5. Other Tektronix Repairs, Modification and Teardowns
INSTRUMENT REPAIRS & MODIFICATIONS (IN ALPHANUMERIC ORDER)
Action Status Problem Model Function
Repair Fixed Bad power lamp 285 Standalone PSU for 7000 TDR S-series plugins
Repair Fixed Bad interlock 571 Digital Curve Tracer
Repair TBD Erratic display 577 Analogue Curve Tracer with 177 Test Fixture
Repair TBD Bad BNCs/DIY batt THS720P 2-ch 100MHz 1GS/s Isolated DSO
Repair TBD Ch1 V pot broken 2455B 4-ch 250MHz Analogue Oscilloscope
Repair TBD No graticule lamps 7104 1.0GHz Mainframe 4-bays
Refurb TBD Backup eproms 7854 400MHz Mainframe 4-bays
Repair Fixed Bad tantalum caps 7A13 100MHz differential comparator, ±10V
Repair TBD Gain maxed out 7A29 1.0GHz 10mV-1.0V 1-channel #5
Repair TBD Dead 7A29 1.0GHz 10mV-1.0V 1-channel #6
Repair TBD Pot shaft broken 7A29P 1.0GHz 10mV-1.0V 1-channel (7912AD programmable)
Repair TBD 2ns shows as 500ps 7B15 1.0GHz timebase with delay, delta readout
Repair TBD Bad trigger 7B87 400MHz timebase / trigger / 7854 1-shot Pre-trig clk
Repair TBD broken contacts 067-0589-00 Rigid Extender for 7000 Series Plugins
Repair Fixed Bad Tr (loose skt) 7S11 Sampler ~50kS/s, 2-200mV/div
Repair TBD Temporal mismatch? 7S12 TDR/Sampler, 20ps/div to 1µ/div
Repair TBD Ch1 noisy Ch2 trig 7S14 Dual Sampler ~20MHz, 100ps/div to 100µs/div
Repair TBD TrTf slow P6015A Probe HV 75MHz 20kV (40kVpk) 1000:1 #1
Repair TBD TrTf 12ns/bad lvl P6201 Probe 900MHz FET input #1
Repair TBD Only 10kHz P6201 Probe 900MHz FET input #2
STANDALONE
285 standalone PSU for 7000 TDR S-series plugins
https://w140.com/tekwiki/wiki/285
The only thing wrong with this was the ac supply power lamp had failed
Lacking an ac LED, I built a miniature ac-dc driven LED into the existing bulb holder.
CURVE TRACERS
571 Digital Curve Tracer
https://w140.com/tekwiki/wiki/571
I bought this curve tracer first and it has always worked well, although I wished it went up to 500V instead of a measly 100V. Mine lacked the bare, crude small transistor socketed pcb module that came with it when new. I built my own using a Textool zero insertion socket wired for 6 pinout variations.
(a) The push switches are getting a little worn now, and may need replacing.
(b) The cover micro-switch didn't engage unless you pressed down on the cover, but it was an easy fix: just a matter of rotating a mechanical adjustment screw inside the case.
577 Analogue Curve Tracer with 177 Test Fixture
https://w140.com/tekwiki/wiki/577
I bought this for avalanche transistor testing because it goes up to 1600V. (repair detail to follow)
(a) 577 PSU dried out electrolytic - replaced
(b) 577 Collector voltage variac rod snapped in transit - replaced
(c) 177 12V adjust pot broken on arrival - repaired
(d) 177 many solder joints crystalised - re-soldered
(e) Various knobs broken - repaired
(f) 177 Added custom lockable safety interlock override switch
(g) Transistor/Fet holder test fixtures are ridiculously expensive, typically around $80 - $100 plus
shipping & import. Initially I made do with DIY SIL sockets and wires to the pins, see near bottom of [References: Avalanche Transistors], but eventually I succumbed and bought one.
(h) The same is true of the safety cover which rarely comes up on eBay and when it does, is usually at least $200 plus shipping. I bought my entire 577+177 for $259 plus $69 shipping and import!
I made my own safety cover out of a $4 eBay China 100mm square acrylic open cube and added my own adjustable safety switch actuator using a nylon screw, although later I won the real deal at a still considerable $100.
THS720P 100MHz dual isolated channel DSO
https://w140.com/tekwiki/wiki/THS720
BATTERY
The original Tek THS7BAT 4-cell NiCd 4.8V 2800mAh battery that came with my 100MHz THS720P DSO still works but doesn't hold its charge for long and will need replacing soon.
The external charger was missing, but the DSO is equipped with its own internal NiCd trickle charger although replacement NiCd batteries, even clones, are ridiculously expensive even on eBay China. That said, there was nothing stopping me from dismantling it, finding out what was inside and replacing it with my own DIY alternative, perhaps using more modern NiMH batteries.
Battery University reveals NiMH trickle charges at 0.05C vs NiCd 0.1C [C is the battery amp/hr rating]: 'A NiMH in a NiCd charger would overheat', furthermore '[NiMH needs] more sensitive charge detection':
https://batteryuniversity.com/learn/article/charging_nickel_metal_hydride
I thought of building my own NiCd pack but C-size 1.2V cells are expensive these days, having been long replaced by NiMH which are only minimally affected by the well-documented memory effect and in more modern batteries, the cadmium crystal growth hazard, both of which shorten NiCd battery life:
https://batteryuniversity.com/learn/archive/memory_myth_or_fact
The THS720P manual says it maintains its signal isolation regardless of being powered by the internal battery or an external 12V 1A supply. It can take up to 15V on this so next I considered using a 12V lead acid battery, but that would be little better than a wall wart power adaptor and I'd prefer if it remained self-contained. Besides, more research revealed others had already built their own DIY packs.
Dated around June 2015, using EBL 5000mAh NiMH batteries:
https://www.eevblog.com/forum/testgear/tek-ths720a-portable-scope-teardowndiscussion/50
'...putting in a 5000 NiMH vs 2800 NiCd Amp-Hr battery is OK as we can probably assume that Tektronix put a quality charger into the scope and it trickle charges at C/10 as (the recommended optimum) That puts the 5000 NiMH at 2800/5000*.1=0.056C very close to the recommendation for trickle charging a NiMH'
Dated around February 2017, this time using 3000mAh NiMH batteries:
https://www.eevblog.com/forum/testgear/tektronix-ths710a-voltage-offset-and-batteries
(Usefully, the above thread also discusses the Cal procedure.)
'The internal charger on these scopes tops up (trickle charges) at 50mA. That's just fine for a C size NiCd or NiMH battery. Modern 3000mAh NiMH batteries can tolerate 50mA trickle without evolving free oxygen (it's reabsorbed). You don't even need to weld or solder the battery terminals for a 720 series battery pack. There is more than ample pressure in the scope housing to provide good continuity.'
'I wrote a blog article on rebuilding a 720 battery pack a while back. The key is using a 8" x 1" piece of heat-shrink tubing to hold the 4 C-NiMH batteries. And DON'T use batteries with flush Plus ends designed for tab welding. Use regular C batteries with protruding plus ends. Oh, and you still need to allow 10-15 hrs to recharge a battery since you are not changing the charging characteristics.'
You can find the article here https://www.hollywoodcontrols.com/php2/TekScopeBattery.php
I'm not sure I believe the figure of 50mA trickle current, since the manual says Tek's internal charger takes about 9 hours to charge the original 2800mAh battery: 2800/50 = 56 hours! Instead, 2800mAh/9 hours = 311mA which is a more likely 0.11C.
For future interest, the bottom left of page 15 of the THS7XX schematic [see links below] has a TL500 PWM controller rigged as a SMPSU and its output feeds into the battery input via what looks like a transzorb. Working out the circuit characteristics should reveal the actual charging current.
For my own peace of mind, I ran a few quick calculations:
If NiMH should not exceed 0.05C then strictly speaking trickle charging should not be more than
0.05 x 5000mAH = 250mAh. The Tek NiCd charger must provide 0.11 x 2800mA = 308mA, a 10% increase on the NiCd 0.1C and a 19% on the NiMH 0.05C with acceptable temperature levels noted in the link above; if the temperature increases linearly by 20%, 30°C might increase to 36°C, and I think that is acceptable.
I decided to build my own 4.8V 5000mAh NiMH battery stack using the construction details provided in the above links. In October 2019 four of the same EBL batteries are on eBay UK for ~£15 (~$20).
I also thought about building my own 4.8V NiMH 0.05C charger. 5000mAh/250mA = 20 hours which is longer than the internal Tek charger; exceeding the 0.05C charge rate shortens the life of the batteries. Battery University (above) warns of the complexity of charging NiMH. By chance I already had a NiMH charger from my jungle trip universal UV lamp/charger project. ADD TO PROJECTS SECTION The charger is a pre-manufactured evaluation board, thankfully devoid of the 'Eco' water-soluble but also electrically-conductive solder flux that had ruined the project and left it abandoned for twenty years.
Useful site with service manuals (Wix won't accept this URL, but copy and paste it to get there):
Basic Circuit description and internal photographs:
https://w140.com/tekwiki/wiki/THS720
The schematic, BOM, Service Manual and other files for the THS7XX from Tekwiki:
http://w140.com/tekwiki/wiki/THS700
Schematic direct link:
http://kazus.ru/forums/attachment.php?attachmentid=25865&d=1315575619
BOM direct link:
http://www.tequipment.net/assets/1/26/Documents/THS700_Supplement.pdf
2400 SERIES
2455B 4-ch 250MHz analogue oscilloscope - STILL TO BE REPAIRED
https://w140.com/tekwiki/wiki/2465B
I picked this up for £235 in 2007. It's a rare member of the 2400 series, of which the 150MHz 2445, 350MHz 2465 and 400MHz 2465/2447 are more readily seen. Thankfully I have not suffered the infamous U800 IC failure and it largely works well, although CRT brightness seems to affect the triggering, which can result in either a good trigger but a dim trace, or vice versa. Fortunately trigger and brightness functionality can be instantly restored by pressing the top left SETUP AUTO button. It boasts manual twin cursor measurement, and automatic measurement of voltage, frequency, time, risetime and falltime but takes a good ten seconds or more to display the results.
a) Channel 1 voltage selection has been erratic since I had too much beer one night and inadvertently swung my HAG lab chair headrest against Ch1's voltage fine control knob, severing it from its shaft. I have a replacement extension spindle, and replacement input channels are only about $24 on eBay, but the instrument is still very usable and indeed in daily use. Given the difficulty of stripping the front panel, repair is a low priority.
MAINFRAMES
7104 1GHz mainframe 4-bays
https://w140.com/tekwiki/wiki/7104
Bought fully operational and freshly calibrated, but the seller packed it badly and eBay's cheap but clumsy Global Shipping almost destroyed it. Amazingly it still worked after all of the damage. It could so easily have had a broken tube. How close it is to the calibrated state I thought I bought is unknown.
(a) The graticule lamps work but do not light up. Their PSU is embedded so deeply, repair is on hold.
(b) Fan surround metalwork crushed in transit. Removed and straightened along with side supports.
(c) Bottom rear foot bent 45 degrees in transit. Metalwork straightened.
(d) Stand missing
(e) 34CFM Fan dated 1997 was a little noisy - replaced with a good quality new 38CFM quieter fan.
At first I was puzzled why this fan blows hot air out, not cool air in but following my
[Projects: Tek 109 1-Shot & LF Mod], I can now see the benefit: a pulsating incoming air stream would cause temperature fluctuations on sensitive components, inducing noise into the circuitry.
7854 400MHz mainframe 4-bays - STILL TO BE REPAIRED
https://w140.com/tekwiki/wiki/7854
(a) Check all caps for ESR, esp. HV PSU.
(b) Cover all EPROM windows to prevent erasure from exposure to light:
http://www.amplifier.cd/Test_Equipment/Tektronix/Tektronix_7000_series_mainframe/7854.html
(c) Although it works ok for now, it would be a good idea to buy EPROM adaptors to fit the ROMS:
http://www.eevblog.com/forum/testgear/tektronix-repair-manuals/msg527971/#msg527971
'Regarding the 7854: Yes, it starts to boot, beeps a few times and some of the horiz/vert trigger
lights flash off. The Busy, ERR, I/O, SRQ and Remote only red LEDs remain on and it halts.
The Vert. and Horiz. mode selectors are unresponsive. Nothing on the screen.'
'That sounds like ROM rot. You can replace the ROMs with EPROMs.
The ROM images and information can be found here:'
http://www.davmar.org/TE/Tek7854/Tek7854-ROMs.zip
I have the older 7854 so I need to look at preserving the Mostek ROMs / copying them.
Here is someone who managed to repair one where the ROMs got damaged:
https://translate.googleusercontent.com/translate_c?depth=1&hl=en&prev=search&rurl=translate.google.co.uk&sl=de&u=http://www.amplifier.cd/smf/index.php%3Ftopic%3D175.0&usg=ALkJrhgWaafCES2YRyJB-xvPeRSaqHgyLA
More info here:
https://translate.googleusercontent.com/translate_c?depth=1&hl=en&prev=search&rurl=translate.google.co.uk&sl=de&u=http://www.amplifier.cd/Test_Equipment/Tektronix/Tektronix_7000_series_mainframe/7854-ROM.html&usg=ALkJrhjFoHQbmtvgO-YJw5JR20qkne361w
Will also need a programmer to read the originals before they die, however see this site
with alternative instructions if there is no room for adaptors (check the 7854 before ordering!):
https://vintagetek.org/7854-mostek-mkb36000-rom-repairs
This place sells a CHEAP $5 adaptor that converts a standard 2764-27512 to Mostek ROM 2364 pinout:
http://store.go4retro.com/2364-adapter
'This adaptor converts the
MOS 2364 or Motorola 68764/68766 pinout to a JEDEC standard 2764-27512 EPROM pinout. The optional
3 high address lines are available via pads complete with provision for pullup resistors (10K).'
http://store.go4retro.com/23xx-adapter
'The 23XX Adapter converts JEDEC standard 28 pin EPROM, EEPROM, or FLASH ROMs to a 2316,2332,
or 2364 socket. Since 23XX ROMs could be produced with active-high or active-low select lines,
the 23XX adapter can be configured for either signal level on any applicable select line.
The optional 5 highest address lines are available via pads complete with provision for
pullup resistors (4.7K).'
7A13 #S/N B246858
On arrival worked to a degree, +/- value intermittent, Vdiv not shown on CRT (but seems to work)
Vdiv main knob has a split down the white line and is loose on its axle former. Glued with epoxy.
C505 O/P amp SM P.112: Sprague 109 wet tant 10µF 50V on +50V supply badly corroded.
Replaced with 10µF 63V elec.
Corrosion from C505 was so bad it had dissolved pcb track from C505 + to +50V supply.
Added a wire from new cap + to +50V. Also added a wire from C505 - to C553 + (where it goes) because both these pads were badly damaged by corrosion.
C553 O/P amp SM P.112: Sprague 109 wet tant 10µF 50V on -50V supply badly corroded.
Replaced with 10µF 63V elec.
Vdiv now fixed at >1V on CRT but switch changes ok.
Main attenuation switch connector loose on O/P amp board. Pushed in.
This plugin now works as far as I can tell. Needs cal.
7A29 - 1.0GHz 10mV-1.0V 1-channel #5 - STILL TO BE REPAIRED
https://w140.com/tekwiki/wiki/7A29
(a) Gain maxed out but fails to reach top of CRT.
7A29 - 1.0GHz 10mV-1.0V 1-channel #6 - STILL TO BE REPAIRED
(a) Completely dead.
Probably the input section as they are renowned for failure.
7A29P - 1.0GHz 10mV-1.0V 1-channel - STILL TO BE REPAIRED
https://w140.com/tekwiki/wiki/7A29P
Bought very cheaply with known fault, fully operational & calibrated.
(a) level control knob plastic spindle snapped off but can be adjusted with a screwdriver.
I'll replace the pot when it really annoys me.
7k SERIES HORIZONTAL PLUGINS
7B15 plugin - STILL TO BE REPAIRED
https://w140.com/tekwiki/wiki/7B15
(a) Timebase needs recalibration (e.g. 2ns/div setting displays waveform at 500ps/div).
7B87 400MHz trigger / timebase / Pre-trig clock for 7854 1-shot #1 - STILL TO BE REPAIRED
https://w140.com/tekwiki/wiki/7B87
eBay auction listed it as Used (eBay definition: 'fully operational') but the description said it was 'As-Is' (a con in itself) and the photo showed it triggering, but this turned out to be a lucky photo of the Cal signal on Auto trigger. When it arrived the same Cal signal was all over the place on Auto-trigger (which ordinarily I would expect, were it not presented as stable in the auction photo), and there was no trace display at all on Normal trigger or Ext trigger. I have no doubt the seller was well aware of this, so photographed it on Auto. My fault for not noticing it was 'triggered' with the Auto trigger switch light on instead of the Normal trigger light.
7k EXTENDER
Tek 067-0589-00 7k rigid extender for 7000 series plugins- STILL TO BE REPAIRED
P6015A 75MHz 20kV (40kVpk) Tr4.67ns 1000:1 HV probe
https://w140.com/tekwiki/wiki/067-0589-00
This page lists the parts to make a DIY version:
DIY Tek 067-0616-00 7k series plugin flexible extender
Not actually a repair, but a cheaper version of the exhorbitant Tek version.
Based on aftermarket pcbs sold by John Griesson, who is also Ecosensory.com:
http://ecosensory.com/tek/tek_7K_flex_sales_blurb.txt
text
7k SERIES SAMPLERS
7S11
https://w140.com/tekwiki/wiki/7S11
Everything worked, but the risetime was around 300ps, which is the last thing you need on a TDR using a ~25ps risetime pulse from an S-52 pulse generator head. Inside the side covers, I found some SIL connectors weren't fully pushed home which I corrected. I found all the discrete transistor and chip pins were in discrete sockets and when I pushed on the tops of the components, they went a bit deeper into their sockets. After doing this, the slow edge disappeared and it worked as well as my other 7S11, with risetime reported on the 7854 at around 47ps with an S-6 ~30ps risetime sampler plugin. Although out of spec, this is fast enough for me. The TDR can be used as a general purpose high speed sampling oscilloscope if an S-53 Trigger Recogniser is utilised.
7S14 100ps Dual sampler plugin - REPAIR IN PROGRESS
https://w140.com/tekwiki/wiki/7S14
(a) Horizontal sweep only scans at about 1cm/s at its fastest.
Initial investigation suggests the tunnel diode oscillator is faulty, probably due to either or both of the TDs out of spec or dead. I want to test them on the Tek 571 curve tracer but I could find no information on how to do this. I don't want to risk destroying them as they are hard to find and expensive at $30 ea just on eBay.
The 7S14 delayed ramp gen has the following tunnel diodes:
Service Manual circuit page P.5A CR459 Tek 152-0169-00 1.0mA 10pF aka GE 1N3712
Service Manual circuit page P.2 CR220 Tek 152-0177-00 10 mA 4pF a Tek special
Service Manual circuit page P.2 CR221 Tek 152-0070-00 0.1mA 3pF aka GE 4JFBD4 (GP tunnel diode) American Microsemiconductor also made it
However there are a plethora of Russian tunnel diodes on eBay (see my tab 'USSR Tunnel Diodes Xref' for specs). I compared as many as I could find and selected three as close as possible to the Tek parts and ordered several to experiment with on the 571. Once I know how to test these safely, I'll check the Tek ones too, and if they are faulty I'll see if I can replace them with the Russian ones.
(b) Both input channels inoperative
Initally both traces were wiggling up and down randomly at slow speed when plugged into a Tek 7834 and its cal fed into the 7S14 inputs, but they were not showing any evidence of the calibration
waveform. A well known fault with the 7S14 is the 4 Hg batteries hidden in the shielded compartments of its input section have run out - see here:
Sure enough this was the case - this 7S14 was sold as 'tested' but these batteries were so badly corroded I'd guess it hasn't been used for 20 years. The original battery was a Mallory Duracell WH3 1.35V 220mAh Hg cell 11.56mm dia x 5.4mm thick. I followed the excellent advice and replaced them with AG10 = LR54 1.5V 60mAh alkalines 11.6mm x 3.1mm dia only because the only coin cell holders I could find would not take the thicker AG13 (LR44) 200mAh. This produced two steady traces but still no calibration waveform visible. At least the level controls work.
(c) CRT readout is momentary
Another strange fault is the readouts (V/div, T/Div) only appear momentarily at the start of a sweep.
Apart from the above, it would seem it works - even the delayed bright spot used to measure waveforms functions, although of course I have no idea of its accuracy.
To be realistically able to repair the 7S14 requires two of Tek's 067-0616-00 7k series plugin flexible extension leads but as with anything so useful, a high price is commanded on eBay and they usually go for at least $150 each. I found a low cost alternative solution available here:
http://ecosensory.com/tek/tek_7K_flex_sales_blurb.txt
P6015A probe #1 - STILL TO BE REPAIRED
(a) Slow rise and fall times.
New text box
µ Ω ± ° ⌠ ⌡ ∫ │ ─ √ φ θ Θ ∂ δ ζ ξ ς λ ψ ω τ µ Ω ∆ Δ ∑ ∏ π Ξ ○ ≠ ³ ² ±