Tektronix 465 Oscilloscope Troubleshooting

[tautech]

I looked at all the tant caps and the only one that looked like it had a small hairline crack is C1113 on the timing circuit board (see picture). No burnt looking caps that I could see.

Also, I checked for AC on the test points. Found the following:

TP1568 -- DC 8.09V -- AC 115mV
TP1548 -- DC 12.87V -- 3.03V
TP1536 -- DC 55.9V -- 0.77V
TP1558 -- DC 5.08V -- 184mV
TP1518 -- DC 110.07V -- 0.72V

The 15, 12 and 5V rails all have excessive ripple IMO.
Check the service manual for the listed max ripple limits.

Something's still pulling the 15V rail down and when that happens and the rail is under stress, ripple measurement rises. Get the 15V rail roughly to spec first before looking at anything else.....other stuff can wait.

[DanielSpokane]

Late night for you bd139 Thanks for the help! Thanks everyone else too!

I pulled one leg of C1113 and tested the 15V TP and it reads 12.7V. Hnmm... any ideas on a next step?

Also, I have a couple caps from a parts radio. They are 100uf 50V. See attached pictures. You think I should replace either C1549 or C1113 with these?... both still have one leg of the orignial cap pulled. I can put back the leg or use these.

[DanielSpokane]

tautech, following are the typical ripple specs. Yes, mine appears to be out of spec. Hopefully if I can get the 15V power supply in spec that will help.

From Table 6-4--Typical Ripple
+55V @ TP 1536 = 4 mV
+15V @ TP 1548 = 2 mV
+5V @ TP 1558 = 2 mV
-8V @ TP 1568 = 2mV
+110V @ TP1518 = 20mV

[tautech]

tautech, following are the typical ripple specs. Yes, mine appears to be out of spec. Hopefully if I can get the 15V power supply in spec that will help.

From Table 6-4--Typical Ripple
+55V @ TP 1536 = 4 mV
+15V @ TP 1548 = 2 mV
+5V @ TP 1558 = 2 mV
-8V @ TP 1568 = 2mV
+110V @ TP1518 = 20mV

Dan, also understand that the rails can be referenced a just one of the rails therefore adjustments can be interlinked.
But for now just get the ones that are way off somewhere near where they should be to see if normal operations return.

Listen to what bd139 says about these, he knows his stuff. 

[bd139]

Those replacement caps are fine. I wouldn’t put the old ones back in because they are guaranteed to fail in the future.  Best to find the actual dead one first. Next step is to either trace all the caps on the 15v rail with the schematics. This is a long job unfortunately. Print out all the schematics and circle them all. They are usually labelled up next to the 15v supply. Then go round and check them all.

As tautech says, all bets are off until that rail is sorted.

Ripple isn’t terribly important to start with. The scope will work with high ripple just the triggering will be dicky and the traces will have a lot of noise on them.

Edit: just a point before I forget. Probably worth looking at the ripple before the regulator as well now. The 15v rail will drop out if it’s not seeing enough volts in as well. If half the bridge rectifier is fluffed or the filter cap is dying it will result in poor regulation.

Do you have another scope?

[DanielSpokane]

Sounds good. I'm going to head to bed now (it's 1AM here) but will do this tomorrow. I will print off copies and make sure to check all capacitors touching the 15V rail. I assume I just lift one leg and test to see if the 15V rail goes from 12.7V to 15V?... if not, then I go to the next one?

Yes I do have another oscope. I have a Siglent SDS 1102X.

Thanks again

[tggzzz]

If the input to a regulator is OK but the output is too low, then there is probably excessive current being drawn somewhere.

A IR camera (or even thermometer) might conceivably help indicate hotspots.

Excess current will cause larger voltage drops along PSU rails. Careful measurement of differential voltages along a PSU track can sometimes help indicate where the current is going. Use a DVM on mV range.

Read the manual to see which PSU rails depend on the regulated 50V rail. Some of them are simply a fixed proportion of that voltage, so if the 50V is X% low then such rails will also be X% low.

[bd139]

Cool. That’s all you have to do yes.

Worth popping a scope on the collector of Q1544 first and seeing what the lowest voltage is before you do this. If this is close to or lower than 15V it will cause regulation loss. At that point C1542 or CR1541 are suspect. Sometimes on these, the capacitor goes high ESR and blows one or more diodes in the bridge rectifier after which it limps.

If you can, also post a picture of the waveform at TP1548 (15v out). This should result in some clues.

There is a possibility that the 55V rail, from which everything is referenced from is misbehaving.

[tautech]

Yes I do have another oscope. I have a Siglent SDS 1102X.

Simple to take screenshots with but you'll have to compress them or change then to another format so that they're smaller than the 1MB/file forum limit.

[DanielSpokane]

Good morning (at least here anyway) and happy Easter

Below is TP 1548 from the scope.

At Q1544 there are 3 leads. Should I scope each of them?

[tautech]

Good morning (at least here anyway) and happy Easter

Below is TP 1548 from the scope.

At Q1544 there are 3 leads. Should I scope each of them?

With 6V+ ripple I'd be looking hard at C1552.

Also:

Also, I checked for AC on the test points. Found the following:

TP1568 -- DC 8.09V -- AC 115mV
TP1548 -- DC 12.87V -- 3.03V
TP1536 -- DC 55.9V -- 0.77V
TP1558 -- DC 5.08V -- 184mV
TP1518 -- DC 110.07V -- 0.72V

The DC value is close enough to agree but the AC not.

[tggzzz]

Below is TP 1548 from the scope.

At Q1544 there are 3 leads. Should I scope each of them?

120Hz => the bridge rectifier diode CR1541 is OK.

I'll bet the flat top is somewhere near 15V, indicating the 15V regulator is working OK. The action of the regulator is to try to keep the voltage at TP1548 the same as the "reference" at the junction of R1543/R1544. It not be 15V if either the input voltage (C1542) drops too low or the reference falls. The reference voltage is simply a fraction of the 55V rail (TP1536).

So, if TP1536 looks OK, then look at the reference voltage, then look at the +ve of C1542 which is the "input" to the 15V regulator.

Quick dirty test to see if C1542 is kaput: temporarily solder a "similar" capacitor across C1542 and see if it makes a difference to the waveforms. "Similar" means adequate voltage, and capacitance not much greater (to avoid excessive inrush current) and not too much smaller. If value is 10% of C1542, then very crudely expect a 10% improvement in ripple.

Caution: the C1542 terminals are close together and could be shorted easily. Keep your scope probe's ground lead well away from that area, since if it touched a non-zero voltage there could be sparks followed by sighs.

[DanielSpokane]

Looking at C1552 and C1542 they are both larger power capacitors mounted to the reverse side of the board. Are you thinking I should solder a capacitor to the bottom side (the visible side) of C1552 and hook my scope up to TP1548 and see if there is any improvement to the ripple? I assume the capacitor should be electrolytic with voltage rating exceeding say 25Volts or so? Any suggestion on the uF rating for this test cap?

[tggzzz]

Looking at C1552 and C1542 they are both larger power capacitors mounted to the reverse side of the board. Are you thinking I should solder a capacitor to the bottom side (the visible side) of C1552 and hook my scope up to TP1548 and see if there is any improvement to the ripple? I assume the capacitor should be electrolytic with voltage rating exceeding say 25Volts or so? Any suggestion on the uF rating for this test cap?

It is a temporary connection to see if the existing capacitor is duff, so make connections any place convenient. Yes, that is probably directly across the end of the terminals that are visible without removing the PCB - and you probably won't be able to put the case on at the same time! (Since you are a beginner, I won't suggest you simply hold a capacitor and touch the leads to the live circuit!)

What's the voltage rating of the existing capacitor? Any parallel capacitor should have the same or higher voltage rating.

Have a look in the manual at the tolerance of the capacitance value: that will give a hint a wide range is acceptable. For the test (cf normal operation) the capacitance is uncritical. The larger the capacitor the easier to see the effect of having a good capacitor. Even if the capacitor is only 10% of the nominal value, you should see some difference; how much difference will depend on how bad the existing capacitor is. But don't use a much higher capacitor value, since the inrush current might be too high. So, start with whatever caps you have to hand, say 10% to 100% of the nominal value.

If you identify problem capacitor(s), you will have to replace them with appropriate alternatives. Removing those caps is a pain, and you'll probably need "adaptors" to match the footprint, but that is a problem for later.

[DanielSpokane]

According to the manual C1552 is rated at 5000uF at 25V.  I looked through my stash and the best I have are the following:

1. 100uF at 50V
2. Sprague powerlytic 36D8084--17000 at 25VDC

Seems like one is too small the other too big?

Also, see picture below. I shined a light behind the board to see where the caps are in relation to the solder points. Looks like there are 3 solder points per cap.

[JDubU]

Here's Jim Williams fixing a Tek 465 (spoiler alert: shorted tantalum cap).

[tautech]

Seems like one is too small the other too big?

Yep but you could try 2x 100's paralleled and tacked on the correct pads......just be sure you get the polarity and position right ! (extender leads permitted) 

Then without any other changes look for an improvement in the ugly sawtooth waveform and grab us a screenshot.
Take your time.....and no mistakes !

[tggzzz]

According to the manual C1552 is rated at 5000uF at 25V.  I looked through my stash and the best I have are the following:

1. 100uF at 50V
2. Sprague powerlytic 36D8084--17000 at 25VDC

Seems like one is too small the other too big?

Also, see picture below. I shined a light behind the board to see where the caps are in relation to the solder points. Looks like there are 3 solder points per cap.

What is the sprague's capacitance?

The existing cap has several can connections plus one other. Look at the schematic and trace out the PCB connections.

Use what you have available.

[DanielSpokane]

The Sprague's capacitance is 17000... I believe. It is big... about half the size of a soda can. The other option is to do as suggested and pair a few spares in parallel and see how much I can add up. ... or order something.   

[DanielSpokane]

Think there is any harm in doing the 17,000uF test cap?

[tautech]

Think there is any harm in doing the 17,000uF test cap?

I wouldn't for only one reason:
Inrush current could damage the bridge rectifier diodes. Their design is such that they only expect to charge 5kuF in however many mains cycles it might take.....times that by 3 and you put some serious load on those diodes......maybe more than they can manage.
Is it worth the risk ? IMHO NO !

[bd139]

Those bridge rectifiers are pretty fragile. I've had a couple of dead ones. They're also difficult to get out.

I wouldn't risk it. I have a couple of 2200uF 35v ones floating around I use for these sorts of things.

[DanielSpokane]

Ok, sounds good, I'll hold off on the 17K

I'll parallel up a few caps and connect to C1552 and will post a picture when I'm done. Any other suggestions?

bd139 on Q1544... which leg is the collector?  I'll put a scope on that too. Sorry I'm not totally familiar with all the terminology. Learning as I go

[tggzzz]

bd139 on Q1544... which leg is the collector?  I'll put a scope on that too. Sorry I'm not totally familiar with all the terminology. Learning as I go

Don't: you could easily short pins together or pull pins from the socket. Put the probe on something connected to the collector.

[tautech]

I'll parallel up a few caps and connect to C1552 and will post a picture when I'm done. Any other suggestions?

None, just one definitive step at a time at this stage while you become more familiar with the scope and its workings.