The Devil in control Flying spot PMT head amp build

[Harry Dalek]

First job today looked into The light transmitter at the base of the BC549 the NBTV reversing video looks distorted (as shown in the first scope picture )it was pretty much the same at the emitter ,I found if i increased the 1uF input cap to any thing higher the video at these test points were pretty much fine again as shown at top of the animated gif wave form trace so i changed that 1uf to a 10uf .
So we know the light transmitter is pretty much be ok now .

Now repeating the PMT test to see what shows up that result is in the animated gif both reversing waveforms show up better the sync pulse not so much and waveform still a bit rounded for my liking .

I think i have to look into the head amp i need to take a reading if possible at the anode and after the first preamp ic .... the distortion must be happening around here .

The PMT is running at around 570 volts above this voltage the distortion shows up in the waveform amplified .....lower voltage not good either seems good around five hundreds ....

I also tried a 1meg pot on the anode to ground no real difference to the waveform on the higher setting area of the pot but i found with this even on the 1meg setting the preamp control was not as good adjusting the level down so i think i will remove this idea ...

I was not expecting so much trouble getting a nice copy of the waveform but it must be possible i will see what results the anode and first preamp ic give i have to take out a panel switch and add another socket to plug into so can out put the signals and keep the case light proof...have some time today give it a go now .

[Harry Dalek]

ARRR this is better now i know where the problem is .

I ended up replacing 2 switches on the panel with 2 sockets which i soldered the PMT anode and second plug pin 6 of the first preamp ic its output pin so i could access the signals keep the case shut and light proof .

Its better to view the video than the animated gif and see what i can see on the scope but even so its a big improvement not using the preamp signal at all ...i forgot to look at what i was viewing both signals on the scope the difference in levels but i think one ic is enough ..i was getting distorted lines on the PMT camera i think i know why now !

So the 3321 file gif and mp4 is top trace the input to the light transmitter bottom direct connection to the PMT Anode .

file 3322 is top trace input video bottom out of pin 6 ic1 preamp
file 3323 top anode bottom pin 6 preamp

Got a bit of noise in the signals on the last scope videos may be its the length of the unshielded wires ..

Worth having a look if interested ...Now i need to look again at the head amp as it looks to me i just need one half of it ,its running on a dual 12 volt supply perhaps 5 volts is all it needs .
Any case the flying spot camera is more doable than it was yesterday

[Steve Anderson]

Harry, are you using the circuit as below? Bearing in mind this was designed for audio, the LF response is lacking, hence the curve on the staircase wave form. Increase the input 1u cap to 100u. (Note the capacitor polarity is the wrong way around). This circuit for the use you are putting it to could be vastly improved without making it more complex. It's a typical internet bodge by someone who has no idea of what they're doing. If the bias on the transistor were properly thought out there would be no need for the trimmer/preset pot.

The input 1u combined with the 1k resistor is -3db down at 160Hz, ideally NBTV should go to 2Hz at -3db, but this is acceptable for telephone quality speech, not NBTV.

A single op-amp arrangement though would be vastly superior.

Steve A.

[Harry Dalek]

Harry, are you using the circuit as below? Bearing in mind this was designed for audio, the LF response is lacking, hence the curve on the staircase wave form. Increase the input 1u cap to 100u. (Note the capacitor polarity is the wrong way around). This circuit for the use you are putting it to could be vastly improved without making it more complex. It's a typical internet bodge by someone who has no idea of what they're doing.

The input 1u combined with the 1k resistor is -3db down at 160Hz, ideally NBTV should go to 2Hz at -3db, but this is acceptable for telephone quality specch, not NBTV.

Steve A.

Yes a few changes Steve the 1uf is now a 10uf for the input cap ....funny enough the first capacitor i tried today was a 100 uf i found any thing above 1uf seemed to improve the results in the circuit i was not sure what the limits were as it still worked up to 1000 uf ...i will increase the capacitor to 100 uf and see if it helps at the PMT end ...wrong way around capacitor i didn't pick that i did try the capacitor both ways around and seemed to work either way but i rather believe you !
I wanted some thing simple to test here for the PMT i am in away glad the curve in the stair case is the transmitter circuit as the reading was taken at the anode so it would other wise be a PMT circuit problem not some thing i want to tackle .
I did change the circuit a touch Steve i know trimmers every where are not the best but i have a 22k after the 100 ohm on the emitter ...to ground and the 1k positive and 10k to base are changed to 100 ohm to positive and 10k to base of the BC548 (resistance across both resistors here in the circuit are running around 500 ohm ) and i have a lm317 adusting the 5 volt supply voltage to it .
Any improvement to the circuit for band width would be good ....i wish was was good at designing circuits i know its more cross your fingers when you pick some thing to make off the internet ...I recall even Steve o had problems with hes Argus monitor build as the magazine circuit had faults and as i found with the SSTV 73 magazine one i first tried .

[Steve Anderson]

Have a look at the waveform at the emitter of the BCxxx, it should be a very close copy of the input waveform without the curve. The same at the collector, though it will be inverted and perhaps a little curved due to the LED, ignore it. If they look good then for the time being this 'test transmitter' is OK. It's the emitter waveform that's important.

Steve A.

[Harry Dalek]

Have a look at the waveform at the emitter of the BCxxx, it should be a very close copy of the input waveform without the curve. The same at the collector, though it will be inverted. If they look good then for the time being this 'test transmitter' is OK.

Steve A.

Yes Steve Klass mentioned to do this last night and that was my first job today a few posts back 3 or 4 scope reading picture first showed it distorted posted there,that made me change the 1 uf cap and the same scope reading then showed the input video at the base and emitter of the BC548 and looked fine with the capacitor change .
The PMT's sure out put a signal may be a fine adjustment of the PMT supply voltage might help the curve looks btw better in the video than the gif which also drops frames .

[Steve Anderson]

Harry, are you using the PMT pre-amp I did for Steve O's colour camera or something else? My version was DC-coupled with no capacitors in the signal path, a response down to DC.

Another thought, how are you generating the -HV for the PMT?

Steve A.

[Harry Dalek]

Harry, are you using the PMT pre-amp I did for Steve O's colour camera or something else? My version was DC-coupled with no capacitors in the signal path, a response down to DC.

Another thought, how are you generating the -HV for the PMT?

Steve A.

The PMT 50 hz voltage multiplier i made for the PMT camera a while back ...deadly but works well ...pictures below but the circuit 10 uf caps HV caps and negative version of this positive circuit ..negative 300 to negative 1600 volts range about 200 volts or touch over per range change ...running this PMT at 570 volts picture shows 580 up and down with mains supply a touch ...a clean supply but dangerous .
BTW also supplies the regulated dual 12 volt supply .

Yes but i had no TL071's using NE5534 yes no caps 10k coupled right to the Anode... my ic change more than likely might be the problem ? i just used it as it worked on the PMT camera when i replaced the TL071's to try ..so not your fault mine just using what i had handy it seems fine first ic output so either the second ic is bad circuit mistake perhaps here i have to check .

[Steve Anderson]

I very much doubt the change to NE5534's from TL071's would make that much difference, it's the bias current in the TL071s is much lower as they are FET op-amps as opposed to bipolar. That is unlikely to be the cause of the non-linearity (the curve), but you never know.

The PMT supply looks OK too.

This is a bit of an odd one. Either the transmitter circuit is still misbehaving or there's a defect in the PMT (let's hope not).

What colour LED are you using?

Steve A.

[Harry Dalek]

I very much doubt the change to NE5534's from TL071's would make that much difference, it's the bias current in the TL071s is much lower as they are FET op-amps as opposed to bipolar. That is unlikely to be the cause of the non-linearity (the curve), but you never know.

The PMT supply looks OK too.

This is a bit of an odd one. Either the transmitter circuit is still misbehaving or there's a defect in the PMT (let's hope not).

What colour LED are you using?

Steve A.

Hi Steve
I sure hope its ok it is a second hand Tube never know what life it had before me ...least i will give it a life again here .
Yes the NE5534 has worked for me before but i was not sure how naughty i was replacing the TL071's with it .

I can see the second ic is the problem when i have time i with replace it with another and another look in case i made a circuit error .

Think the effect at the moment if i feed the video from the pmt to a monitor would be a sight bending of the grey bars .

I have a 3 colour Led for testing but i have really only looked into the Green Steve as thats the colour of the scope trace i want to use for a flying spot camera ,the last PMT testing all the colours Red was best as far as waveform i saw out of the PMT green second and Blue was the worst ...
I was thinking before i tested the signal out at the anode it would be very weak but its not the case.

[Steve Anderson]

PMTs are very sensitive to blue, far much less for green and as good as totally deaf to red. Change to the blue LED. I have a suspicion you're overloading/saturating the PMT with too much light. The output of the first op-amp should only be 100mV or so, if higher you're running the risk of damaging the tube.

If you can see the blue light from the LED in a darkened room (and I mean totally dark) and your eyes have become dark-adapted - that's far too much light.

These things can detect a single photon depending on the design. We can't.

Steve A.

I'm also wondering about the 'transmitter circuit'. I assume the author used a red or IR LED which have a far lower forward voltage than a blue one. This requires thinking about.

[Harry Dalek]

PMTs are very sensitive to blue, far much less for green and as good as totally deaf to red. Change to the blue LED. I have a suspicion you're overloading/saturating the PMT with too much light. The output of the first op-amp should only be 100mV or so, if higher you're running the risk of damaging the tube.

Hi Steve

With the other PMT red was giving the best results and blue was the worst which at the time did surprise me i was expecting the reverse since blue is closer to the uv side than red i can repeat the experiment on this pmt.
I think PMT's depending which some must be sensitive to red to some degree as Steve O would not of got hes 3 colour PMT camera to work .
i will check more if my case is 100% light proof before saying it is ... as i know it does not take much to let in a stream of photons but i have not noticed a problem since i last looked at it .
I am also running the tube well below normal operating range 950 volts I get over loading effect if its over 700 high 500,s seems right at the moment .
Pretty much i can't see the led light up at all operating the transmitting circuit.
I have the 2 Leds in parallel one in the case and one on the circuit board so i have an indicator of whats happening inside the case ,i will check what i am running the circuit on via the regulator .
Perhaps i should of run it on 5 volts and just have a higher resistance on the transistors emitter .

If you can see the blue light from the LED in a darkened room (and I mean totally dark) and your eyes have become dark-adapted - that's far too much light.

When i can i will do a night test but i know i can at the moment run the led at very low levels invisible to my eyes ...only tried Green this time around so far i am interested in the other colours but it has to detect green .

These things can detect a single photon depending on the design. We can't.

When i got a scratch on my painted black Nipkow case last camera i could tell right away as the image started to become very snowy staticy looking washed out due to the extra light ...i was looking them up and it does say extremely sensitive detectors of light in the ultraviolet, visible, and near-infrared ranges of the electromagnetic spectrum so red is in there ....an infrared led makes me want to test this !

Steve A.

I'm also wondering about the 'transmitter circuit'. I assume the author used a red or IR LED which have a far lower forward voltage than a blue one. This requires thinking about.

The link to the circuit below Steve he used a red led yes i have to readjust the circuit to red blue or green led but playing with green had not had to readjust it for these PMT tests

http://www.burchills.info/sound_via_lig ... light.html

OH also i will feed the audio direct to the preamp and see what i get out of ic 1 and 2 and take the pmt factor out .

[Klaas Robers]

I fear that the sensitivity for blue was great and the PMT was saturated from the far to bright light. For red the sensitivity is low, so it was less saturated and you had the idea that it was working better. So see that you can lower the light output (emitter resistor large, more than 22k) and start with the lowest brightness. You just see it giving light, if the room is darkened.

[Harry Dalek]

I fear that the sensitivity for blue was great and the PMT was saturated from the far to bright light. For red the sensitivity is low, so it was less saturated and you had the idea that it was working better. So see that you can lower the light output (emitter resistor large, more than 22k) and start with the lowest brightness. You just see it giving light, if the room is darkened.

Hi Klass
I will repeat the experiment bypassing the preamp with just a Anode reading from the PMT that way the waveform can't be effected by any thing here
It could be the case thinking back my testing was flawed with the first test i was expecting Blue to give the better result at that time ,i didn't have as much control of the 3 colour led as i do now , all those factors into account now on hindsight i will see as you and Steve suspect the result should be the other way around .
I will see about changing the 22k trimmer to a higher resistance .

[Harry Dalek]

Steve and Kass Test results........ i had time this afternoon to redo the Led transmitting tests did all 3 colours i matched my on board 3 colour led to the led in the testing case..... same colour leds in parallel.
Increased the BC548 emitter resistance from 22k to 75k trimmer .
Now you both were 100% right Blue is the most sensitive and red the least ,
Klass i pretty much know where i went wrong the first time i just swapped over colour leds without adjusting them light levels all different overloaded it on blue as well as it being a poor circuit for the job.
What i did today was reconnect the transmitting circuit to the 12 volt supply used the on board LM317 regulator dropped that to 5 volts and adjusted the BC548 emitter and base resistor trimmers for each led change each time .. i had control over the low light level from zero to the eye to as around in the light level in the picture of the leds viewable .
Now this is where it gets interesting .
I was running that light transmitter before today just under 3 volts being cautious.
Today i stated off on 5 volts but its a touch under that just for better control of the leds light level range with the Base and emitter trimmers .
I really took on board Steve and Klass your warnings on the Led light levels and on starting my testing this afternoon i made sure both leds on the circuit board and in the light proof case were both indicating the same light levels and i started off with no light visible to the eye in the dark i hooked every up for the first test .
I got the PMT anode test of the reversing wav on the scope but it was rounded on the staircase of the signal and bottom half missing i was thinking perhaps the level is to low and i started to increase the led level with the base and emitter controls on the transistor and the waveform started to become better not 100% a copy of the input but pretty good ...problem if its a problem the leds are at that light level in the photos !
When i tried Blue the signal was overloading badly waveform very large and distorted ..i stated all leds off no light visible ..but this time i had to drop the PMT voltage to 300 volts the other 2 colours were fine at 580 volts but i still had to increase the light level as before to the led being viewable for a correct waveform at 300 volts .
Scope settings same as in photo all Led colours .
I am running the PMT well below its normal operating voltage i can only think this is the reason this is working as it does ...
The green Led test waveform worked very well ..







youtu.be/9Z0MxPcJrek

youtu.be/5TE5PJ6QA9o

youtu.be/VzzmKRlq-ZU