The trigger circuit in my Tektronix 454 has stopped working; in the "normal" position, I get no trace, rotating the trigger level control does nothing, and the trigger light does not come on. In the "auto" position, I do get a trace. Any ideas before I go diving in? I'm using the built-in calibrator as a signal source.

The 454 uses Nuvistors in the input amplifiers and the trigger generator. If you have a tubetester that can handle Nuvistors check the one in the trigger generator or swap it with one from the channel B amplifer. I have seen this more than once with the 454 considering it just stopped trigering.
Good Luck.
Jerry
W2JI

Later 454s use FETs, but let's not put the cart before the horse. Since NORM and AUTO are working like they're supposed to, we can eliminate the sweep and bright-baseline auto circuitry. Let's eliminate some possibilities first.

See if it will trigger on CH1 only, CH2 only or COMP. If not, see if will EXTernally trigger or trigger on LINE. If the scope will trigger on all but one of those four options, the triggering circuit is fine and the problem lies in one of the small circuits that lie between the source (CH1 or CH2 or COMP) and the trigger circuitry, whichever isn't working. I don't have a schematic here, but these usually involve only one transistor in most Tek scopes. I've a feeling that isn't the problem though.

If none of those trigger the scope, the problem is in the triggering circuit itself. First, wiggle all the transistors of the A trigger circuit in their sockets just in case it's developed an intermittent. If that fails to get things going, it's time to chase signals.

Set both the "A" and "B" the trigger sources to EXTernal and feed the same signal into both with the TRIG LEVel set to zero with a positive slope. The reason for sourcing both trigger inputs is that the neat thing is that you have this twin of the "A" trigger circuit with the "B" trigger, so you can compare signals even if the schematic doesn't have waveforms for every little point. Use another scope to follow the line signal through the trigger circuitry using the schematic as a guide. If you con't have another scope, nothing says that you can't use the 454 itself to look at itself. It's not like it's doing an appendectomy on itself, or that you're going to create a huge "feedback" loop, but no harm will come of using CH1 with a 10X probe to look at the signals. Now, they'll be free-running and not triggering, but that's better than nothing. At least you can compare with the "B" trigger. You can mess with sweep speed and the VAR TIME/DIV to try to somewhat sync the waveform or make it stand still enough to see something of the waveshape. You could even add another probe and use CH2 in a dual-trace mode to compare both trigger circuits at the same time. But of course, access to another working scope would be easiest.

My guess is that the problem will be something simple.

As I recall, if you have the trace positions set about mid-screen vertically (input set to ground is good); and set-up for DC triggering, you definitely should get one trigger and blink of the triggered light by swinging the trigger level control back & forth between + & -.
Best to try that at medium sweep speed.

Agree with Dean and willitrun, both of whom give good advise. Like Dean I'd suspect one of the trigger interface circuits. By a methodical process of elimination, you can narrow the problem down to just a very few components before you turn the first cabinet screw.

I know to some I'm like a broken record about this, but TEK produced an excelent troubleshooting guide for their analog scopes. However, the process applies to any analog scope, and, the methodology outlined on pages 3/4/5 should be the beginning of any trouble shooting session.

Here is an update: following Dean's advice, I could not get the scope to trigger on anything. Chasing the signal through the circuit with my other oscilloscope, I found that I am losing the signal in Q664; I can get a good signal at the base, and adjust it with the trigger level control, but there is nothing coming out of the collector. I'll have to pull it out and check it, along with Q667 and Q670 which are also part of that circuit; those are soldered in.

Do those tunnel diodes ever go bad?

Tunnel diodes can go bad, but suspect the transistors first. It takes a curve tracer to check a tunnel diode. Don't even try it with a multimeter. If the transistors seem OK and you're down to suspecting the tunnels, swap them out with the ones from the "B" trigger circuit. If you never use the "B" sweep in triggered mode (few people do), you can stop there. Otherwise, you're on a parts hunt.

On reviewing the circuit description in the manual, apparently Tek is using the tunnel diode to generate a negative-going pulse at the base of Q664, which I am not getting. I'm going to try the one from the "B" trigger generator to see if that fixes it. All the transistors in the circuit check OK, and they all DO plug in!

Not easy to see a tunnel diode pulse ,what is the bandwidth of your test scope ?

It's fixed! The trouble WAS the tunnel diode, just as I suspected; I subbed the one from the "B" trigger generator. Maybe I can pick up a junker 453 or 454 for parts, as it seems that those TDs are scarce.

Are the TDs socketed? If so, beware that any junker may already have them pulled! If they're soldered, you're probably OK. Nearly every portable scope from 1975 and on has TDs and most are socketed. The 465 is more plentiful than the older 400's. I'm not sure of the specs, but the 465 TDs may work (as a set) if the scope trigger circuit gets recalibrated. If it's only the last two digits of the 9-digit part number that's different, you're likely OK.

They are soldered in on this model. I found listings on the web for a couple of guys who deal in Tektronix parts, so they may have one.

They're relatively scarce, but they do show up.

You may have some success with a Russian TD. I haven't tried one and, from what I've seen, there's been mixed success, but it's a possibility.

Google for the Tek PN 152-0310-00 or the equivalent 1n3758. Looks like Talonix has it in stock, but you probably don't want to look at the price. I've bought from them; they're reputable.

It still amazes me you can find NOS parts for the 454.

Edit: looks like Tucker has one too.

Russian 5ma tunnel dodes are available on that auction site if you look for part #GI304A.

I think there's lots of NOS parts for the 453/A and 454/A because Tek extended their normal product support "life" by about six years before being declared obsolete. Back then, a product would be supported by Tek for 12 years after last appearing in the catalog. After that, it was obsolete, and repairs would be by the hour and if a product couldn't be fixed, the customer was still liable for the costs. Parts would no longer be produced and typically, only the various Service Centers would have any remaining stock until they ran out. The fact that there were some military contracts for those helped the NOS thing, too.

Sphere's (and other sites) evidently have an error on their cross-reference. The 1N3758 is a standard rectifier. However, the 1N3717 is a 4.7mA tunnel diode and should work. My research show that the Tektronix 152-0310-00 is a 4.7-5.3 (nominal 5) mA tunnel diode with 7pF max capacitance. I found that information here: http://www.docstoc.com/docs/82530193/Tunneldiodesummary
The GI304A (Russian) tunnel diode has a max C of 20pF, so, some compromises may be in order. It (the GI304A) is also available on amazon.

Sphere (and others) has an error in their cross reference because the number is taken straight from the Tektronix 454 manual.

Thought it was odd that it was coming up as a rectifier.

Consolidated Electronics shows the 1n3717 for $13 and change. You could also buy it from American Microsemiconductor on Amazon for $40.

The 453 used a TD3a which is 18pf, so the Russian part might be more compatible with that one.

Thanks, PaulAm. I didn't have my Tek manuals handy because I was at work.

Which diodes did you end up ordering?

Looking through my junk box, I had picked up some 3i306J Russian diodes which are spec'd a 4.5-5.5ma.

If you're asking me, I have a couple of flavors; 10 mA (I'll look up the part number tonight) on hand and 5 mA (GI304A) on order. I also suspect that one could use a UJT with just a bit of creative design, but, that takes time, testing, and (heaven forbid!) thought.