applications, that difference in start-up may be an issue - it would depend on an assessment. So it may not be as simple as using it instead of an ss zener (which is effectively instantaneous).
Metal was for early devices: they quickly switched to plastic (TO92), then 1N4148-like diode case, TDA1550 for example.
The graph is a bit subtle and misleading: it is a ratio of the delta's, not the voltage, and the max delta is 3.2mV (for 33V device), 0.0001% in fact.
In addition typical tempco is zero.
This means in practice that the start-up is instantaneous, just like zeners, which is normal because you were not going to wait several minutes for your TV to stabilize each time you switched it on
Because they are ordinary and conscpicuous (they look like ordinary diodes or transistors), you can easily miss them.
I was able to pick a bag of TDA1550's for very cheap many years ago
Last edited:
I also was puzzled somewhat first when reading varicap regulator and wondered how a varicap could regulate at all. But then it became clear that the IC's mentioned are regulators for varicap tuning diodes, mainly in TV sets, often with a so called Preomat. They provided a highly stable and temperature compensated voltage of -33 Vdc from which the tuning potentiometers were fed.
Nowadays they're almost obsolete.
Best regards!
Last edited:
At least that tells me where to look to salvage a couple- the tuner. I have some color TV's in odd corners of the garage waiting to buy the disposal tags from the city.
Last edited:
I'd be not too sure about that. At least in most German TV's these could be found in the PSU section, rather than next to the front end.
Best regards!
Best regards!