Wanted - Tektronix SG5010

Les Orr

If you have a Tektronix SG5010 that you would like to sell, please contact me so we can discuss it. Thank You.

Tek 2445B Display problem

Rick_T <hstjepkema@...>

I have a tek 2445B that shows a display problem:
The intensity control does not seems to work ok and the display shows 4 x 2 dots in the upper halve of the display and the same pattern in the lower halve of the display.
The dots are seperated by 2 divs in horizontal direction and 1div in vertical direction.
Also some digital noise between the dots is vague visible.

I already replaced all caps in the power supply and the voltages are ok.
Also the damage caused by leaking caps on the A5 board has been taken care of.
My final thought was that the Z-Axis IC (U950) could be the problem. So I bought a spare via ebay and replaced it.
But no luck. The problem persists. If you use the scope in XY mode, the dots are gone.
Any new idea's to solve this issue are really appreciated!

PS. Is there some paperclip function to add a picture to this message? Or do I just upload in under photo's?

Tektronix DSO TDS-2024


Hi Good day everyone, I've just join this forum and I am looking for some help on repairing my Tektronix TDS-2024 DSO. The problem that I am facing with the DSO is as Follows:
The scope will power on but it come up with the display reading distorted with vertical bar lines and have double image. I would like to upload some pictures but don't know how to do that on this site.
I open the scope to investigate and check some voltages and notice that there are two 35 volts coming off the PSU to the main acquisition board and one of them is missing. I think it the 35VL voltage if I am not mistaken.
Does anyone have a schematic for this model scope?

Best Regards,

Re: Tektronix 2230 - no readout



I've just had a similar issue with my 2230 - very faint and fuzzy image on the CRT, but Beam Find displays a solid rectangle. I found that R886 (180K) at the bottom of the Focus divider chain had gone open circuit for no obvious reason. You can check for this with simple resistance checks between the -8.6V rail and the -2KV Cathode supply rail. You can also check the focus voltage on pin4 of the CRT with a proper HV meter - around 1½KV is fine, but 2KV indicates a fault.

Hope this helps,


Photo Avalanche pulser, 85 volts with RC snubber updated #photo-notice Notification <noreply@...>

The following photos have been updated in the Sampling with 3S2 album of the group.

By: Charles <charlesmorris800@...>

Re: 3T77 tunnel diodes (again)


Agreed, but thread drift (if not outright hijacking) doesn't seem to stop anyone else. For example, check out the thread I started on sampling head diode replacements (and in which I protested more than once about the drift!)... Besides, a lot of useful info has come from you and Ed already ;) There's so many pages now, it might as well stay here. Maybe I can modify the title to include pulser design (after all, I WAS originally planning a fast pulser with a tunnel diode). :biggrin:

Anyway... today I tried a piece of poly RG-213 for the charge line. No noticeable effect. Also hung a 10 pf cap and 100 ohm trimmer off the end of the charge line per Jim Williams. I can tune the ripples at the top of the pulse somewhat, but they won't go away.

The avalanche still exhibits two modes, but the higher one is not as stable and has a narrow range of bias voltage. I'd like to enhance that mode of operation since the slow front slope goes away completely, and most of the back porch although there are still very noticeable ripples. Maybe I need to try another 2N2369, or 2N2501, or perhaps a BFR505 (avalanches at 30 volts, too) although those only come in SOT23 packages.

Re: Cleaning Scratchy Bourns MOD pots

Thomas Garson


How to get into Bourns type 81 or 81/82, 5/8" (or similar AB) Mod pots that do not have screws holding them together:
They are held together with four plastic rivets (pins) that protrude through the pots body from the rear cover. Their ends are melted down to "cap" the rivets. I use a Dremel tool with a small carbide burr (think dental tool) to carve the ends of the rivets away just enough to release them. The rear cap can then be removed by sliding it back. Cut off the remaining rivet stubs with flush cutters and carefully carve/drill holes in the back cover where the pins originally protruded. 2-56 screws and nuts can then be used to hold it all back together at end of operation.

I've repaired or cleaned many of these as they are often found in high end pro audio products. I went to the trouble of using my small CNC mill to machine a drill jig from a small chunk of aluminum to aid in drilling the screw holes in the back cover.

I've never run across a mod pot in which the wiper assembly was broken but its a big world....

I have used carbon paint and water thin toughened CA adhesive to repair ones that had a body crack which made the element discontinuous. (Yes: Broken mod pots! Musicians can be very tough on their gear.) Carefully apply the paint to the element prior to using the CA on the body to prevent the CA from flowing into the crack in the element. Quickly apply a small clamp to hold it tightly together after applying CA. Clamp can be safely removed after about 5 minutes. If this fix works (not 100% certain), once the conductive pain has dried the pot generally operates as new.

This is very touchy work, to say the least, but sometimes alternatives to repair are just not available: Think custom ordered factory build of PCB mounted multi-gang pot assemblies.

Thomas Garson
Aural Technology, Ashland, OR
By my calculation, the dynamic range of the universe is roughly 679dB,
which is approximately 225 bits, collected at a rate 1.714287514x10^23 sps.

On 3/13/21 12:56 AM, Martin wrote:
Hi all,
I'm repairing a PS503 module. It has two of these MOD pots, for course and fine voltage control.
Mine do not scratch, but one of the fine controls isn't connected to the slider anymore. When I unscrew the knob, the axis disappears inside the outer axis...
Has anyone experience with this type of mechanical failure? Can it be repaired??

Re: Modifying W Plugin to use 6DJ8 Tubes

Dave Wise

Correction to previous post number 179890. With 8416 I see 28MHz not 33MHz. I was counting graticule minor divisions wrong.

Re: 475B intermittent horizontal


On Sat, Mar 13, 2021 at 12:17 AM, Ondrej Pavelka wrote:

It's a brain fart
Hi Ondrej,
Just couldn't help myself...
As we're all aware, errors in subject titles, especially wrong model numbers, can be annoying, often negatively affecting future searching. Not applicable in this case of course.


Re: 3T77 tunnel diodes (again)

Albert Otten

On Sat, Mar 13, 2021 at 04:24 AM, Ed Breya wrote:

A very good more modern (only forty years or so old) was originally called a
Bratz generator inside Tek, then later adapted to become a TM500 plug-in. I
forget which, but someone will know.
Hi Ed ,
You mean the High Amplitude Pulse Generator which after several modifications became available as the PG509.
Fascinating stuff. Some years ago I also tried several transistor types as avalanche candidate.
Those modifications show that it was not so easy to generate a satisfactory pulse shape!

Hi Charles,
This is drifting far away from the original topic. To address a larger public you might rename this topic or start a new topic with appropriate title.


Re: Cleaning Scratchy Bourns MOD pots


Hi all,

I'm repairing a PS503 module. It has two of these MOD pots, for course and fine voltage control.

Mine do not scratch, but one of the fine controls isn't connected to the slider anymore. When I unscrew the knob, the axis disappears inside the outer axis...

Has anyone experience with this type of mechanical failure? Can it be repaired??


Re: 3T77 tunnel diodes (again)


Thanks again for the detailed information :)

I don't have any good coax (although I did try another piece of poly RG-58 without a PL-259 on the end. No change). Seems like it's time to invest in a few feet of .141 semi-rigid line! I'll try lifting one emitter resistor first.

No BNCs, only GR874 connectors in the fast signal path. The S-4 has an SMA input, but I don't have an attenuator with SMA connectors (and its unobtainium diode hybrid assembly is more electrically fragile than the S-2 or S-1).

I tried the series resistor, ferrite bead, and reducing the coupling cap to 5 pf. Basically no change to the pulse artifacts. But I did discover a very interesting phenomenon - the pulse is not on or off... two basic modes depending on the adjustment of the avalanche power supply. The pulse amplitude gets significantly larger (almost full screen i.e. 16 volts!) and the back step becomes smaller, while the ripples get larger. (Photos in album). Not only that, it avalanches with less voltage (down to 65) although the pulse becomes smaller and the shape "spreads". Whereas before I needed 90-100. Does that mean anything useful?

A Tale of Two Hypcons - Repair Report 7904A and Signal Standardizer

Tom Norman

There are a few posts here in TekScopes about the Tek Hypcon connectors, mostly around cleaning. It may be luck of the draw, but I've now had a total of 6 items that were not functioning properly, and were ultimately fixed with the disassembly and cleaning of a hypcon connected item.

A recent 067-0587-02 signal standardizer purchase had an issue with the bandwidth check leveling circuit fixed this way, as well as an apparently thermal issue with the vertical channel switch in a 7904A. The latter presented as a fairly rapidly decreasing gain on the right vertical plugin slot, which compressed the gain check display of the standardizer by about half over the course of the first 30 to 60 seconds of operation. The issue was with the scope, not the standardizer as the display was fine with the standardizer in the LH slot. After much use of IPA and ultimately freeze spray (to no benefit), I decided to just clean the dang hypcon, board and channel switch hybrid. This fixed the issue. The peak detect hybrid connector was at fault with the leveling on the standardizer.

So if you are experiencing apparently thermal related issues in an instrument that makes use of the hypcon connected hybrids, it might be reasonable to suspect the connector in addition to other discrete components or the hybrid itself. So far, I've been relieved to find that the elastomer used in the connector appears to not be degrading at this point ... at 35+ years of service.

While on the topic of hypcon connectors, I got some advice/help from John Addis while trying to fix an 11A34 amplifier. During that repair, he mentioned that some of the earlier hypcon frames had some issues with bending, that would unload some of the connections in the connector. This apparently was fixed by adding nylon to the recipe for the frame, and his email seems to indicate that some of these may have made it into the wild (not confirmed). So if you encounter a particularly troublesome connector, try swapping them out (if there are others in the unit). It's possible that the problem might be mechanical, and might be fixed accordingly.


Tektronix gain adjuster adapters for sale on eBay provide useful connectors for 500 extensions

John Williams

These little gain adjuster adapters for setting the main vertical amplifier gain in 530 540 and 550 mainframes show up on eBay frequently and are not super pricey. I have a couple now. While not used for their intended purpose very much, they can be a source of useful connectors for making up extender cables and other times you might want a male and/or female 16 pin set. The connectors themselves are scarce and expensive new. It is easy to disassemble the adapter and reuse the connectors in some other application. Not a bad bit of kit to have in a drawer.

Re: 3T77 tunnel diodes (again)

Ed Breya

When you're looking at the app notes, keep in mind that the transistor is directly driven from ground level, affecting both turn on and off characteristics. A more complicated but much cleaner way is to use transformer drive. You should look at other designs too, for reference. The Jim Williams one is just one of many that have been done over the years. The oldest charge line-avalanche pulser as a product I ever saw was from the 1960s, from an outfit called Millivac, as I recall. I got it long ago, but have lost it over the years. It was capable of 1 nSec pulse width, and got wider as charge line was added. A very good more modern (only forty years or so old) was originally called a Bratz generator inside Tek, then later adapted to become a TM500 plug-in. I forget which, but someone will know.


Re: 3T77 tunnel diodes (again)

Ed Breya

Hi Charles,

The simplest simplification is to view the transistor as a constant resistance in avalanche. It's quite a bit more complicated than that, since you can also view it as a constant voltage drop in certain conditions, with very low effective on resistance. In reality, it's somewhere in between, and variable. Either way, there's a mismatch at both cables. One thing you can easily try is to temporarily lift one of the 100 ohm emitter resistors - and be ready for a much larger signal - and watch what happens at the end of the pulse. That tail step may get relatively smaller, or may even go below zero. If the tail does not get relatively smaller, then suspect the charge line. A poor line will have too much loss. Remember the reflection is always reduced compared to the forward (it has to make two trips), so can never cancel it perfectly. You should see a remarkable difference in going from "regular" coax to the good stuff, especially if you use a fairly long line. I wouldn't be surprised if that 20% you mentioned represents mostly cable loss. So anyway, try what's easier first - if you have some good cable (I always use 141 semi-rigid for this kind of stuff), then do that right away. Also, no BNCs at these edge speeds - SMA or hard-wired is best.

You can experiment with various shunt termination and series resistors in different places, and gradually figure out how to pad out the mismatches to where they're acceptable. Note that the same pattern is repeated following the first tail, but smaller. That's a series of the same reflections back and forth through the entire system (virtually forever), but they eventually get too small to see. The trick is to tweak things up to get the cleanest cancellation from the very first reflection, then the rest will quickly diminish. It can actually be a lot of fun to experiment on this sort of thing - until it becomes a chore.

Once you get the reflections working right, you can work on the edges.


Re: 3T77 tunnel diodes (again)


Also, I re-read AN72 (the other Jim Williams fast pulser application note). He put ferrite beads on the avalanche transistor base feed, a 100 ohm series resistor, and only a 5 pf coupling cap, and his very short (1.2 ns) pulse has that back porch too, although much less noticeable than mine. I think I'll play with that area next.

Re: 3T77 tunnel diodes (again)


Another look at AN94 shows that the large reverse base current during the avalanche may be important to the drive circuitry. I used a 7404 hex inverter through 10 pf to the base, which has a 12K carbon comp resistor to ground.
Jim Williams found that that reverse current caused noticeable artifacts in the output pulse, even after putting ferrite beads on the connection to the base. I also don't have any Schottky diodes handy, or I'd add one as he did. Lots of possibilities here besides the PL259!
RF stuff always reminds me of Osborn's Law: Variables won't, constants aren't.

Re: 3T77 tunnel diodes (again)


Thanks for the tips, Ed.

I am always concerned about blowing sampling head diodes, especially since they are a RPITA to replace without OEM parts available. I did successfully replace the diodes in an S-1 with SOT-323 packages and the response time seems at least as good as stock. That will be harder (electrically) to do on the 75 ps head, and of course the 25 ps S-4 uses a very different technique and I lucked into the last hybrid assembly that Walter had. Anyway, I always put my GR 10X attenuator in front of it...

I had assumed that the small ripples in the 5 ns/div picture (Avalanche pulser, in Sampling with 3S2 album) were due to various mismatches and parasitic inductance/capacitance. Jim's design includes a very small 1 turn inductor between the charge line and the collector, and a trimmer cap + resistor snubber from collector to ground. These were undoubtedly determined by experiment, as you also noted. At this point, I'm primarily interested in the fastest possible rise and fall times for testing my sampling heads so some irregularities in the flat top aren't important (yet...) ;)

But this is a large nearly level step at exactly 20% of the pulse height. Would the mismatch situation you describe be that severe? It occurs to me that it might be that dangling PL-259 connector... they're actually about 36 ohms, not 50. Plus some stray C...

Re: 3T77 tunnel diodes (again)

Ed Breya

Charles wrote: "Not sure where that reflection step in the falling edge is coming from - is that an artifact of the open-end coax charge line? Plenty of (small) ripples too. I have two 100 ohm resistors with very short leads at the 2N2369 emitter, and of course everything else is 50 ohm (about 3 ft of coax to the sampler head)."

The problem with avalanche pulsers is that nothing is truly 50 ohms throughout the cycle. Each piece has a certain resistance at a certain time. If you're using a classic charge-line and avalanche transistor setup, you have a 50 ohm line at the collector, for twice the transit time, when the reflection from the open end ideally turns the transistor off. The transistor in avalanche typically has around 50 ohms (maybe 10 to 100, and not necessarily constant) on resistance. When no pulse is being delivered, the load sees the 50 ohms at the emitter, and the emitter resistor sees the 50 ohm output cable to the load, and the load is 50 ohms (all ideally speaking), so fine and good, when nothing's happening. Now fire the transistor, and it becomes say 50 ohms, so you have the charge line with say 50-60 V on it, looking into the transistor's 50 ohms plus 25 ohms (the emitter resistor parallel with the output cable), or about 75 ohms - a mismatch, but not too bad. The load sees the emitter resistor paralleled with the transistor plus the charge line, so 50 parallel 100, or 33 ohms - again, a mismatch, but not too bad. There's no perfection available in this simple circuit, but you can clean it up a lot by tweaking the values and circuit in a number of ways, to pad out the mismatches to some extent. It should be done experimentally.

The nice thing about avalanche pulsers is that there's usually plenty of signal compared to what's needed (or safe), so you can afford to make some heat, and get a more civilized output. The large raw output can be dangerous to sampling and microwave gear inputs, so be careful to attenuate the final output to a safe level, considering that a simple change in the circuit, parts or conditions can cause a very large level.

The charge line should be RF/microwave grade - silver plated conductors and PTFE dielectric, for best results. You can make it tens of nSec long if desired, for longer pulse duration. You can run faster by having less line charging resistance, and more DC power available. The transistor power dissipation will be the limiting factor - the longer the pulse, and the higher the rep rate, the harder it has to work.The transistor may carry a half an amp or so, well over 10 W while in avalanche. The duty factor needs to be low enough for it to not overheat and burn out. If you're using a TO-18 package, it will dissipate better. Don't check the temperature with the fingertip method while it's running - you'll get the wrong impression of the temperature, and maybe an RF burn too.


6261 - 6280 of 186256