Page 1 of 2

Need electronic expert advice for a project

Posted: Tue Jun 05, 2018 5:53 pm
by FrancoisLanciault
Hi,

I want to recreate the electronic inside a sophisticated light pen for a vintage computer. Nothing related to the QL hence the "Off-Topic" section.

I have the original blue print of the electrical schematic. However, there are two or three things that can be interpreted one way or the other and a section that make no sens (from what my limited knowledge of electronics tells me). The circuit is made of a few transistors, diodes and passive components. No IC.

I have already ask around here at my job and got conflicting answers. And nobody took the time to look and understand what the circuit was actually doing. I know that there are specialized boards dedicated to electronic but I though I would ask here first as I know some of you are real experts!.

Any taker ?

I will post the schematic and my questions in a followup email if any of you thinks he may help.

Thank you.
Francois Lanciault

Re: Need electronic expert advice for a project

Posted: Tue Jun 05, 2018 6:04 pm
by Pr0f
pop up the circuit, I am fairly sure there are folk on here who can decipher the operation of your light pen...

Re: Need electronic expert advice for a project

Posted: Tue Jun 05, 2018 7:57 pm
by Dave
I'm game, too.

Re: Need electronic expert advice for a project

Posted: Wed Jun 06, 2018 7:41 am
by Pr0f
One quick question - as this could be a showstopper for your project...

What monitor or TV do you intend to use?

Light pen will only work correctly with a CRT based TV or Monitor, as the LCD doesn't have a 'light spot'.

Re: Need electronic expert advice for a project

Posted: Wed Jun 06, 2018 6:06 pm
by FrancoisLanciault
Pr0f wrote:One quick question - as this could be a showstopper for your project...

What monitor or TV do you intend to use?

Light pen will only work correctly with a CRT based TV or Monitor, as the LCD doesn't have a 'light spot'.
Well, not only it needs a CRT monitor, but it needs the specific high resolution monitor that was made for this computer. Moreover, it needs the light pen controller board option installed in the monitor. I have both obviously :-) What I don't have is the light pen itself and the electronic inside the light pen body. The good news is that the light pen electronic is quite simple. Not like the light pen controller board with it's 4 layer PCB and 50+ chips.

For those interested, the light pen is for the HP-9845, a very powerful workstation first released in 1978. With dual 16bit processors, integrated 80 col printer and up to 1.5Meg of memory, it had 80286 computing power years before the Intel equivalent. The light pen however only works with the enhanced version released in 1980. The monitors were not dummy monitor but were taking care of vector and other graphic primitive generation, alpha character generation with flashing and underscore, light pen handling etc. In fact, when HP released the color version of the monitor in 1980, the monitor had more computing power then the computer itself.
9845C.jpg
But back to the light pen schematic. It should be attached to this post.
lpscematic.jpg
In this schematic, as in all schematics of that era I guess, connections are symbolized by big fat dots. When a wire pass over another without being connected, the line kind of "bridges over" the other as can be seen at the right of diode CR3. But in this schematic, there are 3 locations where lines just go over each other, without fat dot that would signify contact nor the little speed bump that would imply no contact.

The first one is easy I guess. I you look at the transistor circled in blue, do I need to understand that both the anode of CR1 (the photocell) and the common point of C1-R1 are attached to the base of the transistor ??

For the area circled in green, is there a contact where the lines cross or not ? As there is no fat dot, I would assume no contact but there is no bump either. I am sure that for the electronic experts, the circuit makes sense for just one of these options. But asking around I got answers like : "The wires are not connected otherwise it does not make any sens" as well as "The wire are obviously connected, otherwise it does not make any sens"...

Finally, for the transistor circle in red: the HP number is 4-392 and an equivalent is not listed in the lower left corner of the drawing. Using the HP cross reference files, I was able to determined the type as being a NPN #2N5088. However, the arrow is missing so I do not know where the collector is. I dug up the original blue print which has a better quality and there is definitely no arrow. Again I hope some electronic expert would be able to correct that missing information by looking at the schematic logic. Finally, still in the red circle, there is also some wire cross over without dot. Some friend of mine told me there should be a dot there but I don't know if I can thrust him. Does the base is only attached to the NC end of the switch with the two resistance (R15 & R17) connecting the emitter and collector with a combined value of 11.96k or does the transistor base connect to both resistances and the switch.

Something that confusing things a lot in my mind, is that the circuit is powered by -12V. Also, I probed the "SW" socket hole on the monitor (without light pen plugged in obviously) and there is = -0.5V on that pin. The OUT and RETURN pins show only low noise values (0 +/- 0.015 Volts)

Going back at the switch logic part of the schematic, I have problem understanding what is going on here, whatever the 2N5088 transistor orientation might be. Everything ends up tied to ground, so I don't see the difference between the switch at the N.O. and at N.C. position. Also the diode orientation does not seems to match the fact that the controller put a minus 0.5 Volt on the SW pin. For information, the pen controller board is working as there are a couple of self diagnostics that the controller board can run on itself and they all come ok. It can even emulate the light pen, without any light pen attached.

Finally, I wrote a small program that monitored the light pen switch button. There is no "button HIT" if nothing is connected to the light pen socket. But as soon as I connect the SW pin to the Ground PIN (a bit like the switch in the N.O. position, but without diode) the program report that the button has been pressed.

Anyway, any information or tip is welcome. I don't want to put together the circuit wrong. Not because I am afraid of damaging some of the light pen component, but because I really don't want to blow up the controller card.

Thank you all for your patience.

François

Re: Need electronic expert advice for a project

Posted: Wed Jun 06, 2018 6:28 pm
by 1024MAK
Pr0f wrote:One quick question - as this could be a showstopper for your project...

What monitor or TV do you intend to use?

Light pen will only work correctly with a CRT based TV or Monitor, as the LCD doesn't have a 'light spot'.
That's not entirely correct. Some light guns just sensed the brightness of the screen and did not attempt to detect the raster beam (bright moving spot of the electron gun(s)). They were only useful with games specially written for them and only worked in a dark or dim room. Oh, and they were a bit rubbish.

Mark

Re: Need electronic expert advice for a project

Posted: Wed Jun 06, 2018 7:09 pm
by 1024MAK
FrancoisLanciault wrote:I you look at the transistor circled in blue, do I need to understand that both the anode of CR1 (the photocell) and the common point of C1-R1 are attached to the base of the transistor ??
Yes, I think these should be joined. CR1 is however a photo-diode.
FrancoisLanciault wrote:For the area circled in green, is there a contact where the lines cross or not ? As there is no fat dot, I would assume no contact but there is no bump either. I am sure that for the electronic experts, the circuit makes sense for just one of these options. But asking around I got answers like : "The wires are not connected otherwise it does not make any sens" as well as "The wire are obviously connected, otherwise it does not make any sens"...
The circuit will not work unless the line between R29, R11, Q6 collector and C6 has a (indirect) connection to the -12V supply, so yes, I think there should be a connection where the lines cross in the green circle.
FrancoisLanciault wrote:Finally, for the transistor circle in red: the HP number is 4-392 and an equivalent is not listed in the lower left corner of the drawing. Using the HP cross reference files, I was able to determined the type as being a NPN #2N5088. However, the arrow is missing so I do not know where the collector is. I dug up the original blue print which has a better quality and there is definitely no arrow. Again I hope some electronic expert would be able to correct that missing information by looking at the schematic logic. Finally, still in the red circle, there is also some wire cross over without dot. Some friend of mine told me there should be a dot there but I don't know if I can thrust him. Does the base is only attached to the NC end of the switch with the two resistance (R15 & R17) connecting the emitter and collector with a combined value of 11.96k or does the transistor base connect to both resistances and the switch.
I agree that the Junction Q7, R15, R17 and switch NC contact makes more sense if joined.
FrancoisLanciault wrote:Something that confusing things a lot in my mind, is that the circuit is powered by -12V.
Using negative supplies was more common in the 1960's and 1970's. At my workplace we used to have a communication system that ran on a -28V DC supply.
FrancoisLanciault wrote:Also, I probed the "SW" socket hole on the monitor (without light pen plugged in obviously) and there is = -0.5V on that pin. The OUT and RETURN pins show only low noise values (0 +/- 0.015 Volts)
Was this with respect to GND? What do you get between SW and -12V?
FrancoisLanciault wrote:Going back at the switch logic part of the schematic, I have problem understanding what is going on here, whatever the 2N5088 transistor orientation might be. Everything ends up tied to ground, so I don't see the difference between the switch at the N.O. and at N.C. position. Also the diode orientation does not seems to match the fact that the controller put a minus 0.5 Volt on the SW pin. For information, the pen controller board is working as there are a couple of self diagnostics that the controller board can run on itself and they all come ok. It can even emulate the light pen, without any light pen attached.
No, I currently don't understand this part of the circuit. Normally I would expect just a resistor and a switch. Plus maybe a capacitor to help with switch contact bounce.
FrancoisLanciault wrote:Finally, I wrote a small program that monitored the light pen switch button. There is no "button HIT" if nothing is connected to the light pen socket. But as soon as I connect the SW pin to the Ground PIN (a bit like the switch in the N.O. position, but without diode) the program report that the button has been pressed.
Sounds like you just need a switch. But check how much current flows in the circuit through the switch.

Mark

Re: Need electronic expert advice for a project

Posted: Wed Jun 06, 2018 11:46 pm
by Pr0f
C1, R1 and CR1 are all connected to the base of the first transistor, C1 and R1 provide frequency sensitive feedback for the 2 transistor amplifier stage.

In the Green circled area, it's missing a blob to join these connections.

Regarding the last area of the circuit, logic tells me the emitter should be on the ground line, but the purpose of the circuit is a little unclear, it may help to know what is on the light pen driver board. I suspect that the switching diode would be biased towards positive, and that in the NC position, the transistor is biased on dropping a voltage across it which raises the SW terminal with respect to ground, in the NO position the diode is shorted to ground, dropping the SW terminal to about 0.6V (maybe less if a shottky diode). That all assumes a positive voltage through a resistor is present on the SW terminal

Re: Need electronic expert advice for a project

Posted: Wed Jun 06, 2018 11:53 pm
by Pr0f
Actually - rethinking the switched transistor - if the switch is in the NC position, the transistor will be biased off, so the voltage at the SW pin is free to rise (depending on the circuitry on the board in the monitor), but as the switch is pressed, the transistor will be briefly biased on, before the switch shorts it out. This may be some sort of soft wave shaping for the light pen switch function.

Re: Need electronic expert advice for a project

Posted: Thu Jun 07, 2018 6:20 pm
by FrancoisLanciault
Thank you 1024MAK and Pr0f, your answers are definitely and better justified than anything I heard so far!

The only remaining mystery is the fact that the switch circuit make sense only if the SW pin has a positive supply even though I have measured a negative voltage.

Here is some description of the controller board I received from an expert in the HP 9845 computer. He is good at logic IC schematic interpretation but could not help me with the light pen board as he, from its own admission, has problem understanding transistor base circuit. Let me know if it match what you think is going on.

Info about the light pen controller from Ansgar:
Let's start with the switch. The switch actually consists of the switch itself, a debouncing cirscuit built up with a LM339 differential comparator, and a latch for capturing the switch state (JK flip flop):
sw.png
Next, have a look on the detect circuit:
There are two signals, OUT and RETURN. RETURN actually is coupled to the threshold. This again is a bit misleading, because the names suggest it exactly the other way around. The threshold is a binary choice you can make from the programmer's side. You can either select high threshold or low threshold. Which simply controls sensitivity of the circuit. The S158 (U45) is used to activate the detection circuit in relation to the interlaced cursor field (a trick which is used to use the cursor field to track the position, see the 98770 manual for a deeper explanation).
detect.png
To sum up the pinouts in works out like this (J24 is the 2x25 edge connector of the control board):

J24 DSub15 Name in schematic

-- 1 n/c
-- 2 n/c
20 3 GND
21 4 -12V
22 5 OUT
23 6 RETURN
24 7 SW
25 8 GND
-- 9 GND
-- 10 GND
-- 11 GND
-- 12 GND
-- 13 GND
-- 14 GND
-- 15 GND
François

P.S.
When I say I measure a negative voltage at the SW pin, it means I check with the black wire of my voltmeter attached to the GND pin #8 and the red wire attached to the SW pin #7.