New head amp design

[Lawnboy]

i was interested in building the new head amp circuit in the latest newsletter, since i have not had much sucess with the design in the handbook. can i use generic transistors in place of the ones specified? i have a BC109C, but the others i do not.

[Lawnboy]

oh also, is the dome photocell really a giant photodiode as the newsletter says? would this amp work for solar cells as well?

[Steve Anderson]

Apart from the BC109C the other transistors can be generic types. The BC212 could be replaced by a BC556/7/8 in any grade, i.e. any suffix A, B or C (or no suffix at all), or a 2N4402/3, almost any silicon PNP device will (should) work. As you're US based also the 2N3904/6.

The 2N3704 could be another BC109 or a BC546/7/8/9, again with any or no suffix, A, B or C, or a 2N2222.

The BC109C could be replaced with a BC547C, BC548C or the super-beta MPSA18 with a typical noise figure of 0.5db which is better than the '109C. If you really wanted to get the noise-floor down paralleled devices like the LM394 or SSM2210 are even better but might require some component values to be changed.

Take heed of the sentence, "Owing to the high gain (around 85db) good RF techniques should be used in the construction of the amplifier." There is no signal-frequency feedback although the DC operating point is to an extent stabilized. Although not shown in the diagram good RF-style decoupling should be used, 10-100nF disc ceramic capacitors from the supply to ground/0V. A ground/earth plane is probably a good investment, as is placing the whole thing in a metal enclosure, especially if you're near any transmitters or other source of RF as mentioned in the article.

Assuming that it does have a gain of 85db and the output is 1V p/p the input voltage is only some 50uV (microvolts). Use metal-film resistors to keep the noise down too, anything with Carbon in its description is going to disappoint. The 22k preset (RV1) should be Cermet, not Carbon for the same reason. I'm tempted to suggest that C2 (10uF) is Tantalum rather than an Aluminium electrolytic.

Also note that it will take several seconds to 'wake up' when switched on each time, this is primarily due to the time-constant of R8 (470k), R2 (150k) and C4 (100uF).

Steve A.

Afterthought... I'm not sure what the dome photocell actually is (anyone know?) but I suspect that a solar cell will have too much capacitance to provide a suitable response. However I have heard that some have been successful in using them, I can't recall who or where.

[AncientBrit]

From head amps I have built in the past I would second Steve's advice about use of ground planes.

In fact I use totally screened boxes made up from small pieces of double sided PCB.

It's fairly easy to run a fillet of solder down the adjoining faces and you can use either tagboard or stripboard for the circuitry within.
Or even free standing components soldered to a few insulated standoff pillars.

I found that a lid helps cuts down any hum pickup.

The dome photocell can poke through a grommeted hole in the side of the box making the unit totally self contained.

Cheers,

Graham

[Steve Anderson]

Or even free standing components soldered to a few insulated standoff pillars. Graham

Often called 'Ugly' (very apt), 'Birdsnest' or 'Bug' construction (it looks like a load of dead bugs).

Example below...often the only way to keep a home-made circuit stable and under control. Not very professional, true, but it's cheap and it works.

Notice the use of small offcuts of PCB as RF bypass capacitors.

More can be found here...a useful site generally as is the forum...especially if you're a ham...

http://www.sm0vpo.com/

Steve A.

[Steve Anderson]

....is the dome photocell really a giant photodiode as the newsletter says? would this amp work for solar cells as well?

I asked Jeremy Jago about the 'dome sensors', as he mentions they aren't quite 'giant'.

They are miniature silicon solar-type cells, like BPW81/A, SFH 2030 etc, ie square-chip, monocrystalline dark blue, but with the considerably larger sensitive area of about 5 mm square. They are in a clear (slightly yellowish) round dome-topped resin encapsulation approx 8mm diameter, like an outsize LED. The base support from which the leadouts emerge is white opal material, perhaps ceramic. Within the package, both gold wire connections are welded to the same upper side of the chip, close together.

From that they sound very similar to an SLD-70 (datasheet attached), but I'm not certain.

As for solar cells I did some rummaging around here and found the following amplifier by Jim Wood, no idea where I got this from, could well have been on this forum.

As I suspected the amplifier needs a rising frequency characteristic to offset the solar cells capacitance. With the components Jim suggests that starts as low as 300Hz! One can only assume that it did work so the idea is plausible. The values of the compensation components will need to be altered to suit the individual solar cell. The value of the resistor between the two 220uF capacitors is unknown, but 1k would be a good starting point.

Steve A.

[AncientBrit]

Some measurements carried out quickly on Club Dome photocells :

3v miniature torch shing into the photocell,
distance 0 through to 100mm,
output substantially constant at 0.47 volt DC when light on axis
Output is polarity conscious

Cheers,

Graham

[gary]

Here is a picture of the blighter.



And here is a somewhat embarrassing video of the output of a 32 line mechanical camera that I lashed up in order to test my automatic timebase correction software. The lash up used (I think) the handbook camera monitor head amp, but had no shielding whatsoever (not even light shielding) so the results were ordinary, and as you can see there was quite a bit of noise from the cassette motor used to turn the disk. The optics used a club bullseye lens as the condenser lens and an old camera lens as the object lens. There was just a voltage regulator speed adjustment for the motor so the speed varied considerably (intentional). This set up required quite a bit of light and I used an AC halogen flood light. At least I think you will agree that the TBC was reasonably successful.

Note that Graham has had much more success using this sensor and has some samples on this forum I think, if not, perhaps he could be arm twisted to share some for our enjoyment.

PS is there any way of setting the order of attachments so they don't all end up at the bottom?

[dominicbeesley]

Wow that correction looks to work very well - it must have been difficult to get it to not get confused by all that motor noise (or did you have clean syncs?)

Dom

[gary]

Well I wouldn't say clean, but they were relatively strong such that there was a strong component around 400 Hz in the spectrum which could be tracked, although I have found the method works pretty well even if there isn't any sync at all as long as the picture content contains strong frame periodic edges, in which case there is still a relatively strong component around the line frequency.

BTW that video was also post processed to remove a fairly strong 100 hz strobing due to using AC lighting.

Also note that whilst the picture is stable there is very strong shearing happening. This was due to too large a window when transforming, there is quite a lot of compromise required there between accurate frequency tracking and accurate picture reconstruction. Like many of my projects, it's something I need to get back to at some stage...

[Lawnboy]

Steve- thanks for all the info on the dome cell and preamp. i had read in some of the early NBTV newsletters that people had some sucess with solar cells. i will look through the archive CD again to see if there is a "newer" or "imporved" version before i start building it. the cells are going to be the heart of my FS scanner.
gary- i am glad that the dome sensor works in a NBTV camera. i was wondering if it had a high enough frequency response for 32 lines. you used the head amp circuit from the handbook? i guess that means the layout diagram i made is no good. (or ive developed a nasty habit of hooking up transistors the wrong way )

[gary]

i will look through the archive CD again to see if there is a "newer" or "imporved" version before i start building it.

If you have the Newsletter archive CD, I would recommend looking up Graham's circuits using this device V24 #1 & #3, V25 #4. He seems to have achieved results I would be very happy with.

i was wondering if it had a high enough frequency response for 32 lines.

Yes the freq response of the dome cell seems to be more than adequate

you used the head amp circuit from the handbook? i guess that means the layout diagram i made is no good. (or ive developed a nasty habit of hooking up transistors the wrong way )

Yes it certainly works, whether it is ideal is another issue. I did nothing to analyse it or improve it because I was just after a quick and dirty "typical" NBTV camera signal to test my software with, so don't take the quality of my video as an indication of the quality of signal that can be achieved using either the dome cell or the head amp circuit - I am sure it can be very significantly improved - but it does show that it does at least work to some degree.

If you are after a better result even than that then you should look at using large sensor area PIN diodes or a Photomultiplier. From my point of view the disadvantage of these devices is cost and availability, but since you live in the US this may not be the case for you.

[Lawnboy]

i also found this collection of preamp designs i was going to eventually try. they all came from old newsletters for use with full-size solar cells and seem somewhat simmilar to each other. to bad i didnt mark which issue they came from.

[Lawnboy]

Apart from the BC109C the other transistors can be generic types. The BC212 could be replaced by a BC556/7/8 in any grade, i.e. any suffix A, B or C (or no suffix at all), or a 2N4402/3, almost any silicon PNP device will (should) work. As you're US based also the 2N3904/6.

Steve, isnt a 2N3904/6 an NPN transistor?

[Steve Anderson]

Steve, isnt a 2N3904/6 an NPN transistor?

Correct, the 2N3905/6 are PNP, the 2N3903/4 are NPN. I goofed by lumping both polarities together.

In both cases the higher-numbered parts have the higher gain, apart from that no difference.

Steve A.