Author Topic: Does this circuit even make sense?  (Read 3634 times)

0 Members and 1 Guest are viewing this topic.

Offline hermitengineerTopic starter

  • Regular Contributor
  • *
  • Posts: 80
  • Country: us
Does this circuit even make sense?
« on: October 26, 2019, 05:54:59 am »
Hello all,

I am attempting to repair an old game console, just for the challenge of it.  The first major challenge, though, was a complete lack of schematics.  So I had to reverse engineer the whole thing.

Anyway, there is a particular circuit in here that does not seem to make sense.  I've pulled the parts and inspected both sides of the board just to verify that I got it correct, and it appears so.  But then... is the circuit even useful?



From what I learned in trade school, the BJT only allows current flow in one direction, right?  So C249 can charge when it turns on, but I see no path for it to discharge.  Is there something I'm missing?

The only possibilities I can think of are:
  • I'm blind and can't see what should be an obvious trace
  • Some troglodyte replaced the transistor with a wrong part
What I do know is that the part number of the current BJT is clearly printed on it.
« Last Edit: October 26, 2019, 05:56:38 am by hermitengineer »
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Re: Does this circuit even make sense?
« Reply #1 on: October 26, 2019, 06:18:24 am »
one possibility is to adjustable capacitance between 4.7nF (Q207 turned off) and 4.7047uF (Q207 turned on)
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline oPossum

  • Super Contributor
  • ***
  • Posts: 1472
  • Country: us
  • Very dangerous - may attack at any time
Re: Does this circuit even make sense?
« Reply #2 on: October 26, 2019, 06:27:28 am »
Collector to emitter current flow is typically better in one direction than the other, but it does go both directions. Some transistors are optimized for bidirectional operation. An audio muting transistor is one example of that.
 
The following users thanked this post: coldfiremc

Offline hermitengineerTopic starter

  • Regular Contributor
  • *
  • Posts: 80
  • Country: us
Re: Does this circuit even make sense?
« Reply #3 on: October 26, 2019, 07:14:52 am »
one possibility is to adjustable capacitance between 4.7nF (Q207 turned off) and 4.7047uF (Q207 turned on)
I agree about the intent, but without a usable drain path, I can't see how the 4.7uF matters in this circuit.  I wonder if it's supposed to be a JFET maybe?

Collector to emitter current flow is typically better in one direction than the other, but it does go both directions. Some transistors are optimized for bidirectional operation. An audio muting transistor is one example of that.
For BJT's, I've always heard that reverse leakage will be in the nano-amp range though.  This circuit is part of the video output generation, and its input is the horizontal sync signal.  So nano-amps through a 4.7uF capacitor probably won't drain it by any noticeable amount.
 

Offline oPossum

  • Super Contributor
  • ***
  • Posts: 1472
  • Country: us
  • Very dangerous - may attack at any time
Re: Does this circuit even make sense?
« Reply #4 on: October 26, 2019, 07:17:59 am »
Reverse leakage would be for BE and BC, not CE.
 

Offline atmfjstc

  • Regular Contributor
  • *
  • Posts: 121
  • Country: ro
Re: Does this circuit even make sense?
« Reply #5 on: October 26, 2019, 04:32:27 pm »
I know it sounds silly, but since you're reverse engineering this off a physical board, maybe double-check that B, C, E have been correctly identified?

Also a bit more context for the components upstream/downstream of the circuit might help.
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Does this circuit even make sense?
« Reply #6 on: October 26, 2019, 04:53:43 pm »
There is a leakage between emitter and collector, it's just very small.

In any case I suspect the circuit shown would make more sense if we could see more context, ie the rest of the immediate circuitry.
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Offline iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: Does this circuit even make sense?
« Reply #7 on: October 26, 2019, 05:01:14 pm »
The 4.7nF/4.7uF caps could be part of a timing RC circuit, when the 3904 is on the RC will be xx usecs, when off xx/1000 usecs.
« Last Edit: October 26, 2019, 05:19:42 pm by imo »
Readers discretion is advised..
 
The following users thanked this post: rs20

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Does this circuit even make sense?
« Reply #8 on: October 26, 2019, 05:04:04 pm »
A little context would be very helpful here, what is the circuit connected to?
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20363
  • Country: gb
  • 0999
Re: Does this circuit even make sense?
« Reply #9 on: October 26, 2019, 09:36:01 pm »
I know it sounds silly, but since you're reverse engineering this off a physical board, maybe double-check that B, C, E have been correctly identified?

Also a bit more context for the components upstream/downstream of the circuit might help.
I agree. There are different variations of pinout for BJTs.

Another thing to check is there are no internal layers inside the PCB which go elsewhere.
 

Offline hermitengineerTopic starter

  • Regular Contributor
  • *
  • Posts: 80
  • Country: us
Re: Does this circuit even make sense?
« Reply #10 on: October 27, 2019, 06:20:11 am »
The pinout question depends:  Can two parts with the same part identifier (2N3904 in this case) have different pinouts?  I would hope for the sanity of a repair guy that they would always be pin compatible.

Anyway, you wanted the whole schematic, so here it is.  The output format is SECAM.  I've tried to understand and identify each stage, but that doesn't guarantee that the descriptions are accurate.
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Re: Does this circuit even make sense?
« Reply #11 on: October 27, 2019, 07:14:57 am »
why the fuck the 74LS624 designer made it so hard to understand what is RNG? Range Input? or Random Number Generator? what a random generator doing in a VCO? :palm:
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 
The following users thanked this post: hermitengineer

Offline hermitengineerTopic starter

  • Regular Contributor
  • *
  • Posts: 80
  • Country: us
Re: Does this circuit even make sense?
« Reply #12 on: October 27, 2019, 07:52:54 am »
why the fuck the 74LS624 designer made it so hard to understand what is RNG? Range Input? or Random Number Generator? what a random generator doing in a VCO? :palm:
Well from what I've seen on my multi-system TV, it almost seems random :D

Well, I just got a used test pattern generator today that can output NTSC, PAL, and SECAM.  The TV responds properly to PAL/D and NTSC/M, but fails to detect SECAM/L over RF.  It can detect it and properly switch through A/V though.  Maybe that's a limitation of the TV.  I wish I could find the manual for it.

Anyway, it appears that the output frequency is determined on the graph page of FC vs RNG.  But yeah, it would be nice to know what "RNG" stands for.
 

Online ataradov

  • Super Contributor
  • ***
  • Posts: 11905
  • Country: us
    • Personal site
Re: Does this circuit even make sense?
« Reply #13 on: October 27, 2019, 07:57:52 am »
But yeah, it would be nice to know what "RNG" stands for.
It says right in the datasheet that it is a "range input".
Alex
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Does this circuit even make sense?
« Reply #14 on: October 27, 2019, 08:17:21 am »
As ataradov says, it's on the datasheet...

So it looks like the transistor is being used to switch the 4u7 cap in and out of circuit to alter the frequency range of the VCO.

Given that the transistor gate connects back to the VSECAM connection and SYN (synchronisation?) pin on the (CPU?) IC it looks like a region selection for PAL/SECAM maybe.
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Re: Does this circuit even make sense?
« Reply #15 on: October 27, 2019, 08:28:22 am »
range what? freq range or amplitude range? do i have to read all those text to understand what is what, and then there's FC, another freq select what? the table only shows their input voltage range.. its my miserable day...
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1855
  • Country: ca
  • Hardcore Geek
Re: Does this circuit even make sense?
« Reply #16 on: October 27, 2019, 10:49:20 am »
range what? freq range or amplitude range? do i have to read all those text to understand what is what, and then there's FC, another freq select what? the table only shows their input voltage range.. its my miserable day...

Huh?  You're complaining that you can't understand the datasheet without reading it?   :-//

The first paragraph (at least of the 1980s TI datasheet) talks about the basic features of the chip, like having Frequency Control and Frequency Range voltage controlled inputs.  These basic features and attributes are then referred to multiple times throughout the datasheet.

Do you expect to just intuit the functionality of the chip through osmosis by waving your hand over the datasheet or something?     ;)
 

Offline hermitengineerTopic starter

  • Regular Contributor
  • *
  • Posts: 80
  • Country: us
Re: Does this circuit even make sense?
« Reply #17 on: October 27, 2019, 11:19:07 am »
range what? freq range or amplitude range? do i have to read all those text to understand what is what, and then there's FC, another freq select what? the table only shows their input voltage range.. its my miserable day...
For that much, you'd start by referring to figure 3.  It gives you a frequency graph with 3 plots for FC=5-RNG volts, so that you can choose your extremes and choose a capacitor to base it on.  Looking at it from the analyst side, use the external capacitor value to find the frequency ranges.  This is 56pF+, depending on the setting of the capacitor, so it's going to be towards the left side of the graph, in the MHz range (as expected).

Figure 1 gives you a 50pF reading with FC on the horizontal axis and plots for RNG=0V to RNG=5V, so you can see what the expected freqs are at that setting.  Figure 2 is for 15pF.  Figures 4, 5, and 6 don't apply to this particular chip.

SECAM blue starts at 4.25MHz and drops to 3.9MHz at bluest.  SECAM red starts at 4.406MHz and moves up to 4.756MHz at reddest.    So naturally the VCO needs to be in that range.

The second capacitor is because there's supposed to be a bandstop filter centered at 4.286MHz, right between the two colors but closer to blue.

So anyway, the 74ls624 should be somewhere in the 3.9-4.75MHz orbit.


So it looks like the transistor is being used to switch the 4u7 cap in and out of circuit to alter the frequency range of the VCO.
But does a 2N3904 provide an adequate discharge route for this?  If not, then the first charge of the capacitor is also its last.  That's why I wonder if it's the wrong part.

Given that the transistor gate connects back to the VSECAM connection and SYN (synchronisation?) pin on the (CPU?) IC it looks like a region selection for PAL/SECAM maybe.[/color][/b]
It's SECAM only, but SECAM is quite complicated in its color sync patterns.  I can't say I fully understand it myself yet, but from what I do understand:

There's a time in the vertical blanking period where 9 color sync pulses are sent out.  They alternate as red and blue, with the last one identifying what the first line represents.  This is one way that SECAM syncs up the colors.  A second way is to provide a smaller pulse on each horizontal sync line.  I suspect that the synchronous counter IC is part of setting up the former.

The pattern of a color sync pulse is to begin at the innermost frequency and sweep out to the outermost in 15us (red) or 20us (blue).  During the blanking interval, the sweep stays at the outermost until that line is over.  Because of the bandstop filter, the amplitude begins low and increases with the frequency sweep, creating what people have termed "green bottles" on the scope trace.

SECAM also uses a four-field interlace instead of the usual two, but I am still struggling to understand that part.  But it probably explains why this circuit is so complex.

Anyway, further reading on what I've been able to find here:

https://radios-tv.co.uk/Pembers/World-TV-Standards/Colour-Standards.html#SECAM
Also, ITU-R  BT.470-6
« Last Edit: October 27, 2019, 11:27:43 am by hermitengineer »
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Re: Does this circuit even make sense?
« Reply #18 on: October 27, 2019, 12:21:21 pm »
range what? freq range or amplitude range? do i have to read all those text to understand what is what, and then there's FC, another freq select what? the table only shows their input voltage range.. its my miserable day...
Huh?  You're complaining that you can't understand the datasheet without reading it?   :-//
maybe i'm used to microchip/atmel way of explaining things... table describing each pins is what is unavail in that VCO DS so i browsed the text looking for RNG, nowhere, how can i relate "freq range" as RNG? it might be FC might it not? anyway it put the guess to the reader.

For that much, you'd start by referring to figure 3.  It gives you a frequency graph with 3 plots for FC=5-RNG volts, so that you can choose your extremes and choose a capacitor to base it on.
its not clear either. ok looking at figure 1 and 2 explained it in somewhat acceptable manner. so i guess your "doesnt make sense" circuit is to control how slow or how fast the freq ramping up or down. or maybe its somesort of diy PLL or feedback control.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline atmfjstc

  • Regular Contributor
  • *
  • Posts: 121
  • Country: ro
Re: Does this circuit even make sense?
« Reply #19 on: October 27, 2019, 01:53:04 pm »
But does a 2N3904 provide an adequate discharge route for this?  If not, then the first charge of the capacitor is also its last.  That's why I wonder if it's the wrong part.

Looking at the broader circuit now, it is clear that there are many other components who can discharge that capacitor while it is in circuit, e.g. the Z209 buffers. So the transistor doesn't need to discharge the cap, it just takes it into and out of the circuit. While it's not in circuit, it will likely retain whatever voltage it had before the transistor cut off, and slowly discharge it through leakage.

If you have an oscilloscope, it would be nice to probe SYNC and the cap to see exactly what is happening.
 

Offline hermitengineerTopic starter

  • Regular Contributor
  • *
  • Posts: 80
  • Country: us
Re: Does this circuit even make sense?
« Reply #20 on: October 27, 2019, 07:31:42 pm »
Looking at the broader circuit now, it is clear that there are many other components who can discharge that capacitor while it is in circuit, e.g. the Z209 buffers. So the transistor doesn't need to discharge the cap, it just takes it into and out of the circuit.
But wait... don't all circuits involving the capacitor necessarily include the transistor?  And we're back to:  Doesn't a 2N3904 (BJT) only allow current flow in ICE direction?  At least, the datasheet doesn't even bother to show current flow in the opposite direction.

If you have an oscilloscope, it would be nice to probe SYNC and the cap to see exactly what is happening.
Oh, now you want real-world measurements?   >:D

SYNC (magenta) vs Q207/C261 junction (yellow)
862254-0

SYNC vs FC pin
862246-1

SYNC vs FC pin with C261 completely removed from the circuit
862250-2

Basically, I see no difference with C261 present or not.
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: Does this circuit even make sense?
« Reply #21 on: October 27, 2019, 09:58:24 pm »
But wait... don't all circuits involving the capacitor necessarily include the transistor?  And we're back to:  Doesn't a 2N3904 (BJT) only allow current flow in ICE direction?  At least, the datasheet doesn't even bother to show current flow in the opposite direction.

A bipolar transistor can be operated with C and E swapped. Parameters like hfe get worse then, but works to some extent. So in a circuit like this, the transistor is driven into saturation by plenty base current, the C is a near short to E for small C currents. If C polarity gets negative in respect to E, the transistor operates in "reverse" mode, and for small enough currents, C is still a near short to E.

So effectively, the transistor connects the capacitor to the respective node if the base is driven, or opens the connection if the base is not driven. A very simple and cheap analog switch. Works as long as the collector current is small enough, this kind of circuit is sometimes seen as a mute switch for audio in / out.
Safety devices hinder evolution
 

Offline hermitengineerTopic starter

  • Regular Contributor
  • *
  • Posts: 80
  • Country: us
Re: Does this circuit even make sense?
« Reply #22 on: October 28, 2019, 04:42:54 am »
But wait... don't all circuits involving the capacitor necessarily include the transistor?  And we're back to:  Doesn't a 2N3904 (BJT) only allow current flow in ICE direction?  At least, the datasheet doesn't even bother to show current flow in the opposite direction.

A bipolar transistor can be operated with C and E swapped. Parameters like hfe get worse then, but works to some extent. So in a circuit like this, the transistor is driven into saturation by plenty base current, the C is a near short to E for small C currents. If C polarity gets negative in respect to E, the transistor operates in "reverse" mode, and for small enough currents, C is still a near short to E.

So effectively, the transistor connects the capacitor to the respective node if the base is driven, or opens the connection if the base is not driven. A very simple and cheap analog switch. Works as long as the collector current is small enough, this kind of circuit is sometimes seen as a mute switch for audio in / out.

Interesting.  This is one thing my semiconductors teacher never mentioned.  Maybe I should do some breadboard experiments to get a feel for it.

Questions to experiment on:  In a reverse current flow, is it still VBE that determines the current flow?  Or is it VBC?  What's the order of gain difference for reverse flow?

BJTs are usually drawn as equal size N regions (for NPN), but as I recall, the collector has a much larger N region.  I guess that's why gain is so much smaller?



Okay, I played with a breadboard with the following circuit:

862532-0

Press the top button and the LED lights, but goes out immediately.  Press the left button while the top button is pressed, and the LED goes out until the capacitor charges.  Release the left button, then release the top button.  LED goes out.  Then press left button and LED lights up until capacitor discharges.

Demonstration of reverse-active mode successful.

But is this what they call reverse-active mode?  Because it says that reverse-active mode requires negative VBE bias.  But the simplified circuit still has forward VBE here?

862528-1

Of course more specific to the logic board in question:  Why does removing the capacitor entirely make no difference?
Obvious answer:  Q207 is dead?
« Last Edit: October 28, 2019, 07:04:55 am by hermitengineer »
 

Offline hermitengineerTopic starter

  • Regular Contributor
  • *
  • Posts: 80
  • Country: us
Re: Next question?
« Reply #23 on: October 29, 2019, 08:51:16 pm »
The part of the diagram before the 74ls624 is two buffers driven by a couple of flip-flops.  Guessing purely by what I know SECAM needs, I referred to this as the color burst strobe.  But after modeling it in a state diagram, that's apparently not it.

Instead, each flip-flop is clocked by a different source:  Z201B by the output of the 74LS624, and Z201A by a 4.45MHz resonant circuit.  When this block is enabled at all, both flop-flops will go high on their respective clock cycles, and when both are high they are (almost) immediately reset simultaneously.  The Q output of the first mixes with the Q output of the second to drive the range voltage, so this has the effect of pulsing the output, which the capacitor(s) will help to smooth.  Because the pull-up resistor is 220K, the quiescent voltage is only around 1V with the buffers detached, so the combination of buffers can easily pull it towards 0V, 2.5V, or 5V (but 5V is only for a propagation delay of time), and 2.5V seems to be the most common.

So, this seems to be a PLL circuit.  Or am I off the mark?
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: Does this circuit even make sense?
« Reply #24 on: October 29, 2019, 09:59:33 pm »
So, this seems to be a PLL circuit.  Or am I off the mark?

Agreed. Looks like the phase detector (that appears kind of similar to the edge based phase detector in a 4046) locks the '624 mid frequency to the crystal frequency using the RNG input, while the SECAM FM is done through the FC input. Q207 / C249 change the loop time constant (I guess it's slow while the visible content of TV line is sent, and fast while or near the sync pulse to allow for whatever burst).

According to the datasheet, nboth RNG and FC have influence on the VCO output frequency, so it looks like the voltage at RNG sets the center frequency (to the crystal frequency) and FC controls the FM deviation.
Safety devices hinder evolution
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf