logo elektroda
logo elektroda
X
logo elektroda
Dostępna jest polska wersja

Czy wolisz polską wersję strony elektroda?

Nie, dziękuję Przekieruj mnie tam

Mystery electronic component from a digital device

TechEkspert  28 3972 Cool? (+31)
📢 Listen (AI):

TL;DR

  • A 14T5750 module from a 286 motherboard turned out not to be a crystal oscillator but a digital delay line with VCC, GND, OUT and multiple taps.
  • It runs from 5V and shifts the input signal through outputs with successive delays, likely to distribute timing edges for RAM or other motherboard timing.
  • The datasheet says the IN-to-OUT offset is about 76ns, and the author also found related models like 14HT5750, 14HT5125, and 14HT5500.
  • Bench tests showed 17mA idle current, rising to 32mA with a 100KHz input, and the output waveform clearly lagged the input.
  • The article also compares classic delay lines from PAL and SECAM TVs, plus their uses in oscilloscopes and early mercury memory.
Black CTC 14T5750 relay with labeled pins on top surface
When I visited an electronics colleague recently, he surprised me by putting a small object on the table and asked what the electronic component could be. The component was working on a board in digital circuits. On the casing you will see the markings VCC, GND and OUT. You might suspect that this is some sort of quartz generator, such as DS32KHz or higher frequency generators such as TCXO . Also visible on the chassis is the designation of an input, perhaps controlling the operation of the circuit, and several outputs marked with consecutive numbers.

Do you know what this electronic component is used for?

I was wrongly betting that it was a quartz generator. On old computer motherboards I often saw several generators with different frequencies, usually enclosed in a metal case. Apparently PLL circuits were not used then and it was easier to use several generators with different frequencies.

Black integrated circuit labeled “14T5750 CTC 9232” with four pins below


After typing the symbol 14T5750 into google, it was possible to find descriptions of 286 motherboards where this component was used:
https://theretroweb.com/motherboards/s/auva-computer-baby-286-bam-12-g1
https://theretroweb.com/motherboards/s/morse-kp286hf

Below are screenshots from the found pages with the undetermined purpose element highlighted. Miniature DC/DC converters are now available in similar enclosures, but this component is not an inverter.

Motherboard from the 1980s–90s with Headland chips and ISA slots


Close-up of 1980s computer motherboard with ISA slots and marked components


Managed to find a pdf describing the parameters of this component:
http://www.datasheet.hk/view_download.php?id=1695449&file=0349%5C14t5750_2726584.pdf

It turns out that the manufacturer supplied more models of this component: 14T5750 14HT5750 14HT5125 14HT5251 14HT5500 14T5451 14T5500

The riddle is explained by the contents of the pdf:

Technical diagram and specifications for Allen Avionics 14-pin DIL device


Diagram and delay specification tables of 14T delay line component


This is a delay line! The circuit is powered by 5V, has an output and several 'taps' with different delays to the input signal. What could such a delay line have been used for on the motherboard of a computer with a 286 processor? I suspect that it was a way of distributing timing signals, with a specific offset/delay of the signal edges. Such an offset may have had applications, for example, in controlling RAM operation. If board schematics with such a line were available, perhaps something could be confirmed.

Let's see if the line still works. We supply the circuit with 5V -> current consumption about 17mA. We apply a 100KHz signal to the input and the current consumption increases to about 32mA. It appears that the embedded circuitry in the line is working.

14T5750 CTC IC connected to breadboard with jumper wires


This is what the IN input signal (yellow trace) and the OUT output signal (blue trace) look like. The action of the delay line is visible.

Oscilloscope screen showing two voltage waveforms in different colors


The offset is approximately 76ns according to the parameters in the description.

Oscilloscope screenshot with two waveforms: yellow and blue, showing amplitude and time measurements


This is what the signals look like on the next "taps" of the line relative to the IN signal:
Rigol oscilloscope screen with four electrical signal waveforms displayed


I was not aware of such a solution for distributing timing signals and delaying digital signals. Although I have seen solutions where gates were used to influence the signal propagation, but then the delay times were shorter.

Have you encountered the use of delay lines in digital devices?

By the way, I am posting pictures of another familiar delay line that was used in analogue TV sets . Specifically, such delay lines were found in PAL decoders. The line delay was 64µs .

Black Soviet chip УЛЗ64-5 and blue module labeled DL-711 PHILIPS


We feed the pulses into the input (yellow trace) and at the output (blue trace) you can observe the pulses appearing with a delay.

Oscilloscope screenshot with two channels: one pulsed signal and one with noise


The delay is 64µs:

Rigol oscilloscope screenshot with yellow digital and blue analog waveforms


Let's take a look inside the 64µs delay line.

Sensor prototype with metal wires on transparent board with black dots


The signal travels a rather complex path in the quartz plate, bouncing at its extremities. Thanks to the multiple reflections, the dimensions of the device can be smaller.

Interior of an analog delay line with signal path marked


Estimating the frequency response with an oscilloscope and tunable generator , the result is quite predictable, the amplitude peaks between 4-5MHz. You can also see that the line loads the generator output, the amplitude of the input signal varies over a small range.

Oscilloscope screen with AM signal and modulating waveform from 1–10 MHz


SECAM decoders also included a delay line which was a long thin wire wound around a tube the size of a ballpoint pen. The delay of such a line was less than that of a 64µs line. Below is an example of a module with a 560ns line.

Circuit board with blue delay line UNITRA TELPOD LD-4 560ns 1kΩ


The delay lines used in televisions are still well recognisable, but with the passage of time, like the 14T5750, they can be a puzzling object for their finder.

Delay lines were used in oscilloscopes with a delayed time base. In the early days of computer science and calculating machines, operational memories were built on the basis of delay lines. Bits circulated in a delay line and were re-transmitted to the line after being received. This is how mercury memory worked.

About Author
TechEkspert
TechEkspert wrote 6928 posts with rating 5391 , helped 16 times. Been with us since 2014 year.

Comments

gregor124 04 Oct 2025 11:58

In digital systems, delay lines were often used to produce the signals necessary to drive dynamic RAMs. [Read more]

TechEkspert 04 Oct 2025 13:02

An interesting stage where the timing of the RAS/CAS signals was determined by the physical parameters of the soldered-in delay lines. A significant technological leap occurred when configuration was done... [Read more]

E8600 04 Oct 2025 14:07

Where can such an electronic component be used in DIY projects? [Read more]

gregor124 04 Oct 2025 14:40

Maybe not like this, but e.g. with more delay to produce an echo effect? Delay lines were e.g. mercury memories in the early days of computers. [Read more]

TechEkspert 04 Oct 2025 18:41

It is rather hard to find a sensible use for these components at the moment, although perhaps I have a poor imagination :) If this delay line can be opened I will also post pictures showing the internal... [Read more]

omron 04 Oct 2025 19:36

ad vocem: ''Specifically, such delay lines were found in PAL decoders'' - this ''64us chrominance delay line'' - was in SECAM decoders too !!! ''In SECAM decoders there was also a delay line which was... [Read more]

TechEkspert 04 Oct 2025 19:49

To date, I cannot write from memory what the 64µs delay line in SECAM and PAL decoders was used for, let alone the 560ns line in a PAL decoder. Maybe one day someone on elektroda.pl will post an article... [Read more]

coberr 04 Oct 2025 20:17

Not surprisingly.... hardly anyone today knows how decoders work - in fact, only the "old guard" ... while the youngsters don't even have a "drive"... :) The line served precisely to delay the luminance... [Read more]

acctr 04 Oct 2025 23:32

It is still used, for higher frequencies the delay is realised by tracks on the PCB . [Read more]

ZbeeGin 05 Oct 2025 08:57

Unfortunately you have simplified this considerably. The colour decoders of older TV sets had two delay lines: - 64µs in the chrominance path, processing colour information, - 0.7µs in the luminance... [Read more]

gregor124 05 Oct 2025 14:50

Delay lines were used in devices for encoding voice information using radios and telephone calls. [Read more]

tytka 05 Oct 2025 16:59

In echo and reverberation circuits, however, it is definitely higher delays that are needed compared to what this line offers. In the old echo and reverberation systems, spring-loaded systems were very... [Read more]

TechEkspert 05 Oct 2025 17:15

@gregor124 I didn't know that, is this some kind of scrambling? In a typical block or stream encryption, I don't see the need for such a line. @tytka I associate echo with the period after spring reverb... [Read more]

gregor124 05 Oct 2025 17:32

Of more interesting applications, for example, they were used in some very high-speed analogue oscilloscopes, so that you could see, for example, the signal with a delay before the trigger signal appe... [Read more]

acctr 05 Oct 2025 17:33

In general, the higher the frequency, whether of a mechanical or electromagnetic wave, the shorter the delays. It may be interesting to note how strings were used by Galileo Galilei to discover the laws... [Read more]

tytka 05 Oct 2025 19:05

I also associate such solutions, And the spring, I had the opportunity to play with quite recently. I had to repair a Pioneer QX-9900, which has reverberation effects realised with such a device. By... [Read more]

gregor124 06 Oct 2025 00:57

This worked on the principle that the signal was split into a number of short time intervals and each section was passed through a different (code-set) delay. The result was that the signal was slightly... [Read more]

minus3db 06 Oct 2025 12:09

Speaking of oscilloscopes - a beautiful and massive (gigantic!) delay line has a Polish oscilloscope made by ZRK OS-150. It consists of several coils of coaxial cable in an outer braid. It looks interesting,... [Read more]

cirrostrato 06 Oct 2025 17:47

The old memory still works, you need from the received signal to reconstruct three colour signals plus luminance (the one sent in each successive line), in successive lines sent one of the two colours... [Read more]

FAQ

TL;DR: It’s a 5 V digital delay line with ~76 ns tap spacing; “This is a delay line!” confirmed by scope tests. [Elektroda, TechEkspert, post #21709529] Why it matters: Helps you identify, test, and repurpose legacy delay-line modules from vintage computers and AV gear.

Quick-Facts:

Quick Facts

What is the mystery component labeled VCC, GND, IN/OUT, and numbered taps?

It’s a digital delay line module. It powers from 5 V and provides several outputs with progressively increased delay relative to IN. Scope shots show discrete tap delays (~76 ns), confirming the function. “This is a delay line!” was verified by measurement during bench tests. [Elektroda, TechEkspert, post #21709529]

How were delay lines used on 286-era motherboards?

They generated precisely shifted timing edges for memory control. Designers used fixed delays to meet DRAM setup/hold for RAS/CAS and address strobes, before flexible BIOS/SPD tuning. This hardware approach locked timing to the soldered part’s propagation. [Elektroda, gregor124, post #21709614]

How do I bench‑test a 14T5750‑type delay line?

Power at 5 V. Drive IN with a square wave (e.g., 100 kHz). Probe OUT and taps on an oscilloscope and measure phase shift versus IN. Expect idle ≈17 mA, rising to ≈32 mA when driven. [Elektroda, TechEkspert, post #21709529]

Can I use this part to make an audio echo or reverb?

No. The delays here are tens of nanoseconds per tap. Audio echo needs milliseconds or more. “In echo and reverberation circuits, higher delays are needed compared to what this line offers.” Use BBD or digital delay ICs instead. [Elektroda, tytka, post #21711019]

What did the 64 µs and ~0.7 µs TV delay lines do?

In PAL/SECAM, a 64 µs chroma line stores a full line of color data for phase correction or sequential decoding. A ~0.7 µs luminance line time‑aligns brightness with color to avoid color fringing at edges. [Elektroda, ZbeeGin, post #21710419]

Are delay lines still relevant on modern boards?

Yes, but often realized as controlled‑length PCB traces at high speeds. Designers tune propagation by routing length and dielectric, achieving required picosecond‑to‑nanosecond delays without packaged modules. [Elektroda, acctr, post #21710330]

Where could a hobbyist reuse this module today?

Use it for pulse alignment, edge‑staggering demonstrations, glitch filtering by time‑selection, or repairing vintage boards. For creative audio, choose BBD or digital delay instead since this module’s delay is too short. [Elektroda, TechEkspert, post #21710035]

What is a BBD (Bucket Brigade Device) in simple terms?

A BBD shifts analog samples through many MOSFET‑capacitor stages, creating millisecond delays for effects like chorus or flanger. Examples include MN3207 (1024 stages) and MN3007 families used in classic pedals. [Elektroda, thereminator, post #21718079]

Did oscilloscopes use delay lines, and why?

Yes. High‑speed analog scopes inserted a delay so users could see pre‑trigger signal content. The main sweep starts, yet the delayed signal is still visible near trigger time. [Elektroda, gregor124, post #21711061]

How is voice scrambling achieved with delay lines?

A scrambler splits audio into time slices and routes each slice through code‑set delays. The receiver applies matching delays to reassemble intelligible speech. The method adds slight latency but is effective. [Elektroda, gregor124, post #21711514]

What’s an example statistic for the 14T5750 tested here?

Measured tap delay was about 76 ns, with current rising from ≈17 mA idle to ≈32 mA when driven at 100 kHz. These values confirm active internal buffering. [Elektroda, TechEkspert, post #21709529]

Any RF or microwave applications of delay lines?

Transmission‑line sections implement ultra‑wideband transformers. A Guanella 1:4 uses matched lines; one performs the transformation and another compensates delay for broadband behavior. [Elektroda, coderemover, post #21715739]

How‑To: Identify a delay line on a salvaged board (3 steps)

  1. Locate a metal can or module labeled VCC, GND, IN/OUT, and taps. 2. Apply 5 V and inject a square wave into IN. 3. Compare OUT/taps to IN on a scope; look for fixed nanosecond offsets. [Elektroda, TechEkspert, post #21709529]

What’s a practical failure or edge case to watch for?

Selecting the wrong tap can violate setup/hold timing and corrupt memory transactions. These modules were chosen to meet strict DRAM constraints. [Elektroda, gregor124, post #21709614]

Why did later PCs drop can‑style delay lines for RAM timing?

BIOS and SPD configuration enabled flexible, per‑board timing without swapping fixed‑delay hardware. PLLs and programmable controllers replaced discrete delay modules. [Elektroda, TechEkspert, post #21709668]
%}