My HP3312A has blown transistors in the output amplifier. These have Motorola part numbers and I cannot find a cross-reference to current production devices.

On the case: Motorola 4-351. K9128.
In the parts reference: Motorola 1854-0351. NPN. TO-18 case. PD=360mW.

Any clues how to cross-reference this part?

Steve

The HP cross refernce documents are available here:

http://www.sphere.bc.ca/test/hpparts.html#cross

BUT, your part is not listed. This is supposed to mean that the part was made exclusively for HP. If you can get a handle on the operating voltage, you can try generic substitutes to see if you can get it working reasonably well. Pay attention to the operating frequency, voltage, and power dissipation. Unless you can find an actual HP replacement part, that is about the best you can do. Sometimes you can locate more complete specs if you do dome digging. There is a good Yahoo group for Tektronix that has that kind of information. I don't know if there is a similar one for HP.

First, the generator is 70's, so we are almost certainly dealing with silicon transistors, not germanium. Second, although you should try to see if you can find out what it is I am almost positive that the operating voltage is not a problem. The power supply probably only puts out 24volts, as that is all that is needed to run transistor circuits. Your generator only puts out ~15Mhz, so even with harmonics most any moderately high frequency transistor should be fine as far as that goes. The TO18 case and PD rating tells me that the thing is not going to have to dissipate much power, and we know it is an NPN device. The good-for-everything NPN transistor, the NTE199, has a PD of 360mw and claims it will work to 90Mhz. However, in my experience HP often used metal cased transistors for shielding reasons, so I would be reluctant to put a TO92 cased part in there. 2N2222A's are cheap and are in a TO18 case. They will do 100Mhz easily and will dissipate ~0.5watt; more than you need. Mouser has them for the outrageous price of $0.43 each. That is my first guess, and really I don't think it matters all that much. I suspect that most any NPN transistor will work at least reasonably well.

Matthew D'Asaro

Yes it is silicon, metal case TO-18. In this application it has 17vdc on the collector, input signal on base, output on emitter. It's compliment on the other side of the output amplifier chain is a 2N4035 configured the same way - except for -17vdc on the collector. Perhaps the NPN equivalent of the 2N4035 will be my first choice - if I can figure out what that is.

Thanks!
Steve

A 2N3302 looks like a pretty good match for an NPN version of a 2N4035.

I have a boxful of 1854- parts but not that one.

Presumably Motorola selected from some ordinary transistor (like a 2222) and put the HP number on them, to save HP the work of culling. I'd try one (or whatever you have on hand) and see if you get lucky.

Alan's suggestion is a good one. I didn't go back to the 2n2222 but if I had, I would have seen that it is also a good match for an NPN version of the 2n4035. A little higher power rating and a little lower Frequency response but pretty good. The 2n2222 should be easy to find as well. The 2N2222A has a better frequency number but it probably doesn't matter that much.

Steve,

I have a bunch of new 2N2222A's in the metal TO-18 can. PM me your mailing address and I'll send you a couple of them gratis.

{over-size pic deleted - resized pic appears later in this thread - Leigh}

This is where the problem is in my 3312a.

At TP10, I should be getting about 24v p-p, and I only get 1.6v p-p.
All Vcc voltages on the collectors are fine.

At the bases of Q507 and Q508, I get around 30v p-p, but nothing at the emitters. I concluded these two transistors were bad, so I pulled them and replaced Q507 with a NTE123A and Q508 with 2N2222.

I fired it up, and was getting a big noisy signal at the emitter of Q507 which was decreasing quickly. As I stared at board for a moment, R531 and R533 were going up in smoke. Test aborted.

Now I could use some help diagnosing why these resistors smoked. Should I be looking at Q509 and Q511 now?, or perhaps the two diodes connecting their bases together?

Thanks, Steve

Your schematic has an awful lot of white space: it's 2859 pixels wide.

Sorry about that - I learn as I go.

I suspect that Q507, Q508, Q509 and Q511 started overheating immediately which is why the signal started to shrink. Eventually they may have shorted out. Check all of those diodes in the picture. They are part of the bias networks. If either string of diodes is open, bad things will happen. You should also check Q507, 508, 509 and 511 to see which ones may have been fried in the process.

If you get it all back together, next time you turn it on keep a finger on the transistors. As soon as you feel it getting hot, turn it off. Don't wait for smoke. If it is overheating, you will have to hook up your meter, turn it on long enough to get a reading, and then turn it off. Record the base and emitter voltages of those transistors.

By the way, delete the pic reference in your first picture post or change it to point to the second picture. People hate these wide threads and you won't get as many responses.

How do I delete the pic reference Tony?

Edit your post. On the top right corner of every post you make is an EDIT button. Click on the EDIT button and change the link in your post to point to the new smaller picture, or remove the link entirely.

I deleted it for you

I'm going to dig a old post since a search for this part on Google lead me here... for future reference the transistor in question is a 2N3904. I was searching for the same info, and finally looked in the HP parts list for a 8552 log amp board I am working on. (First place I should have looked.) What is odd is that the same part may not show the actual part number depending on what equipment it was used it.

In general, I've noticed that part cross references that are missing in the Sphere lists often are found in HP service manuals.

Pete

In the early 70s, complimentary pairs were very commonly used, the most common pairs, depending on power output were, 2N3904/3906;2N2222A/2N2907;2N3055 and Mot MJ2955. There were other pairs to be sure, as well as those for consumer products, but for Hi-Rel stuff, those mentioned were typically at the top of the list.

What a lightweight! Here's ALL the transistors listed under the NSN for the 1854-0351, available 24/7 at WEBFLIS (a DoD Logistics product......works well with Tek/HP/Fluke and Wavetek unknowns...)

EDIT: CLEANED UP THE LIST....

A127142-03T
JAN2N3904A
JHP133-833
PS101436
P00347001
P00347099
Q0247
Q0247
RELEASE5128
SF93007
SM-A-595830-6
SM-B-606019-3
SM-B-632280-2
SMB596115-4
SMB596133-2
SMB605951-2
SMB618102-2
SMB618297-2
SMB618852-2
SMB692746-3
SMB692851-3
SMB692989-3
SPS3319
SS2076
TG47
V131494
007003100
033055
0402-3904
084-20115-056
1-958000-101
10000080
101000210
101377
101377
10212257-1
10508100
110-334-16
112-0100-001
12047-0062
13B23904-20
13222E1019
15-09524-00
150-0390
1509524-00
151-0190-00
151-0190-02
151-0190-02
152008
16762172-002
174-70464-43
1824-0215
1854-0205
1854-0215
1854-0351
1854-0352
1854-1028
1854-1177
2N2920A
2N3904
2N3904A
20-034F00M00R00
20-701019
24146
253082-4
294-312
31-3904-0000
310003
330-002
345-212-013
38239040-1001
400345-01
4820-3904
505-094
531017-002
531017-00248
59-001-0001
60141-1
603-T92N-3904D
60414-1
654004
731421-02
8-729-139-04N
803631N
82-10003T
847102-0070
847102-0070
853037
914046M
914046M


BTW, SPS3319 is the contract transistor to be marked 4-0351 for 28480 (HP) by Motorola (04713) - all others are "growth" numbers within the NIIN, meaning the equivalency was established either by DSCC or another organization somewhere in the last 39 years. Note the many HP equivalents too......

As most of the old timers know, the major lab grade equipment builders did a lot of hand picking and testing from semiconductors that existed, or had the manufacturers do that for them. Same thing was done with tubes actually. Lots of times the hand picking would be to guarantee the absolutely best performance at the absolutely most difficult set of parameters the instrument would ever be called upon to produce, hence the hand selecting.

Think matching tubes for a push-pull circuit.... same thing with transistors. Or a single one hand picked because it's upper or lower frequency response was better than the actual part specification. Or stability .. or or or ... something mysterious. Or long term drift over time, or temperature range, or sunspot activity, or tidal turbulence. Or you name it.

Or..... HP and Tek just did this so people would have to buy parts from THEM at outrageous prices. Although sometimes the factory parts were surprisingly low priced. Not often though.

Selected transistors show just that within WEBFLIS - and Tek delineated a "selected" device with a suffix of > -00. All base equivalency with Tek ended with -00 for semiconductor devices. Matched devices had a different prefix, usually 153-

-00 = Base equivalency (151-0188-00 =2N3906, nothing required beyond base specs)
-01 = Selected device - their Semiconductor Reference Guides (in-house, but loose in the interweb..) detail the selection criteria, usually a tight spec in Vbe or Gain.

Fluke tried to keep selection to a minimum, but most of their JFETs have a selection criteria.

It's HP Zener diodes that usually were HP-only due to 3-decimal place accuracy. 5.81V Zener, anyone?