Clifton Laboratories 7236 Clifton Road  Clifton VA 20124 tel: (703) 830 0368 fax: (703) 830 0711

E-mail: Jack.Smith@cliftonlaboratories.com


 

Home
Updates
Prior Products - no longer available
Documents
Book
Modulation
Loads
Lattice Crystal Filter
IMD Measurements
Using LP100 Coupler
Prototyping
Software Updates
K2 Measurements
Oscillator Noise Measurements
Bypassing
Capacitor Voltage Change
K2 Freq Stability
Cohn Crystal Filter
Receiver AGC Curves
K2 RX Sensitivity
Canned Osc Phase Noise
K2 Interface
K2 Filter
Surface Mount Assembly
TL750L Low Dropout Regulator
Swordfish DDS
Swordfish GLCD Module
Bessel Nulls
AM Modulation
Z10000 with FT-920
Z100 Tuning Aid
Dayton 2007
Softrock Lite 6.2
Header Adapter
Carbon Composition Resistors
Thermometers
Hakko FT-800 Thermal Wire Stripper
Heat Sinks
Diode Turn-On Time
Bill Hewlett and his Magic Lamp
Tektronix P6022 Current Probe
1N400x Diode Family Forward Voltage
Temperature Chamber
Diode Vf vs If
Ferrite Transformers
6 dB Hybrid Combiner
Type 43 Ferrite B-H Curve
K3 IF Bandpass Filter
Estimating Q of Ferrite Cores
Z10000 Buffer Amp
Z10010 Bandpass Filter
Using Softrock as a Panadapter for the K2
Signal Generator Phase Noise & Elecraft K2
Audio Transformer Data and Modeling
Measuring 60 Hz Frequency
Compact Fluorescent Lamp
Z10000-U Buffer Amp and FT-1000MP
WJ-8617B Receiver Impressions
Weather in Clifton VA
Radio Intelligence Example
Diodes for RF Probes
PIC Development Boards and Programmers
Elecraft K3 and Panadapters
Elecraft K3 AGC and S-Meter
Elecraft K3 Noise Blanker and Crystal/DSP Filtering
Jackson Harbor Press VLF Converter
Elecraft K3 Receive Audio
Headphone Impedance

 

 
TL750L Low Dropout 12V Regulator Oscillations

My Z93 RF/IF prototype PCB powers three AD8007 amplifiers with Texas Instruments TL750L12 12 volt low dropout regulators. These devices are similar to the familiar 78L12 TO-92 packaged regulators, but will maintain a stabilized 12V output with an input only 0.6V greater than the output voltage. The 78L12, has a 1.7 V dropout voltage, which makes the TL750L an attractive choice as it permits stabilized 12 V with only 12.6 V input.

Working with the prototype board, however, shows that the TL750L is not a direct "drop in" replacement for a 78L12 regulator, and unless the output is properly bypassed, the TL750L12 will oscillate.

TI's TL750L12 datasheet can be found at http://focus.ti.com/lit/ds/symlink/tl750l05.pdf.
 

Bypassing typically used for 78Lxx series regulators will not work for TL750L12 low dropout regulators. Do NOT use this circuit!

My prototype design uses three TL750L12 regulators, and here's how the +13 V input voltage looked when first built:
 
+13 V input bus with three TL750L12 regulators, improperly bypassed.

There's a 2 V peak-to-peak oscillation at around 50 KHz superimposed on the 13 V bus. And, this is with a 4700 μF capacitor across the +13 V input. The +13V input is supplied by an HP 6205C laboratory-grade variable DC power supply.
 
Oscillation on +13 V input is at about 50 KHz.

The TL750L12's output also has a high frequency oscillation, at about 1 V peak-to-peak.
 
TL750L12 Output voltage, with oscillation.

To say the least, I was not expecting this, as I've never seen similar instability with the 78Lxx regulators I've used in many other projects.

As usual, it pays to read the data sheet before designing in the part. If I had, I would have seen the following caution:

The TL750L, TL751L series are low-dropout regulators. This means that capacitance loading is important to the performance of the regulator because it is a vital part of the control loop. The capacitor value and its equivalent series resistance (ESR) both affect the control loop and must be defined for the load range and temperature range. Figure 1 shows the recommended range of ESR for a given load with a 10-μF capacitor on the output.

The load current in my design is around 30 mA for each TL750L12, so Figure 1 says that the output 10 μF capacitor should have an equivalent series resistance (ESR or RS) below 0.4 ohms.

My junkbox has a few 10 μF /16v tantalum capacitors, but do they have the requisite low ESR? If you have an RLC bridge that measures C and D, you can easily compute the ESR. (If you don't have a bridge, or an ESR meter, there are other methods of determining the ESR.)

We know that D or "dissipation factor" is defined, in terms of series capacitance and series resistance, as:

D = ωRSCS

Solving for Rs:

 RS = D/ωCS

ω = 2πf where f is the test frequency

Here's what I found when measuring three 10 μF capacitors, with a General Radio 1658 Digibridge, test frequency 1000 Hz:

Nominal Value Type Measured CS Measured D Computed RS (Ohms)
10 μF / 16 V Tantalum radial 9.8 μF 0.0250 0.406  
10 μF / 10 V Tantalum radial 10.28 μF 0.0364 0.564
10 μF / 50 V Alum. Electrolytic 9.70 μF 0.1177 1.931
         
Leaving aside the 10 V rated capacitor—definitely not a good choice for a 12 V power supply rail—the first capacitor tested will work, although its Rs is up against the recommended maximum 0.4 ohms. The aluminum electrolytic's RS grossly exceeds the recommended 0.4 ohm value, and hence is not usable for this purpose.

Applying one 10 μF / 16 V tantalum capacitor to each TL750L12 regulator's output completely removed the instability. The output voltage shows essentially zero noise and the input voltage shows about 25 mV high frequency hash coupled into the oscilloscope probe from nearby digital circuitry.
 

+13 V input after applying 10 μF / 16 V tantalum capacitors to outputs of all three TL750L12 regulators.

The noise is mostly coupled into the oscilloscope probe from digital circuitry on the board.

 
Recommended bypassing for TL750L12. The 10 μF tantalum capacitor must have Rs < 0.4 Ω and should be installed physically close to the regulator.