Pushing TS555 to Mhz region

Thread Starter

decltype

Joined Mar 3, 2021
3
Hello. A novice searching for guidance here.
I'm trying to assemble a simple schematic that'll give me 1-2 Mhz (actually 1.65) square signal. After fruitless attempt at classical multivibrator with 2 transistors, I moved on to ICs. First tried NE555, but after some research I figured that it can't go higher than a few hunderd kHz.
But TS555 specification states its advantage of going up to a couple of Mhz. Datasheet features this formula: \[ f = {1.44 \over (R_a + 2R_b) * C} \]
I assembled it on a breadboard:
layout.png
For measuring the frequency, I use GM328A. It may not be the most accurate device, but it should handle this and unfortunately I don't have access to a proper oscilloscope.
Sadly, the maximum number I saw on the screen was about 250kHz. I tried a range of capacitors and resistors. Even tried different voltage out of curiosity: 4.5, 9, 24. Also, tried adding some resistance load. Tested on 2 different TS555 chips, both give the same result.
Any advice to what I am missing here?
 

AnalogKid

Joined Aug 1, 2013
9,079
Decrease C1 to 10 nF, and recalculate Ra and Rb. If you want a symmetrical square wave, use this circuit:

1614778826483.png

NOTE: Above 1 MHz, the type and construction of the capacitor will affect its performance. Choose one with a high self-resonant frequency, and use good point-to-point assembly techniques.

ak
 

Ian0

Joined Aug 7, 2020
1,656
Don't forget the propagation delays (quoted at 100ns in the ST datasheet). That will add 200ns to the time period, and if you want 1.65MHz (600ns period) then you will need to calculate the capacitors for 400ns.

Is there a reason you want to achieve this with a 555, and not some other more suitable device? Like a 74HC14 for instance? or a crystal oscillator?
 

Thread Starter

decltype

Joined Mar 3, 2021
3
Thank you for all the replies.

Use a smaller capacitor for C1 and larger resistors for Ra and Rb.
Bertus
I used 1 pF capacitor (lowest I could find). With it I managed to get to 1.2 Mhz with Ra ~5 kOhm and Rb ~460 kOhm. I guess the formula doesn't hold at these high ratings.

Decrease C1 to 10 nF, and recalculate Ra and Rb. If you want a symmetrical square wave, use this circuit:
NOTE: Above 1 MHz, the type and construction of the capacitor will affect its performance. Choose one with a high self-resonant frequency, and use good point-to-point assembly techniques.
ak
Also tried with 10nF, got much lower output frequencies. I couldn't find spec for my caps, but from what I read on the web, for 1pF ceramic capacitors, it should be in terms of Ghz, so should be fine.

Don't forget the propagation delays (quoted at 100ns in the ST datasheet). That will add 200ns to the time period, and if you want 1.65MHz (600ns period) then you will need to calculate the capacitors for 400ns.

Is there a reason you want to achieve this with a 555, and not some other more suitable device? Like a 74HC14 for instance? or a crystal oscillator?
That's a good catch, I've incorporated 200ns into calculations, but unfortunately it still won't go over 1.2Mhz. I may play with it some more later, maybe I'll get something.
Answering your question, there's no particular reason for using TS555. I saw it being used in a similar project (ultrasonic atomizer), but the author needed ~130 kHz output for his piezoelectric disk.
 

BobTPH

Joined Jun 5, 2013
3,020
Was that a plug-in breadboard, which is often what's being referred to?
Yep, the standard white plastic with many holes to accept components and wires.

Of course I did not loop the wires up four inches above it in a hopeless tangle, I used the correct length jumpers laying right on the suface. The inductor was 4 or 5 turns of hookup wire formed by winding around a pencil. I could tune it by spreading or compressing the coil.

I might just try to reproduce it. It was a single transistor, pn2222 as, I think, a Colpitts oscillator, since it needed only 1 coil.

Bob
 

Audioguru again

Joined Oct 21, 2019
2,881
Most people looking at this thread are thinking about an old powerful NE555 but this TS555 is a Cmos one and is blown up since its max operating voltage is 16V and he tried it with 24V.

Its datasheet shows its Discharge pin has a 1mA output but the 450 ohm and 190 ohm resistors used are WAY too low a value.
With a 5V supply its output logic high current is 100 times less than an old NE555.
It cannot drive the high stray capacitance of the rows of contacts and wires all over the place on a solderless breadboard at radio frequencies.
 

Thread Starter

decltype

Joined Mar 3, 2021
3
Did you see this in the datasheet:
If you mean the last line, yeah, tried those values too.

Its datasheet shows its Discharge pin has a 1mA output but the 450 ohm and 190 ohm resistors used are WAY too low a value.
With a 5V supply its output logic high current is 100 times less than an old NE555.
It cannot drive the high stray capacitance of the rows of contacts and wires all over the place on a solderless breadboard at radio frequencies.
So the conclusion here is that breadboard is too parasitic for this thing to work? Will try soldering it then.
 

Audioguru again

Joined Oct 21, 2019
2,881
An oscillator is not useful without a load. What resistance will be your load? The Cmos 555 might drive 1.2k ohms if the supply is 10V and the output is 6Vp-p.
 

Ian0

Joined Aug 7, 2020
1,656
The TS555 datasheet says that Discharge has a 50mV saturation voltage at 1mA output but doesn't specify the maximum output current. Texas's TLC555 has a 30mV saturation voltage at 1mA and a 150mA limit. That suggests to me that the discharge pin has a MOSFET with an Rds(on) of 30 ohms or 50 ohms.
TS555 has a "typical" maximum astable frequency of 2.7MHz, specified with RA = 470 Ω, RB = 200 Ω, C = 200 pF, VCC = +5 V, but specifies no minimum.
TLC555 has a typical maximum astable frequency of 2.1MHz, specified with the same component values. But Texas also specifies a minimum maximum astable frequency (which sounds contradictory) of 1.2MHz, suggesting that the SGS-Thomson part might have a similar lower limit. So perhaps, 1.2MHz is as good as it gets.
 

dendad

Joined Feb 20, 2016
3,796
Does it have to be a 555?
Years ago, I built this VCO using a 4046 chip as part of my development test setup. Use a multi turn pot.
4046VCO.jpg
It runs from about 200Khz to 2Mhz in one range.
Adding resistors to the ends of the pot will limit the frequency range if you like.
But, if you need more accuracy and stability, an Arduino driving an Si5351 board or an AD9850 DDS (more expensive) can work wonders.
 

MrChips

Joined Oct 2, 2009
23,099
Just as a test I fired up a TI TLC555CP @ 5V with just a 100Ω timing resistor on a solderless breadboard. It runs at 1.75MHz.
 

whershey

Joined Mar 12, 2017
1
The TS555 datasheet says that Discharge has a 50mV saturation voltage at 1mA output but doesn't specify the maximum output current. Texas's TLC555 has a 30mV saturation voltage at 1mA and a 150mA limit. That suggests to me that the discharge pin has a MOSFET with an Rds(on) of 30 ohms or 50 ohms.
TS555 has a "typical" maximum astable frequency of 2.7MHz, specified with RA = 470 Ω, RB = 200 Ω, C = 200 pF, VCC = +5 V, but specifies no minimum.
TLC555 has a typical maximum astable frequency of 2.1MHz, specified with the same component values. But Texas also specifies a minimum maximum astable frequency (which sounds contradictory) of 1.2MHz, suggesting that the SGS-Thomson part might have a similar lower limit. So perhaps, 1.2MHz is as good as it gets.
TI's data sheet for the LMC555CN says "LMC555 CMOS Timer Industry's Fastest Astable Frequency of 3 MHz"
Clearly they are proud of their specs. You might want to give this a try.
 

Ian0

Joined Aug 7, 2020
1,656
TI's data sheet for the LMC555CN says "LMC555 CMOS Timer Industry's Fastest Astable Frequency of 3 MHz"
Clearly they are proud of their specs. You might want to give this a try.
Fmax=3MHz is a "typical" spec. It's by no means guaranteed. They don't give a minimum!
If it's typical, then half will manage over 3MHz and half won't.

It's very easy to fall into the trap of thinking all CMOS 555s are alike. They are all different. Different maximum speeds, different output sink currents, different output source currents, different discharge currents, different reset thresholds.
Only the trigger-pin and the threshold-pin thresholds at 33.3% and 66.6% of Vdd respectively are consistent.
 
Top