HP1743A Oscilloscope Rescue

[MelJon66]

Greetings All,

Some years ago I purchased a non-functioning HP1743A oscilloscope for literally pennies from a local seller. This appears to be similar to a HP1740A but with a digital counter in the top right corner which looks like it's associated with the delayed sweep.

The seller said the vertical amplifier hybrid chip (A5A1) had blown so he had fitted a new one but that also failed. He said they were almost unobtainable and very expensive so it was beyond economical repair. It sounded like too much of a challenge for my limited skills so it has languished in dry storage since then waiting for the right time to tackle it. On a whim I have just had a look at it to see if it can be rescued. Lo and behold there is a trace with full vertical deflection! If I put pressure on the connector to board A5 the trace becomes a blur extending full screen height so perhaps the seller was mislead into thinking the hybrid was at fault.

There are still a few obvious problems. The trace is sharp and clear but does not go across the full width of the screen. The sweep is non-linear with the waveform being compressed at two points either side of centre. The far left, right and centre are not compressed so the traces look very odd. This compression probably causes the narrow display. So far all I have been able to do is check the ramp output waveform at test point 1 on the Main Sweep board A8 which looks as it should. So, I suspect some bad capacitors or poor connections between A8 and the CRT horizontal deflection pins.

The output from A8 goes across a wire link to board A17, which is shown on the chassis parts diagram in the 1743A service manual (page 6-5) but no schematics are given for any of the boards. The manual was downloaded from the Agilent website. Even the parts list (almost illegible anyway) ends at board A14 so there is nothing useful for board A17 at all. Either all the pages beyond 6-23 are missing in the Agilent download file or possibly it has to be used in conjunction with another manual. I have downloaded the 1740A manual but despite being a much more comprehensive manual the 1740A does not appear to have a board A17 so it doesn't help. These models are not as similar as I thought.

Does anyone know if there is a better manual available for the HP1743A or know which model it is most similar to?

Cheers
Mel

[MelJon66]

Errata - I should have said that the manuals were downloaded from the Keysight website.

Mel

[Chris55000]

Hi!

K*** have really excelled themselves this time! Not only it is the usual atrocious quality, it is also missing the entire content of sections 4 (Theory), 7 (Manual Changes) and 8 (Troubleshooting, Schematics and Board Layouts!)

They have recently introduced a much more hostile corporate policy towards any so-called "non-professional" who dares to even LOOK at, let alone handle, or use, their hallowed pieces of expensive gear - I wonder if they're purposely going through all their downloads and deleting all the good copies of old manuals and making sure all schematics, theory, change information is removed from what they do leave on there? - it certainly seems like it !

Leave it with me for a little while, I have got them and I will start posting them in Forum-Sized Sections for you shortly once I've got a suitable PDF editor up and working!

Chris Williams

[RogerEvans]

Have a look at artekmanuals.com, they have the HP1743A service manual including schematics at $7.50 for a PDF. Dave has a really good reputation in the community and on my experience the scanned manuals are excellent quality and you usually get the file within a day or so.

Have you tried wiggling the beam finder switch/button several times, it could be that part of the switching that limits the X deflection amplitude has become stuck in the on position. Probably more likely is a fault in the X amplifier or a failure of one of the low voltage stabilised supplies (standard advice is always check the low voltages for voltage and ripple before going any further). Obviously a decent manual is a necessity for most trouble shooting.

Best of luck with the repair, keep us all informed how you get on.

Roger

[MelJon66]

Hi All,

Here are some photos to illustrate the compressed waveform issue. Note the liberal spattering of fine paint dots on the front of the 'scope. It appears to have been standing in the corner with the screen upwards while the previous owner was decorating the room. Judging by the colour of the dots the walls were blue and the ceiling white ! It obviously wasn't his most valued possession. It's only emulsion so it will come off with some careful wiping. I'll get it working properly before I spend time on that.

Photo 1 shows a nice sine wave produced when the whole waveform is to the left or right of the compressed bits.

Photo 2 shows the same waveform positioned in the centre. It will not stretch to full width. The left and right ends are fine. The centre has been stretched suggesting that the sweep speeds up there. There are two short sections either side of centre where the sweep slows right down compressing those sections of the waveform. The line under the trace is channel 2 with grounded input. You can just see the short compressed sections showing up as brighter parts of the trace.

Photo 3 shows the ramp output signal at test point 1 on board A8. It looks OK to me.

Cheers
Mel

[MelJon66]

Thanks to Chris and Roger,

Chris, I have downloaded the manuals from arktekmanuals as Roger suggested. It will save time for both of us but sincere thanks for offering to spend time to scan the missing sections. The small cost is worth it to avoid struggling with the rubbish that Keysight are making available.

Cheers
Mel

[cmjones01]

The fact that the distortion is symmetrical about the centre, and there's no distortion in the Y axis, gives me some hope. A biasing problem in the latter stages of the X amplifier could cause such distortion. Otherwise I'd be worrying that it had been dropped and the CRT had suffered internal damage.

Chris

[factory]

Originally Posted by RogerEvans

Have a look at artekmanuals.com, they have the HP1743A service manual including schematics at $7.50 for a PDF. Dave has a really good reputation in the community and on my experience the scanned manuals are excellent quality and you usually get the file within a day or so.

Have you tried wiggling the beam finder switch/button several times, it could be that part of the switching that limits the X deflection amplitude has become stuck in the on position. Probably more likely is a fault in the X amplifier or a failure of one of the low voltage stabilised supplies (standard advice is always check the low voltages for voltage and ripple before going any further). Obviously a decent manual is a necessity for most trouble shooting.

Best of luck with the repair, keep us all informed how you get on.

Roger

Not the same scope, but this is effect dirty contacts in the beam finder switch had on my 180A, note the trace horizontal position control is (nearly) centred. I've had issues with dirty beam finder switch in a few old HP scopes.


Re:Artek manuals, make sure the manual covers the serial prefix of the equipment you have, can be a PITA it doesn't & you find differences in the board layouts & circuits.

David

[MelJon66]

Hi Roger,

Lots of knob twiddling has taken place but with no effect on the trace non-linearity and compression.

I visually checked the power supplies before switching on and the voltages are good. But, there are some minor historic repair bodges such as piggy backing and daisy chaining resistors that do not fill me with confidence. Some people will do anything to avoid spending time getting access to the right side of the PCB. The supplies appear to differ only slightly from what is shown in the 1740A manual but I will wait for the 1743A copy before I can be sure I have the correct values before I change anything.

Cheers
Mel

[MelJon66]

Thanks Chris (cmjones01),

The distortion isn't really symmetrical as the compressed sections move around at faster timebase settings. I have now noticed that with a grounded input and timebase settings of 10us or faster the trace does stretch across the full screen but the trace still has bright compressed spots at various positions depending on timebase setting.

I note that the beam finder line has even brightness throughout it's length. Would that happen if the CRT was damaged?

Cheers
Mel

[cmjones01]

Originally Posted by MelJon66

The distortion isn't really symmetrical as the compressed sections move around at faster timebase settings. I have now noticed that with a grounded input and timebase settings of 10us or faster the trace does stretch across the full screen but the trace still has bright compressed spots at various positions depending on timebase setting.

If the distortion moves around the screen with different timebase speeds that's an even better sign that it's electronic in nature and not a mechanical problem in the tube.

I note that the beam finder line has even brightness throughout it's length. Would that happen if the CRT was damaged?

Probably not. With a damaged CRT I'd expect any trace, even the beam finder one, to show distortion and bending as you move it around the screen with the X and Y position controls. It sounds like that's not happening, so the chances seem good that the CRT is OK.

Chris

[Radio Wrangler]

Move the trace up and down across the face of the tube using a Y shift control. Does the X distortion or size vary?

If not, then it's a good sign of a usable CRT.

By the way, the 1740 uses the same CRT as the 180.

David

[MelJon66]

Hi David,

The pattern of varying brightness on the trace stays the same when shifting the trace up and down. That sounds like good news for the CRT .

Thanks
Mel

[Diabolical Artificer]

Have a look at this video of mine here - https://www.youtube.com/watch?v=oezw1THwKL0 it's a different scope and a valve type scope but shows a similar issue which was caused by a leaky cap and excessive ripple if I remember right.

A good start point with fault finding old scopes is to first check your power supply DC voltages as well as the AC ripple of the same rails, anything over 1v RMS is suspect. There is a brilliant Tektronix book here - https://usermanual.wiki/Document/Tro...490300137/view which you might find useful.

Andy.

[Radio Wrangler]

You have a good, undistorted sawtooth from the timebase. So it is time to follow that waveform, stage by stage to the X plates themselves, and see where it goes sour.

You can also input a signal, like a sine into both channels in X/Y and you should see a straight diagonal line. This will show up distortion in the X amp. Use the shift pots to move things around to explore. You can also see if things change with frequency.

David

[PJL]

I would agree with David, If the sawtooth is good then the fault must lie between the takeoff of the sweep output and the X plates. If it's symmetric around the centre that sounds like a form of crossover distortion.

[MelJon66]

Thanks to all for your contributions,

I now have the manual and a quick glance confirms that it does seem to cover everything I need so I will be able to make more progress soon.

Andy, yes I have seen your video and I will be looking at all your 500 series videos when I build up enough courage to tackle my Tek 543. On the 1743A the power supply voltages are all good and the ripple is no more than 15mV peak except the 120V rail which is about 100mV. The 120V ripple is unlike the other rails, being generally very flat with some very high spikes (100mV) so I do lack confidence in it. Some of the historic repair work on that rail is not to my liking but I think its ok for now and I will work on it later.

David and PJL, I agree and now I have the manuals I can follow the circuit through to the CRT. I have worked out which boards to check through and the manual gives good example waveforms at the test points. The test points on these are excellent clip points and access to them seems pretty good.

One bit of good news is that I have solved the vertical deflection issue that was the original problem reported by the seller. Amazingly, given that this 'scope is over 40 years old there was no solder on one of the pins on the connector to vertical amp board A5. It's the third pin from the left on the attached photo. There is no track to this pin on the other side so the only contact between tracks and connector pin are what you can see. The original plating has never had any solder on it and was completely dry. The (gold?) plating must have been making adequate contact all these years but every time the 'scope was moved the plating at the contact points must have been wearing thin and eventually contact was lost. The seller was sure the hybrid chip was damaged and must have missed the obvious. I have soldered the connection and the vertical deflection is now working well and the trace is rock steady and pin sharp.

Cheers
Mel

[MelJon66]

Hi All,

The 1743A is up and running .

A couple of 47.5K resistors on the horizontal output amp board A11 had gone bad. R17 was open and R11 was at 1.1Meg. They are the only ones of that value on the board and there was no sign of heating so I assume they have just aged disgracefully. Luckily A11 is one of the easiest boards to remove. Disconnect two connector pins (to the CRT) and pull A11 off its connector pins to board A7 and job done. The plug in boards do make access for repairs pretty easy.

The trace is now excellent - see attached photo. But, there is still a bit of work to do. The more I look at the 120V supply rail the less I like it. I now see that there should be a transistor that provides current limiting but it has been removed. No current limiting then! I will avoid using it until the work on the power supply is completed and the 120V rail ripple is down from the current 100mV to the <20mV specified in the manual.

Also, there are lots of switching problems. When I turn on it's pot luck whether it gets stuck in Chop mode so I usually get two dotted line traces whether I like it or not. The line switch (on/off) often doesn't switch off either. Many of the front panel switches need attention and the front panel needs a good clean up to remove the paint spatters. Finally, it needs setting up and calibration. I don't have any equipment that goes anywhere near 100MHz but I will do what I can.

I'm very happy with progress and thanks again to everyone for their help so far.

Cheers
Mel

[Radio Wrangler]

Good progress!

Once you get used to it, you'll find the 1740 is a very comfortable scope to use, and it'll become your daily user. The 1743 just seems to be a light refresh and someone either thought adding an LED display was sexy, or just cheaper than a mechanical turns counter dial

You should also find that it will trigger on 2mm of trace. But don't tell anyone, it's a secret.

David

David

[MelJon66]

Hi David,

If the price reflected the true cost of the additions the digital interval timer must have been a hugely expensive extra. The attached photo shows an extra logic board (A18) mounted across the top of the other PCB's. There is another extra board (A17 - Time Division Decoder) that is piggy backed onto the normal A8 board below. All this appears to be just to support the interval timer . A18 obstructs access to the other boards so I hope it is worth it when I get the chance to use it.

Perhaps a big industrial customer demanded this feature and HP obliged. I can't think of a more likely justification for such additional complexity.

Cheers
Mel