BK Precision 1248 NTSC Color Bar Generator Repair

[Bushougoma]

I figured I'd post my experience fixing a BK Color Bar generator I mostly use for adjusting focus on CRTs I'm servicing. As with many of my repairs the faults are NEVER common.

I had this issue since I initially bought it when displaying the cross hatch or color bars they were never straight at the top of the screen and the intersections where the lines met were kinked.



Naturally I thought maybe it was a capacitor problem since this unit had a lot of old axial caps in it but after replacing them the problem still persisted so I put it on the back burner for awhile. This didn't affect it's usefulness but it always annoyed me.

The unit itself is pretty simple thanks to the National Semiconductor MM5322 color bar generator IC which basically does all the work. It has a simple 4 bit binary input to select the various test patterns and outputs a composite video signal which is sent out via the front banana jack and is also sent to a discrete RF modulator 300 ohm twin lead not even coax . But I don't even bother using RF.

Here is a look inside:



The power supply is bare bones a small transformer that's half wave rectified, filtered, and sent to a zener diode to provide voltage regulation for the IC and surrounding discrete components.

So one day I decided to dig deeper and find out what was up the first thing I did was measure the supply voltage to the IC. I measured 12.6 volts so I looked at the zener diode it was a 1N4742 (a 12 volt zener) so no issues there. I didn't have a schematic but the board was relatively simple with few components so I started checking the transistors and resistors nothing seemed out of the ordinary the voltage drops were good on the transistors and no resistors were out of spec.

My gut was saying for some reason low supply voltage so I pull up the datasheet for the MM5322 to see what the required supply voltage was and it had a range from 12 - 19 volts DC they certainly don't make ICs with a wide range like this anymore.

http://pdf.datasheetcatalog.com/datasheets/270/493036_DS.pdf

So that left me stumped the zener is a 12 volt zener doing it's job and the IC can function down to 12 volts so what is going on here?

At that point I decide it's time to look for a schematic and I'm glad I did or I probably wouldn't have figured this out. The Internet was a no go luckily the guys at BK have a section on the site where they post schematics for older equipment and if it's not there you can request schematics. So I sent a request and a day later I had my schematic  .

 BK Precision 1248 Color Bar Generator Schematic.pdf (113.93 kB - downloaded 114 times.)

So I check over the schematic and low and behold something catches my eye the zener diode in the schematic is a 1N4745 (a 16 volt zener) and NOT a 1N4742 (a 12 volt zener). The current limiting resistors were also different a 56 ohm resistor for the 16 volt zener in the schematic vs the 130 ohm resistor on my board for the 12 volt zener.

So before I got ahead of myself I check the board over to see if someone at some time worked on this unit and substituted the wrong value zener and resistor but no the joints are wave soldered and have never been reworked. It came this way from the factory!

So I desolder one of the legs of the zener to isolate the internal power supply from the circuit:



and clip my power supply into the circuit (red to the VCC pin of the IC and black to a convenient ground point):



and send 16 volts to the board and lo and behold. It works PERFECTLY!



Just for fun I reduced the voltage by 1 volt while watching the output and slowly but surely the issue returns.

15 Volts:



14 Volts:



13 Volts:



12 Volts (lost vertical sync and picture began to roll because the original zener voltage was 12.6 it seems it was just on the verge of losing sync at 12.6 volts):



So I install the proper zener a 1N4745 and the proper 56 ohm current limiting resistor per the schematic and it now works great.



All I can figure is someone at BK decided to "update" the design and run the IC at a lower supply voltage which is fine I have the utmost confidence that the IC will run at that voltage as it's within the permissible range on the datasheet. However the surrounding discrete circuitry was sized for a 16 volt supply and I verified that ALL of the resistors surrounding the transistors were EXACTLY the same value as those on the schematic which called for a 16 volt supply.

Short of having an identical unit to compare the supply voltages to without the schematic there is no way I could have figured this out. But these are the problems that I always run into NEVER simple to find .

Oh and as an added bonus to make this very antiquated piece of gear more useful I made an adapter cable to convert the mini banana jacks on the front panel to an RCA connector so I could get composite video out of this thing as it outputs a much nicer signal than RF. Man those mini banana connectors were hard to get a hold of.

[Renate]

That's interesting, but at some point it did work at 12V.

The question is, is the visual distortion from the dividers miscounting or
is it from a crunching of the sync due to a voltage offset/headroom problem?
I'd rather see a scope capture of the signal on 12V vs 16V.

Shooting in the dark: maybe a coupling cap is bad and the bias is going too low and the sync is getting crunched when the supply is lower.

[Bushougoma]

I've attached the schematic if want to look at the circuit.

[Renate]

Oh, sorry, I didn't notice the link for the schematic.

Do you have a scope?
I see that the pot for the video has positive and negative polarity.
Set the supply at 12V and look on the scope with the knob fully CCW vs fully CW.
They should be identical (except upside-down).
I'll bet that when it's upside-down (sync going positive) that the sync pulses are fully grown
and that when it's right-side up the sync pulses are stunted.

Mmm, maybe C24? Leaky and low capacitance?
Low capacitance would trash the V sync.
Leakage could cause the bias to shift and stunt the sync.

C25 having low capacitance would make the V sync badly tilted too.

[Renate]

Bump. Aw, give us an update, please.

[Bushougoma]

Both of those caps are electrolytic. All of the electrolytics were replaced and it made no difference at 12.6 volts still same symptoms.

I do remember when it was on 12.6 volts unless the video knob was cranked all the way positive the picture rolled and even then it produced that terrible picture. So I believe you're correct the sync pulses were stunted.

Now on 16 volts it has next to no effect the only difference I can see is in the color bars if it's turned all the way positive faint vertical white lines appear in them at the midpoint they're gone so that's where I leave it. I can also turn it all the way negative and don't lose vertical sync anymore.

[Renate]

Well, the output is 1V P-P.
It really can't be necessary to juice the thing with 16V?
That last follower/inverter stage is clearly not in the linear region.
Something's wrong there, could be a bad/leaky/missing-collector transistor.
Do you have a scope? It's easy enough to check.
Midpoint you should have no signal at all, so something's wrong.
All the way negative it should be upside-down and drive your monitor crazy.

Heck, you should be able to connect an ear phone to the video output.
It should buzz, but drop to silence at the midpoint before buzzing again.