laser range finder.. receiver fluctuation

[danishkaleem]

i am making a laser range finder.

the receiver i have made is showing a lot of fluctuations in the received pulse. (10-20 ns)

can anyone tell me a solution to it.

i am using two stage amplification..

in first stage:

transimpedance amplifier:
op amp: AD829

in 2nd stage:

voltage amplifier
op amp: LM6361


the photodiode used is S1223 from HAMATSU..


Danish.

 

[FvM]

I don't see how the pulse propagation delay could vary. I guess, you mean that the detected pulse arrival time after the comparator varies?

 

[danishkaleem]

yeah FvM .. i mean that the time of flight is not constant for the same target...

on the oscilloscope.. if i see the transmitted pulse and the received pulse.. then the time delay b/w the 2 pulses is not constant but keeps on fluctuating...

and its the received pulse which shows fluctuation otherwise the transmitted pulse is constant...

 

[keith1200rs]

You need to trace the signal through the system to find the source of the jitter, starting with the laser. A reverse biased PIN photodiode with a low resistance load can check the laser driver.

Noise in the circuit can introduce jitter.

Firing high power lasers can disturb the receiver due to the high currents involved and affect short range readings.

Showing your schematics may be helpful.

Keith

 

[danishkaleem]

i have already done this thing Keith... the signal is jitter free till the output of MOSFET driver (EL7104) for laser diode (SPL LL85).

then at the receiver end after the PIN photodiode (S1223) there is the 1st stage amplifier which do not introduce much of noise.. but after that the 2nd stage amplifier, alot of jitter is introduced at its output.

heres the schematic of my receiver circuit:


- - - Updated - - -

should i replace the 2nd stage amplifier with some other amplifier..?? can u suggest some good choice?

or any other suggestion?

 

[keith1200rs]

You wouldn't normally connect your photodiode between the inverting & non-inverting inputs of the transimpedance amplifier. The photodiode would normally be reverse biased for speed. R4 just increases the noise and I wouldn't include it.

You transimpedance amplifier is presumably generating a negative going pulse but your LM6361 is running from a single supply so cannot amplify negative going pulses. Maybe your photodiode is connected so the transimpedance amplifier gives you positive pulses?

I haven't used the LM6361 that I remember, but I would expect the resistors used to be a lot lower (although the LM6361 isn't very fast) and with a gain of 51 your bandwidth will be around 1MHz so I would be surprised if you see a laser pulse at all. You need a lot more bandwidth.

You also need some good decoupling and PCB layout.

Keith

 

[danishkaleem]

thanku keith for your concern...

well actually with the same setup ( except that i am using two supplies for LM6361 ) i am able to see the laser pulse.. but the fluctuations are still there..

i would appreciate if u can give me details about the schematics of the receiver you are talking about... ( in which photodiode is reversed biased and also not connected b/w inverting and non-inverting inputs )

i hope that will help to remove fluctuations??

by the way is it possible to get a constant received pulse.. i mean a constant delay b/w the transmitted and received pulse.. because with the passage of time it looks like a dream to me!!:!::!::-(

also can u tell me how can i improve the range.. because once i move the target beyond 2m. the received pulse almost vanishes! ..

thanku,
Danish.

 

[keith1200rs]

Danish,

The transimpedance amplifier configuration I would expect would be something like
**broken link removed**

Vbias would be positive, so it reverse biases the photodiode which improves speed and reduces capacitance.

The S1223 is not a good choice for a laser rangefinder. It is far too slow and high capacitance.

I am not sure the AD829 is a good idea either. It has very high current noise.

You don't show any capacitance across the feedback resistor of the transimpedance amplifier so I am surprised it isn't oscillating. Maybe it is but you will need a fast oscilloscope to see it.

It would be useful to see the waveforms you are getting from the transimpedance amplifier and subsequent amplifier.

Range is partly down to optics. I assume you have a collimating lens on the laser and a receiving lens for the PIN of some sort - what are their specifications?

Keith

 

[danishkaleem]

well thanks for your reply keith... but unfortnately the circuit which you have shown me is showing only oscillations!! nothing else!! :-(

however the previous schematics which i posted are showing me the received pulse!!



now this circuit is giving me a good response.. i mean these pulses are good enough to start and stop the timer of TDC chip.. but the time delay between the start and the stop pulses is not constant!!

due to which i am unable to calibrate the system.. furthermore this fluctuation is ruining the resolution of the system!! :-(

 

[keith1200rs]

I assume the laser pulse is the yellow trace and the received pulse is the green one? If so, the received pulse is negative going and the remainder of the pulse (which goes positive) is the overload recovery? Your laser pulse looks like it is over 100ns long. 100ns corresponds to 15m return path. So, any received pulse within 15m is likely to be disturbed by the laser firing pulse unless everything is well isolated and screened. I assume the laser current is around 10A so it takes a bit of care to ensure that current doesn't affect the receiver otherwise you need to have a minimum range of 15m or so.

Transimpedance amplifiers will usually oscillate unless you get the feedback capacitance right.

Keith

- - - Updated - - -

What does the signal on the receiver look like if you block the light to the photodiode when you fire the laser?

Keith

 

[FvM]

I would expect some explanation from which circuit nodes the waveforms are acquired.

Apart from this point, we see a 100 ns pulse (probably the laser current, as Keith guessed) producing an almost rectangular pulse of nearly 500 ns length, so apparently saturation takes place. In this case, jitter won't be surprizing, because any small intensity variation translates into a time shift. To get ns or even sub-ns resolution, the receiver channel must be linear and either implement automatic gain control or a constant fraction trigger.

 

[danishkaleem]

well yes the yellow pulse is the transmitted pulse (acquired from the laser driver output just before the laser diode) and the green pulse is the received pulse (acquired after the second stage amplification) ...

and i require the rising edges for the operation of start stops timers of the ACAM TDC GP2 chip..

i have tried to do enough isolation.. made the PCBs.. i have been using different power supplies for the transmitter, receiver and range computer with all the grounds common...

and yes the laser driver i am using gives a maximum output current of 4A... but using separate power supplies would help it up i guess..

FvM i want sub ns-resolution.. can i do somthing for it??


- - - Updated - - -



on the left side is receiver, on right is transmitter... at the back is the range computer....

- - - Updated - - -



i have put up a covering on the photodiode on the left side to avoid the effect of ambient light....

- - - Updated - - -



this is another output with yellow transmitted pulse and green received pulse...

- - - Updated - - -

keith once i block the receiver the received pulse distorts and no more remains a rectangular pulse but it becomes a small peak...

 

[danishkaleem]

keith do u mean that the rectangular part of the received pulse is just the overload recovery?? i am using the rising edge of the rectangular part to stop the timer.. i am not considering the negative going pulse at all!! :shock:

 

[keith1200rs]

Yes, I believe the rectangular pulse is overload recovery and it is the initial negative going pulse that is the "real" pulse, although you need to confirm that isn't simply pickup from the current driving the laser.

Your construction method isn't great for something dealing with high current pulses and very small signals where you want one to not interfere with the other. I would expect everything to be screened; PCBs to have good ground planes and you need some optics! That is after you fix the problems of, unsuitable photodiode (too large and slow) and unsuitable transimpedance amplifier (high current noise and input bias current). The leads to the laser need to be as short as possible. I don't know what the bundle of black wire is for (testing?) but you need to be very careful when probing the laser drivers and the oscilloscope grounding when doing so - it will usually introduce some crosstalk to the received side.

You need to read some of the Osram application notes if you haven't already done so - I have attached some.

Your layout is very spread out for something dealing with high currents and high speeds. That will introduce considerable inductance which will severely limit the performance. You need some optics. If you calculate the sort of range you can get with no optics it will be very short. With good optics you may get a decent single pulse from a few tens of metres away (with some better electronics).

Construction needs to be a lot more rigorous. This is something that works and has a screened receiver and decent, split ground planes:



While it also has a screened laser driver, that is more for EMC purposes as the receiver screening prevents crosstalk.

Keith.

 

[FvM]

I presume that the "negative going pulse" is caused by the unsuitable first amplifier stage with "floating" photodiode and zero bias.

Time resolution will be effectively set by the ratio of receiver speed to relative noise level. Both sides apparently need improvement in the present design.

As discussed in the profound Osram application note, an amplitude independent trigger (or "constant fraction" trigger) is required.

 

[danishkaleem]

As discussed in the profound Osram application note, an amplitude independent trigger (or "constant fraction" trigger) is required.

can u please help me with this thing?? i am not sure about it..

- - - Updated - - -

Keith..

that bundle of black wire is to produce delay..

because the TDC chip has two stop channels and it needs 3.5ns b/w the two stop channels... so i am using the same stop pulse for two stop channels and this bundle of wire is giving me a delay of 9 ns..

danish

- - - Updated - - -

can u suggest me any other op amp for transimpedance amplifier with low current noise..

 

[keith1200rs]

that bundle of black wire is to produce delay..

It is a transmitter that will interfere with the receiver (as well as causing unwanted EM radiation).

If you need a delay there are decent ways of doing it - a piece of wire isn't a delay line. The best way would be fibre optic cable. Alternatively you can put a digital delay in your transmit pulse generation circuit. Alternatively, a length of coax, correctly terminated, is a good delay line. Whichever method you use it should be where the signal is small and current is low - not in series with the laser.

It would be worthwhile checking your laser pulse width with a fast PIN photodiode and 50 ohm resistor, reverse biased.

For a decent opamp for the transimpedance amplifier, have a look at the OPA355.

Keith

 

[danishkaleem]

hello keith!

for your concern...

can u tell me what thing will determine the minimum range of the laser range finder? u once mentioned something about it.

actually the TOF keeps on increasing after 50cm.. but before that it gets reversed... can u explain this phenomenon.?

 

[keith1200rs]

Minimum range depends on how well the laser driving circuitry is isolated from the receiver and the length of the pulse. It also depends on the optics. Separate transmit/receive optics targeted at a distance will not necessarily see and light at short distances. It can be difficult to prevent pulses from the laser driver (up to maybe 75A) from affecting the receiver. If so, your minimum range will be determined by the length of that disturbance - typically the laser pulse width.

Until you sort out your transmitter/receiver circuit I don't think you can take much notice of the values out of the TDC. They probably depend as much on amplitude as time. Also, you have to be careful in the design of any comparator for creating the trigger pulse to ensure that you don't introduce an amplitude related shift.

Keith.