testing for hcpl 4562

[fstspec]

Hi

I need to test whether whether this IC will function properly by imputting a 5, 10, 25Mhz sine wave at 0.2mVp-p. After connecting as shown below, i did not get any output sine wave.

Can anyone clarify with me what are the proper testing procedures.

 

[Nigel Goodwin]

I would suggest you use a scope to follow the signal through, for a start check on the input and output of the opto-isolator.

 

[RadioRon]

The circuit does not show any input source of a sine wave, so first you need to attach a sine wave generator to the input. Then set the generator to a low frequency, like 100KHz for example, with amplitude of about 0.25 volts. Next, use an oscilloscope and probe the input to make sure there is indeed an input signal of the correct amplitude. If that is OK, probe the input to the optoisolator (the HP part) to see if the signal appears there. If it does, then look at the output of the HP part on the right side, at the collector of the output transistor.

Continue to follow the signal path through the circuit to see if the signal is coming through. When you find a point where the signal should be but it isn't, then you conclude that there is something wrong just before that point.

 

[kchriste]

I need to test whether whether this IC will function properly by imputting a 5, 10, 25Mhz sine wave at

I don't think the circuit will work well into the Mhz range without some supply decoupling caps on the power rails.
If you don't have a scope, check the bias voltages. Check that you have about 1.2V at the base of Q1 and apx 0.5V at the emitter.
Adjust R4 and see if the voltage at the emitter of Q4 varies. Then set it for apx 1/2 VCC2
Now try your signal again.

 

[fstspec]

I got the sine wave at the output at 100Khz. But after tuning to higher frequency (5Mhz), the wave becomes distorted. In the circuit, I have replaced the diode IN4150 with IN4148 as there are no more such transistors in my lab. Need to clarify whether it has any impact on the output as the fast switching diode is needed only at the input stage.

 

[fstspec]

If that is OK, probe the input to the optoisolator (the HP part) to see if the signal appears there. If it does, then look at the output of the HP part on the right side, at the collector of the output transistor.

Continue to follow the signal path through the circuit to see if the signal is coming through. When you find a point where the signal should be but it isn't, then you conclude that there is something wrong just before that point.

Need to check with you, what do you mean by the HP part?

 

[fstspec]

I don't think the circuit will work well into the Mhz range without some supply decoupling caps on the power rails.

Hi, I am using LM7809 and LM7805 voltage regulators to generate the +5V and +9V. I have used 0.1uF caps between supply and ground; output and ground. Do I still need decoupling caps at the power rails.

Another question, what are the use of decoupling caps.

 

[fstspec]

If you don't have a scope, check the bias voltages. Check that you have about 1.2V at the base of Q1 and apx 0.5V at the emitter.
Adjust R4 and see if the voltage at the emitter of Q4 varies. Then set it for apx 1/2 VCC2

I check the voltages already. I have 1.146V and 0.446V at the base and emitter of Q1.
I tried adjusting the value of R4 from 51 to 521 ohm, the voltage at emitter of Q4 remains the same at 1.7V.

Pls advise

 

[fstspec]

I did check for all the biasing voltages for my circuit, i think the reason why the Vout is not biased at 0.5 of Vcc2 is because base of Q4 is 2.47V.

Any ideas how should i modify the circuit so that it is operating at 0.5Vcc2

 

[RadioRon]

Need to check with you, what do you mean by the HP part?

Oh, sorry. I meant the HCPL 4562 opto coupler. This part is from Agilent who used to be called Hewlett Packard or HP.

 

[RadioRon]

I got the sine wave at the output at 100Khz. But after tuning to higher frequency (5Mhz), the wave becomes distorted. In the circuit, I have replaced the diode IN4150 with IN4148 as there are no more such transistors in my lab. Need to clarify whether it has any impact on the output as the fast switching diode is needed only at the input stage.

This diode is mainly just for ensuring that DC bias to Q1 is at least 0.7 volts. The 1n4148 diode should be fine there.

What is the appearance of the distortion? When I suggested 0.25 volts at 100 KHz, I meant this to be a starting point and it seems to have worked OK. If the distortion that appears as you increase frequency is peak clipping, then perhaps your gain is increasing with frequency and you simply have to back off the input level to something lower than 0.25 volts to get a better sine wave output. Can you describe the distortion, or provide a picture of it at about 1 MHz?

 

[RadioRon]

I did check for all the biasing voltages for my circuit, i think the reason why the Vout is not biased at 0.5 of Vcc2 is because base of Q4 is 2.47V.

Any ideas how should i modify the circuit so that it is operating at 0.5Vcc2

After a quick look at the circuit, I'm wondering why adjustment of R4 didn't shift the Q4 emitter voltage at all. It is important that you check this when there is no input signal at all, so make sure the input is disconnected. Since it appears that the Q1 collector current should go as high as 8.6 mA when R4 total is 51 ohms, I think this should be enough to shift the base voltage at Q2. Can you measure the variation in Q2 base and Q2 emitter voltages as R4 is adjusted from lowest to highest resistance?

 

[kchriste]

Hi, I am using LM7809 and LM7805 voltage regulators to generate the +5V and +9V. I have used 0.1uF caps between supply and ground; output and ground. Do I still need decoupling caps at the power rails.

If the caps are more than an inch or two away from your video/RF portion of the circuit, then yes.

Another question, what are the use of decoupling caps

It prevents the signal from appearing on the power supply leads which leads to all kinds of havoc and strange behaviour.

I took a look at the last page of this doc:
**broken link removed**
It seems that R4 is supposed to adjust the gain of the circuit and that you select R11 so that the current through Q4 is at an optimum as defined by formula (8)