Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
- TekScopes
- Messages
Search
|
Re: Calibration and full checkout needed - Tek 7000 series
John Griessen
On 5/4/19 8:44 AM, David Berlind wrote:
For horizontal, it sounds like a 7A22 and 7A26 would be good to throw on the mix. For the vertical, I have 7B53A's. Also, why would a 200Mhz 7A26 make sense in a 100Mhz 7603 (if it does)?There are plenty of 7A26's out there, so skip that rare cheapo plugin. because I don't need storage which is also why the 7633 goes too, vs one of the 7603's).7633 is valuable for its variable persistence. |
|
Re: Again a 7854 power supply beast to be repaired...
Hi Dave and all,
I made some progress! Found Q143 shorted. Replaced with another 151-0220-00. Now the trace is focused but turning the intensity control fully ccw the trace do not disappears but remains somewhat brilliant. Checking waveform at TP143 and following the service manual setting (H plugin set to 2us/div) I see a negative pulse train whose amplitude varies turning the B intensity control cw or ccw. But I do not observe the square wave as shown in the schematic. (waveform 3). Moreover, the trace seems to be chopped randomly and turning on the readout it appears as random vertical bars moving around. Seems that something is not synchronized? 130 V supply look good. When powering up the scope, I observe a bright point at the extreme left of graticule, after which it behaves as described. I've not yet checked Q173 base voltage. As soon as possible I will check all around Q173 if you think it's worth it. Thank you all for help and suggestions up to here! Max |
|
Re: Again a 7854 power supply beast to be repaired...
If the +130v supply is good, check the voltage on Q173's base. It should be 82vdc.
toggle quoted message
Show quoted text
If it is low check C171 for leakage. If C171 is OK, check CR177, CR173, and Q173 for leakage and shorts. Then check all the R's around Q173. Dave 5) With another scope I checked the output of the Z-axys board (board A21, |
|
Re: [OT] Best glue to repair lifted trace?
I use Circuit Works CW2500 epoxy to repair traces on a PCB and overcoat the result.? Its a good epoxy but somewhat expensive.? All epoxy "glues" will melt when soldering so plan accordingly.? In your case I would 1.? remove the connector, 2.? use CW2500 to put the trace back on the board, 3.? scrape mask off a strip of the conductor after its repaired, 4.? solder a small wire or strip of copper tape to the repaired trace, 5. Put the connector (better, a new one) back on the board, 6. use CW2500 to overcoat the repaired trace.
toggle quoted message
Show quoted text
Amazon and Newark sell CW2500 but I have no idea what the shipping might be to Austria. On Saturday, May 4, 2019, 02:06:46 PM CDT, cheater cheater <cheater00@...> wrote:
Good warning. Thanks. On Sat, May 4, 2019 at 6:20 PM Mark Goldberg <marklgoldberg@...> wrote:
|
|
Re: [OT] Best glue to repair lifted trace?
Good warning. Thanks.
toggle quoted message
Show quoted text
On Sat, May 4, 2019 at 6:20 PM Mark Goldberg <marklgoldberg@...> wrote:
|
|
Re: Again a 7854 power supply beast to be repaired...
OK Harvey,
I did as you suggested, putting a V and H plugin. Here the results: 1) Trace is moveable in both horizontal and vertical position. So I put the trace off-screen to avoid CRT damage. 2) Readout is visible. 3) All the display is highly blurred 4) Focus and astigmatism works but in a barely manner because the the display is at very very high intensity. 5) With another scope I checked the output of the Z-axys board (board A21, diagram 13) test point TP183: the output is stuck at approx. +120V. No pulses or other artifacts are visible. Moving the A, B or readout intensity has no effect. Max Max |
|
Re: Again a 7854 power supply beast to be repaired...
On Sat, 04 May 2019 09:22:37 -0700, you wrote:
O.K. I checked power rails and found a short on -15V, pin 7 of P82 connector (sense).Check to see if the astigmatism control works, if focus works (do not focus down to a full intensity spot). Do this with at least a V and H plugin. That's a sanity check on the V and H frame amplifiers. You ought to see readouts, too. Spot should be moveable with the V and H controls, so you can move it off the screen while you check the power supply voltages (just because it comes up doesn't mean that there isn't still a bad cap somewhere). The good news is that you have high voltage on the scope. The question is if it's right, and the control voltages to the CRT are ok. That's what I'd do to start with. Harvey
|
|
Re: Again a 7854 power supply beast to be repaired...
Craig Sawyers
Oh dear - the dreaded one pin out problem. We've all done that at some point. I've got away with it
toggle quoted message
Show quoted text
until I did it on my 577 curve tracer and took out most of the silicon. Fixing it is a "work in progress". Hopefully there won't be too horrendous damage. Craog -----Original Message----- |
|
Re: Again a 7854 power supply beast to be repaired...
O.K. I checked power rails and found a short on -15V, pin 7 of P82 connector (sense).
NOTE: some other people was attempted to repair this poor 7854, so...they left behind a disaster with no success. After careful inspection I found the P132 connector on the motherboard mounted wrong, shifted one pin left! I pulled out P132 connector and reinsert it in the correct way, carefully matching all the pins. Power up and... woah! It started! All the lights up! BUT: now the bad surprise. On the CRT I see a BIG unfocused and very, very brilliant spot centered on the screen (no plug-ins are inserted). The INTENSITY control has no effect (fully counterclockwise). To avoid damage to the crt I turned it off immediately. I turned the scope back on, just to perform a quick check on the low voltage regulator board A22: all voltages are perfect, all in range. And now? Z board damaged? :-( Where to check? Max . |
|
Re: [OT] Best glue to repair lifted trace?
Silicone will rip with mechanical stress. Epoxy is much stronger.
Regards, Mark W7MLG On Sat, May 4, 2019 at 8:56 AM Brad Thompson <brad.thompsonaa1ip@...> wrote:
|
|
Re: [OT] Best glue to repair lifted trace?
cheater cheater wrote on 5/4/2019 10:15 AM:
That's the plan, unfortunately I can't find an acceptable source ofHello-- Do you have automobile-supply stores in Austria such as this one... <> If so, you might investigate whether they sell high-temperature silicone gasket sealant such as... This might do the job. 73-- Brad? AA1IP |
|
Re: [OT] Best glue to repair lifted trace?
I really don't understand this insistence on 'high temperature' epoxy, as I said, re-seat the connector and if it has no retention itself then clamp it or put a bit of epoxy away from the solder area to hold it , then solder, then epoxy.
toggle quoted message
Show quoted text
On 5/4/2019 3:15 PM, cheater cheater wrote:
That's the plan, unfortunately I can't find an acceptable source of |
|
Re: Calibration and full checkout needed - Tek 7000 series
On Sat, 04 May 2019 09:44:43 -0400, you wrote:
John and Harvey, this really great information I wish I had BEFORE I gotWith the 76xx series you do need a dual timebase plugin, since there are only 3 slots. You're mixing up the letters. For 7000 series, the nomenclature goes as follows (not all inclusive) and there's an exception. 7/letter/two numbers/optional letter N 7 for 7000 series A for vertical plugins B for horizontal plugins CT (the exception) for the curve tracer (7CT1N) D for digital L for spectrum analyzers S for sampling T for timebase for samplers (7T11) the N suffix indicates that readout circuitry was deleted from the plugin. Bandwidth limiting can be useful. The 7A22 is also differential, which gives additional capability. If you want full bandwidth 75 Mhz, then the 7A15 is useful, Because you get the scope frame limiting the bandwidth, not the plugin. You should get the full 100 Mhz frame bandwidth. The further advantage is that if you ever go to a faster mainframe, you don't need to buy plugins for it (or not as many). I'd recommend keeping at least one storage scope. See the thread about the 7854 power supply. The advantage of the 466 is that it's more portable than any 7000 frame, or, shall we say, it was designed for it. If you want the versatility of the 7000 series and the storage capability, I'd keep the 7633. The third rule of acquisition states that it's easier to keep something than to try to find one again. You'll find that having spares, say you have a lot of 7A26, and still have some 7A18, well, the 7A18 can be be the "if something else breaks....." plugin. Since it takes a scope to fix a scope (very frequently), I'd not be without a spare. I can see situations where, if working on two things, you'd like a separate scope rather than have to tear down and rebuild a setup. I generally work in two or three locations, two of the locations have scopes, all have power supplies, two have meters. One is the full lab. I'm working on microprocessor based, FPGA enhanced, sensor and graphics platforms, so the power supplies are needed, scopes may be, meters can be, etc.... Might be a nice gift to a maker group, but not the ones that charge high admissions fees... (those IMHO are strictly profit centers, and are more interested in making money than providing facilities...). If you can find someone who is really working on electronics in high school or college, that could help as well... Can't think of a thing wrong with that. Some people adore the old tube scopes, and if you have one that works, they're pretty rare. They get bought at high prices and robbed for the tubes. Had you considered making a small "history of oscilloscopes" exhibit? This restricts you to the 7000 series for right now, but yes, there are ways. Note that the signal standardizer substitutes for a constant amplitude signal generator for the frame, but not for any plugin. Pay attention to the minimum frequency limit on the standardizer, they mean it. You are going to get close. As pointed out, any DMM is better (more accurate) than a scope. So you could use a decent DMM as a DC voltage "standard" to at least look at voltages on the scope. While you'd like the 1-2-5 sequence that the PG506 provides, if you can find a decade divider, that would allow you to relatively easily dial in known voltages (from a known voltage) to check most of the ranges on a plugin. You could also simply make a series of voltage dividers with 1% resistors that would give you a convenient sequence. You're pretty much calibrating the scopes to the DMM, not elegant, but should be good for performance verification. You'd want a decent signal source, DC and AC, though. Timing? Assuming you're happy with working in digital, buy a packaged oscillator, 5 volts is fine, 50 Mhz is ok. Divide by 10 with decade counters, use 7490's (which are /5 /2) wire them as /5 first, then /2. not that it's super important, but it makes the /10 composite output symmetric. What you're looking for is to divide by 10, then take that div10 output and tap off a div2 and div 5 for each. Doing that, you could also use 74LS192 decade counters, and that gives you the requisite 1,2,5 timing for most scopes. That handles horizontal timebase stuff. So yes, you can make stuff if you want. These won't be as nice (or as full featured) as the stuff made for the scope calibration, but they'll easily do performance verification. If you had a synthesized signal generator, it would substitute for the TTL stuff. All depends on what you have and are willing to make. Harvey
|
|
Re: Again a 7854 power supply beast to be repaired...
On Sat, 04 May 2019 07:22:34 -0700, you wrote:
Good news!Overload. You had bad caps on the power supply, you likely have bad caps in the scope. 1) remove as many connections to the PS as possible. Check resistance from the power lines to ground. 2) if you can isolate to a section, check that one out, of course. Do all of this without plugins. 3) you can try plugging in each section to see if it holds down the supply. 4) another scope can be used on the rails to see if the power supply rail comes up or stays flat. 5) digital scope is by far the best for this, if you have one. Divide and conquer works here. Check individual boards (with power disconnected) for ESR and resistance on electrolytics from power to ground. Be especially suspicious of the little gum-drop tantalums. If they have a substantial series resistor (100 ohms or better, they really won't hold down the supply unless it's a significant voltage). Harvey
|
|
Re: Again a 7854 power supply beast to be repaired...
Hi
toggle quoted message
Show quoted text
A regular ticking means that you have a short on one of the power rails in the mainframe. George G6HIG On Saturday, May 4, 2019 3:25 PM, unclebanjoman <mmazza@...> wrote:
Good news! I found another couple of leaky electrolytic. So I decide to change ALL electrolytic capcitors on inverter and control/rectifier board, using Nichicon PWM-R 105 deg (low ESR). Replaced the two big capacitors 950uF 200V (one of these was severely leaky) with new Nippon Chemi-Con 1000 uF 200V. I built a decent dummy lod using the Dennis Tillman's scheme. Finally the SMPS started up! All voltages was O.K.! Start-up is practically instantaneous and the the waveforms are in good agreement with those reported in the service manual. Now the bad news..... I reconnected the SMPS to the mainframe and... nothing happens. No lights, no trace, nothing. I hear a regular tick tick (it was irregular and random before the cure). So I'm sure that the SMPS is correctly functioning but the service manual does not explain much about how to proceed in this case. Any idea/suggestion? Max |
|
Re: Again a 7854 power supply beast to be repaired...
Good news!
I found another couple of leaky electrolytic. So I decide to change ALL electrolytic capcitors on inverter and control/rectifier board, using Nichicon PWM-R 105 deg (low ESR). Replaced the two big capacitors 950uF 200V (one of these was severely leaky) with new Nippon Chemi-Con 1000 uF 200V. I built a decent dummy lod using the Dennis Tillman's scheme. Finally the SMPS started up! All voltages was O.K.! Start-up is practically instantaneous and the the waveforms are in good agreement with those reported in the service manual. Now the bad news..... I reconnected the SMPS to the mainframe and... nothing happens. No lights, no trace, nothing. I hear a regular tick tick (it was irregular and random before the cure). So I'm sure that the SMPS is correctly functioning but the service manual does not explain much about how to proceed in this case. Any idea/suggestion? Max |
|
Re: [OT] Best glue to repair lifted trace?
That's the plan, unfortunately I can't find an acceptable source of
toggle quoted message
Show quoted text
high temperature resistant, electrically insulating epoxy in Austria. The closest is in Germany, but that's still 10 euro shipping. On Sat, May 4, 2019 at 2:54 PM Tony Fleming <czecht@...> wrote:
|
|
Re: Calibration and full checkout needed - Tek 7000 series
John and Harvey, this really great information I wish I had BEFORE I got started acquiring stuff (though all is not lost). To be clear, in terms of acquiring additional scopes or plug-ins, today I already own 2 7603s, a 7603R, a 7633, a 466 (to Harvey's points about the 46x family), a 547(1A4), and a T935. My vertical 7000 plug-ins are a mixture of 7A18's (which get me to 4 channels on a 7000) and 7A15's (the military version). For horizontal, it sounds like a 7A22 and 7A26 would be good to throw on the mix. For the vertical, I have 7B53A's.
toggle quoted message
Show quoted text
I'm curious about your thoughts on the 75Mhz 7A15. Its sensitivity isn't the 10uV of the 7A22, but with its 10x switch, it can apparently go down to 500uV. Also, why would a 200Mhz 7A26 make sense in a 100Mhz 7603 (if it does)? I love the Hantek idea. For DMMs, I have a Fluke 87V and Agilent u1253b already. So I have no immediate needs there. Ultimately, I want to thin and enhance (ie: plug-ins) my herd down to only that which is necessary (and good enough calibrated) for my applications (again tube amps and radios). At 100Mhz without the flexibility of the plug-ins, my feeling is the 466 is out (also because I don't need storage which is also why the 7633 goes too, vs one of the 7603's). I haven't played with the 35Mhz T935 yet but really see no need for it. Portability is not an issue for me. Technically, the 50Mhz 547 seems unjustifiable to keep. Although the TekWiki mentions a cascading option that, if I read it correctly, gets its sensitivity down to 200uV of sensitivity (not as good as the 7A22 for low freq, but better than the 7A15). But I haven't read into it yet. Maybe I have it wrong. In the end, I just may keep the 547 for its sheer beauty. Finally, the big question (bringing it back to the thread) for me is how badly do I need the stupid money TM500-related calibration gear or is there a short cut with something else that gets me close enough for my application now that I have the signal standardized plug in. As always, thank you for your feedback. On May 3, 2019 9:48:48 PM "John Kolb" <jlkolb@...> wrote:
If you go with the 7000 series mainframe, I'd recommend the 7A26 and |
|
Re: [OT] Best glue to repair lifted trace?
Aha! okay, it looks like the power pins are surface mount so there must be something designed to help add mechanical retention to the PCB as well?
toggle quoted message
Show quoted text
Either way looks like a real easy thing to fix as it aint exactly a fine pitch multi-pin connector! I would say step one is to clean things up which I think involves (carefully!) unsoldering the remaining pin from the trace so you can get to things. Then I would scrape the solder resist off the copper over the via area(s) and tin it ready for the repair. If there are retention features, and they are not busted that will make things easier to re-seat the connector body to the board, if not or if they are broken, you could try and find a replacement from Farnell/RS/Digi/whoever but failing that use a bit of epoxy to secure it enough to allow you to repair and re-solder the pins to the traces. I would be using copper foil for that, tin and sweat it on. I have found that the tacky copper foil sold for RF shielding works well for this as (to my surprise) the glue does not stop you getting a nice bead of solder round the edge and the tackiness holds it in place as you do it! Then add more epoxy to secure the body if you think it needs it and yopu're done. I really don't think you will find it too difficult to do. Good luck! Adrian On 5/4/2019 1:23 PM, cheater cheater wrote:
Here you go! |
|
Re: [OT] Best glue to repair lifted trace?
Glue it back - remove all debris first.
toggle quoted message
Show quoted text
After that I would reinforce the connections with a wire - solder onto existing traces. I'm thinking that there is a problem with the connector - why would it brake in the first place? On Sat, May 4, 2019 at 7:23 AM cheater cheater <cheater00@...> wrote:
Here you go! |
to navigate to use esc to dismiss